universidad politÉcnica de madrid escuela tÉcnica...

UNIVERSIDAD POLITÉCNICA DE MADRID

ESCUELA TÉCNICA SUPERIOR DE INGENIEROS INDUSTRIALES

REGULATORY PROPOSALS FOR THE DEVELOPMENT OF AN EFFICIENT IBERIAN ENERGY FORWARD MARKET.

PROPUESTAS REGULATORIAS PARA EL DISEÑO DE UN

MERCADO IBÉRICO A PLAZO EFICIENTE DE LA ENERGÍA.

PhD THESIS TESIS DOCTORAL

Álvaro Capitán Herráiz

Ingeniero Industrial (UPM) & Mecánico (KTH, Suecia) y MSc Internacional en Ingeniería Energética Sostenible (KTH, Suecia)

Director:

Carlos Rodríguez Monroy Doctor Ingeniero Industrial

Ingeniero Industrial Lcdo. en Ciencias Económicas y Empresariales

Lcdo. en Derecho Lcdo. en Sociología y Ciencias Políticas

2014

Tribunal nombrado por el Magnífico y Excelentísimo Sr. Rector de la

Universidad Politécnica de Madrid

Presidente: D. Sergio Martínez Gonzalez

Secretaria: Dª. Rosa María de Castro Fernández

Vocal: D. David Robinson

Vocal: Dª. Gema Rico Rivas

Vocal: D. Fermín Pedro Moreno García

Suplente: D. Santiago Chivite Fernández

Suplente: D. Egbert Rodríguez Messmer

Realizado el acto de lectura y defensa de la tesis el día 14 de julio de 2014, en

Madrid.

Calificación:

PRESIDENTE LOS VOCALES

LA SECRETARIA

To those who have guided me in my educational process, especially to my parents Ramiro and Angelines. To my family, my sustainable energy source: Sonsoles, Arturo, Amalia, Lucas and grannie “tata” Angelines.

ACKNOWLEDGEMENTS

I would like to thank to all my colleagues in the Spanish Energy Regulator (currently Comisión Nacional de los Mercados y la Competencia, CNMC, formerly Comisión Nacional de Energía, CNE) who have provided me with a great support for the completion of my PhD Thesis. In particular, I am very grateful to all my mates in the Energy Derivatives department – we have learnt together since year 2007 in this exciting and fast developing world of energy derivatives –, to my boss in the Gas Markets department during years 2004-2006, Javier Notario Torres – a privilege to learn from his wisdom and knowledge about gas markets, the gas price formation mechanisms, and gas regulation in general –, to the Librarian, José Antonio Sánchez Montero – who always shows a great interest in my research work and provides me with many sources of information related to energy markets –, and to the technical specialists Fermín P. Moreno García – providing me with many insights to build this research in a comprehensive way to understand the evolution of the Spanish electricity sector –, and to Javier Rincón García, who gave me a great guidance for the econometric research. Finally, I thank especially my PhD Thesis Director, Carlos Rodríguez Monroy, for all his support during the whole PhD programme and his confidence and stimulus in publishing in prestigious peer reviewed energy journals.

Regulatory proposals for the development of an efficient Iberian energy forward market, Á. Capitán Herráiz

i

GENERAL INDEX FIGURES INDEX…………………………………………………………………….vii TABLES INDEX………………………………………………………………………ix RESUMEN……………………………………………………………………………..xi ABSTRACT………………………………………………………………………….xiii KEY WORDS / PALABRAS CLAVE………..…………………………………….xv LIST OF ABBREVIATIONS………..………………………………………….…xvii

CHAPTER 1. INTRODUCTION AND METHODOLOGY .................................... 1

1.1 The research question…………………………………………………………….1

1.2 The PhD thesis goal……………………………………………………………....1

1.3 Structure of the PhD Thesis………………………………………………………3

1.4 Fundamentals of Energy Forward Trading……………………………………..4

1.5 The basics of the Iberian Power Futures Market and interrelated market

mechanisms.………..…………………………………………………………………5

1.5.1 The main features of the the spot market, adjustment markets, and ancillary services .......................................................................................... 9

1.6 Methodology of the research performed………………………………………11

1.6.1 Description of tests related to market price efficiency ........................ 11 1.6.1.1 Definition of the Ex-post Forward Risk Premium ...................... 11

1.6.1.2 Description of tests regarding forward risk premium ................. 13

1.6.1.3 Description of tests regarding cointegration analysis of energy prices and analysis of the clean spark spreads……………………………17

1.6.2 Description of the regression model built related to liquidity of energy markets…………………………………………………………………………...19 1.6.3 Description of the hedging efficiency analysis through the net position ratio………………………………………………………………………………..21

1.6.3.1 The relationship between volume and open interest ................. 22

1.6.3.2 The fundamentals of the net position ratio methodology……….22

1.6.3.3 Description of test regarding net position ratio .......................... 23

CHAPTER 2. LITERATURE REVIEW .............................................................. 25

2.1 Introduction……………………………………………………………………....25


ii

2.2 Literature related to market price efficiency …………………………………..25

2.3 Literature related to liquidity of energy markets………………………………30

2.3.1 Academic research ............................................................................ 30 2.3.1.1 Supervision Reports in European Energy Markets .................... 31

2.4 Literature related to ratios measuring the hedging efficiency……………….33

2.4.1 Literature regarding commodity derivatives and application of the Iberian energy derivatives market .............................................................. 33

2.4.1.1 Literature review of the hedge ratio ........................................... 33

2.4.1.2 Literature review of energy markets about analysis of the open interest………………………………………………………………………….34

CHAPTER 3. OVERALL ASSESSMENT OF THE IBERIAN ENERGY DERIVATIVES MARKET AND RELATED REGULATION ....... 37

3.1 Introduction…...............................................................................................37

3.2 The current electricity policy context…………………………………………...38

3.2.1 The subsidised coal fired generation with indigenous coal ................ 38 3.2.1.1 The impact of the recognised price of the indigenous coal fired generation in the Spanish forward price formation .................................. 39

3.2.2 The moratorium to renewables .......................................................... 41 3.2.3 The extension of life cycle of power plants ........................................ 41

3.2.3.1 The effect of the German Nuclear moratorium on power prices 42

3.2.4 The mitigation of large cost deficits in the electricity sector ............... 42 3.2.4.1 Policy recommendations by the National Regulatory Authority . 42

3.2.4.2 The first measures taken by the Government ........................... 43

3.2.5 The introduction of household hourly tariffs ....................................... 43 3.3 Evolution of the trading efficiency………………………………………………45

3.3.1 Volume comparison between Iberian forward trading mechanisms ... 46 3.3.2 Competition in the power futures market ........................................... 49 3.3.3 Key trading drivers in OMIP continuous market ................................. 49 3.3.4 Comparison with the most developed European Exchanges ............. 50

3.4 Evolution of the price efficiency………………………………………………...51

3.4.1 Comparison of the ex-post forward risk premia in the Iberian power futures market and the CESUR auctions.................................................... 52 3.4.2 Economic impact of CESUR auctions in the energy cost of the last resort supply rates ...................................................................................... 53


iii

3.5 Energy policy considerations……………………………………………………54

3.5.1 The need for increased post-trade transparency from the power futures market operator............................................................................... 54 3.5.2 The necessity for trade repositories for a comprehensive oversight by regulators .................................................................................................... 55

3.6 Results.………………………..…………………………………….…………….57

CHAPTER 4. EVALUATION OF THE FORWARD RISK PREMIUM ............... 59

4.1 Introduction……………………………………………………………………….60

4.2 Comparison of the ex-post forward risk premium with some relevant

international energy markets………………………………………………………..61

4.2.1 Test 1 results ...................................................................................... 61 4.2.2 Test 2 results ...................................................................................... 66 4.2.3 Test 3 results ...................................................................................... 67

4.2.3.1 Test 3.1 results ......................................................................... 67

4.2.3.2 Test 3.2 results ......................................................................... 68

4.2.4 Test 4 results ...................................................................................... 73 4.3 Comparison of the ex-post forward risk premia in the Iberian power forward

contracting mechanisms…………………………………………………………….75

4.3.1 Some introductory facts…………………………………………………..75 4.3.2 Analysis of the forward risk premium .................................................. 78

4.4 Results…………………………………………………………………………….83

CHAPTER 5. EVALUATION OF THE LIQUIDITY DEVELOPMENT ................ 87

5.1 Introduction……………………………………………………………………….87

5.2 The basics to assess the liquidity development of the Iberian Power Futures

Market…………………………………………………………………………………88

5.2.1 The derivatives listed in OMIP ............................................................ 88 5.2.2 OMIP market makers as liquidity boosters ......................................... 90

5.2.2.1 The effects of the market maker agreements in the bid ask spread reduction ..................................................................................... 91

5.2.3 Comparison with the most mature European power futures markets.95 5.3 Analysis of the drivers developing the continuous market managed by

OMIP…………………………………………………………………………………..96


iv

5.3.1 Evolution of the traded volumes in the continuous market ................. 96 5.3.2 The enrollment of trading members……………………………………..96 5.3.3 The discounts in OMIP trading fees ................................................... 98 5.3.4 The regression model for the continuous traded volumes ................ .98

5.3.4.1 The regression results……………………………………………...98

5.3.5 Correlation analysis ......................................................................... 101 5.4 Efficiency recommendations………………………………………………..…101

5.4.1 The three-layers liquidity pyramid .................................................... 101 5.4.1.1 The basic layer ........................................................................ 102

5.4.1.2 The intermediate layer ............................................................. 103

5.4.1.3 The top layer ........................................................................... 105

5.5 Results…………………………………………………………………………...106

CHAPTER 6. EVALUATION OF THE FORWARD PRICE FORMATION THROUGH THE GENERATION COST ASSESSMENT ......... 109

6.1 Introduction……………………………………………………………………...109

6.2 Evaluation of the Forward Generation Costs………………………………..111

6.2.1 Correlation Analysis ......................................................................... 111 6.2.2 Cointegration Analysis of Energy Prices .......................................... 118 6.2.3 Analysis of the Clean Spark Spreads with Forward Prices .............. 120

6.3 The first renewable trading and clearing mechanisms in the Iberian

Electricity Forward Market…………………………………………………………123

6.3.1 Renewable auctions in Latin America .............................................. 124 6.3.1.1 The Peruvian case .................................................................. 124

6.3.1.2 The Brazilian case ................................................................... 124

6.3.2 The first mechanisms in the Iberian electricity market ..................... 126 6.3.2.1 The Contract for Differences derived from CESUR auctions in Spain………………………………………………………………………….126

6.3.2.2 The auctions for the sale of the special regime production in Portugal ………………………………………………………………………128

6.4 Results…………………………………………………………………………...130


v

CHAPTER 7. EVALUATION OF THE HEDGING PERFORMANCE BASED ON OPEN INTEREST AND CLEARED VOLUMES ...................... 133

7.1 Introduction……………………………………………………………………...133

7.2 Analysis of the net position ratio of the Spanish electricity derivatives…..134

7.3 The net position ratio and the prudential oversight of the systemic risk….136

7.4 Results…………………………………………………………………………...137

CHAPTER 8. RESULTS, CONCLUSIONS AND FUTURES LINES OF RESEARCH ............................................................................. 141

8.1 Regulatory recommendations…………………………………………………141

8.2 Further research………………………………………………………………..152

CHAPTER 9. REFERENCES……………………………………………………..159

ANNEX: LIST OF PUBLICATIONS………………………………………………179


vi


vii

FIGURES INDEX

Figure 1.1. Evolution of resulting Weighted Average Equilibrium Price in OMIP Call Auctions versus Average Underlying Spot Price (€/MWh)……..............................................13

Figure 3.1. Evolution of the Spanish prompt year (“Y+1”) base load power futures settlement price versus the underlying spot price, the French base load power futures settlement price and the forward GTCC generation costs (€/MWh)...................................................................................................................40

Figure 3.2. Evolution of traded and cleared volumes in OMIP-OMIClear, traded volumes in CESUR auctions, and cleared volumes in MEFF Power (TWh)..............................47

Figure 3.3. Evolution of the ex-post forward risk premia in the Iberian energy derivatives exchange and in CESUR auctions...........................................................................52

Figure 4.1. OMIP Risk Premia in different quotation periods with different Reference Prices.......................................................................................................................62

Figure 4.2. Comparison of OMIP Settlement Prices: Quarterly Contracts versus Weighted Average Monthly Contracts (*) and underlying spot prices (OMEL)........................64

Figure 4.3. OMIP Forward Risk Premia distinguishing Reference Prices per approach to maturity....................................................................................................................69

Figure 4.4. Powernext Forward Risk Premia distinguishing Reference Prices per approach to

maturity........................................................................................................................................69 Figure 4.5. Nord Pool Forward Risk Premia distinguishing Reference Prices per approach to

maturity........................................................................................................................................70

Figure 4.6. NBP Gas Forward Risk Premia distinguishing Reference Prices per approach to

maturity........................................................................................................................................70

Figure 4.7. Brent Forward Risk Premia distinguishing Reference Prices per approach to

maturity........................................................................................................................................71

Figure 4.8. EEX ARA Coal Forward Risk Premia distinguishing Reference Prices per approach

to maturity....................................................................................................................................71

Figure 4.9. Delivered Energy (MWh) per MIBEL Forward Contracting Mechanism....................76

Figure 4.10. Evolution of OMIP Auction and Continuous Volumes of Month Contracts..............77

Figure 4.11. Evolution of OMIP Auction and Continuous Volumes of Quarter and Year

Contracts.....................................................................................................................................77

Figure 4.12. Forward Risk Premia of OMIP Month Futures Contracts........................................79

Figure 4.13. Forward Risk Premia of OMIP Quarter and Year Futures Contracts......................79

Figure 4.14. Forward Risk Premia of OMIP Futures: Global weighted average versus weighted

average of month contracts.........................................................................................................81 Figure 4.15. Quarterly Forward Risk Premia of MIBEL Forward Contracting Mechanisms........82

Figure 5.1. Evolution of OMIP Closing Spreads: Monthly Futures contracts quoting from July 3rd 2006 to July 2nd 2007...............................................................................................92


viii

Figure 5.2. Evolution of OMIP Closing Spreads: Quarterly & Yearly Futures contracts quoting from July 3rd 2006 to July 2nd 2007……...................................................................92

Figure 5.3. Evolution of OMIP Closing Spreads: Monthly Futures contracts quoting from July 3rd 2007 to November 20th 2008....................................................................................93

Figure 5.4. Evolution of OMIP Closing Spreads: Quarterly & Yearly Futures contracts quoting from July 3rd 2007 to November 20th 2008…...........................................................93

Figure 5.5. Enrollment of OMIP trading members …..................................................................97

Figure 5.6. The three-layers liquidity pyramid ….......................................................................102

Figure 6.1. Evolution of power (OMIP), gas (TTF), LNG import prices in Spain and emission (ICE EUA) forward prices (month maturity)............................................................116

Figure 6.2. Evolution of power (OMIP), gas (TTF) and emission (ICE EUA) forward prices (quarter maturity)....................................................................................................117

Figure 6.3. Evolution of power (OMIP), gas (TTF) and emission (ICE EUA) forward prices (year maturity).................................................................................................................117

Figure 6.4. Daily evolution of CSS built with M+1 power and gas contracts and Spanish month LNG index..............................................................................................................121

Figure 6.5. Daily evolution of the CSS built with the prompt quarter power and gas contracts.................................................................................................................121

Figure 6.6. Daily evolution of the CSS built with the prompt year power and gas contracts.................................................................................................................122

Figure 7.1. Evolution of OMIP-OMIClear and MEFF Power net position ratio per delivery month……………………………………………………………………………………..135


ix

TABLES INDEX

Table 1.1. Iberian Regulated Forward Contracting Mechanisms within the MIBEL Framework complementing the OMIP call auctions………….....................................................21

Table 3.1. Correlation coefficients of the Spanish prompt year base load power futures settlement price with the underlying spot price, the French prompt year power futures and the forward GTCC Generation Costs....................................................40

Table 3.2. Evolution of traded volumes in OMIP, CESUR auctions, OTC and demand (TWh).......................................................................................................................49

Table 3.3. Correlation coefficients between OMIP continuous traded volumes and the monthly evolution of key trading drivers..................................................................................50

Table 3.4. Comparison of the Iberian power futures market with the most developed European energy derivatives exchanges..................................................................................51

Table 3.5. Economic impact of the electricity purchased by the Spanish last resort suppliers in

CESUR auctions..........................................................................................................................53 Table 3.6. Main European legislative pieces impacting on energy derivatives

trading..........................................................................................................................................55

Table 4.1. Costs assessment of Energy purchased in OMIP Call Auctions by Spanish Distribution Companies. Distinction per Forward Risk Premium nature..................62

Table 4.2. Costs assessment of Energy purchased in OMIP Call Auctions by Spanish Distribution Companies. Distinction per contract type..............................................63

Table 4.3. Basic Statistics of Fall & Underlying Spot Prices of Monthly Futures Contracts during period Aug.06-Jul.08................................................................................................66

Table 4.4. Basic Statistics of Fall & Underlying Spot Prices of Quarterly Futures Contracts during period Q4.06-Q2.08..................................................................................................66

Table 4.5. t-Student test regarding null hypothesis of no existence (“zero value”) for the Forward Risk Premium...........................................................................................................68

Table 4.6. Regression results regarding compliance with Bessembinder-Lemmon's Hypothesis...............................................................................................................73

Table 4.7. Spanish Regulated Forward Contracting Mechanisms within the MIBEL Framework complementing the OMIP call auctions....................................................................75

Table 4.8. Analysis of OMIP Forward Risk Premia: basic statistics............................................80

Table 4.9. Average OMIP Forward Risk Premia per delivery year for global and maturity

weighted average series..............................................................................................................81

Table 4.10. Basic statistics of the quarterly Forward Risk Premia for the MIBEL forward contracting mechanisms…........................................................................................83

Table 5.1. Derivatives listed in OMIP: basic features and liquidity diagnosis…..........................89

Table 5.2. Market maker agreements within the Iberian Power Futures Market…………..…….91


x

Table 5.3. Evolution of OMIP Closing Spreads: Futures contracts quoting from July 3rd 2006 to

November 20th 2008....................................................................................................................94

Table 5.4. Comparison of the main European Power Derivatives Exchanges with data of year

2008……………………………………………………………………………………………………....95

Table 5.5. Regression model results of traded energy in OMIP continuous market……............99

Table 6.1. Correlation Matrix between Wholesale Energy Prices…..........................................113

Table 6.2. Correlation between Gas Prices (TTF, in €/MWh) and Oil Prices (Brent, $/Bbl)......115

Table 6.3. Dickey-Fuller’s Test for Analysis of Unit Root Variables in Energy Log Price Series.....................................................................................................................118

Table 6.4. Unitary Root Analysis of the Residue in Regression OMIP M+1 versus Fuels in

Columns....................................................................................................................................119

Table 6.5. Regression Results OMIP M+1 versus Fuels shown in columns.............................119

Table 6.6. Comparison of Annual Average CSS per Maturity (“M+1”, “Q+1”, “Y+1”) and CSS

built with Spanish LNG Monthly Index.......................................................................................122


xi

RESUMEN El mercado ibérico de futuros de energía eléctrica gestionado por OMIP

(“Operador do Mercado Ibérico de Energia, Pólo Português”, con sede en Lisboa), también conocido como el mercado ibérico de derivados de energía, comenzó a funcionar el 3 de julio de 2006. Se analiza la eficiencia de este mercado organizado, por lo que se estudia la precisión con la que sus precios de futuros predicen el precio de contado. En dicho mercado coexisten dos modos de negociación: el mercado continuo (modo por defecto) y la contratación mediante subasta. En la negociación en continuo, las órdenes anónimas de compra y de venta interactúan de manera inmediata e individual con órdenes contrarias, dando lugar a operaciones con un número indeterminado de precios para cada contrato. En la negociación a través de subasta, un precio único de equilibrio maximiza el volumen negociado, liquidándose todas las operaciones a ese precio. Adicionalmente, los miembros negociadores de OMIP pueden liquidar operaciones “Over-The-Counter” (OTC) a través de la cámara de compensación de OMIP (OMIClear). Las cinco mayores empresas españolas de distribución de energía eléctrica tenían la obligación de comprar electricidad hasta julio de 2009 en subastas en OMIP, para cubrir parte de sus suministros regulados. De igual manera, el suministrador de último recurso portugués mantuvo tal obligación hasta julio de 2010. Los precios de equilibrio de esas subastas no han resultado óptimos a efectos retributivos de tales suministros regulados dado que dichos precios tienden a situarse ligeramente sesgados al alza. La prima de riesgo ex-post, definida como la diferencia entre los precios a plazo y de contado en el periodo de entrega, se emplea para medir su eficiencia de precio. El mercado de contado, gestionado por OMIE (“Operador de Mercado Ibérico de la Energía”, conocido tradicionalmente como “OMEL”), tiene su sede en Madrid. Durante los dos primeros años del mercado de futuros, la prima de riesgo media tiende a resultar positiva, al igual que en otros mercados europeos de energía eléctrica y gas natural. En ese periodo, la prima de riesgo ex-post tiende a ser negativa en los mercados de petróleo y carbón. Los mercados de energía tienden a mostrar niveles limitados de eficiencia de mercado. La eficiencia de precio del mercado de futuros aumenta con el desarrollo de otros mecanismos coexistentes dentro del mercado ibérico de electricidad (conocido como “MIBEL”) –es decir, el mercado dominante OTC, las subastas de centrales virtuales de generación conocidas en España como Emisiones Primarias de Energía, y las subastas para cubrir parte de los suministros de último recurso conocidas en España como subastas CESUR– y con una mayor integración de los mercados regionales europeos de energía eléctrica.

Se construye un modelo de regresión para analizar la evolución de los volúmenes negociados en el mercado continuo durante sus cuatro primeros años como una función de doce indicadores potenciales de liquidez. Los únicos indicadores significativos son los volúmenes negociados en las subastas obligatorias gestionadas por OMIP, los volúmenes negociados en el mercado OTC y los volúmenes OTC compensados por OMIClear. El número de creadores de mercado, la incorporación de agentes financieros y compañías de generación pertenecientes a grupos integrados con suministradores de último recurso, y los volúmenes OTC compensados por OMIClear muestran una fuerte correlación con los volúmenes negociados en el mercado continuo. La liquidez de OMIP


xii

está aún lejos de los niveles alcanzados por los mercados europeos más maduros (localizados en los países nórdicos (Nasdaq OMX Commodities) y Alemania (EEX)). El operador de mercado y su cámara de compensación podrían desarrollar acciones eficientes de marketing para atraer nuevos agentes activos en el mercado de contado (p.ej. industrias consumidoras intensivas de energía, suministradores, pequeños productores, compañías energéticas internacionales y empresas de energías renovables) y agentes financieros, captar volúmenes del opaco OTC, y mejorar el funcionamiento de los productos existentes aún no líquidos. Resultaría de gran utilidad para tales acciones un diálogo activo con todos los agentes (participantes en el mercado, operador de mercado de contado, y autoridades supervisoras).

Durante sus primeros cinco años y medio, el mercado continuo presenta un crecimento de liquidez estable. Se mide el desempeño de sus funciones de cobertura mediante la ratio de posición neta obtenida al dividir la posición abierta final de un contrato de derivados mensual entre su volumen acumulado en la cámara de compensación. Los futuros carga base muestran la ratio más baja debido a su buena liquidez. Los futuros carga punta muestran una mayor ratio al producirse su menor liquidez a través de contadas subastas fijadas por regulación portuguesa. Las permutas carga base liquidadas en la cámara de compensación ubicada en Madrid –MEFF Power, activa desde el 21 de marzo de 2011– muestran inicialmente valores altos debido a bajos volúmenes registrados, dado que esta cámara se emplea principalmente para vencimientos pequeños (diario y semanal). Dicha ratio puede ser una poderosa herramienta de supervisión para los reguladores energéticos cuando accedan a todas las transacciones de derivados en virtud del Reglamento Europeo sobre Integridad y Transparencia de los Mercados de Energía (“REMIT”), en vigor desde el 28 de diciembre de 2011. La prima de riesgo ex-post tiende a ser positiva en todos los mecanismos (futuros en OMIP, mercado OTC y subastas CESUR) y disminuye debido a la curvas de aprendizaje y al efecto, desde el año 2011, del precio fijo para la retribución de la generación con carbón autóctono. Se realiza una comparativa con los costes a plazo de generación con gas natural (diferencial “clean spark spread”) obtenido como la diferencia entre el precio del futuro eléctrico y el coste a plazo de generación con ciclo combinado internalizando los costes de emisión de CO2. Los futuros eléctricos tienen una elevada correlación con los precios de gas europeos. Los diferenciales de contratos con vencimiento inmediato tienden a ser positivos. Los mayores diferenciales se dan para los contratos mensuales, seguidos de los trimestrales y anuales. Los generadores eléctricos con gas pueden maximizar beneficios con contratos de menor vencimiento. Los informes de monitorización por el operador de mercado que proporcionan transparencia post-operacional, el acceso a datos OTC por el regulador energético, y la valoración del riesgo regulatorio pueden contribuir a ganancias de eficiencia. Estas recomendaciones son también válidas para un potencial mercado ibérico de futuros de gas, una vez que el hub ibérico de gas –actualmente en fase de diseño, con reuniones mensuales de los agentes desde enero de 2013 en el grupo de trabajo liderado por el regulador energético español– esté operativo. El hub ibérico de gas proporcionará transparencia al atraer más agentes y mejorar la competencia, incrementando su eficiencia, dado que en el mercado OTC actual no se revela precio alguno de gas.


xiii

ABSTRACT

The Iberian Power Futures Market, managed by OMIP (“Operador do Mercado Ibérico de Energia, Pólo Português”, located in Lisbon), also known as the Iberian Energy Derivatives Market, started operations on 3 July 2006. The market efficiency, regarding how well the future price predicts the spot price, is analysed for this energy derivatives exchange. There are two trading modes coexisting within OMIP: the continuous market (default mode) and the call auction. In the continuous trading, anonymous buy and sell orders interact immediately and individually with opposite side orders, generating trades with an undetermined number of prices for each contract. In the call auction trading, a single price auction maximizes the traded volume, being all trades settled at the same price (equilibrium price). Additionally, OMIP trading members may settle Over-the-Counter (OTC) trades through OMIP clearing house (OMIClear). The five largest Spanish distribution companies have been obliged to purchase in auctions managed by OMIP until July 2009, in order to partly cover their portfolios of end users’ regulated supplies. Likewise, the Portuguese last resort supplier kept that obligation until July 2010. The auction equilibrium prices are not optimal for remuneration purposes of regulated supplies as such prices seem to be slightly upward biased. The ex-post forward risk premium, defined as the difference between the forward and spot prices in the delivery period, is used to measure its price efficiency. The spot market, managed by OMIE (Market Operator of the Iberian Energy Market, Spanish Pool, known traditionally as “OMEL”), is located in Madrid. During the first two years of the futures market, the average forward risk premium tends to be positive, as it occurs with other European power and natural gas markets. In that period, the ex-post forward risk premium tends to be negative in oil and coal markets. Energy markets tend to show limited levels of market efficiency. The price efficiency of the Iberian Power Futures Market improves with the market development of all the coexistent forward contracting mechanisms within the Iberian Electricity Market (known as “MIBEL”) – namely, the dominant OTC market, the Virtual Power Plant Auctions known in Spain as Energy Primary Emissions, and the auctions catering for part of the last resort supplies known in Spain as CESUR auctions – and with further integration of European Regional Electricity Markets.

A regression model tracking the evolution of the traded volumes in the continuous market during its first four years is built as a function of twelve potential liquidity drivers. The only significant drivers are the traded volumes in OMIP compulsory auctions, the traded volumes in the OTC market, and the OTC cleared volumes by OMIClear. The amount of market makers, the enrolment of financial members and generation companies belonging to the integrated group of last resort suppliers, and the OTC cleared volume by OMIClear show strong correlation with the traded volumes in the continuous market. OMIP liquidity is still far from the levels reached by the most mature European markets (located in the Nordic countries (Nasdaq OMX Commodities) and Germany (EEX)). The market operator and its clearing house could develop efficient marketing actions to attract new entrants active in the spot market (e.g. energy intensive industries, suppliers, small producers, international energy companies and renewable


xiv

generation companies) and financial agents as well as volumes from the opaque OTC market, and to improve the performance of existing illiquid products. An active dialogue with all the stakeholders (market participants, spot market operator, and supervisory authorities) will help to implement such actions.

During its firs five and a half years, the continuous market shows steady liquidity growth. The hedging performance is measured through a net position ratio obtained from the final open interest of a month derivatives contract divided by its accumulated cleared volume. The base load futures in the Iberian energy derivatives exchange show the lowest ratios due to good liquidity. The peak futures show bigger ratios as their reduced liquidity is produced by auctions fixed by Portuguese regulation. The base load swaps settled in the clearing house located in Spain – MEFF Power, operating since 21 March 2011, with a new denomination (BME Clearing) since 9 September 2013 – show initially large values due to low registered volumes, as this clearing house is mainly used for short maturity (daily and weekly swaps). The net position ratio can be a powerful oversight tool for energy regulators when accessing to all the derivatives transactions as envisaged by European regulation on Energy Market Integrity and Transparency (“REMIT”), in force since 28 December 2011. The ex-post forward risk premium tends to be positive in all existing mechanisms (OMIP futures, OTC market and CESUR auctions) and diminishes due to the learning curve and the effect – since year 2011 – of the fixed price retributing the indigenous coal fired generation. Comparison with the forward generation costs from natural gas (“clean spark spread”) – obtained as the difference between the power futures price and the forward generation cost with a gas fired combined cycle plant taking into account the CO2 emission rates – is also performed. The power futures are strongly correlated with European gas prices. The clean spark spreads built with prompt contracts tend to be positive. The biggest clean spark spreads are for the month contract, followed by the quarter contract and then by the year contract. Therefore, gas fired generation companies can maximize profits trading with contracts of shorter maturity. Market monitoring reports by the market operator providing post-trade transparency, OTC data access by the energy regulator, and assessment of the regulatory risk can contribute to efficiency gains. The same recommendations are also valid for a potential Iberian gas futures market, once an Iberian gas hub – currently in a design phase, with monthly meetings amongst the stakeholders in a Working Group led by the Spanish energy regulatory authority since January 2013 – is operating. The Iberian gas hub would bring transparency attracting more shippers and improving competition and thus its efficiency, as no gas price is currently disclosed in the existing OTC market.


xv

KEY WORDS / PALABRAS CLAVE

Energy regulation; Power futures; Market supervision; Market efficiency; Forward risk premium; Risk Management; Energy derivatives

Regulación energética; Futuros de energía eléctrica; Supervisión del mercado; Eficiencia del Mercado; Prima de riesgo a plazo; Gestión del riesgo; Derivados de energía


xvi


xvii

LIST OF ABBREVIATIONS

• ACER: Agency for the Cooperation of Energy Regulators • ACM: Dutch Energy Regulatory & Competition Authority • ADF: Augmented Dickey-Fuller’s Test • AOC: Spanish Gas Virtual Trading Point • ARA: Amsterdam-Rotterdam-Antwerp coal import harbours • ARIAE: Association of the Ibero-American Energy Regulatory Agencies • ARIS: ACER REMIT Information System • ARMA: AutoRegressive Moving Average • BB: Best Bid (the most expensive purchase) • BME: Spanish Bourse and Markets • BO: Best Offer (the cheapest sale) • BOE: Spanish Official Gazette • CAE: Portuguese Energy Purchasing Contracts • CCGT: Combined Cycle Gas Turbine • CCP: Central CounterParty • CEER: Council of European Energy Regulators • CESUR: Energy Contracts for the Last Resort Supply • CfD: Contracts for Differences • CFTC: U.S. Commodity Futures Trading Commission • CMVM: Portuguese Securities Market Commission • CNE: Spanish Energy Commission • CNMC: Spanish Commission of Markets and Competition • CNMV: Spanish Securities Market Commission • COB: California-Oregon Border • COT: Commitments of Traders report • CPUC: California Public Utilities Commission • DERA: Danish Energy Regulatory Authority • DG ENER: European Commission’s Directorate-General for Energy • DG TREN: European Commission’s Directorate-General for Transport and

Energy • DJ-UBSCI: Dow Jones-UBS Commodity Index • DTe: Dutch Office of Energy Regulation • EEC: European Commodity Clearing • EEX: European Energy Exchange • EFET: European Federation of Energy Traders • EI: Swedish Energy Market Inspectorate • ENTSO-E: European Network of Transmission System Operators for Electricity • ENTSO-G: European Network of Transmission System Operators for Gas • EPE: Energy Primary Emissions • EPEX: European Power Exchange • ERSE: Portuguese Regulatory Entity for the Energy Services • EU: European Union • EUA: European Union Allowances • EU ETS: European Union Emissions Trading System • FERC: Federal Energy Regulatory Commission • FIFO: First In First Out • FTB: Futures Base load • FTR: Financial Transmission Right


xviii

• GARCH: Generalized AutoRegressive Conditional Heteroscedasticity • GDP: Gross Domestic Product • GSE: Italian Manager of Energy Services • GTCC: Gas Turbine Combined Cycle • GTS: Spanish Gas TSO • HHI: Herfindahl-Hirschman index • ICE: InterContinental Exchange • ISO: Independent System Operator • LIFFE: The London International Financial Futures and Options Exchange • LNG: Liquefied Natural Gas • MEFF: Spanish Financial Futures Market • MEI: Portuguese Ministry of Economy and Innovation • MIBEL: Iberian Electricity Market • MiFID: EU Directive on Markets in Financial Instruments • MINECO: Spanish Ministry of Economy and Competitiveness • MINETUR: Spanish Ministry of Industry, Energy and Tourism • MISO: North American Midwest Independent System Operator • MITyC: Ministry of Industry, Trade and Commerce • MoU: Memorandum of Understanding • NBP: National Balancing Point • NMa: the Netherlands Competition Authority • Nord Pool: The Nordic Power Exchange • NordReg: Nordic Energy Regulators • NVE: Norwegian Water Resources and Energy Directorate • NYMEX: New York Mercantile Exchange • OECD: Organisation for Economic Co-operation and Development • OFGEM: British regulatory agency for electricity and gas markets • OMIE: Iberian Market Operator, Spanish Pool • OMIP: Iberian Market Operator, Portuguese Pool • OMIClear: Clearinghouse of the Iberian Market Operator, Portuguese Pool • OPEC: The Organization of the Petroleum Exporting Countries • OTC: Over-The-Counter • OTE: Czech electricity and gas market operator • PJM: Pennsylvania/New Jersey/Maryland Power Market • PRN: Settlement Price (Negotiation Reference Price) • PTEL: Portuguese Electricity Spot Price Index • PVPC: Voluntary Price for the Small Consumer • REE: Spanish Electricity System Operator • REMIT: Regulation on Wholesale Energy Market Integrity and Transparency • REN: Portuguese Electricity System Operator • SFE: Sydney Futures Exchange • SIMEX: The Singapore International Monetary Exchange • SPEL: Spanish Electricity Spot Price Index • STEM: Swedish Energy Agency • TSO: Transmission System Operator • TTF: Dutch Title Transfer Facility • TWC: Tradable White Certificate • UMM: Urgent Market Message • US: The United States of America • VPP: Virtual Power Plant • WTI: West Texas Intermediate crude


1

CHAPTER 1. INTRODUCTION AND METHODOLOGY

This chapter describes the key research question, the goals of the research, the structure of the document and the methodology employed. Furthermore, the description of the diverse research tests is provided.

1.1 The research question

The key research question can be formulated as follows: Has the Iberian power futures market performed efficiently during its first years of existence? In order to answer accurately this key question, the following questions have also been analysed: Are the Iberian power future prices good predictors of the future spot price?, in other words, has the Iberian electricity forward risk premium measured ex-post been larger than in other comparable international energy markets? Have there been remarkable price arbitrage opportunities with the related forward contracting mechanisms in the Iberian electricity market? Has the liquidity of the Iberian power futures market evolved properly based on sound liquidity drivers? Have there been policy measures interfering in the fair price formation of the Iberian power futures market? Can the supervision of the forward price mechanisms in the Iberian electricity market be strengthened? Do the Iberian electricity futures prices provide a robust price reference for the generation companies to assess the profitability of their combined cycle gas turbine units? Can renewable market mechanisms be introduced to better reflect the renewable generation effect in the forward price formation and contribute to a more sustainable energy system? Are the Iberian power futures market and the Spanish and Portuguese clearing houses useful instruments for the Iberian market participants to satisfy their hedging needs?

1.2 The PhD thesis goal

The PhD thesis aims to assess by means of quantitative and qualitative analyses the performance of the Iberian power futures market in its first years of existence, taking into account its interrelationship with other forward market mechanisms, namely, the non-organised market (known as the Over-The-Counter or “OTC” market) and the auctions catering for part of the last resort supplies (CESUR auctions, where CESUR stands in Spanish for “Contracts of energy for the last resort supply”). The Iberian power futures market managed by OMIP (Iberian Market Operator, Portuguese Pool) started on 3 July 2006. The market efficiency is analysed and regulatory proposals are formulated to streamline the market performance of the Iberian electricity forward market.

In order to perform the research, the thesis is structured in different goals, namely:


2

• Goal 1: An analysis of the evolution of the ex-post forward risk premium of the Iberian power futures market is done compared to other international energy markets as well as other electricity forward market mechanisms. The ex-post forward risk premium is defined as the difference between the futures price with a given maturity (i.e. applying for a specific supply period or time horizon) and the resulting average spot price in that time horizon. Diverse publications during years 2008-2010 cover this research goal (Capitán Herráiz and Rodríguez Monroy, 2008a; 2008b; 2009b; 2010a; 2010b).

• Goal 2: The Iberian power futures market is composed of two trading modes: “auctions” (basically diverse auctions are arranged to comply with Spanish and Portuguese electricity regulatory pieces in order to foster liquidity in this emerging market) and “continuous” (the main mode). An analysis of the monthly evolution of the traded volumes in OMIP continuous market is done by means of a regression model detecting the key factors (i.e trading drivers) for the development of such a market. Diverse publications during years 2011-2013 cover this research goal (Capitán Herráiz and Rodríguez Monroy, 2011a; 2013a).

• Goal 3: The evolution of the Iberian power futures market is analysed in terms of liquidity compared to the main European Energy exchanges (i.e organised markets, in the case of spot electricity markets they are usually known as power pools). An analysis of the main regulatory features with incidence in the price formation of this market is provided. Regulatory proposals for the supervision of this market in order to increase its efficiency are formulated. A publication in year 2012 covers this research goal (Capitán Herráiz and Rodríguez Monroy, 2012b).

• Goal 4: The analysis of the price formation in the Iberian power futures market is completed through its comparison with the forward estimation of the generation costs of the gas fired power plants (i.e. the Combined Cycle Gas Turbines, usually known as “CCGT” plants). Therefore, the forward prices of the fuel (natural gas) and the futures prices of the CO2 emission allowances are considered in the calculation of the forward estimation of the generation costs. The price differential obtained as the difference between the electricity futures price and the generation cost assessment (internalizing the CO2 price) is commonly known as the “clean spark spread”. Due to the relevance of the renewable energy sources in the generation mix of both Spain and Portugal, a first assessment of the existing Iberian forward market/settlement mechanisms based in the renewable generation is also done, and regulatory reflections are also provided to consider in the most efficient manner the renewable generation in the Iberian electricity forward price formation. A review of pioneering renewable generation auction mechanisms in Latin American countries is provided to gain insights for their potential implementation in the Iberian electricity market. Diverse publications during years 2012-2013 cover this research goal (Capitán Herráiz and Rodríguez Monroy, 2012a; 2013c).

• Goal 5: The efficiency in terms of the hedging needs of the market participants in the Iberian electricity forward market is measured through the analysis of an


3

original indicator termed the “net position ratio”. Such a ratio is obtained through the net final position for a given maturity (i.e. the final open interest for a given contract, reflecting the final open positions by the market participants) divided by the total cleared volume in the clearing house for such a contract. Diverse publications during years 2011-2013 cover this research goal (Capitán Herráiz and Rodríguez Monroy, 2011b; 2012b; 2013b).

1.3 Structure of the PhD Thesis

The PhD Thesis is structured in different chapters revolving around the goals mentioned above, in particular:

• Chapter 1 “Introduction and Methodology” provides a basic introduction to understand the fundamentals of the wholesale electricity markets, composed of spot and forward markets. Both market segments (i.e the segment closed to delivery and the segment linked to the use of derivatives (e.g futures and options)) are closely interrelated as their price formation is influenced by each other. Therefore, a comprehensive supervision taking into account such interactions is requested (see e.g. European Union, 2011). A brief description of the existing Iberian forward market mechanisms and the spot market (as well as the adjustment market and the ancillary services close to real-time, i.e. the physical delivery) is provided to understand the research themes developed in the succeeding chapters. The description of the research methodology explaining the tests performed in those subsequent chapters are also provided.

• Chapter 2 “Literature review” provides the theoretical basis from academic and regulatory literature consulted to perform the different research tests described in Chapter 2.

• Chapter 3 “Overall Assessment of the Iberian Energy Derivatives Market and Related Regulation” provides a general overview of the research topics and is related to Goal 3 described above.

• Chapter 4 “Evaluation of the Forward Risk Premium” provides a quantitative and qualitative analysis of a key indicator for measuring the price efficiency (the ex-post forward risk premium) and is related to Goal 1 described above.

• Chapter 5 “Evaluation of the Liquidity Development” analyses the evolution of the traded volumes in the continuos trading mode of the Iberian power futures market as a key liquidity measure and is related to Goal 2 described above.

• Chapter 6 “Evaluation of the Forward Price Formation through the Generation Cost Assessment” analyses the evolution of the price differential obtained as the difference between the Iberian power futures prices and the forward estimation of the generation costs with a Combined Cycle Gas Turbine (the main technology fixing the marginal price in the Spanish day-ahead market when the wind mills production is low). The first renewable trading/settlement


4

mechanisms within the MIBEL are also analysed. This chapter is related to Goal 4 described above.

• Chapter 7 “Evaluation of the hedging performance based on open interest and cleared volumes” analyses the evolution of the final open positions in the Spanish derivatives contracts when the trading period of a given contract expires compared to the accumulated cleared volumes for that contract either in the Portuguese clearing house (OMIClear) or in the Spanish clearing house (MEFF Power, currently called BME Clearing). Such a research intends to create a methodology useful for supervisory purposes in order to understand the trading and hedging behavioural patterns of the market participants. This Chapter is related to Goal 5 described above.

• Chapter 8 “Results, conclusions and futures lines of research” summarizes the main findings of the research, provide regulatory recommendations, and suggests further research topics in this emerging and fast developing area.

• Chapter 9 “References”.

• Annex “List of publications”.

1.4 Fundamentals of Energy Forward Trading

The liberalization process in the energy sector and the determination of spot prices through market mechanisms have produced the development of forward contracting. Derivatives are widely used, as their main function is the risk management (Martín Martínez and Villaplana Conde, 2009). Derivatives are financial instruments whose characteristics and value depend upon the characteristics and value of an underlying product (e.g. the electricity spot price) (NordREG, 2010). Forward markets play an important role as a mechanism for transferring risk and in the process of gathering information that leads to price discovery. If two agents who are exposed to the same risk but hold opposite positions (e.g. a power generation company (usually a seller) and a large industrial firm (usually a buyer)) wish to reduce their exposure to future spot price fluctuations, they will be interested in doing a forward transaction. Those agents less/more tolerant to risk may be willing to pay/receive a premium to reduce/increase their risk exposure. Forward trading can take place in futures markets (subject to regulation) or in OTC markets. To reduce counterparty search costs and increase liquidity, the OTC markets can be organized around brokers. In the case of futures markets, the central clearing house bears the counterparty risk, as it becomes the seller to every buyer, and the buyer to every seller (Villaplana and Cartea, 2012). Hedging eliminates financial difficulties produced by adverse price movements. Hedging can have a positive effect on the financial stability of utilities that use forward markets to protect their spot positions (Furió and Chuliá, 2012). Every market participant should assess the benefits of participating in forward markets (e.g. for portfolio hedging or arbitrage/speculation gains) against the inherent costs (e.g brokerage/clearing house fees, and credit line/collateral requirements).


5

As indicated by Köhlner (2014a), the decision to clear (i.e to register the OTC transactions in a clearing house) or not belongs to the market participant, taking into account all the costs, time and effort for both options (i.e. the use of the clearing house or pure OTC trading not centrally cleared and settled). Regarding the use of central counterparties (CCP, the abbreviation commonly used for the clearing houses) there are direct costs (e.g. clearing fees (annual fees and variable fees per registered kWh) and the cost of the collateral pledged with the clearing house). However, regarding OTC trading, there are many hidden costs (and to some extent, complex back-office procedures) that should be considered in such an economic assessment, e.g.: the establishment and management of the rule book by negotiating European Federation of Energy Traders (EFET) master agreements with all the counterparties (this can take more than one year); risk analysis to establish individual counterparty limits; risk metrics to monitor on a frequent basis; counterparty confirmations and market conformity checks to prevent fraud or improper behaviour; adequate staff resources (lawyers, risk and compliance officers, and a back-office team). The collateral requirements are optimised in the energy derivatives clearing houses through margin offsets considering all the cleared market participant’s transactions in energy derivatives (e.g. power, natural gas, coal, white (i.e. energy efficiency) certificates, et cetera) and CO2 emissions. The clearing house can be an interesting option for those market participants interested in trading through an organised market place in an anonymous way. It might seem reasonable to keep pure OTC trading for taylor-made transactions (i.e. not standardised contracts) and employ the CCP to reduce multiple risk exposures and cash flows to one net position with such trusted counterparty with sound credit rating.

1.5 The basics of the Iberian Power Futures Market and interrelated market mechanisms

Since its beginning in July 2006, the Iberian Power Futures Market managed by OMIP (Iberian Forward Market Operator), within the framework of the Iberian Electricity Market (MIBEL), has experienced a continuous development, in terms of number of participants and liquidity. MIBEL stands for Iberian Electricity Market, composed of the Spanish and Portuguese markets. A common matching algorithm in the spot market based on market splitting is applied since 1 July 2007.

During the first three years of that futures market, the main amount of traded energy in OMIP was driven by compulsory call auctions according to national regulations aimed at fostering the MIBEL. The International Agreement of Santiago de Compostela, signed by the Portuguese and the Spanish governments on October 1, 2004, related to the constitution of the MIBEL between the Spanish kingdom and the Portuguese republic, indicates that OMIP compulsory call auctions would serve as a transitory mechanism to foster the liquidity of the continuous market managed by OMIP. The Spanish Distribution Companies and the Portuguese Last Resort Supplier with more than 100,000 clients were obliged to purchase in these auctions, in order to


6

partly cover their portfolios of end-user regulated supplies. Such an obligation comprises 5% of their regulated supplies, for the 2nd half of year 2006, as agreed by MIBEL Council of Regulators in the Évora Summit (November 2005), and published in the corresponding legislation (Spanish Order ITC/2129/2006 and Portuguese “Portaria” 643/2006), and 10% for year 2007 onwards, as agreed in the Badajoz Summit (November 2006), and published in Spanish Order ITC/3990/2006 and Portuguese Dispatch 780/2007 (for 1st half of year 2007), Spanish Order ITC/1865/2007 and Portuguese Dispatch /2007 of 29 June 2007 (for 2nd half of year 2007 and 1st half of year 2008), Spanish Order ITC/1934/2008 and Portuguese Dispatch 19098/2008 (for 2nd half of year 2008), and Spanish Order ITC/3789/2008 and Portuguese Dispatch 125-A/2009 (for 1st half of year 2009) (Fernández Domínguez and Xiberta Bernat, 2007; Capitán Herráiz and Rodríguez Monroy, 2009b; MITyC, 2006a, 2006b, 2007c, 2008c, 2008d; Ministério da Economia e da Inovação (Portugal), 2006, 2007a, 2007b, 2008, 2009).

Since November 18th 2008, the Iberian Power Futures Market has the EU Regulated Market status, according to Directive 2004/39/EC of the European Parliament and of the Council of April 21st 2004 on Markets in Financial Instruments (MiFID), following the registration with the Portugese Securities Market Commission (Comissão do Mercado de Valores Mobiliários, CMVM) on October 30th, 2008. Whereas OMIP works as Market Operator of the MIBEL Derivatives Market, OMIClear acts as the clearing house. There are two trading modes coexisting within OMIP: the continuous market (default mode) and the call auction. In the continuous trading, anonymous buy and sell orders interact immediately and individually with opposite side orders, generating trades with an undetermined number of prices for each contract. Buy orders with the highest prices and sell orders with the lowest prices are executed first. In the call auction trading, a single price auction maximizes the traded volume, being all trades settled at the same price (equilibrium price). The call auction algorithm is based on the maximum tradable volume and minimum price criteria, following a First In First Out (“FIFO”) allocation method. Additionally, OMIP trading members may settle Over The Counter (OTC) trades through OMIClear, either registrating their transactions by themselves or through a broker, experiencing this activity a remarkable growth in the last quarter of year 2008 due to the difficulties of holding credit lines in the global financial turmoil during those years. OMIP trading sessions are composed of the following time windows: pre-trade phase happens between 8 a.m.-9 a.m.; auction phase between 9a.m.-9:10a.m.; continuous trading phase between 9:10 a.m.-4:30 p.m.; and pre-close phase between 4:30 p.m.-6:30 p.m. During the first years of existence of the futures market, in the first four Wednesdays of each month, the auction phase was extended until 10a.m. as the compulsory call auctions where Spanish Distribution Companies and the Portuguese Last Resort Supplier were to purchase regulatorily fixed volumes occured (OMIP-OMIClear, 2008a). The last procurement auctions for the Spanish Distribution Companies were held on 2009 and for the Portuguese Last Resort Supplier on 2010. Since 1 April 2013, the time window for the auction phase disappears, thus the continuous phase spans from 9 a.m-4:30 p.m. Auctions can now be held in any of the three phases (OMIP, 2012b; 2013).


7

As stated by Martín Martínez and Villaplana Conde (2009), the regulatory development in Spain and Portugal towards the establishment of last resort tariffs competing with the liberalised market fosters the electricity derivatives use as hedging instruments for the market participants. In this sense, since June 2007 other forward contracting mechanisms based on compulsory auctions have been created, namely: Virtual Power Plant (VPP) Auctions (in Spanish, the so-called “Emisiones Primarias de Energía” or EPE auctions), and the Last Resort Supply Auctions (in Spanish, the so-called “Contratos de Energía para Suministros de Último Recurso” or CESUR auctions). The coexistence of these instruments boosted the liquidity of energy derivatives for risk management purposes, both within OMIP market as well as in the OTC market, the latter already active since 1999. The Royal Decree 1634/2006, of 29 December, established the five first EPE auctions with physical delivery (MITyC, 2006c). Such auctions are regulated by Resolution of the Energy General Secretariat of 19 April 2007 (MITyC, 2007b). These auctions, call options regarding the “virtual” capacity of the Spanish incumbent generators (Endesa and Iberdrola) were auctioned following a multi-round ascending clock algorithm (i.e. electronic auctions) in order to mitigate market power, as previously done in other European and North-American markets. The successful bidders pay the option price (the premium) and are thus are granted with the right to access to this power capacity. When the option expires, if the successful bidder wants to execute the option, he pays for the strike price. This price was previously fixed as these options were of European nature.

In the financial literature, a call option, is usually defined as a financial contract in which the buyer has the right, but not the obligation to buy an agreed quantity of a particular commodity or financial instrument (the underlying) from the seller of the option at a certain time (the expiration date) for a certain price (the strike price). The seller is obliged to sell the commodity or financial instrument if the buyer so decides. The buyer pays a fee (the premium) for this right. The opposite case is the put option, where the buyer has the right, but not the obligation, to sell the asset at a specified price (the strike), by a predetermined date (the expiry or maturity) to the seller. The majority of options are either European or American (style) options, both known as "vanilla options". Whereas a European option may be exercised only at the expiration date of the option, i.e. at a single pre-defined point in time, an American option may be exercised at any time before the expiration date. Options where the payoff is calculated differently are called "exotic options" (Hull, 2005).

The Spanish EPE auctions are regulated by the Ministry of Industry, Trade and Commerce (MITyC), supervised by the Spanish Energy Regulator (Comisión Nacional de Energía or CNE), and managed by an independent entity (CNE, 2009). MITyC does not exist currently. Its energy functions are taken over in November 2011, with the change of Government, by the Ministry of Industry, Energy and Tourism (MINETUR). The CNE is extinguished due to the merging of diverse Spanish Regulatory Authorities under the so-called “Spanish Commission of Markets and Competition” (CNMC). The CNMC started its functions on October 7, 2013 (MINECO, 2013).

The equilibrium price in the Spanish EPE auctions is determined when the demand equals the targeted volume at a price bigger than the reserve price. In auctions theory, the reserve or reservation price, commonly used in such market


8

mechanisms, is the highest price a buyer is willing to pay for goods or a service, or the smallest price at which a seller is willing to sell a good or service (Steedman, 1987). In the case of the Spanish EPE auctions, the reserve price is the minimum price of the premium allowed to the bidders.

The Royal Decree 324/2008, of 29 February, establishes a calendar for the sixth and seventh EPE auctions, in this case both settled by differences (i.e. pure financial settlement without physical delivery), being celebrated in September 2008 and March 2009 (MITyC, 2008a). These two auctions are regulated by Resolution of the Energy General Secretariat of 13 May 2008 (MITyc, 2008b). On the other hand, the CESUR auctions are a forward contracting mechanism for the Spanish distribution companies and the Portuguese last resort supplier, complementing their procurements in the OMIP call auctions as well as in the spot market. The CESUR auctions have facilitated the entry of new agents and foster the development of coexisting forward contracting mechanisms. The CESUR auctions, as stated in Order ITC/400/2007, of 26 February, contribute to the price valuation of the energy component included in the last resort tariffs, and intend to prevent further undesirable regulated tariff deficits (MITyC, 2007a). The CESUR auctions are also iniatially regulated by MITyC (afterwards, by MINETUR), supervised by CNE (afterwards, by CNMC), and managed by an independent entity. They are also electronic though based on a multi-round descending clock algorithm. In CESUR auctions, the equilibrium prices are determined when the selling volumes equal the targeted volume. From the ninth CESUR auction –held on June 25, 2009– onwards, the distribution companies no longer acquired energy but the last resort suppliers, as the latter have taken over the regulated supplies of the former since July 1, 2009, and the settlement is pure financial, as indicated in MITyc (2009c). Only the 5 Spanish last resort suppliers participate on the demand side since the 9th auction. A detailed description of the regulatory features and a synopsis of the results of these auctions are provided by MIBEL Regulatory Council (2009).

In Portugal, four VPP auctions have been held. They were managed by OMIP and cleared and settled by OMIClear during the years 2007 and 2008 (MIBEL Regulatory Council, 2009; Capitán Herráiz and Rodríguez Monroy, 2010a). The first two auctions were celebrated by initiative of REN Trading, a subsidiary company of the Portuguese System Operator (Redes Energéticas Nacionais, SGPS, S.A., known as “REN”). The third and fourth auctions were promoted by REN Trading and by EDP Power Generation Management, a subsidiary company of the main Portuguese utility. These auctions released energy traditionally captive to Energy Purchasing Contracts (known in Portuguese as “CAE”, Contratos de Aquisição de Energia) established between the incumbent and certain power plants. The products auctioned were call options of European nature with monthly and quarterly maturity, covering prompt months and quarters. They were base load products with underlying price the spot electricity price of the Spanish zone.

Extensive analysis of the EPE and CESUR auctions held until year 2008 is provided by Federico and Vives (2008). They conclude that the EPE dimension (i.e. the offered capacity by the incumbents) is still limited and these auctions have dealt so far short duration contracts, thus rendering relatively ineffective regarding market power reduction goals. They also conclude that CESUR auctions are unlikely to have a strong


9

pro-competitive impact in the market, as there is no obligation to participate. Nevertheless, they think that the introduction of longer contracts in CESUR auctions would improve the price valuation of the energy component of the last resort tariff. On the other hand, a good legal analysis of the start of the Iberian power futures market is provided by Pereira da Costa and Lage (2006). They find that a different fiscal policy for the diverse instruments traded in OMIP could damage the liquidity development of the financial products with a heavier fiscal burden (forwards) and therefore they propose a change in the Portuguese securities law for the proper functioning of the Iberian power futures market.

To complete the whole overview of the electricity derivatives traded within the MIBEL, the hedging instruments for cross-border trading (i.e derivatives to manage the price spread between the spot prices in the Spanish and Portuguese interconnection) are briefly described in Chapter 5 and the OMIP auctions for the sale of the Portuguese special regime production (i.e. renewable sources and cogeneration) are described in Chapter 6.

1.5.1 The main features of the the spot market, adjustment markets, and ancillary services

As stated above, the derivatives traded in the futures market, OTC or through CESUR auctions have all of them as underlying instrument the electricity price in the day-ahead market (commonly known as the spot price). In order to understand the physical spot markets, comprising both day-ahead and intraday markets, as well as other real-time market mechanisms, a brief description of such physical markets close in time to the physical delivery of the electricity traded is provided in this subsection.

As indicated by the Iberian electricity spot market operator (OMIE, 2013b), the

electricity market is the set of transactions resulting from the trading by the market participants in the sessions of the day-ahead and intraday markets, from the application of the System Technical Operation Procedures, as well as in the forward market as previously described. Physical bilateral contracts concluded by buyers and sellers are incorporated in the production market once the day-ahead market has closed. Market participants are companies authorised to participate in the electricity production market as electricity buyers and sellers. Entities that are authorised to engage in the market are electricity producers, last resort suppliers, resellers, retailers, direct consumers and companies or consumers resident in other countries that are authorised to participate as resellers. Producers and direct consumers may participate in the market as market participants or sign physical bilateral contracts. The economic management of the electricity market is entrusted to the market operator OMI-POLO ESPAÑOL, S.A. (“OMIE”). The spot market is composed of the day-ahead market and the intraday market managed by the market operator. Additionally, there are other market segments managed by the System Operator (Red Eléctrica de España, REE) to adjust the generation and the demand (in particular, the resolution of technical


10

constraints, the ancillary services and the deviation management). The processes in the production market are as follows:

• Most transactions are carried out in the day-ahead market. In this market, a

daily auction matching the suppy and demand for each of the 24 hours of the following day is performed. The auction results in 24 different equilibrium prices, one per delivery hour. All the matched supply and consumption offers are cleared at that equilibrium price. The result ensures that the maximum interconnection capacity with the external electricity systems (i.e. Portugal, France and Morocco) is not exceeded, considering physical bilateral contracts that affect those international interconnection points.

• The intraday market is an adjustment market composed of 6 individual auctions

corresponding to subsequent sessions approaching the delivery day. Therefore, the first session is the longest one (27 hours, i.e. 27 individual equilibrium prices are obtained from the matching process), followed by shorter sessions (24 hours in the 2nd session; 20 hours in the 3rd session; 17 hours in the 4th session; 13 hours in the 5th session; and 9 hours in the 6th session). This is different to the typical intraday market design in the rest of European countries, consisting on continuous trading mode.

• Resolution of technical constraints. Once the day-ahead market session has been held and national physical bilateral bids have been received, the System Operator evaluates the technical viability of the operating schedule of the production units in order to guarantee the safety and reliability of supply on the transmission network. If the aggregation of the day-ahead matched quantities and physical bilateral contracts does not respect the maximum exchange capacity between electricity systems or the mandatory security requirements, the technical constraints solution procedure is applied. This procedure consists, firstly, of the modification of purchases or sales from external electricity systems responsible for this excess in interconnection exchanges and, secondly, of the allocation of the power from the production units.

• The purpose of ancillary services and deviation management is to ensure that

energy is supplied under established conditions of quality, reliability and security and that production and demand are balanced at all times. The System Operator incorporates regulating band ancillary services in the viable daily schedule after the day-ahead market sessions have been held. After every intraday market session, the system operator manages any deviations in real time using ancillary services and the deviation management procedure.

Another good description of the physical markets close to real time, regarding the monitoring tasks of the Spanish National Regulatory Authority can be found in Section 2.2 of CEER (2011) as well as in MIBEL Regulatory Council (2009). Since 1 July 2007, two MIBEL price areas coexist. They have the same spot price with lack of congestion but different price when congestion occurs in the interconnection between Spain and


11

Portugal. In this case, the market splitting algorithm is applied (MIBEL Regulatory Council, 2009). It is important to remark that the price formation between the energy derivatives market and the spot market is interlinked and therefore the proper supervision of such wholesale energy markets should cover in a comprehensive manner all those market segments, as indicated in the European Regulation in wholesale energy market and transparency, commonly known as “REMIT” (European Union, 2011).

In line with the process of integrating the Electricity Wholesale Markets in the EU, OMIE changes the Iberian Day-Ahead Market gate closure time to 12:00 h on October 15, 2013. This change is a necessary step towards a single European electricity wholesale market (OMIE, 2013a). In this sense, the Price Coupling of Regions (PCR) initiative started successfully operations on 4 February 2014, providing a single matching algorithm (named “Euphemia”) for almost all the power exchanges in Europe. The following European power exchanges are involved in such an initiative: Anglo-Dutch APX and Belgian Belpex (both exchanges belong to the same corporate holding), French EPEX Spot, Italian GME, Norwegian Nord Pool Spot, Spanish OMIE, and Czech OTE. The Iberian day-ahead market is transitorily coupled in the PCR initiative in a synchronous way (i.e. the transmission capacity between France and Spain is still only offered via explicit auctions) but it is intended to be fully coupled in May 2014 (OMIE, 2014b). Analogously for the intra-day segment, on 10 February 2014, the Power Exchanges APX, Belpex, EPEX Spot, Nord Pool Spot and OMIE signed a cooperation agreement for a common European cross border intraday solution. In addition, an early start agreement was signed with Deutsche Börse AG for the delivery of a technical system (OMIE, 2014a).

The PCR initiative will provide price convergence amongst the involved spot markets. Such a convergence will also be reflected in the forward markets, improving overall the wholesale price formation in Europe (i.e. price efficiency).

1.6 Methodology of the research performed

This section explains the research methodology, basically composed of different empirical tests and a regression model.

1.6.1 Description of tests related to market price efficiency

1.6.1.1 Definition of the Ex-post Forward Risk Premium

The research presented in Chapter 4 is focused on the analysis of the forward risk premium in the Iberian power futures market comparing different settlement price criteria and comparing the magnitudes of such risk premium with other European power markets and even other fuel markets of interest. There are some studies regarding market efficiency based on the evaluation of forward risk premium. Some of those studies are based on theoretical “ex-ante” analysis by modelling forecasted spot


12

prices. Other studies use empirical data and evaluate “ex-post” the differences between the futures and spot prices. This research represents an empirical analysis using the “ex-post” forward risk premium.

The “ex-ante” forward risk premium (“Δex-ante”) can be matematically expressed as indicated in Equation (1.1):

Δex-ante = Ft,T - Et(ST) (1.1)

Where Ft,T refers to the futures power price observed on day “t” for delivery over period “T”, and Et(ST) refers to expected spot price on day “t” for delivery over period “T”.

The “ex-post” forward risk premium (“Δex-post”) can be mathematically expressed as indicated in Equation (1.2):

Δex-post = Ft,T - Average(ST) (1.2)

Where Average(ST) refers to average spot price for delivery over period T.

In this research, the considered futures contracts are baseload and with monthly and quarterly maturity. Three European power markets are considered, with all their prices in €/MWh: OMIP (Iberian Market), Powernext (French Market), and Nord Pool (Nordic Market). The considered fuel markets correspond to oil (InterContinental Exchange (ICE) Brent futures and spot Over the Counter (OTC) assessments by Platts (the Brent crude is the main reference in European markets as it relates to the North Sea production; in the United States, the main crude reference is the West Texas Intermediate (WTI)); only monthly futures are analysed, expressed in US$/Bbl), natural gas (ICE monthly futures and OTC quarterly Platts’ assessments, all related to the British National Balancing Point (NBP), and expressed in GB pence/therm), and coal (European Energy Exchange (EEX) Amsterdam-Rotterdam-Antwerp (ARA) coal futures, related to the underlying Argus McCloskey weekly spot index, expressed in US$/t; the ARA coal is the main coal reference used in Europe as ARA relates to the main harbours in North West Europe).

As different monetary units and energy units are used (original units for each market), Forward Risk Premium expressed in percentage over the futures price is preferred when comparing all these markets. Such an expression is matematically written as indicated in Equation (1.3) (e.g. Furió and Meneu, 2010):

Δex-post % = [Ft,T - Average(ST)] / Ft,T (1.3)

The selected period for the tests regarding forward risk premium corresponds to the first two years of existence of OMIP market, which started on 3rd July 2006. Therefore, the monthly contracts span from August 2006 to July 2008, and the quarterly ones from Q4-06 to Q2-08.


13

1.6.1.2 Description of tests regarding forward risk premium

1.6.1.2.1 Test 1: Assessment of OMIP Auction Equilibrium Prices

The main purpose of this test is to detect if the equilibrium prices of the compulsory auctions for the purchases by the Spanish distribution companies and the Portuguese last resort suppliers are upward biased.

As explained in Section 1.5 and in the introduction of Chapter 3, the Spanish Local Distribution Companies (and the Portuguese Last Resort Supplier) are obliged to purchase during the second half of year 2006 5% of their regulated power supplies (10% from year 2007 onwards) in OMIP call auctions. If they do not comply with such obligations, each national regulation establishes different penalties. Due to that fact, those companies have satisfactorily purchased their required amounts in all the OMIP call auctions. According to the legislation mentioned in Section 1.5 (“Orders ITC”), the cost of the energy purchased by the Spanish Distribution Companies in the OMIP call auctions is recognised through the resulting equilibrium price of each call auction.

Since the start of OMIP (in terms of quarterly periods, from Q4-06), all the auctioned settled contracts have experienced positive forward risk premia until October 2007 (in terms of quarterly periods, until Q4-07), when a trend change is appreciated and negative risk premia become dominant during quarters Q4-07 and Q1-08. The forward risk premium is again positive along Q2-08, as shown in Figure 1.1.

Figure 1.1. Evolution of resulting Weighted Average Equilibrium Price in OMIP Call Auctions versus Average Underlying Spot Price (€/MWh).

2530354045505560657075

FTB

M A

ug-0

6

FTB

M S

ep-0

6

FTB

M O

ct-0

6

FTB

M N

ov-0

6

FTB

M D

ec-0

6

FTB

Q4-

06FT

B M

Jan

-07

FTB

M F

eb-0

7

FTB

M M

ar-0

7

FTB

Q1-

07

FTB

M A

pr-0

7

FTB

M M

ay-0

7FT

B M

Jun

-07

FTB

Q2-

07

FTB

M J

ul-0

7

FTB

M A

ug-0

7

FTB

M S

ep-0

7

FTB

Q3-

07

FTB

M O

ct-0

7FT

B M

Nov

-07

FTB

M D

ec-0

7

FTB

Q4-

07

FTB

M J

an-0

8

FTB

M F

eb-0

8

FTB

M M

ar-0

8FT

B Q

1-08

FTB

M A

pr-0

8

FTB

M M

ay-0

8

FTB

M J

un-0

8

FTB

Q2-

08

FTB

M J

ul-0

8

€/M

Wh

OMIP Call Auction Weighted Average Price (€/MWh) Underlying Spot Price (€/MWh)

Source: OMIP, OMIE

Due to the scarce alternation of different signs of the forward risk premium, Test 1 considers two periods (“positive” premia and “negative” premia) in order to assess for each period the cost of the purchased energy by distribution companies. Test 1 is also completed with another test in which monthly contracts are separately considered from quarterly contracts. In both tests, 3 different reference prices are employed:

• Resulting Auction Equilibrium Price (“Feq”): this is the price recognised to the distribution companies, as stated above. It is calculated as the


14

weighted average price of all the volumes acquired by the distribution companies in the call auctions celebrated for each futures contract.

• Average Futures Price for all the quotation period (“Fall”): this is the average price of all the Daily Settlement Prices published by OMIP along the whole quotation period of the futures contract. The algorithm employed by OMIP for determining the Daily Settlement Price, based on the traded prices and the bid-ask spread, is described in Section C.6 (Settlement Price Calculation) of OMIP OMIClear Operational Guide (version of June 2008) (OMIP-OMIClear, 2008a). OMIP is the Market Operator and OMIClear is the Clearinghouse. The algorithm can be summarised as follows:

• The Settlement Price for a futures contract is the last traded price if it is within the closing bid-ask spread.

• If the last traded price during the trading session is not situated in the closing bid-ask spread, the settlement price is the bid or ask price closest to the last traded price.

• If there is no traded price during the trading session, the settlement price is the average of the bid-ask corresponding to the closing spread. If there is no traded price during the trading session, and no closing bid-ask spread, the settlement price corresponds to the settlement price of the previous trading session.

• Nonetheless, when OMIP does not rely on the resulting price due to scarce negotiation of the contract, OMIP consults a Price Committee and the daily price is obtained from representative quotations of the OTC market. Additionally, OMIP often employs the arbitrage criterion between a quarterly contract and their comprised monthly ones, to obtain the settlement prices by using weighted averages among these 4 contracts. It also applies arbitrage criterion between a calendar contract and their comprised quarterly ones. This is due to the fact that as other forward market mechanisms coexist with OMIP call auctions (OTC, EPE auctions and CESUR auctions, as previously described in Section 1.5), the most traded contracts in OMIP are the prompt months and quarterly ones (quarterly contracts with the same maturity as those from EPE and CESUR auctions), being the settlement prices of the least traded contracts in OMIP obtained through this arbitrage criterion.

• Average Spot price (“S”): this is the average price resulting from the Spanish power pool day-ahead prices and corresponding to the whole delivery period of the considered futures contract. This power pool is managed by OMIE (Market Operator of the Iberian Energy Market, Spanish Pool, known traditionally as “OMEL”).


15

1.6.1.2.2 Test 2: Analysis of Basic Statistics of Futures & Spot Prices

The statistics of the price series related to spot and futures markets of relevant international energy markets are analysed. The purpose of this test is to compare such data with the Iberian power futures prices.

In particular, basic statistics (Average, Median, Maximum, Minimum, Standard Deviation, Asymmetry Coefficient, and Kurtosis) for the monthly and quarterly futures contracts and their underlying average spot prices are provided in order to compare all the energy markets considered. The data set is comprised of the arithmetical mean values for the settlement prices of each futures contract during its quotation period. For the corresponding spot price, arithmetical mean for the underlying delivery period is calculated.

The median is the numerical value separating the higher half of a data sample, a population, or a probability distribution, from the lower half. The standard deviation shows how much variation or dispersion from the average exists. A low standard deviation indicates that the data points tend to be very close to the mean (also called expected value). The asymmetry coefficient renders the asymmetry of a distribution. This function characterizes the asymmetry degree of a distribution against its average value. A positive asymmetry indicates a unilateral distribution extending towards more positive values. Conversely, a negative asymmetry indicates a unilateral distribution extending towards more negative values. Kurtosis is any measure of the "peakedness" of the probability distribution of a real-valued random variable. It characterizes how elevated (positive value) or flattened (negative value) a distribution is, compared to the normal distribution (Hogg and Craig, 1995).

1.6.1.2.3 Test 3: Analysis of Δex-post % Magnitudes

1.6.1.2.3.1 Test 3.1: Assessment of Forward Risk Premium Existence

The main purpose of this test is to assess the existence of the forward risk premium. Therefore, the average values of the positive and negative forward risk premium are compared within each market.

For all the markets considered, distinguishing between monthly and quarterly futures contracts (Fall), a t-Student test is performed to detect, for each market, if the positive Δex-post % and the negative Δex-post % have the same average value (i.e. if the risk premium tends to 0, there would not be evidence of its existence). The t-Student test is done considering two tails and heteroscedasticity.

A t-Student test is any statistical hypothesis test in which the test statistic follows a t-Student distribution if the null hypothesis is supported. It can be used to determine if two sets of data are significantly different from each other, and is most commonly applied when the test statistic would follow a normal distribution if the value of a scaling term in the test statistic were known. When the scaling term is unknown and replaced by an estimate based on the data, the test statistic (under certain conditions) follows a t-Student distribution. On the other hand, the null hypothesis refers to a general or default position: that there is no relationship between two measured phenomena. Therefore, rejecting the null hypothesis would imply that there are grounds for believing


16

that there is a relationship between two phenomena. The null hypothesis (often denoted H0) is generally assumed true until evidence indicates otherwise (Student, 1908: Fisher Box, 1987).

A collection of random variables is heteroscedastic if there are sub-populations that have different variabilities from others. The "variability" could be quantified by the variance or any other measure of statistical dispersion. Note that the standard deviation, defined above, is the square root of the variance (Greene, 1993).

1.6.1.2.3.2 Test 3.2: Comparison of Futures Behaviour towards Maturity

The main purpose of this test is to assess the behaviour of the futures contracts against the time proximity to the settlement period (i.e the time to maturity). A comparison between monthly and quarterly contracts is performed.

For all the markets considered, distinguishing between monthly and quarterly futures contracts, and per approach to maturity (all quotation period (“Fall”), 3rd last month of quotation (“FM-3”), 2nd last month of quotation (“FM-2”), and last month of quotation (“FM-1”)), different magnitudes are compared:

• Assessment of similar behaviour between Monthly and Quarterly Contracts.

• Quantitative comparison of Δex-post % between Monthly and Quarterly Contracts.

• Quantitative comparison of Δex-post % between Periods with positive or negative values.

• Correlation between Future Series (Fall versus FM-3, FM-2, or FM-1).

• Samuelson’s hypothesis (1965): “Volatility increases as Futures contracts approach maturity”.

• Increasing convergence to spot price (less Δex-post % in absolute value) with maturity.

1.6.1.2.4 Test 4: Bessembinder’s & Lemmon’s hypothesis compliance

Test 4 is performed to assess the compliance with Bessembinder & Lemon hypothesis. Such researchers conclude in year 2002 that the forward risk Premium decreases with the variance of the spot prices and increases with the asymmetry of the spot prices.

For each futures contract type (monthly and quarterly, distinguishing between Fall, FM-3, FM-2, FM-1) of the three considered European Power Markets, testable hypothesis from Bessembinder and Lemmon (2002) is checked through Equation (1.4) by using Δex-post as the power prices in the 3 markets are commonly expressed in €/MWh:

Δex-post = Ft,T - Average(ST) (1.4)


17

The testable hypothesis is as follows:

“The Forward Risk Premium decreases in the variance of spot prices and increases in the skewness of wholesale prices”. The skewness refers to the asymmetry of the spot prices. In order to test the hypothesis, linear regression is applied according to Equation (1.5):

Δex-post = α + β*VAR(ST)+γ*ASIM(ST)+εT (1.5)

Where α is a constant, β and γ are coefficients, VAR(ST) reflects the variance of spot prices, ASIM(ST) represents the non-standardised Asymmetry Coefficient (“skewness”) of spot prices (it is the Asymmetry Coefficient multiplied by cubed Standard Deviation of Spot Prices), and εT is an error term.

Good compliance should render negative β, positive γ, with significant values for their t-statistics, as well as a high value of R-squared statistic. For the t-Student tests, a level of confidence of 95% with 2 tails is considered.

1.6.1.3 Description of tests regarding cointegration analysis of energy prices and analysis of the clean spark spreads

The research presented in Chapter 6 analyses the relationship between the Iberian power prices with oil, gas, and emission prices. Correlation analyses are performed splitting the data series in different time intervals. Additionally, cointegration analyses and an assessment of the evolution of the clean spark spreads are done as described below.

1.6.1.3.1 Cointegration tests of energy price series

Cointegration is a statistical property of time series variables. Two or more time series are cointegrated if they share a common stochastic drift. In particular, if two or more series are individually integrated but some linear combination of them has a lower order of integration, then the series are said to be cointegrated. A common example is where the individual series are first-order integrated (i.e. I(1)) but some (cointegrating) vector of coefficients exists to form a stationary linear combination of them. For instance, a commodity spot price (e.g. the electricity case) and the price of its associated futures contract move through time, each roughly following a random walk. Testing the hypothesis that there is a statistically significant connection between the futures price and the spot price could now be done by testing for the existence of a cointegrated combination of the two series. If such a combination has a low order of integration, in particular if it is I(0), this can signify an equilibrium relationship between the original series, which are said to be cointegrated. Therefore the possible presence of cointegration must be taken into account when choosing a technique to test hypotheses concerning the relationship between two variables having unit roots (i.e. integrated of at least order one I(1)) (Granger and Newbold, 1974; Murray, 1994).


18

Cointegration analyses of energy logarithmic price series are done running Stata and E-views statistical packages. The Augmented Dickey-Fuller’s (ADF) test is employed. This is a test for a unit root in a time series sample. It is an augmented version of the Dickey–Fuller test for a larger and more complicated set of time series models. The ADF statistic, used in the test, is a negative number. The more negative the result, the stronger the rejection of the hypothesis of unit root existence at some level of confidence. Basically, the Dickey-Fuller’s test checks if a unit root is present in an autoregressive model (Fuller, 1976; Said and Dickey, 1984).

Therefore the ADF test is run for the existence of unit root variables of times series for log daily prices of prompt month contracts regarding electricity, oil and gas prices. In particular the prompt month contract of the Iberian power futures (OMIP M+1), the oil futures (Brent M+1), and the gas forward contract in the Dutch national balancing point (TTF M+1) are used. The Dutch national balancing point, known as the Title Transfer Facility (TTF), is the most liquid price reference in Continental Europe. Log monthly prices of average Brent spot prices in the last 6 months are also analyzed as such structure is used in the Spanish gas last resort tariffs and in long term gas contracts (CNE, 2012e). The null hypothesis for existence of unit root is rejected at a significance level of 1% or less, otherwise the null hypothesis cannot be rejected and the series is diagnosed as non-stationary.

The number of lags in the ADF test is determined following the Phillips-Perron’s criterion. The Phillips–Perron’s test is also a unit root test. It was performed by Peter C. B. Phillips and Pierre Perron in 1988. It is used in time series analysis to test the null hypothesis that a time series is integrated of order 1. It builds on the Dickey–Fuller test mentioned above. Like the augmented Dickey-Fuller test, the Phillips-Perron test addresses the issue that the process generating data might have a higher order of autocorrelation than is admitted in the test equation. Whilst the ADF test addresses this issue by introducing lags of the dependent variable as regressors in the test equation, the Phillips–Perron test makes a non-parametric correction to the t-test statistic (Phillips and Perron, 1988).

The same unit root test has been applied to the first difference of the variables. If the first difference proved to be stationary, the diagnosis of the test would produce a result known in the statistics language as integrated of Order 1 (i.e. “I(1)”). Cointegration results are obtained based on unitary root analysis for the residue of the regression for OMIP M+1 as dependent variable. Three regressions are built in which the single independent variables are Brent M+1 (daily), TTF M+1 (daily) and Brent spot 6 month rolling average (monthly, thus OMIP M+1 monthly average is used as a dependent variable in its regression). Finally, the regression results for the 3 cases (OMIP M+1 versus fuels) are analysed, employing the coefficient of determination (R2) statistics. This coefficient indicates how well data points fit a statistical model. Its main purpose is either the prediction of future outcomes or the testing of hypotheses, on the basis of other related information. It provides a measure of how well observed outcomes are replicated by the model, as the proportion of total variation of outcomes explained by the model (Steel and Torrie, 1960).


19

1.6.1.3.2 Tests for evaluation of the clean spark spreads

The Clean Spark Spread (CSS) is obtained as the difference between the power futures price and the forward generation cost with a gas fired combined cycle plant taking into account the CO2 emission rates. It is analysed for the prompt contracts, i.e. month (M+1), quarter (Q+1) and year (Y+1) contracts, as expressed in Equation (1.6.):

CSS = PPower – ((PGas x ηCCGT) + (PCO2 x ECCGT)) (1.6)

where “CSS” is the Clean Spark Spread, “PPower” is OMIP power futures price, “PGas” is TTF forward gas price, “ηCCGT” is the CCGT thermal efficiency, “PCO2” is the emission futures price, and “ECCGT” is the CCGT emission rate (in tCO2/MWh). The CCGT forward generation cost is calculated as the sum of the gas forward price – divided by the thermal efficiency of the power plant – and the price of the CO2 allowance multiplied by the emission rate. A CCGT thermal efficiency of 55% and an emission rate of 0.37 tCO2/MWh are employed (Abadie and Chamorro, 2009). The CSS daily evolution for each series with data spanning from July 3 2006 to November 30 2011 is analysed compared to the electricity futures prices and the gas forward prices. Furthermore, the analysis is strengthened assessing the average values per year.

1.6.2 Description of the regression model built related to liquidity of energy markets

The analysis performed in Chapter 5 assesses the efficiency of the Iberian power futures market focused on another cornerstone: the liquidity. The employed data set is robust, as it covers the first four years of existence of this market (from July 3, 2006, to June 30, 2010). Such an ample data set facilitates the detection of the most significant traded volume drivers and on the other hand, the identification of the products that still show poor performance (i.e. illiquidity). These findings allow the formulation of policy recommendations for streamlining the efficiency of this market. A regression model using Ordinary Least Square methodology is estimated to assess the effect of twelve selected drivers (the independent variables) for the following key liquidity measure (the dependent variable): the evolution of the energy traded in the continuous market. The research is also reinforced by means of a correlation analysis of the independent variables with the dependent variable. As market players trade essentially energy derivatives to hedge their supply commitments (Universia, 2011) – in the case of electricity suppliers, such hedges through forward contracting are established to secure their retailing margin (Bartelj et al., 2010) –, the analysis of the traded volume drivers allows to determine if the Iberian power futures market is growing properly to consolidate its original role as key hedging vehicle. Other typical liquidity measures, suggesting potential research through other econometric models for further analysis of this still emerging market, are the bid-ask spreads (STEM, 2006), the open interest


20

(Lucia and Pardo, 2008), the volatility and sensitivity of prices to additional demand (Krishna, 2008), and the resilience (Newbery et al., 2003). The term resilience, whose original use comes from Physics, is widely used in other scientific areas, as e.g. Finance, related to the robustness of the financial institutions or even the whole financial system (González-Páramo, 2011). In the case of liquid energy markets, prices must be resilient to large orders (Newbery et al., 2003). In the Iberian power futures market, the resilience is mainly secured by the market rules, as the market operator establishes maximum price variation limits. The orders exceeding such limits cannot be matched (OMIP, 2011b). Additionally, The Spanish energy regulator (CNE, now CNMC) performed weekly monitoring of the evolution of the power futures prices and the OTC prices, by means of charts of OMIP settlement prices and OTC weighted average prices (CNE, 2011a; 2012a). Afterwards, instead of weekly bulletins, more descriptive monthly supervisory reports are published, see e.g. CNE (2012d) and CNMC (2014). As deviations between the values of both price series are not usual, the resilience is also found in the OTC market, strenghtening the forward market integrity.

Equation (1.7) shows the regression model composed of twelve variables (constant a0 is included) theoretically explaining – according to the main features of the market described in Section 1.4 and Section 3.3 – the monthly evolution of the traded volumes in the continuous market (GWh_Contt).

GWh_Contt= a0 + a1 * Nr_Integrt + a2 * Nr_Non_Integrt + a3 * Nr_Financt + a4 * GWh_auctt + a5 * GWh_OTCt + a6 * GWh_OTC_clt + a7 * Nr_trading_sessionst + a8 * Abs(FMt+2,t-St) + a9 * Expans_Contt + a10 * Mkt_Makerst + a11 * VPP_CESURt + a12 * Cont_Comiss_Disct + ε t (1.7)

The twelve variables were selected by means of multicollinearity tests. Multicollinearity refers to a situation in which two or more explanatory variables in a multiple regression model are highly linearly related (in the case of just 2 variables, this pehenomenon is known as collinearity). Perfect multicollinearity exists when the correlation between two independent variables is equal to 1 or -1. In practice, perfect multicollinearity is rarely found in a data set. Usually, multicollinearity arises when there is an approximate linear relationship among two or more independent variables (Farrar and Glauber, 1967). Error term is expressed through ε t. The variables are: Nr_Integrt, Nr_Non_Integrt, and Nr_Financt showing respectively the increasing number of integrated, non integrated, and financial trading members enrolled as depicted in Figure 5.5 – a bigger amount of participants should contribute to increase the trading activity; GWh_auctt, GWh_OTCt, and GWh_OTC_clt showing respectively the traded volumes in OMIP auctions, in the OTC market, and the cleared OTC volumes by OMIClear – the exchange liquidity channels (diverse trading modes and the possibility to clear external trades) should attract liquidity (Meeus, 2011); Nr_trading_sessionst indicates the number of OMIP sessions per month, as months with more trading sessions should experience a more sustained trading activity; FMt+2,t-St is a forward risk premium measuring in absolute value, for each month t, the difference between the arithmetical average of OMIP settlement prices for the monthly SPEL baseload futures with delivery 2 months later (FMt+2,t) and the average underlying spot price during that


21

month t (St) – shall the premium be significant, the agents could exploit price arbitrages close to the expiration of the contract, increasing the traded volumes in the continuous market; Expans_Contt, Mkt_Makerst, VPP_CESURt, Cont_Comiss_Disct are dummy variables recording respectively the expansion of the continuous trading phase (“0” until December 2007, “1” since January 2008 onwards) – longer trading periods should contribute to larger trading activity –, the existence of market maker agreements (according to Table 1.1, “1” for the months with agreements in force, “0” otherwise) – the active quotation of the market makers should facilitate the matching of orders –, the celebration of regulated auctions, namely CESUR auctions and Spanish and Portuguese VPP auctions as briefly described in Section 1.5 (according to Table 1.1., “1” for the months with celebration of such auctions, “0” otherwise) – the arbitrage opportunities with those auctions should facilitate the trading in the continuous market –, and the existence of discount campaigns (“1” for the months with campaigns, “0” otherwise) – as previously mentioned, such campaigns intended to foster the trading activity. Good compliance should render positive values for coefficients a1 to a12 with significant values for their t-statistics, as well as high value of R-squared statistic. For the t-Student test, a level of confidence of 95% with 2 tails is considered. The number of observations is 48 (monthly values during 4 years).

Table 1.1. Iberian Regulated Forward Contracting Mechanisms within the MIBEL Framework complementing the OMIP call auctions.

Celebration Products Celebration Products Celebration Products1st June 13, 2007 1st June 19, 2007 1st June 26, 20072nd Sept. 13, 2007 2nd Sept. 18, 2007 2nd Sept. 21, 2007

3rd Dec. 11, 2007 3rd Dec. 18, 2007 3rd Jan. 16, 2008

Baseload: months "M+1", "M+2";quarters "Q+1", "Q+2", "Q+3"

4th March 11, 2008 4th March 13, 2008 4th March 7, 2008 Baseload: quarters "Q+1", "Q+2"5th June 10, 2008 5th June 17, 20086th Sept. 23, 2008 6th Sept. 25, 2008

7th Dec. 16, 20087th March 24, 2009 8th March 26, 2009

9th June 25, 200910th Dec. 15, 200911th June 23, 2010 Baseload and peak: quarter "Q+1"

Baseload and peak:quarter "Q+1"

Baseload and peak: quarter "Q+1"; quarter "Q+2"

Portuguese VPP Auction

Baseload: quarter "Q+1"Baseload and peak:

quarter "Q+1";six month "(Q+1)+(Q+2)";

year "(Q+1)+(Q+2)+(Q+3)+(Q+4)"

Baseload: quarter "Q+1"

Spanish VPP ("EPE") Auction CESUR Auction

Baseload:quarter "Q+1";

six month "(Q+1)+(Q+2)"Baseload and peak:six month "(Q+1)+(Q+2)";

year "(Q+1)+(Q+2)+(Q+3)+(Q+4)"

Source: Capitán Herráiz and Rodríguez Monroy (2013a)

1.6.3 Description of the hedging efficiency analysis through the net position ratio

The research performed in Chapter 7 assesses the hedging performance in the Iberian forward electricity market. Aggregated data from the Portuguese and Spanish clearing houses for energy derivatives are considered. The hedging performance is measured through a “net position ratio” obtained from the final open interest of a month derivatives contract divided by its accumulated cleared volume. This original approach is built based on volume and open interest data. Therefore, existing literature is considered. Additionally, the essence of this ratio is compared with the traditionally


22

used “hedge ratio” in financial research. Extensive literature research is provided in Section 2.4.2., apart from identified key literature pointed out in Section 1.6.3.1.

1.6.3.1 The relationship between volume and open interest

Lucia and Pardo (2008) perform a research with stock data on the measurement of speculative and hedging activities in futures markets from volume and open interest data. They indicate that a necessary condition to be a hedger is to have a spot (or forward) commitment that involves a risk exposure. The speculators are outright position-takers, including the day traders, holding their positions for less than one trading day. The daily open interest determines the number of outstanding contracts at the end of a trading day (entered contracts but not yet liquidated). The open interest equals the number of outstanding long positions (or equivalently, short positions) at the end of the day. It increases whenever neither of the two traders involved in a contract is closing out a position and decreases when both parties are closing out a position. It remains the same when only one trader is closing out a position, being this trader replaced by another one. The daily trading volume reflects somehow movements in the speculative activity. The daily open interest captures hedging activities as it excludes all intraday positions taken by day traders, mainly inspired by speculative reasons. Hedgers tend to hold their futures market positions longer than speculators.

Regarding analysis of volume and open interest data for the European CO2 market, Mansanet-Bataller et al. (2012) study the evolution of the speculative activity in futures carbon markets. A comparison of the three first phases in the European market is provided. The first phase covers years 2005-2007. The second phase covers years 2008-2012; the third phase covers years 2013-2020. The fourth trading period may possibly run from 2021 to 2028. They find a high degree of speculative behavior when listing the contracts for the first time, for every Phase; a higher level of speculation in the first quarter of each year (which could be explained by the increase of the number of informed traders in the market during these months, in relation with the specific schedule of deadlines that characterizes the EU Emission Trading Scheme); Phase II of the EU ETS seems to be the most speculative phase to date; and the front contract concentrates the majority of the speculative activity every year. An analysis of the price evolution of the CO2 futures contracts can be found in Chapter 6 (see Section 6.2.1).

1.6.3.2 The fundamentals of the net position ratio methodology

The net position ratio employed in Chapter 7, being the first research performed for the Iberian power futures market based on this particular methodology, intends to measure the relation between the cleared volume and the open interest in that market. Therefore, it is built as the division of the open interest presented at the end of the last trading session of each month contract and the accumulated cleared volumes in the clearing house for that specific contract (and the corresponding part of the cleared volumes from quarter and year contracts containing that month) at that last trading session (i.e. when such a contract does no longer quote, as its trading period expires


23

and its settlement period begins). This original ratio enlarges the variety of existing ratios based on open interest and traded volume data as reviewed by Lucia and Pardo (2008), mentioned above. Additionally, it provides a complementary alternative to the traditionally used hedge ratio, described in Section 2.4.2.

1.6.3.3 Description of test regarding net position ratio

OMIP-OMIClear and MEFF Power final open interest (i.e. the value published for the last session in which a given contract is quoted) are studied for each month contract. Only Spanish month contracts are considered: OMIP-OMIClear base load and peak futures and MEFF Power base load swaps. The division of the final open interest and the accumulated cleared volume is used as a net position ratio to measure the potential interest of the traders in these contracts for risk management (i.e. hedging by means of final open positions, in case they trade the same amount with the same nature afterwards in the spot market (e.g. spot purchases in case of long derivatives positions)). The data set covers the first five and a half years of existence of the Iberian power futures market (i.e. from 3 July 2006 to 31 January 2012).


24


25

CHAPTER 2. LITERATURE REVIEW

2.1 Introduction

This chapter shows the academic and regulatory literature consulted for the different test performed in the research. In particular, section 2.2 provides the literature review related to the tests performed in Chapter 4 (tests regarding forward risk premium) and Chapter 6 (test regarding cointegration analysis of energy price series and analysis of the clean spark spreads); section 2.3 provides the literature review related to the tests performed in Chapter 5 (analysis of the liquidity of the Iberian power futures market through a regression model); and section 2.4 provides the literature review related to the tests performed in Chapter 7 (analysis of the hedging efficiency through the net position ratio).

2.2 Literature related to market price efficiency

A short literature review of market price efficiency is provided, focused on energy markets, and especially, in power markets compared to other commodities and to other financial markets. Market efficiency mainly refers in this context about how well the future price predicts the spot price.

Cointegration tests as well as tests for measuring if the forward price is an unbiased forecast for cash price for commodity and power markets show that Futures Markets are efficient in the long-term, but not in the short-term, even if risk neutrality is neglected and a risk premium is assumed. In practice, the hypothesis claiming that forward price is an unbiased forecast of future cash price (“Efficient Market Hypothesis”) is usually rejected (Engel, 1996).

According to statistics and econometric research, many commodity futures markets existing since the middle of the 19th Century are not efficient. Power Markets are considerably younger than Commodity Markets due to the deregulation trend in the 90’s. Power Markets differ from other markets since electricity storage is very limited. There are many studies for the US and European Power Markets, analysing the behaviour and interactions of their different regional markets. Nord Pool is the oldest Power Exchange. It was founded as a Norwegian Power Exchange in 1993 (“Statnett Marked AS”) and became a Nordic Power Exchange in 1996 as the Swedish System Operator (Svenska Kraftnät) became one of the main shareholders. In 1998 Finland enrolled to this market, followed by Western Denmark in 1999, and by Eastern Denmark in 2000. In the Nordic market, a significant part of power production is from hydro reservoirs, playing such hydropower inventory an important role in the pricing of electricity (STEM, 2006).

Regarding energy markets, Serletis (1992) examines the effects of maturity on future price volatility and trading volume for 129 energy futures contracts traded in New York Mercantile Exchange (NYMEX) in the beginning of the 90’s. The results provide support for the maturity effect hypothesis theoretically demonstrated by Samuelson


26

(1965). In the applied Serletis’ research, energy futures prices become more volatile and trading volume increases as futures contracts approach maturity. As the majority of the studies testing Samuelson hypothesis are applied for US Futures markets, Allen and Cruickshank (2002) prefer to research with commodity futures on the Sydney Futures Exchange (SFE), the London International Financial Futures and Options Exchange (LIFFE), and the Singapore International Monetary Exchange (SIMEX). They also find evidence of Samuelson hypothesis in the majority of the contracts analysed. Samuelson assumed that competitive forces in the futures market cause spot and futures price to converge at expiry. Futures price volatility can be associated with the amount of information available in a market. Little information is known regarding distant contracts compared to contracts closer to expiration. Thus as maturity approaches, the amount of information reflecting the fundamentals of the spot asset increases, causing large changes in the futures prices and consequently intensifying price volatility. These results are important to many participants in futures markets as volatility has implications for hedgers and speculators who try to optimise their positions with respect to the level of price variability. In markets compliant with Samuelson hypothesis, speculators may find it beneficial to trade in contracts closer to expiry as greater volatility implies greater short-term profit opportunities. Hedgers would benefit from trading in longer dated contracts as lower volatility would require fewer hedges.

Regarding US Power markets, there are many studies comparing different regional markets. Arciniegas et al. (2003) detect that the Pennsylvania/New Jersey/Maryland (PJM) Power Market and the California Power Market are more efficient than the New York Power Market. They find that efficiency has risen with the maturity of the markets, as players have learnt to take advantage of arbitrage opportunities. They also find that a multi-settlement scheduling system leads to higher efficiency. A multi-settlement system implies that the prices and quantities established in market phases prior to dispatch are binding forward contracts. Their study is built with hourly prices of day-ahead and real-time markets. They consider that a market is efficient when all the relevant and ascertainable information is fully and immediately reflected in market prices. Therefore all players are well-informed and adjust their market strategies to profit from arbitrage opportunities. Their literature review confirms that important differences in market structure and in the organisation of forward markets across US states may explain differences in market efficiency. California was the only market where competing scheduling coordinators ran the forward markets. In PJM and New York markets, the Independent System Operator (ISO) who centralized the day-ahead market competed with bilateral markets. An ISO can use its power as grid manager to favour its day-ahead market undermining its efficiency. In the summer of 2000 in California, one player (Pacific Gas and Electric Company) tried to exercise monopsony power, being this a possible cause of the large differences between the forward and spot prices in that period. PJM is the most liquid market in the East coast due to its bigger number of participants attracted because of lower transaction costs and by PJM reputation of delivering transparent and reliable information. New York power market lacked transparency in delivering information. They claim that more aggressive competition should lead to faster learning and more efficiency. In PJM, where utilities could fix prices through long term contracts, price volatility was less than


27

in California, where utilities were not allowed to access to long term contracts and more than 90% of the power was purchased in the spot market. The lower volatility in energy prices may lead to fewer arbitrage opportunities and a more efficient and stable market. Good interconnection of regional markets brings efficiency and stability in the involved power markets. Additionally, disparity in the competition level of international power markets arises from differences in market design.

Avsar and Goss (2001) study market efficiency for the PJM and the California Power Markets and cannot reject the Efficient Market Hypothesis for the period July 1998-March 1999, but cannot accept it for the whole data period. They find remarkable learning effects from market agents. Additionally, market efficiency is linked to market maturity. In this sense, market players in power markets seem to learn faster than in oil markets, for instance, increasing its efficiency with time (Walls, 1999). Bessembinder and Lemmon (2002) consider that electricity cannot be economically stored and therefore, arbitrage-based methods are not applicable for pricing power derivative contracts. They build an equilibrium model implying that the forward power price is a downward biased predictor of the future spot price if expected power demand is low and demand risk is moderate. The equilibrium forward risk premium, understood as the bias in the forward as a predictor of the delivery-date spot, increases when either expected demand or demand variance is high, due to positive skewness induced in the spot power price distribution. Optimal forward positions for power producing and retailing firms depend on forecast power demand and on skewness of power prices. The premium in forward power prices is positively related to expected demand, and is large during summer.

Shawky et al. (2003) attribute the high price volatility of US deregulated power markets to the nature of how electricity is produced and consumed, inelastic demand, seasonal effects and nonstorability of electricity. They investigate the empirical relation between daily spot and futures electricity prices traded on NYMEX and delivered at California-Oregon Border (COB) during years 1998-1999. They find that the behaviour of the electricity market is consistent with efficient markets. Due to the unique features of electricity as a nonstorable commodity and the relatively few players on the generation and wholesale demand sides, they find that electricity futures differ significantly from other commodities as the former present larger estimates of forward risk premium and larger hedge ratios. The hedge ratio is defined as the ratio of the position taken in the futures contracts that will exactly offset the size of the exposure in the spot market. The larger forward risk premium may be required to bring equilibrium to a futures market where supply and demand conditions are much volatile and may also be caused by limited participation of financial players. Due to the unique characteristics of electricity, the price volatility in the spot market is many times higher than in the futures market. Positive shocks to spot prices have significantly more impact on both current and futures values of electricity than shocks to futures prices. Shocks to both spot and futures returns appear to be relatively short-lived (half-life of 4-5 days) before they converge to their long-run equilibrium.

Longstaff and Wang (2004) perform an empirical analysis of forward prices in PJM power market with hourly data set of spot and day-ahead forward prices. They find remarkable forward risk premia in power prices and obtain results consistent with


28

Bessembinder’s and Lemmon’s model. They find that premia vary systematically through the day and are related to agents’ measures of economic risk (volatility of unexpected changes in demand, spot prices and total revenues). They conclude that PJM forward prices are determined rationally by risk-averse economic agents, not applying such finding to other power markets more exposed to market abuse.

Ullrich (2007) streamlines Bessembinder’s and Lemmon’s model considering the constrained capacity, allowing the model to reproduce price spikes by using reasonable parameter values. He finds that the forward risk premium decreases/increases in spot price variance when the expected spot price is low (i.e. less than the retail price)/high (bigger than the retail price), because of retailers’ hedging needs. The enlarged PJM data set from Longstaff’s and Wang’s empirical research supports these model predictions.

Regarding European power markets, the largest number of studies exists for Nord Pool, the most developed power market in Europe since its foundation in 1993 (e.g. Byström, 2003). Byström concludes that traditional simple price hedging models are almost equally efficient as the most advanced ones. Therefore, hedging at Nord Pool (or whatever Power Futures Market) does not request more advanced models than from other financial markets though underlying product features differ noticeably from other financial or commodities products.

Lucia and Schwartz (2000) analyse Nord Pool spot, futures and forward prices during years 1998-1999 and conclude that the seasonal systematic pattern of spot electricity prices throughout the year is of crucial importance in explaining the shapes of the futures and forward curve. They detect that a simple sinusoidal embedded in their one and two factor models captures the seasonal pattern of the futures and forward curve. Their models include a deterministic component reflecting remarkable regularities in the behaviour of electricity prices. They find that volatility of Nord Pool spot system price is consistently different between cold and warm seasons. They detect that transportation constraints for electricity make electricity contracts and prices highly local, i.e. strongly dependent on the local determinants of supply and demand (e.g. local generation plants, local climate, and local uses of electricity). Additionally, regulatory issues such as market rules and market structure may also impact on prices behaviour in competitive electricity markets and on their differences across countries.

The research works regarding European markets are usually focused on the Regional Integration of the Power Markets (e.g. Armstrong and Galli, 2005; Zachmann, 2005). Armstrong and Galli study European wholesale spot power prices and detect a price convergence between the price differences. Zachmann also finds a price convergence during the period covering years 2002-2004 between Dutch and German wholesale power prices but not between East Danish and German prices. He concludes that it is necessary to overcome the bottlenecks in the physical interconnection capacity in order to achieve an integration of the European Power Market.

Regarding research focused on the forward risk premium, Karakatsani and Bunn (2005) classify half-hourly trading periods in two clusters (peak and off-peak) discovering a systematic diurnal reversal in the forward premium nature for the British


29

power market after the market reforms in year 2001. The reversal can be explained by the asymmetric positions of generators and suppliers towards risk and its intraday variation, due to the heterogeneity of the power generation mix and to market design specificity (introduction of penal balancing prices and abolishment of capacity payments and uniform pricing). Ex-ante forward premia, built upon predictice intervals and based on expected spot prices, are similar to ex-post premia but sensitive to assumptions on agents’ spot price model, information set and learning scheme. Redl et al. (2009) assess the key parameters of forward electricity prices and the relationship between forward and future spot prices by means of an empirical analysis of electricity prices at European Energy Exchange (EEX) and Nord Pool energy exchanges. They find that price formation in the considered markets is influenced by historic spot market prices rather than fundamental modelling approaches yielding a biased forecasting in the long-term contracts price formation. They conclude that market and risk assessment measures of market participants and supply and demand shocks can partly explain the futures-spot bias and that inefficiencies in the analysed forward markets cannot be neglected. Redl and Bunn (2011) find, through analysis of forward and spot data in the German power market (EEX), that the forward risk premium in electricity is a function of fundamental, behavioural, dynamic, market conduct and shock components. This premium is influenced by the gas prices, the oil price volatility, the generators’ market power and the power scarcity. They suggest to use this premium as a market monitoring indicator as higher premium could be caused by market concentration. Therefore, they conclude that the considerations of the scale and determinants of the forward premium are as important as the market power effects in spot market price formation when evaluating the efficiency of wholesale power trading. Cartea and Villaplana (2008) build a model for wholesale power prices explained by two state variables (demand and capacity) and calculate the forward premium. They perform empirical research embracing PJM, England & Wales, and Nord Pool markets. They find that, depending on the market and the period under study, the volatility of capacity and the market price of capacity risk could either put upward or downward pressure on forward prices. They also find that the forward premium follows a seasonal pattern, being positive in the months of high volatility of demand and close to zero or even negative in the months of low volatility of demand. Furió and Meneu (2010) perform theoretical and empirical research (based on OTC prompt month forward prices and spot prices of the Spanish power market) and find that the ex-ante forward premium varies with the expected demand in tight market conditions, and the ex-post forward premium depends on the unexpected variations in demand and hydro capacity. They also find a positive relation between the Spanish spot prices and the CO2 emission allowance prices. The implications derived from Bessembinder’s and Lemmon’s model are supported by their data.

Conclusions from existing studies measuring the efficiency of futures markets vary considerably. Reviewed literature shows no uniformity regarding the results provided by the existing measuring methods. The selected method can slightly bias the results. Additionally, the most advanced models may question previous results from older and simpler models. More advanced models tend to confirm market efficiency but older ones may be prone to reject it. In general, it seems that commodity, energy, and even power markets are not especially efficient (STEM, 2006).


30

2.3 Literature related to liquidity of energy markets

2.3.1 Academic research

Fusaro (2011) researches about the emerging energy derivatives markets since the 80’s (oil and gas) and 90’s (power) describing their specificities compared to the original pure financial markets. He defines liquidity as a characteristic of a market where there is a high level of trading activity. He provides an overview of the development of North American, European and Asian oil, gas, and power markets. He foresees the liquidity growth in those markets due to the development of energy derivatives. Mork (2001) researches about the deregulated European power markets in the 90’s envisaging the emergence of financial markets for electricity. He focuses in three case studies: United Kingdom, Norway and Switzerland. He indicates that the Nordic region has developed a quite advanced and well-functioning electricity trading pool, with widespread financial trading and good liquidity. In order to create healthy financial power markets, the choice of pool model, including spot, adjustment and forward markets, will make or break liquidity. Newbery et al. (2003) analyze the inadequate liquidity level of the Dutch electricity market. They indicate that liquid markets enable the immediate execution of standard orders, exhibit prices that are resilient to large orders, and present low transaction costs due to the high amount of active participants and traded volumes. Based on this analysis and on the Monitoring Report on the Dutch Wholesale Electricity Market, 2006, prepared by the Office of Energy Regulation (DTe) for the Netherlands Competition Authority (NMa), Krishna (2008) examines the existing measures of liquidity in that market adopted by NMa. He defines liquidity as the ability of an asset to be instantly converted into cash without any significant movement in the price. He detects the liquidity improvement caused by the decline of the level of bid-ask spreads in the period 2006-2007. He also detects a positive, though insignificant relationship between the volatility of electricity prices and the level of liquidity in the Dutch electricity markets in that period. Hence, the volatility in prices in this electricity market during this period is not necessarily a sign of illiquidity. Batlle et al. (2007) indicate the positive effects of the market maker agreements for enhancing liquidity. Their research, applied to the French balancing power market, proposes to introduce such agreements in order to improve efficiency. Market makers are needed when the structure of the traders is such that liquidity does not arise naturally. A market maker is an exchange member obliged to make a continuous two-way price, creating bid and ask prices for a given security. It generally maintains inventory and stands ready to buy and sell at the quoted price to keep a functioning market and fostering liquidity. Due to such a dynamic and active quotation service, they constitute a key factor in attracting new actors in the organized markets (Batlle, 2007). Otherwise, the new actors should be mainly obliged to contract OTC brokerage services to find counterparty, or establish bilateral contracts. Meeus (2011) analyzes the regulation of the European power exchanges, distinguishing between “merchant” and “cost-of-service-regulated”. He indicates that the exchanges can benefit from a positive network externality, as liquidity attracts liquidity. The liquidity supporting measures fostered by regulators, forcing international traders, Transport


31

System Operators, or incumbent generation companies to trade on the incumbent power exchange, improve the liquidity of that exchange. However, due to the natural monopoly features of the exchanges, they reinforce the dominant position of the incumbent power exchange, which can be problematic. Molzahn and Singletary (2011), in their research about the Financial Transmission Rights (FTR) in the North American Midwest Independent System Operator (MISO) auction market, indicate that the large speculator profits are potentially concerning. On the other hand, they find support to the argument that speculators confer useful liquidity benefits. Regardig liquidity of emission trading programs and international emission markets, an overview can be found in Haites and Missfeldt (2004). For the case of Tradable White Certificates (TWC), Mundaca (2008) describes the liquidity increase due to the implementation of a European scheme to improve energy efficiency.

2.3.1.1 Supervision Reports in European Energy Markets

2.3.1.1.1 Monitoring of market liquidity by the European Commission

The Directorate-General for Energy (DG ENER) of the European Commission performed in year 2008 –at that time named DG TREN, as it also covered transport issues– two analyses of EU wholesale energy markets. The first one (European Commission DG TREN, 2008a) evaluates the factors impacting on current and future market liquidity and efficiency. EU wholesale energy markets were relatively underdeveloped. Wholesale power markets were significantly more advanced than natural gas ones. Progress was not uniform and there were large variations in market liquidity and efficiency across the EU. Natural gas, power and CO2 trading is a deregulated activity with a large and growing proportion taking place in the opaque OTC market. Exchange prices set a benchmark for spot prices. OTC is seen as more flexible, cheaper, and offering more specialized products. There is a strong inverse relationship between the levels of market concentration and the degree of liquidity. Improvement in supply and demand data transparency is a quick-win. The second report (European Commission DG TREN, 2008b) analyzes the historical data of EU wholesale electricity, natural gas and CO2 markets. For power exchanges during 2002-2007, a clear increase is found in traded volumes, market participants, and price correlations. There is a negative relation between an increase in market participants and volatility. Price volatility remains notably high on spot power trading. Derivative contracts are higher in volume, less volatile but more concentrated on one exchange (the German based European Energy Exchange, EEX). The substantial OTC growth (traded volumes doubling since year 2006) is caused by trading in forward physical markets (1% of total volumes for pure financial trades). The growth of physical connection capacities as well as the market coupling initiatives have improved liquidity and price signals.


32

2.3.1.1.2 OFGEM Monitoring of the liquidity of British Energy Markets

The British regulatory agency for electricity and gas markets (OFGEM) is analyzing since 2009 the liquidity of the British wholesale power and gas markets. The liquidity (especially in the electricity market) was of concern (OFGEM, 2009). The low liquidity acts as a barrier to new entry into generation and supply markets and may be a source of competitive disadvantage to small suppliers. Liquid markets provide investment signals to market participants and reduce price manipulation. The low liquidity is a function of interrelated factors: the period of rapid growth in vertical integration, which followed the collapse of Enron in 2001-2002 and the exit of active market participants; the regulatory risk; the undermined confidence in market competitiveness; the complex balancing market arrangements, acting as an entry barrier for smaller parties and non-physical participants; the price volatility; the volumes migrated towards the more liquid gas market; the lack of power exchange based trading; and the harder credit arrangements and collateral requirements after Enron collapse and in the current financial crisis. OFGEM (2010) provides a further consultation of policy options, some of them to be implemented in case current market initiatives fostered by power exchanges and brokers do not get liquidity gains. The options aims to increase the trading of the large vertically integrated utilities by applying self-supply restrictions, obliging them to trade with small/independent suppliers, compelling them to act as market makers, and establishing mandatory sales auctions.

2.3.1.1.3 Analysis of the Nordic Derivatives Market by the Swedish Energy

Authority and the Nordic Energy Regulators

The Swedish Energy Agency (in Swedish, “Statens energimyndigheten”, STEM) conducted an extensive research of the Nordic Electricity Derivatives Market in order to see how to improve its efficiency STEM (2006). Analysis of the liquidity evolution of the Nordic power exchange (Nord Pool) was done, detecting that the entry and further exit of Enron and other American players clearly affected its liquidity. The financial players reckon 1/3 of the amount of members and traded volume, and they are the group experiencing bigger enrollment in the last years. The distribution companies are the most active players, followed by financial agents. Two further types of agents are distinguished, though less active: the big four producers and small producers. The contracts for differences (CfD) are very illiquid contracts, showing larger bid-ask spreads. Many agents would desire bigger transparency for OTC trades not cleared through Nord Pool. Comparison with other energy derivatives exchanges confirms that liquidity is poor for contracts whose maturity date is far. A more updated report about the evolution of the Nordic financial electricity market is provided in 2010 by the association of Nordic Energy Regulators (NordReg). NordReg is composed by the Danish Energy Regulatory Authority (DERA), the Finnish Energy Market Authority (“Energiamarkkinavirasto”), The Icelandic Energy Authority (“Orkustofnun”), the Norwegian Water Resources and Energy Directorate (NVE) and the Swedish Energy


33

Market Inspectorate (EI). As a general conclusion, the Nordic electricity derivatives market works well (NordReg, 2010).

2.4 Literature related to ratios measuring the hedging efficiency

2.4.1 Literature regarding commodity derivatives and application of the Iberian energy derivatives market

This section provides useful insights regarding commodity derivatives that can be useful for further research of the still developing Iberian power futures market. Firstly, a literature review of the traditional hedge ratio employed in financial literature is provided. Afterwards, a literature review of key research based on open interest analysis related to energy derivatives markets is presented.

2.4.1.1 Literature review of the hedge ratio

The hedge ratio is the ratio of the position taken in the futures contracts that will exactly offset the size of the exposure in the spot market. Chen et al. (2003) review the different theoretical approaches to deriving the optimal futures hedge ratio, based on minimum variance, mean-variance, maximized expected utility, mean extended-Gini coefficient, as well as semivariance. They discuss diverse methods of estimating the hedge ratio, ranging from ordinary least squares to heteroscedastic cointegration methods. Under martingale (i.e. the expected futures price change is zero) and joint-normality conditions for the futures and spot prices, different hedge ratios are the same as the minimum variance hedge ratio. Otherwise, there is no single optimal hedge ratio superior to the remaining ones. They indicate that the random coefficients model employed by Grammatikos and Saunders (1983) for currency futures can, in some cases, improve the effectiveness of hedging strategy, but that technique does not allow for the update of the hedge ratio over time.

Regarding oil futures, greater financialization – i.e. the massive expansion of the financial layer of oil due to the explosion in the derivatives permitting speculation – raises the hedge ratio (Fattouh and Mahadeva, 2012a, 2012b).

Regarding electricity futures, Shawky at al. (2003) focus on the futures contracts traded on the New York Mercantile Exchange (NYMEX) for delivery at the California-Oregon Border (COB) hub for the period 1998-1999. Using a GARCH specification, they estimate a minimum variance hedge ratio. In econometrics modelling, GARCH stands for Generalized AutoRegressive Conditional Heteroscedasticity. If an autoregressive moving average model (ARMA model) is assumed for the error variance, the model is a GARCH model (Bollerslev, 1986). The mean hedge ratio (1.63) is bigger than for other commodities (e.g. Baillie and Myers (1991) using a


34

similar GARCH model obtain an estimation of 0.07 for beef, 0.25 for coffee, 0.61 for corn, 0.38 for cotton, 0.50 for gold, and 0.76 for soybeans). The reason behind the bigger electricity hedge ratios is the non-storable nature of electricity, the presence of a relatively few market players, and a bigger spot price volatility compared to the other commodities.

Further research is encouraged regarding the estimation of the optimal hedge ratio for the Iberian energy derivatives market, considering OMIP future prices and the underlying spot prices, to seize the magnitude of the resulting ratio compared to other commodities markets.

2.4.1.2 Literature review of energy markets about analysis of the open interest

Buchananan et al. (2001) provide a method of predicting direction of spot price movements in the natural gas market for the month succeeding from market participants’ net positions in the New York Mercantile Exchange (NYMEX) futures market. The Commitment of Trade reports collected by the Commodity Futures Trading Commission (CFTC) act as a proxy for ex ante forecasts of futures price movements. These weekly reports contain the net long and short positions for each contract held by large hedgers, speculators, and nonreporting or small traders. While the natural gas contract did not begin trading until the early 1990s, it has grown faster than most commodities. They find that the position of the larger speculator contains valuable information for predicting the direction and magnitude of subsequent price changes. Regarding the relation between MIBEL spot and futures prices, Ballester et al. (2012), with data from year 2006 to year 2011, find causal relationships in the short term from futures prices to the geometric average of the forward curve (obtained with OTC data published by Reuters) as a proxy for spot prices, and empirical evidence of unidirectional Granger causality from one-month-ahead- and one-quarter-ahead futures prices to spot prices. The Granger causality test is a statistical hypothesis test for determining whether one time series is useful in forecasting another (i.e. predictive causality). A time series X is said to Granger-cause Y if it can be shown, usually through a series of t-tests and F-tests on lagged values of X (and with lagged values of Y also included), that those X values provide statistically significant information about future values of Y (Granger, 1969). An F-test is any statistical test in which the test statistic has an F-distribution under the null hypothesis. It is most often used when comparing statistical models that have been fitted to a data set, in order to identify the model that best fits the population from which the data were sampled. The name was created by George W. Snedecor, in honour of Sir Ronald A. Fisher, who initially developed the statistic as the variance ratio in the 1920s (Lomax and Hahs-Vaughn, 2013).

Bolinger et al. (2006) analyse the price dynamics of NYMEX US natural gas futures contracts. They indicate that the open interest serves as a liquidity proxy. Beyond the first 36 of the 72 gas futures contracts listed, the remaining futures are scarcely traded (their open interest drops to zero). Although the traded volume may be a better measure of liquidity, it is also more sensitive to the particular day chosen. A


35

larger part of the open interest is due to commercial traders. In the US natural gas futures market, those hedgers account for 60-75% of the open interest. In the case of the Iberian power futures market, the contracts with higher traded volumes (Spanish base load futures for the two prompt months and quarters and the prompt year) are as well the unique contracts with remarkable open interest values. Therefore, the liquidity in this emerging market is more concentrated than in more mature markets as NYMEX. Neither the Iberian power futures market operator and Portuguese clearing house (OMIP-OMIClear) nor the Spanish clearing house for energy derivatives (MEFF Power, whose name was changed to BME Clearing on 9 September 2013) do not publish statistics regarding long and short positions associated to commercial traders (i.e. energy companies) or to financial entities. The publication of such statistics via periodical market monitoring reports from each clearing house – or alternatively by the regulatory agencies having access to centrally cleared data – would provide more post-trade transparency, increasing the confidence in the market and thus its efficiency. As indicated in BME (2014b), up until September 9 2013, MEFF Sociedad Rectora de Productos Derivados S. A. (MEFF) ran both the activities of exchange and central counterparty (CCP) for the Spanish Financial Derivatives, the repo transactions on Public Debt (MEFFREPO), and power derivatives (MEFF Power). EMIR (European Market Infrastructure Regulation, Regulation (EU) 648/2012), obliges to separate the exchange activities from the CCP (European Union, 2012). Due to that, MEFF Sociedad Rectora del Mercado de Productos Derivados (in short, MEFF Exchange) acts since that date as the exchange and a new company (BME Clearing) as the CCP taking over the clearing house activities previously performed by MEFF.

Sanders et al. (2004) examine the CFTC’s Commitments of Traders reports for crude oil, unleaded gasoline, heating oil, and natural gas futures contracts. A positive correlation between returns and positions held by noncommercial traders (i.e. funds), and a negative correlation between commercial (i.e. hedgers) positions and market returns, are found. For both groups, returns lead positions. Commercials are net sellers the week following an increase in prices, and noncommercials are net buyers. The publication of long and short positions per agents’ type for MIBEL derivatives data centrally cleared would serve to assess if their evolution is also driven by the evolution of returns based on available prices (e.g. OMIClear and MEFF Power publish their settlement prices daily). Any anomaly in the usual evolution of those relationships might warn the supervisory authorities for excessive speculation in the context of some special event, e.g. a new regulated forward contracting auction.

Regarding crude oil market, Fleming and Ostdiek (1999) detect a strong inverse relation between the open interest in crude oil futures and spot market volatility. When the open interest is greater, the volatility shock of the spot market associated with a given unexpected increase in futures trading is much smaller. Therefore the trading of futures contracts improves depth and liquidity in the underlying market. Further research is encouraged with MIBEL spot and futures data to assess the impact of the development of the open interest on the spot market volatility. However, the main factors affecting the volatility of the Iberian spot market are the abundant wind generation and the regulatorily fixed prices for the remuneration of coal power plants burning indigenous coal. This subsidised price is applied since February 2011 by the


36

Spanish government for the sake of security of supply (MITyC, 2011a). Whereas the volatility is increased by the former due to the intermittent nature of wind generation, it is decreased by the latter as such fixed prices act as a price cap in the spot offers of thermal power plants limiting as well the price fluctuations in the futures market.


37

CHAPTER 3. OVERALL ASSESSMENT OF THE IBERIAN ENERGY DERIVATIVES MARKET AND RELATED REGULATION

CHAPTER SUMMARY

The efficiency of the Iberian Energy Derivatives Market in its first five and a half years

is assessed in terms of traded volumes, cleared and settled volumes in its clearing

house, and price. The continuous market shows steady liquidity growth. Its volume is

strongly correlated to that of the Over The Counter (OTC) market, the amount of

market makers, the enrolment of financial agents and generation companies belonging

to the integrated group of last resort suppliers, and the OTC cleared volume in its

clearing house. The ex-post forward risk premium has diminished due to the learning

curve and the effect of the fixed price retributing the indigenous coal fired generation.

This market is quite less developed than the European leaders headquartered in

Norway and Germany. Enrolment of more traders, mainly international energy

companies, financial agents, energy intensive industries and renewable generation

companies is desired. Market monitoring reports by the market operator providing post-

trade transparency, OTC data access by the energy regulator, and assessment of the

regulatory risk can contribute to efficiency gains.

3.1 Introduction

This research analyses the efficiency of the Iberian Energy Derivatives Market, the power futures market managed by OMIP (“Iberian Market Operator, Portuguese Pool”), by means of the evolution of two key indicators: (i) the traded volumes (reflecting the liquidity development); and (ii) the ex-post forward risk premium (reflecting the price efficiency). The analyses described in this chapter are based on the research performed by Capitán Herráiz and Rodríguez Monroy (2012b). Further research about the ex-post forward risk premium and the traded volume evolution is provided in Chapters 4 and 5 respectively. This market started operating on July 3, 2006. The derivatives data set in this research considers the first five and a half years (i.e. traded volumes until December 31, 2011).

The evolution of the Spanish electricity market has been influenced by the introduction of new regulation favouring the coal based generation with indogenous coal. Section 3.2 provides a snapshot of the Spanish electricity policy influencing the wholesale power market. This section is focused on the main measures affecting the


38

performance of the electricity forward market. The evolution of the traded volumes in the power futures market compared to the rest of existing forward contracting mechanisms is presented in Section 3.3. The evolution of the ex-post forward risk premium is tracked in Section 3.4. Section 3.5 provides policy considerations derived from the analyses performed in the previous sections with the aim of contributing to higher efficiency levels. Section 3.6 summarises all the results found in the research.

3.2 The current electricity policy context

3.2.1 The subsidised coal fired generation with indigenous coal

The Royal Decree 134/2010, of 12 February 2010, compelling the Spanish coal power plants to burn indigenous coal at a regulated price for the sake of security of supply, in force on February 26, 2011, fixes the generation costs of 10 coal power plants. For year 2011, the variable cost of the coal plant with biggest production is around 53 €/MWh, being the weighted average price of the 10 plants, according to their maximum allowed volumes, equal to 55.30 €/MWh. As a consequence, this policy measure sets a price threshold in the merit order curve of the spot market (MITyC, 2010a, 2011a). As indicated in CNE (2012b), this mechanism produced in the second quarter of year 2011 a smoother merit order curve in the day-ahead market. It reduced the peak price and increased the off-peak price. Therefore, the price signal needed for the efficient consumption disappeared. As the day-ahead market became more expensive, the suppliers not belonging to the incumbents purchased in year 2011 the major part of their energy in the subsequent intraday markets. A brief description of the day-ahead and intraday markets, as well as the market mechanisms for solving technical constraints and ancillary services is provided in Section 1.5.1 above. These markets, composed of 6 sequential auctions, resulted in average in year 2011 1.7 €/MWh cheaper than the day-ahead market. From August 2011, the prices in the day-ahead market increased due to more expensive offers of Gas Turbine Combined Cycles (GTCC). This was caused by higher Iberian gas prices, motivated by a larger demand of liquefied natural gas by Japan, in the aftermath of Fukushima nuclear accident. Therefore, in the second half of year 2011, the Spanish spot price was more expensive –with the exception of November– than in the main European power markets (Nordpool (Nordic countries) and European Power Exchange “EPEX” (France, Germany, Austria and Switzerland)).

The following subsection analyses the main drivers of the Spanish electricity forward price formation and the impact of the Royal Decree 134/2010.


39

3.2.1.1 The impact of the recognised price of the indigenous coal fired generation in the Spanish forward price formation

The GTCC generation accounts in Spain in 2009 for 24.2% of the installed capacity and for 29% of the electricity gross generation, being the main technology in the generation mix (Moreno and Martínez-Val, 2011). However, due to the effect of the recognised price for the coal fired generation with indigenous coal, the GTCC production passed from 63 TWh in year 2010 to 50 TWh in year 2011 (the load factor, obtained as the ratio between the equivalent working hours and the total amount of hours in a year, diminished from 30% to 22%) (CNE, 2012b). The Spanish electricity forward prices are influenced by the spot prices, the fuel prices –especially the gas prices– and the international electricity forward prices (Villaplana and Cartea, 2011). Figure 3.1 shows the evolution of the Spanish prompt year base load futures contract, compared to the underlying spot price (OMIE spot price; OMIE stands for “Iberian Market Operator, Spanish Pool”), the French prompt year base load futures contract traded in the European Energy Exchange (EEX), and the forward estimation of the GTCC generation costs. Monthly average values are used for each energy price. The settlement price of the Spanish prompt year power futures stabilises around 53 €/MWh since March 2011 influenced by the entry in force of Royal Decree 134/2010.

The clean spark spread is obtained as the difference between the Spanish power futures price and the forward GTCC generation cost taking into account the CO2 emission rates (e.g. Abadie and Chamorro, 2009). For the gas prices, Platts assessments of prompt year base load forward contract traded OTC (“Over The Counter”, i.e. out of organised markets) in the Dutch virtual trading point (TTF, “Title Transfer Facility”) are used. The Spanish gas market lacks of price transparency, as physical swaps for balancing purposes are arranged amongst participants without disclosing the price (Honoré, 2011; CNE, 2012c). Therefore a liquid reference in continental Europe has been taken instead. For the CO2 emissions, European Union Allowances (EUA) futures settlement prices in the InterContinentalExchange (ICE) are considered. The GTCC generation cost is calculated as the sum of the gas forward price, divided by the thermal efficiency of the power plant, and the price of the CO2 allowance multiplied by the emission rate. As in Abadie and Chamorro (2009), a thermal efficiency of 55% and an emission rate of 0.37 tCO2/MWh are employed. As shown in Figure 3.1, in the majority of the months, the clean spark spreads are positive (generation profits). There are some periods with negative spreads (second and third quarters of 2008 and 2011).

The Spanish prompt year futures price tends to be smaller than the neighbouring French power futures price. There are only few months with Spanish futures prices bigger than French ones, being the price difference smaller than in the opposite case. During year 2011, the Spanish price exceeds more often the French price influenced by the price level bound to Royal Decree 134/2010.

The Spanish prompt year futures price tends to be higher than the underlying spot price (“contango”). However, in year 2008, the spot price was often bigger than


40

the futures price, especially in the second half of the year (“backwardation”). At that time, the futures prices were influenced by downward prognoses due to the global financial turmoil. During year 2011, the Spanish spot and futures prices present smaller differences influenced by the price level set by Royal Decree 134/2010.

Figure 3.1. Evolution of the Spanish prompt year (“Y+1”) base load power futures settlement price versus the underlying spot price, the French base load power futures

settlement price and the forward GTCC generation costs (€/MWh).

10

30

50

70

90

Jul.2

006

Sep.

2006

Nov

.200

6

Jan.

2007

Mar

.200

7

May

.200

7

Jul.2

007

Sep.

2007

Nov

.200

7

Jan.

2008

Mar

.200

8

May

.200

8

Jul.2

008

Sep.

2008

Nov

.200

8

Jan.

2009

Mar

.200

9

May

.200

9

Jul.2

009

Sep.

2009

Nov

.200

9

Jan.

2010

Mar

.201

0

May

.201

0

Jul.2

010

Sep.

2010

Nov

.10

Jan.

2011

Mar

.201

1

May

.201

1

Jul.2

011

Sep.

2011

Nov

.11

€/M

Wh

OMIE Spot OMIP Y+1 GTCC Y+1 EEX French power futures Y+1

Source: OMIP-OMIClear, OMIE, EEX, Platts and ICE, adapted by authors

Table 3.1 shows high correlation coefficients of the Spanish prompt year base load power futures price with the GTCC generation cost, the French prompt year base load power futures price, and the Spanish power spot price. The correlation is obtained through monthly values for the data set spanning from July 2006 to November 2011 (65 observations).

Table 3.1. Correlation coefficients of the Spanish prompt year base load power futures settlement price with the underlying spot price, the French prompt year power futures and the forward GTCC Generation Costs.

Prompt Year Forward GTCC Generation Costs 0.94French Power Prompt Year Futures Price 0.87Spanish Power Spot Price 0.82

Source: OMIP-OMIClear, OMIE, EEX, Platts and ICE, adapted by authors


41

3.2.2 The moratorium to renewables

The deployment of renewable energy sources in the last 20 years has been outstanding in Spain (Moreno and Martínez-Val, 2011). However, the Royal Decree-Law 1/2012, of 27 January, freezes temporarily the grant of new feed-in tariff contracts to contain the large tariff deficit, as a portion of this deficit is caused by large incentives to the renewable energy sources. For instance, the market premium to solar technologies in 2010 exceeded 2,000 million € (MINETUR, 2012a). This Royal Decree-Law could affect negatively about 4,500 MW and 550 MW of wind and solar photovoltaic power projects respectively, as well as other special regime (renewables and cogeneration) projects (Energy Market Price, 2012). Nonetheless, a significant amount of renewable generation (ca. 11,000 MW) officially registered at the entry into force of the Royal Decree-Law can be installed, not being affected by the moratorium (Ordóñez, 2012). García Breva (2012) warns about the regulatory risk damaging the development of Spanish renewable energy sources, due to continuous changes in the special regime regulation. After the entry into force of the key regulation (Royal Decree 661/2007, of 25 May 2007, regulating the special regime production), a new Decree has been introduced every six months generating uncertainty to investors.

The effects of maintaining this moratorium should be analysed cautiosly by policy makers, due to the key contribution of the renewable energy sources in diminishing the CO2 emissions of the generation mix and the reduction of spot prices. Gelabert et al. (2011) study ex-post the effects of special regime generation on Spanish wholesale electricity prices, with data for years 2005-2009. They find that a marginal increase of 1 GWh of special regime generation is associated with a reduction of ca. 4% in wholesale power prices.

The special regime power plants profiting of a feed-in-tariff scheme can sell the electricity at a fixed price. Those installations affected by Royal Decree-Law 1/2012 may now be interested in hedging their production in the futures market or OTC.

3.2.3 The extension of life cycle of power plants

Whereas the GTCC and renewable sources in Spain are quite modern, as they have mainly been installed in the last 20 years, the nuclear plants – built in the nuclear boom of the 70’s and 80´s – and the coal power plants are older. Continuous debate is created about the extension or closure of the ageing Spanish nuclear plants. No new nuclear power plants are envisaged. Such new investments are surrounded of several uncertainties (Linares and Conchado, 2010). The effects on power prices due to the extension of life cycle of power plants are different depending on the generation technology. In the case of nuclear power plants, the extension produces smaller spot and forward prices, due to the base load nature of this technology. On the other hand, Sensfuß et al. (2008) detect that the extension of life cycle of old thermal power plants


42

(coal and fuel oil) creates upward price pressure in the wholesale power prices due to higher inefficiency of such aged plants.

3.2.3.1 The effect of the German Nuclear moratorium on power prices

After the Fukushima nuclear disaster on March 11, 2011, the German government discussed the shutdown of nuclear power by 2022. This moratorium was officially announced on March 14, 2011, stopping 7 nuclear reactors (5.3 GW). On May 21, 2011, the German government decided to shut down progressively all the nuclear reactors by 2022. The German power market joined the Trilateral Coupling in the Central West European Region – formed by the French, Belgian, and Dutch markets – on November 9, 2010. The market coupling algorithm provides a unique spot price in the region in the absence of congestion in the interconnection. Therefore their price spreads (both spot and futures) diminished with the introduction of the algorithm. However, the German moratorium has produced a smaller spot price convergence between the German and French prices (the French prices were 1.8 €/MWh cheaper in the first half of year 2011) and the inversion in the price level of the year futures (the German forward prices have become more expensive) (CRE, 2011).

3.2.4 The mitigation of large cost deficits in the electricity sector

3.2.4.1 Policy recommendations by the National Regulatory Authority

The Spanish national regulatory authority CNE (2012c) indicates the existence of a structural deficit in the Spanish electricity sector since a decade, as the recognised costs for regulated activities have been (and are) higher than the incomes from regulated prices paid by consumers. The deficit at March 6, 2012, reckons 21,812 million € (23,312 million € including the established deficit for year 2012). In order to tackle this deficit, CNE (2012c) suggests short term (i.e. urgent) measures, as the cost review of regulated activities and the delay of planned investments, the utilities’ cession of part of the generated debt to third parties, the pass through of the System Operator’s retribution and the large consumers’ interruption cost (they would be considered as energy costs rather than access costs), the charge of the capacity availability payments and generation investment incentives on the market participants instead of the consumers, the elimination of the financing of the insular costs by the mainland power system, the financing of feed-in tariff costs through the incomes of the CO2 auctions, and the temporary lamination of the premiums for the concentrating solar power plants already registered but still not working. As medium term measures, CNE (2012c) suggests to review the unit costs and the asset retribution basis (i.e. update of the incentive based regulation), the attribution to the CNE (now CNMC) – instead of the


43

Ministry of Industry – for establishing the methodology of retribution of the regulated activities, and the introduction of renewable auctions.

3.2.4.2 The first measures taken by the Government

The Royal Decree-Law 13/2012, of 30 March, diminishes the costs in the electricity and gas systems. It aims that the third party access rates are sufficient to cover the regulated costs, taking into account the foreseen ex-ante deficit for 2012. The following measures stated in this Royal Decree-Law may affect the performance of the electricity forward market: (i) The capacity payments are reduced 80 million € (this can produce higher spot prices in the peak hours, increasing the forward risk premium embedded in the forward prices; in this sense, the risk premia increase when the generators’ profits are less stable (Fabra, 2009)); (ii) The retribution to the interruption services by large consumers is reduced 60 million € (such consumers may now be incentivised to contract instead firm services, producing an upward pressure in the spot and forward prices; additionally, a bigger demand on firm services may produce a change in the energy tariff cost structure to reflect this need properly); (iii) The recognised payments to the coal fired generation with indigenous coal will be limited to 50 million € (the reduction of recognised incomes to older coal plants may increase the spot and forward prices, in line with Sensfuß et al. (2008)) (MINETUR, 2012b, 2012c).

In order to understand the capacity payments mentioned above, they are a regulated payment for the generation plants providing capacity. They include 2 services: the long term investment incentive and the medium term generation availability. The CNE (now CNMC) has assessed the introduction of a capacity market mechanism to streamline this issue (CNE, 2012b, 2012c, 2012h).

Regarding the neighbouring country, the Portuguese government also envisages immediate energy policy reforms. Those measures will also focus on strict control of the energy systems costs in a first phase, followed by their reduction in a second phase. For instance, the state costs related to the supervision of the electricity and gas concessions will have to be paid by the consumers through the end-user tariffs (Cabral, 2012).

3.2.5 The introduction of household hourly tariffs

As indicated by CNE (2012c) and MINETUR (2012b), a review of the cost components is needed to ensure that the Third Party Access rates cover the costs of the electricity sector. Additionally to this, changes in the domestic end-user tariffs are needed to take into account the benefits of smart grids and smart meters, helping to a massive demand response that increases the efficiency of the whole power system (Conchado and Linares, 2010). Moreno (2011) indicates that hourly rates and smart meters are widely used by large consumers since the liberalisation of the electricity


44

sector in 2008. Such a research detects the positive effect of liberalisation correcting inefficient trends of electricity demand in Spain. Time-of-use tariffs for household consumption would be adequate if such prices are not set by the government (e.g. for consumers with power exceeding 2 kW) and they are liberalised. The gradual introduction of smart meters in 25 millions of households (i.e. power less than 15 kW) in Spain, as envisaged by the Plan of meters substitution, has not become a reality yet. This plan should be completed by 2018 (MITyC, 2009b). Cabeza (2011) indicates that in 2014 half of those meters would be installed and the remote system would be working. Regulatory developments establishing a national plan for active management of the power demand are needed to take full advantage of these devices. The major Spanish utilities have developed pilot projects of smart grids in cities where their distribution companies have installed smart meters. See e.g. the projects developed by Endesa in Málaga (Endesa, 2013) and by Iberdrola in Castellón (Iberdrola Distribución Eléctrica, 2013).

Joskow and Wolfram (2011) indicate that the introduction of time-varying electricity prices for households in the United States (US) is slow despite of decreasing device costs and improved functionality for meters and automated demand response technologies. They think that most utilities will begin offering alternative tariffs while leaving flat-rate pricing the default option. Regarding the US implementation of demand response and advanced metering, the US Federal Energy Regulatory Commission (FERC) performed and extensive survey in the middle of 2012. FERC published a report with the results in December 2012 (FERC, 2012). Shioshansi (2013) reflects about the results and findings of such a research. He indicates that the survey comprises 3,349 companies, of very different nature and size. Around 60% answered to FERC consultation. These big magnitudes reflect the complexity of this issue. The leader state in the estimated penetration of advanced metering is District Columbia (87.1%) followed by California (70.5%), whereas the national average falls to 22.9% (a goal of 50% is envisaged for 2015-2016). FERC defines demand response as “changes in electric use by demand-side resources from their normal consumption patterns in response to (a) changes in the price of electricity, or (b) to incentive payments designed to induce lower electricity use at times of high wholesale market prices or when system reliability is jeopardized”. FERC estimates the demand response potential at 72 GW, a 25% rise from the 2010 survey – that is more than 9% of US peak demand. Shioshansi (2013) criticizes this optimistic figure as demand peaks on different parts of the network at different times. The market operators with bigger demand response potential are those who have introduced wholesale auctions in which load is identified as a capacity resource, introducing “demand bidding & buyback” (i.e. “negawatts”). The current leaders are PJM and MISO. FERC identifies the following 4 programs accounting for 80% of the total reported potential peak reductions in the US: (i) Load as capacity resource; (ii) Interruptible load; (iii) Direct load control; and (iv) Time of use. FERC reports a little over 20 GW of peak demand reductions from “demand response resources” in 2012, “representing use of 31% of the total reported peak load potential.” Regarding time variable pricing, FERC survey exposes the scarce advance: residential time-of-use programs were offered by 151 respondents (of a total of 1,900) and only 28 entities reported offering real-time pricing. Therefore, there is substantial room for improvements. The leader states in time of use


45

tariffs and real time pricing for residential customers are Arizona and California, due to their favourable regulatory frameworks. In the former, 1/3 of residential users of the 2 largest utilities have voluntarily engaged in time of use tariffs. In California, the debate is in extending the critical peak pricing already used for large industrial and commercial consumers to household ones, but legal and regulatory challenges are currently assessed by the regulator CPUC (California Public Utilities Commission). FERC identifies several barriers to demand response implementation, being the main ones: (i) limited number of retail customers on time-based rates; (ii) measurement and cost-effectiveness of (load) reductions; (iii) lack of uniform standards for communicating demand response price signals and usage information; (iv) lack of customer engagement; and (v) lack of demand response forecasting and estimating tools. Shioshansi (2013) concludes that the demand response hurdles (as well as massive integration of renewable energy sources) are put everywhere (not just in the US case) by the large generators who view load as a “given” and generation as what has to be adjusted to meet the variable load. Even FERC’s chairman (Jon Wellinghoff) criticizes that outdated perspective of treating demand to be whatever it wants to be and ramping generation up and down to match it, no matter the costs. The latter defends that demand could be adjusted when and if it is cheaper to do rather than generation, e.g., through time variable pricing or other multiple ways of managing load. Such alternatives would be easier, more environmentally friendly, and much cheaper.

As a final conclusion regarding the impact of smart grid innovative technologies in the price formation, the massive deployment of dynamic pricing of electricity can reduce the price spikes. The price differences between peak and off-peak periods would then fall. Therefore, average spot prices would decrease as well as the forward risk premium.

The first bet by the Spanish Government in the billing based on real –time pricing from the wholesale electricity market applied to the last resort end users’ tariff (now such a tariff is called Voluntary Price for the Small Consumer, PVPC, introduced by the electricity sector reform at the end of year 2013, through the Law 24/2013, of 26 December) has entered into force on 1 April 2014 by means of the Royal Decree 216/2014, of 28 March, establishing the PVPC calculation methodology and its contractual legal regime (BOE, 2013; MINETUR, 2014). Further research is encouraged to assess the potential savings from this new billing mechanism to end users compared to the application of forward pricing mechanisms (e.g. the CESUR auctions) as in the previous years.

3.3 Evolution of the trading efficiency

The trading efficiency of the Iberian energy derivatives market is analysed according to its main structural features. A basic overview of the main features of this market is provided in Chapter 1 (Section 1.5). Regarding the trading efficiency analysis, the volume development of the Iberian energy derivatives market is compared with the rest of existing forward market mechanisms. Its concentration basic figures are compared with those of the generation and supply activities. The key trading drivers of


46

OMIP continuous market are identified. Additionally, this futures market is compared with the main European benchmarks.

3.3.1 Volume comparison between Iberian forward trading mechanisms

The evolution of OMIP traded volumes has to be seen in conjunction with the dominant non-organised OTC market –in Spain, financial trading either done bilaterally or with the intermediation of brokers, being a portion of them cleared and settled through clearing houses– and with the volumes matched in the regulated auctions for the Spanish last resort supplies (“CESUR” auctions). CESUR stands in Spanish for “Energy Contracts for the Last Resort Supply”. Although the first CESUR auction was celebrated in June 2007, covering part of the regulated supplies by the distribution companies, the last resort supplies – managed by new companies, the so-called “last resort suppliers” belonging to the same vertical group of those distribution companies – substituted such regulated supplies since July 1, 2009, according to Royal Decree 485/2009 (MITyC, 2009a).

Figure 3.2 shows the evolution of the cleared and settled volumes (in TWh) in OMIClear (OMIP clearing house), and the matched volumes in CESUR auctions in the first five and a half years of the Iberian energy derivatives market (i.e. data from July 3, 2006, until December 31, 2011). It additionally shows the cleared volumes in another clearing house (MEFF Power, whose current name since 9 September 2013 is BME Clearing), which started operation on March 21, 2011, for the clearing and settlement of OTC power trades with Spanish underlying spot prices. MEFF stands in Spanish for “Spanish Financial Futures Market” (BME, 2011; 2012).

There are two market modes in OMIP futures market: the continuous market and auctions. Whereas the former is the main mode, the latter has performed a key role in the development of the liquidity in OMIP, as the Spanish distribution companies and the Portuguese last resort supplier were obliged to purchase energy in such auctions until July 2009 and July 2010 respectively (Capitán Herráiz and Rodríguez Monroy, 2009a; 2009b). Furthermore OMIClear permits the clearing and settlement of OTC volumes by OMIP trading members, either bilaterally or through one of the registered brokers. As shown in OMIP (2014a), at the beginning of year 2014, five brokers (widely known in the European wholesale energy trading) are registered in OMIP, namely: CIMD – Corretaje e Información Monetaria y de Divisias, S.V., S.A.; ICAP Energy, AS; Spectron Energy Services Limited; Tradition Financial Services Ltd.; and Tullet Prebon (Europe) Limited.

In the period June 2007-December 2011, 17 CESUR auctions have been celebrated sourcing part of the last resort supplies. The last resort suppliers purchase the remaining part in the spot market. Prior to each CESUR auction, they have to communicate to the Ministry of Industry their load forecast. Since March 2011, they hedge the part of their forecast not covered by the CESUR auction at the same equilibrium price by means of a contract for differences mechanism regulated by Royal


47

Decree 302/2011. The counterparties are the special regime facilities selling their energy in the spot market (day-ahead and intraday auctions) and receiving a fixed price (“regulated rate”). As the equilibrium prices of the CESUR auctions are usually bigger than the underlying spot price, the rents obtained from this mechanism help to mitigate the part of the tariff deficit generated by the regulated rate recognised to such special regime facilities (BOE, 2011; CESUR, 2012).

Figure 3.2. Evolution of traded and cleared volumes in OMIP-OMIClear, traded volumes in CESUR auctions, and cleared volumes in MEFF Power (TWh).

0

6

12

18

24

Jul.2

006

Sep.

2006

Nov

.200

6

Jan.

2007

Mar

.200

7

May

.200

7

Jul.2

007

Sep.

2007

Nov

.200

7

Jan.

2008

Mar

.200

8

May

.200

8

Jul.2

008

Sep.

2008

Nov

.200

8

Jan.

2009

Mar

.200

9

May

.200

9

Jul.2

009

Sep.

2009

Nov

.200

9

Jan.

2010

Mar

.201

0

May

.201

0

Jul.2

010

Sep.

2010

Nov

.10

Jan.

2011

Mar

.201

1

May

.201

1

Jul.2

011

Sep.

2011

Nov

.11

TW

h

OMIP Call Auction OMIP Continuous Market OTC cleared by OMIClear OTC cleared by MEFF Power CESUR Auction Source: OMIP-OMIClear (2012), CESUR (2012), BME (2012)

OMIP traded volumes in the first two years – i.e. from July 2006 to June 2008 – were led by compulsory auctions. Since that moment until the end of 2009 the continuous volumes reached a similar size to the auction ones. Afterwards, the continuous market volumes kept growing. During 2010, the scarce auction volumes were generated by compulsory auctions of peak futures for the Portuguese last resort supplier. The underlying price of such peak futures was the spot price of the Spanish price area. OMIP peak futures are still very illiquid. The month with record of continuous volumes in the analysed period was March 2011 (4.86 TWh). The OTC cleared volumes also reached a record in that month (5.68 TWh) and maintained a growing trend, influenced by the strong OTC trading development (OMIP-OMIClear, 2012).

Since March 21, 2011, OTC power trades with Spanish underlying spot prices can also be cleared and settled by MEFF Power. Although the number of enrolled members in MEFF Power is growing fast (25 at the end of December 2011, of which 13 are also active in OMIP, which has 38 members at that date) the registered volumes until the end of December 2011 were still small (3.8 TWh) compared to the OTC


48

registered volumes in OMIClear in the same period (24.5 GWh) (BME, 2012; OMIP-OMIClear, 2012). The amount of registered members in OMIP and in MEFF Power follows growing, in the second case at a faster pace. At the beginning of 2014, there are 48 trading members registered in OMIP (OMIP, 2014a) and even more (59) in MEFF Power (BME, 2013). Such a growth has produced an increase in the traded/cleared volumes, producing a record during year 2013: in October 2013, 5,258 GWh were traded in OMIP continuous market, 4,334 GWh were cleared in MEFF Power, and 39,053 GWh were traded OTC (CNMC, 2014b). The cleared volumes in MEFF Power during year 2013 reached 33.5 TWh, four times the volumes registered in year 2012 (BME, 2014a). Regarding the agents active in the OTC market, Alba Ríos and Moreda Díaz (2010) indicate that the Spanish OTC market is composed of around 25 players (gas and power companies, financial entities, large consumers and commodity traders). The number of participants in the CESUR auctions is around 30 (CNE, 2011c).

Whereas OMIClear is preferred by the market participants for the settlement of large maturities (month, quarter and year) futures contracts, MEFF Power is preferred for short maturities (day and week) contracts (so far base load swaps of the Spanish zone). The different preferences by the market participants in terms of derivative type and maturity make that both clearing houses behave complementarily. For the sake of systemic risk mitigation, a larger portion of centrally cleared OTC trades is always welcome by the supervisory authorities (IOSCO, 2011). That goal may be better reached through dynamic competition regarding tariffs and services between the existing clearing houses, as such a competition may attract more OTC volumes.

Table 3.2 shows the traded volumes in OMIP, in CESUR auctions and OTC, as well as the cleared and settled OTC volumes in the 2 clearing houses. Additionally, it shows the Spanish mainland demand at busbar. Values for year 2011, year 2010, and the whole period covered by this research (July 2006 – December 2011) are provided. The traded volumes in OMIP (auction and continuous), CESUR auctions, and OTC grow in year 2011 compared to year 2010. Conversely, the OTC registered volumes in OMIClear and the demand diminish in year 2011. The CESUR volumes are only 7% bigger than OMIP continuous volumes in that year. The OTC market has experienced a steady growing trend, summing up in the whole period ca. 863 TWh. The first two years in which the OTC volumes are bigger than the demand (i.e. a “churn” ratio bigger than 1) are the years 2010 and 2011. The churn is defined as the ratio of traded volume of a commodity to throughput or generated output, or some other measure denoting physical consumption (OFGEM, 2009). The accumulated OTC volumes are 10.3 times bigger than OMIP continuous volumes and 4.5 times bigger than CESUR matched volumes. Only a minor part of the whole OTC volume is centrally cleared: 10.4 % by OMIClear and 1.4% by MEFF Power (OMIP-OMIClear, 2012; BME, 2012; CESUR, 2012; Intermoney, 2012; REE, 2012).


49

Table 3.2. Evolution of traded volumes in OMIP, CESUR auctions, OTC and demand (TWh).

Data in TWhYear 2011 Year 2010

July 2006 - December

2011OMIP Auctions 1.32 0.59 57.22OMIP Continuous 32.87 25.19 83.79CESUR auctions 35.31 27.27 191.26Total OTC 298.20 278.84 863.47OTC registered at OMIClear 27.08 29.46 89.74OTC registered at MEFF Power 3.76 - 3.76Spanish mainland demand at busbar 255.18 260.61 1,423.44 Source: OMIP-OMIClear (2012), BME (2012), CESUR (2012), Intermoney (2012) and REE

(2012)

3.3.2 Competition in the power futures market

According to the futures market operator, this market became less concentrated during year 2010 and the market share of the three biggest players accounted for 40% in the last months of 2010 (OMIP, 2011a). This share compares well against the generation and supply concentration, according to 2010 figures: in Spain, the 3 largest generators’ market share, based on the installed capacity, covers around 59%. That figure decreased compared to the previous year, and is smaller than in Portugal: in Spain, for year 2009, the 3 largest generators’ market share, based on the installed capacity, covers around 67%, and in Portugal around 75% (Capgemini, 2010). In the Spanish electricity retail market, discarding the last resort supplies, the market shares of the 3 biggest companies add up to 74% in energy and 90% in customers (CNE, 2011b).

3.3.3 Key trading drivers in OMIP continuous market

Correlation analysis of the traded volumes in OMIP continuous market during each month serves to identify key trading drivers for such a market. The correlation coefficients of those drivers are shown in Table 3.3, namely: the OTC volumes, OMIP market makers active at each moment (in December 2011, 4 market makers provided quotations for prompt month, quarter, and year base load futures contracts with the spot price of the Spanish zone as underlying price), the enrolment of financial agents, the enrolment of generation companies belonging to the integrated group of the last resort supplier and the OTC volumes cleared and settled by OMIClear. Regarding the amount of market makers, this figure has stabilized afterwards, not growing with the enrolment of new trading members in OMIP. For instance, on January 2014, there are three market makers active in OMIP (OMIP, 2014b).


50

Table 3.3. Correlation coefficients between OMIP continuous traded volumes and the monthly evolution of key trading drivers.

Correlation between OMIP continuous volumes and:

Correlationfactor

OTC volumes 0.88OMIP market makers 0.82OMIP financial agents 0.80OMIP vertically integrated generation companies 0.76

OMIClear OTC registered volumes 0.70 Source: OMIP-OMIClear (2012) adapted by authors

3.3.4 Comparison with the most developed European Exchanges

This section provides a comparison with the main figures of the main European energy derivatives exchanges, located in Norway and Germany, with data corresponding to years 2010 and 2011. Capitán Herráiz and Rodríguez Monroy (2010a) also provide a comparison among those energy exchanges with data related to year 2008 and 2009. Furthermore, Chapter 5 provides some comparison (with data related to years 2008) including the French energy exchange. The research performed by Capitán Herráiz and Rodríguez Monroy (2010a), mentioned above, provides a description of the status quo of the Latin American Energy Derivatives Markets. They found that there is no sound initiatives regarding Latin American Energy Derivatives Markets, as the typical hedging instruments in the local power markets are the bilateral contracts, and the hydrocarbon markets are still mainly driven by former monopolies. The successful experiences regarding power derivatives within MIBEL may be employed to boost the potential energy derivatives markets in Latin America. An harmonised approach led by ARIAE similar to the EU regional initiatives regarding a single EU energy market, an appropriate climate for investors, the promotion of incentives for dynamic traders, and expansion of the existing financial commodities markets in Latin America embracing energy derivatives products would also help to fulfil this worthy goal.

Despite of the steady development in terms of members and cleared volumes in the Iberian power futures market, its figures are still far from the most mature European energy derivatives exchanges (Nasdaq OMX Commodities – traditionally known as Nord Pool, headquartered in Norway – and EEX – headquartered in Germany). Table 3.4 shows the traded volumes and OTC registered volumes, as well as the number of participants in the three exchanges in the year 2011, compared to the previous year.


51

Table 3.4. Comparison of the Iberian power futures market with the most developed European energy derivatives exchanges.

Year2011

Year2010

Year2011

Year2010

Year2011

Year2010

Traded volumes (TWh) 34.2 25.8 1,028.2 1,286.7 498.6 496.4OTC registered volumes (TWh) 27.1 29.5 695.1 803.1 576.8 711.9Total cleared volumes (TWh) 61.3 55.2 1,723.3 2,089.8 1,075.4 1,208.3Number of participants 38 32 356 361 172 157

OMIP-OMIClear

Nasdaq OMX Commodities

EEX

Source: OMIP-OMIClear (2012), Nasdaq OMX Commodities (2011, 2012) and EEX (2010;

2012b), adapted by authors

Due to OMIP emerging nature, it is the only exchange showing in year 2011 bigger total cleared volumes and members compared to year 2010. The smaller total cleared volumes in EEX in year 2011 are due to the uncertain trading environment, in which traders prefer shorter term maturities (EEX, 2012a). Both the German and Nordic markets present quite larger churn ratios than in the Spanish case (around 8 and 7 respectively, considering OTC and exchange based trading) (OFGEM, 2009). Their exchanges perform quite more efficiently than OMIP, as liquidity exists for all maturities regarding electricity base load products and the other derivatives traded (gas, CO2 and coal (EEX)). In Nasdaq OMX Commodities, some liquidity exists for electricity contract for differences hedging the price risk of the Nordic price areas. In EEX, liquidity of peak futures is well developed, and there are also off-peak contracts, though their liquidity is small. Additionally those markets have a much wider range of trading members, with participation of many municipalities and large industrial consumers, balancing the participants’ structure and theoretically providing more robust price signals (EEX, 2012b; NASDAQ OMX Commodities, 2012).

3.4 Evolution of the price efficiency

The price efficiency is tracked through the ex-post forward risk premium, obtained as the difference of the average forward price during the quotation period of the contract and the resulting average spot price during the delivery period (see e.g. Redl et al., 2009; Redl and Bunn, 2011; Capitán Herráiz and Rodríguez Monroy, 2009b; Furió and Meneu, 2010; and Villaplana and Cartea, 2011). A quantitative and qualitative analysis of the Spanish forward electricity prices based on this concept is provided in detail in Chapter 4. The mathematical definition of the ex-post forward risk premium appears in Section 1.6.1.1.


52

3.4.1 Comparison of the ex-post forward risk premia in the Iberian power futures market and the CESUR auctions

Figure 3.3 shows the evolution of the ex-post forward risk premium in OMIP and in the CESUR auctions. For OMIP prices, the average settlement price of the prompt quarter base load futures contract in its last month of quotation is employed. The CESUR price corresponds to the equilibrium price of the prompt quarter base load forward contract in the auction celebrated immediately before its delivery.

Figure 3.3. Evolution of the ex-post forward risk premia in the Iberian energy derivatives exchange and in CESUR auctions.

-11

-9

-7

-5

-3

-1

1

3

5

7

9

11

13

15

17

19

21

Q4-06

Q1-07

Q2-07

Q3-07

Q4-07

Q1-08

Q2-08

Q3-08

Q4-08

Q1-09

Q2-09

Q3-09

Q4-09

Q1-10

Q2-10

Q3-10

Q4-10

Q1-11

Q2-11

Q3-11

Q4-11

Delivery Period

€/M

Wh

OMIP Settlement Price - Spot Price (€/MWh) CESUR Price - Spot Price (€/MWh)

Source: OMIP-OMIClear (2012), CESUR (2012), and OMIE (2011) adapted by authors

The price efficiency has improved with the development of the futures market, as the forward risk premia evolve towards smaller values. However, positive values tend to dominate. Capitán Herráiz and Rodríguez Monroy (2010b) find, as developed in Section 4.3, that the analysis of OMIP forward risk premia (data from July 2006 to February 2010) shows differences per contract maturity, allowing arbitrage gains through combined trading of month, quarter and year futures contracts. Arbitrage opportunities also arise between OMIP and CESUR due to the differences in their equivalent premia. The large positive premia during 2009 and the first quarter of 2010 are caused by rising forward prices and decreasing spot prices in the Iberian energy market. The former were pushed up by uncertainty in global financial markets, tight credit conditions and price increases in energy commodity prices. The latter were


53

decreased by the depressed economic situation and low prices in the power pool with reduced demand and strong penetration of renewables. The smaller spot prices were also influenced by large rainfalls and the effect of gas take-or-pay contracts (Alba Ríos and Moreda Díaz, 2010).

The positive forward risk premia in CESUR auctions are usually bigger than in OMIP. The smaller premia during year 2011 are influenced by new regulation – the mechanism solving restrictions for security of supply described in Section 3.2.1.1 – affecting the spot price formation and accordingly the forward price levels.

3.4.2 Economic impact of CESUR auctions in the energy cost of the last resort supply rates

Table 3.5 shows the economic impact on the energy costs of the last resort supply rates – based on CESUR equilibrium prices – due to frequent positive forward risk premium both in base load and peak products. The data span from the 9th auction (celebrated in June, 2009) until the 16th (in September 2011).

Table 3.5. Economic impact of the electricity purchased by the Spanish last resort suppliers in CESUR auctions.

Auction Product Type Capacity(MW)

Energy(MWh)

Auction price(€/MWh)

Spot price(€/MWh)

Premium(€/MWh)

Economicimpact (€)

baseload 4,800 10,598,400 42.00 35.05 6.95 73,658,880peak 670 530,640 47.60 38.55 9.05 4,802,292baseload 5,000 11,045,000 45.67 32.87 12.80 141,376,000peak 670 530,640 51.31 36.72 14.59 7,742,038baseload 4,800 10,363,200 39.43 25.38 14.05 145,602,960peak 540 414,720 43.70 29.99 13.71 5,685,811baseload 4,800 10,483,200 40.49 34.97 5.52 57,867,264peak 600 468,000 44.52 39.20 5.32 2,489,760baseload 4,000 8,832,000 44.50 44.07 0.43 3,797,760peak 536 424,512 50.48 49.01 1.47 624baseload 4,000 8,836,000 46.94 43.33 3.61 31,897,960peak 392 310,464 53.00 48.22 4.78 1,484,018baseload 4,000 8,636,000 49.07 45.22 3.85 33,248,600peak 306 235,008 53.99 48.66 5.33 1,252,593baseload 4,000 8,736,000 51.79 48.12 3.67 32,061,120peak 406 316,680 55.13 51.25 3.88 1,228,718baseload 3,600 7,948,800 53.20 54.23 -1.03 -8,187,264peak 688 544,896 56.63 58.62 -1.99 -1,084,343baseload 3,800 8,394,200 57.99 52.01 5.98 50,197,316peak 458 357,240 63.00 58.58 4.42 1,579,001

Total (MWh & €), Weighted Average (€/MWh) 98,005,600 46.83 40.83 5.99 587,324,516

9thQ3-09

Q4-09

10thQ1-10

Q2-10

16th Q4-11

11th Q3-10

12th Q4-10

13th Q1-11

14th Q2-11

15th Q3-11

Source: CESUR (2012) and OMIE (2011) adapted by authors

The purchase of the Spanish last resort suppliers in CESUR auctions, with delivery until the end of year 2011, has resulted 587 million € more expensive than the same amount of electricity valued at the spot price, due to the dominant positive forward risk premium. The only delivery period with negative premium is the third


54

quarter of 2011, resulting in a saving of 9.3 million € compared to the spot price valuation.

3.5 Energy policy considerations

The following recommendations are drawn from the facts exposed in Section 3.2, and the results of the analyses done in Sections 3.3 and 3.4 with the aim of increasing the efficiency of the Iberian power futures market.

3.5.1 The need for increased post-trade transparency from the power futures market operator

OMIP-OMIClear has published quarterly bulletins with basic statistics until the end of 2009. For the sake of post-trade transparency, its market surveillance department could publish monthly (or even much worthy, on a weekly basis, e.g. as a newsletter) reports analysing key oversight and liquidity indicators. Such market monitoring reports could aggregate statistics of the volumes traded for different groups of trading members (e.g. financial agents, companies belonging to Iberian integrated energy groups – i.e. the traditional “incumbents” –, energy companies not belonging to such incumbents, and even market makers) to show which are the most active members. Indication of the number of active companies effectively trading would indicate how dynamic this market is. Concentration indexes (e.g. Herfindahl-Hirschman index “HHI”) would provide worthy information. As in CNE (2011c), the “HHI” index is built as the sum of squares of the market shares considering all the market players. Therefore, it ranges between 0 (perfect competition) and 10,000 (monopoly). CRE (2012) states that concentration is small with values lower than 1,000 and large with values higher than 1,800. Valid references for this kind of monitoring reports are Nasdaq OMX Commodities monthly reports, and EEX market monitoring semiannual reports (NASDAQ OMX Commodities, 2012; EEX, 2009a).

OMIP-OMIClear could provide more transparency regarding the net positions of OMIP trading members by including open interest statistics per different traders’ categories in the suggested market monitoring report. A good reference is provided by the U.S. Commodity Futures Trading Commission (CFTC), in its disaggregated “Commitments of Traders” (COT) weekly report, showing a breakdown of each Tuesday’s open interest for markets in which 20 or more traders hold positions equal or above the reporting levels established by CFTC. The 2 categories of the legacy (i.e. basic) COT reports are very indicative: commercial (hedgers) and non-commercial (speculators). For each category, the COT report shows long and short positions, changes from previous report, percentages of open interest, numbers of traders, and concentration of positions held by the largest four and eight traders (CFTC, 2011). Basically, there are two basic types of position in derivatives trading: long and short. Regarding long position when a trader buys a derivatives contract (e.g. futures or


55

options) that he is not short (i.e. he did not sell similar contract before), he is opening a long position. Conversely, when a trader sells a derivatives contract that he is already long (i.e. he bought similar contracts before), he is closing a long position. When a trader is “long”, he wins when the price increases, and loses when the price decreases. Regarding short position, when a trader sells a derivatives contract that he is not long, he is opening a short position; when he buys a derivatives contract that he is already short, he is closing a short position. When a trader is “short”, he wins when the price decreases, and loses when the price increases. In general, “net position” is the difference between total open long (receivable) and open short (payable) positions in a given financial asset held by an individual. Another key defition regarding the management of short and long positions is the “open interest”, also known as open contracts or open commitments. It refers to the total number of derivative contracts not settled in the immediately previous time period for a specific underlying security. A large open interest indicates more activity and liquidity for the contract (Chance and Brooks, 2010).

Another worthy structure could be the one provided by the Administrator of CESUR auctions, as it indicates the CESUR matched amounts for 3 separated agents’ types: (i) companies owning power plants in Spain; (ii) companies headquartered in Spain; (iii) integrated business groups owning a last resort supplier (OMEL, 2011).

3.5.2 The necessity for trade repositories for a comprehensive oversight by regulators

The bulk of the Iberian power forward trading is OTC not cleared through a central counterparty. Furthermore, the agents can close their open positions registered in clearing houses through other trades established OTC or through other market mechanisms, e.g. OMIP continuous market or CESUR auctions. In order to have a full picture of the market, enabling the national regulatory authorities to perform a comprehensive oversight, all the energy transactions would have to be reported to such authorities. Current European legislative initiatives, shown in Table 3.6, are pointing out in that direction.

Table 3.6. Main European legislative pieces impacting on energy derivatives trading.

Legislative piece

Published on: Scope

REMIT Dec. 8, 2011 Regulation on wholesale Energy Market Integrity and TransparencyMiFIR Oct. 20. 2011 Draft Regulation on Markets in Financial Instruments and amending EMIRMiFID II Oct. 20, 2011 Draft reviewed Directive on Markets in Financial InstrumentsMAR Oct. 20, 2011 Draft Regulation on insider dealing and market manipulation (Market Abuse)EMIR Jul. 27, 2012 Regulation on OTC derivatives, central counterparties and trade repositories Source: European Union (2011, 2012) and European Commission (2011a, 2011b, 2011c)


56

The Regulation on wholesale energy market integrity and transparency (REMIT) entered in force on December 28, 2011, prohibiting insider trading and market manipulation practices in the European wholesale electricity and natural gas markets (European Union, 2011). The increase in existing transparency levels will produce more competitive and liquid markets, reducing costs to market players and on end-users’ supplies (Gensler, 2012). Discussion among all the stakeholders is desired for the proper implementation of these legislative pieces, in order to prevent troublesome overlapping and flaws in this regulatory orchestration, providing more trading confidence to the participants and a sound supervisory framework for the authorities involved. REMIT envisages transaction reporting and inclusion of trading orders (i.e. submitted bids (purchase orders) and offers (sale orders)). The definition of the reporting mechanism is currently under development by the European Commission. This can take up to 18 months (ACER, 2011b). Such a mechanism should include non-standardised contracts (e.g. reporting could be done every time the price varies according to an indexation formula or once providing the details of such formula) and the orders from the opaque OTC trading to avoid supervisory gaps. The reporting of bilateral contracts – e.g. standard contracts as those defined through the European Federation of Energy Traders (EFET) templates – would contribute to price discovery of intra-group transactions (this is especially relevant in a market where the generation and retailing activities could be dominated by a single integrated company, as it is the case in the Portuguese electricity market). The access to all the OTC data will help to prevent excessive speculation impacting on the prices (Chilton, 2012a).

Despite of long lead times for the implementation of the European legislative pieces cited above, national energy regulators can also access to OTC data by own initiative, if stated in their national law. This is the case of the Spanish energy regulator (CNE, now CNMC). As stated in CNE (2011b), this regulator has access to limited information over OTC power transactions (volumes and transaction prices, through the information voluntarily submitted by the main brokers). On March 5, 2011, the Law of Sustainable Economy was published in the Spanish Official Gazette. The 5th final disposition of this Law modifies the Securities Market Law, enabling the information exchange between CNMV (the Spanish Financial Services Authority, empowered to request OTC power data in the Spanish financial market) and the remaining entities composing the MIBEL Regulatory Council.

MIBEL stands for “Mercado Ibérico de Electricidad” (Iberian Electricity Market). The MIBEL Regulatory Council is composed of the national energy and securities regulatory agencies, namely, Spanish CNE (“Comisión Nacional de Energía”) –since October 7, 2013, CNMC (“Comisión Nacional de los Mercados y la Competencia”), the Spanish CNMV (“Comisión Nacional del Mercado de Valores”), Portuguese ERSE (“Entidade Reguladora do Sector Energético”) and Portuguese CMVM (“Comissão do Mercado de Valores Mobiliários”), as established in the “Agreement between the Portuguese Republic and the Kingdom of Spain relative to the constitution of an Iberian Electrical Energy Market” – the so-called MIBEL Agreement – signed by the respective governments, on October 1st 2004, further modified on 18 January 2008 (BOE, 2004; 2008). The four entities composing the MIBEL Regulatory Council signed on May 17, 2011, a Multilateral Memorandum of Understanding (MoU) for the efficient coordination


57

in the MIBEL supervision, facilitating the OTC data collection and supervision (MIBEL Regulatory Council, 2011). All the key information regarding MIBEL regulatory issues can be found at their own website (www.mibel.com).

On the other hand, the Royal Decree-Law 13/2012, of March 30, transposes the Directives for the internal electricity and gas markets, i.e. Directive 2009/72/EC and Directive 2009/73/EC (MINETUR, 2012b; European Commission, 2009a, 2009b). These two Directives together with other three Regulations constitute the third legislative package for an internal EU gas and electricity market (ACER, 2011a). That Royal Decree-Law modifies the Law of the Electricity Sector (Law 54/1997, of November 27) allowing the Ministry of Industry, the CNE and the Spanish Competition Authority (now both merged at CNMC), as well as the European Commission to access during at least 5 years to the data of all the transactions of electricity supply contracts as well as electricity derivatives concluded with wholesale customers and Transmission System Operators (TSOs). Likewise, this Royal Decree-Law modifies the Law of the Hydrocarbons Sector (Law 34/1998, of October 7) allowing the same aforementioned authorities to access during at least 5 years to the data of all the transactions of gas supply contracts as well as gas derivatives concluded with wholesale customers and TSOs, the underground storage system operators, and the Liquefied Natural Gas (LNG) system operators. The Spanish Government intends to perform a reform in the gas sector during year 2014, and therefore a new hydrocarbons law would be enacted. On the other hand, during year 2013 the electricity reform was perfomed and thus the current electricity sector law is the Law 24/2013, of 26 December (BOE, 2013).

3.6 Results

The traded volumes in the continuous market of the Iberian Energy Derivatives Market – the power futures market managed by OMIP – grow steadily since its start on July 3, 2006. They are correlated to the OTC volumes, the amount of active market makers in OMIP, the enrolment in OMIP of financial agents and generation companies of vertically integrated groups, and the OTC cleared volumes by OMIP clearing house (OMIClear).

The ex-post forward risk premium, obtained as the difference of the average forward price during the trading period and the underlying spot price during the delivery period, has diminished due to the agents’ learning curve and the effect of the fixed price retributing the indigenous coal fired generation. As a result of frequent positive forward risk premia, the electricity purchase by the Spanish last resort suppliers in CESUR auctions, with delivery between July 2009 and December 2011, has resulted 587 million € more expensive than the electricity cost valued at the spot price. Analysis of the economic impact of the contract for differences mechanism defined in Royal Decree 302/2011 is suggested for additional research. This mechanism helps to mitigate the tariff deficit caused by regulated prices for special regime generation.

Further research, based on literature review about regulatory risk, is suggested to quantify the increase in the Iberian forward risk premium due to regulatory risk

http://www.mibel.com/


58

provoked by intensive introduction of policy measures in Spain and Portugal with urgent nature. Due to time pressure, such measures may lack a robust cost-benefit analysis. The balance between the benefits and cost of imperfect regulation should overcome the costs of the related market failures. The outcomes of the suggested potential research could be very useful for energy regulators streamlining the existing regulation and supervising the performance of the wholesale energy markets.

Finally, despite of the dynamic trading environment, the Iberian Energy Derivatives Market is still much less developed than the European leaders (Nasdaq OMX Commodities and EEX). Its overall efficiency could increase with wider enrolment – there is still much room for international energy companies, financial agents, energy intensive industries and renewable generation companies –, post-trade transparency actions through market operator’s monthly oversight reports, and stronger supervision of regulators through integral data access from all trading venues. In this sense, the analysis of short and long positions per trading member crossed with the evolution of the open interest can be very useful for energy regulators performing market oversight once they access to the details of all the transactions of energy derivatives according to the EU Regulation on Energy Market Integrity and Transparency (“REMIT”).

The access to such REMIT data by the National Regulatory Authorities through a centralised system from the Agency for Cooperation of Energy Regulators (ACER) is envisaged for the first half of year 2015, i.e. six months after the entry into force of the REMIT implementing acts regarding trading and fundamental data collection from market participants. These implementing acts, published under the form of European Commission Regulation, will be presumably adopted during the second half of year 2014, after the adoption of key European financial legislative pieces (namely, a Regulation and a Directive related to markets in financial instruments (known as MiFIR and MiFID II), and a Regulation of market abuse (known as MAR)). The key center piece in the European financial legislation is MiFIR/MiFID II and once they are approved, the other affected pieces will be adopted (i.e. MAR and REMIT Implementing Acts, the latter due to the strong interrelationship between energy and financial (commodity derivatives) legislation) (European Parliament, 2014).


59

CHAPTER 4. EVALUATION OF THE FORWARD RISK PREMIUM

CHAPTER SUMMARY

Price efficiency is analysed for the Iberian Power Futures Market and other European

Power Markets, as well as other fuel markets through evaluation of ex-post Forward

Risk Premium, obtained as the difference between the forward price and the realised

spot value during the delivery period of the forward instrument. The equilibrium price

from compulsory call auctions for distribution companies within the framework of the

Iberian Power Futures Market (OMIP) is not optimal for remuneration purposes as it

seems to be slightly upward biased. In the period considered (July 2006 (start of OMIP)

to July 2008), monthly futures contracts behave similarly to quarterly contracts.

Average risk premia have been positive in power and natural gas markets but negative

in oil and coal markets. Different hypotheses are tested regarding increasing volatility

with maturity and regarding Forward Risk Premium variations (decreasing with

variance of spot prices during delivery period and increasing with skewness of spot

prices during delivery period). Enlarged data sets are recommended for stronger test

results. Energy markets tend to show limited levels of market efficiency. Regarding the

emerging Iberian Power Futures Market, price efficiency is improved with market

development of all the coexistent forward contracting mechanisms and with further

integration of European Regional Electricity Markets. A complementary analysis

through the ex-post Forward Risk Premium is also performed for the diverse forward

contracting mechanisms within the Iberian Electricity Market (the so-called “MIBEL”).

Those mechanisms are the OMIP call auctions, the OMIP continuous market, the

auctions catering for the last resort supplies (commonly known as CESUR auctions)

and the Spanish Virtual Power Plant auctions (VPP, also known as EPE). The data

embraces in this case forward and spot prices from July 2006 to February 2010.

Differences are found amongst the premia of the diverse forward contracting

mechanisms, providing arbitrage opportunities. Large positive premia are observed

since 2009 due to price uncertainty and low spot prices.


60

4.1 Introduction

An analysis of the Efficiency of the Iberian Power Futures Market is done to assess the situation of this emerging market. This information is of special interest both for all MIBEL market participantas and for MIBEL Regulatory Committee in charge of MIBEL market supervision. In order to perform this analysis, the article is separated in two sections corresponding to two different studies. Briefly, the first section (Section 4.2) assesses the economic results derived from different reference prices and compares the price evolution with other relevant international energy markets. The second section (Section 4.3) compares the price evolution with other interrelated forward contracting mechanisms in the Iberian electricity market. The analyses provided in this chapter are based on the research performed by Capitán Herráiz and Rodríguez Monroy (2008a; 2008b; 2009b; and 2010b).

In the first study (Section 4.2), the evolution of the ex-post forward risk premium of this market, obtained as the difference between the average settlement price of a futures contract (or any equivalent forward instrument) and the resulting average spot price during delivery, is quantitatively and qualitatively compared choosing diverse reference prices. Briefly, it is considered either the settlement prices of all the trading sessions of the Iberian power futures market or the equilibrium prices of the compulsory purchase auctions for the Spanish electricity distribution companies and the Portuguese last resort supplier. Additionally, the Iberian electricity price evolution is compared with the evolution of other relevant international energy markets energy markets (Brent, natural gas and coal). The data set covers the first two years of this market (i.e data from July 3, 2006, until July 31, 2008). The results of the diverse tests performed, described in Section 1.6.1.2, are thus provided in Section 4.2. Briefly, Section 4.2.1 “Test 1 Assessment of OMIP Auction Equilibrium Prices” assesses if the price formation in OMIP call auctions is satisfactory; Section 4.2.2 “Test 2 Analysis of Basic Statistics of Futures & Spot Prices” compares the price evolution of various energy markets; Section 4.2.3 “Test 3 Analysis of Ex-post Forward Risk Premium magnitudes” assesses the Forward Risk Premium Existence and compares the Futures behaviour towards maturity of these energy markets; and Section 4.2.4 “Test 4 Bessembinder’s & Lemmon’s hypothesis compliance” analyses the compliance of EU electricity futures markets (OMIP, Powernext and Nord Pool) regarding the hypothesis derived from seminal research based upon an equilibrium model by Bessembinder & Lemmon (2002), claiming that the Forward Risk Premium decreases in the variance of spot prices and increases in the skewness of wholesale prices.

In the second study (Section 4.3), the evolution of the price efficiency of the diverse forward contracting mechanisms within the MIBEL is analysed as well by means of the ex-post forward risk premium. A comparison of such premia from the MIBEL derivatives market (i.e. the futures market managed by OMIP), and the VPP and CESUR auctions is done. The Virtual Power Plant (VPP) auctions are also known in Spanish as “EPE” (Energy Primary Emissions) auctions. The data set covers approximately the first three and a half years of this market (in particular, data from July


61

3, 2006, until February 28, 2010).In order to have a comprehensive overview of all these market mechanisms, Section 4.3.1 “Some introductory facts” provides a snapshot of the trading development (in terms of traded/delivered volumes) of those mechanisms and Section 4.3.2 “Analysis of the Forward Risk Premium” describes the comparative analysis of their forward risk premia.

Finally, Section 4.4 “Results” summarises the findings from both studies and proposes future developments and recommendations related to this topic to mitigate potential inefficiencies.

4.2 Comparison of the ex-post forward risk premium with some relevant international energy markets This section shows the results of the tests described in Section 1.6.1.2.

4.2.1 Test 1 results

Test 1 assesses the evolution of the equilibrium prices in OMIP compulsory auctions for the Spanish distribution companies and the Portuguese last resort supply against the evolution of the prices in OMIP continuous market.

Figure 4.1 shows the evolution of Δex-post % according to the three reference prices stated in the Methodology (Section 1.6.1.2.1) built with daily spot and futures prices spanning from 3 July 2006 until 31 July 2008.

Briefly,the three reference prices considered are the resulting auction equilibrium price (“Feq”), the average futures price for all the quotation period (“Fall”), and the aerage spot price (“S”),

Two different kind of futures contracts are considered: monthly (the first contract considered has a time horizon or delivery period covering the month of August 2006, and the last one covers July 2008); and quarterly (the first contract covers the delivery period of the fourth quarter of year 2006 (Q4-06) and the last one covers the second quarter of year 2008 (Q2-08)).

Table 4.1 shows the economic results for each of the two observed periods (as shown in Figure 4.1) for the Forward Risk Premium. The first period relates to positive forward risk premia: in the case of monthly contracts, it covers the months between July 2006 and September 2007, and between April 2008 and July 2008. In the case of quarterly contracts, it covers the quarters between the fourth quarter of year 2006 (Q4-06) and the third quarter of year 2007 (Q3-07), as well as the second quarter of year 2008 (Q2-08). Conversely, the second period relates to negative forward risk premia: it covers a quite much shorter period: from October 2007 to March 2008, and in case of quarterly contracts, from the fourth quarter of year 2007 (Q4-07) to the first quarter of year 2008 (Q1-08).


62

Figure 4.1. OMIP Risk Premia in different quotation periods with different Reference

Prices.

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

FTB

M A

ug-0

6FT

B M

Sep

-06

FTB

M O

ct-0

6FT

B M

Nov

-06

FTB

M D

ec-0

6FT

B M

Jan

-07

FTB

M F

eb-0

7FT

B M

Mar

-07

FTB

M A

pr-0

7FT

B M

May

-07

FTB

M J

un-0

7FT

B M

Jul

-07

FTB

M A

ug-0

7FT

B M

Sep

-07

FTB

M O

ct-0

7FT

B M

Nov

-07

FTB

M D

ec-0

7FT

B M

Jan

-08

FTB

M F

eb-0

8FT

B M

Mar

-08

FTB

M A

pr-0

8FT

B M

May

-08

FTB

M J

un-0

8FT

B M

Jul

-08

Risk Pr. Future All Risk Prem. Auct. Eq. Price-30%

-20%

-10%

0%

10%

20%

30%

40%

FTB

Q4-

06

FTB

Q1-

07

FTB

Q2-

07

FTB

Q3-

07

FTB

Q4-

07

FTB

Q1-

08

FTB

Q2-

08

Risk Pr. Future All Risk Premium Auct. Eq. Price

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

FTB

M A

ug-0

6FT

B M

Sep

-06

FTB

M O

ct-0

6FT

B M

Nov

-06

FTB

M D

ec-0

6FT

B M

Jan

-07

FTB

M F

eb-0

7FT

B M

Mar

-07

FTB

M A

pr-0

7FT

B M

May

-07

FTB

M J

un-0

7FT

B M

Jul

-07

FTB

M A

ug-0

7FT

B M

Sep

-07

FTB

M O

ct-0

7FT

B M

Nov

-07

FTB

M D

ec-0

7FT

B M

Jan

-08

FTB

M F

eb-0

8FT

B M

Mar

-08

FTB

M A

pr-0

8FT

B M

May

-08

FTB

M J

un-0

8FT

B M

Jul

-08

Risk Pr. Future All Risk Prem. Auct. Eq. Price-30%

-20%

-10%

0%

10%

20%

30%

40%

FTB

Q4-

06

FTB

Q1-

07

FTB

Q2-

07

FTB

Q3-

07

FTB

Q4-

07

FTB

Q1-

08

FTB

Q2-

08

Risk Pr. Future All Risk Premium Auct. Eq. Price

Source: OMIP, OMIE

Table 4.1. Costs assessment of Energy purchased in OMIP Call Auctions by Spanish Distribution Companies. Distinction per Forward Risk Premium nature.

Period MWh € Fall €Feq € Spot

Positive Forward Risk Premium:

Jul.06 to Sep.07 & Apr.08 to Jul.08

(Q4-06 to Q3-07 & Q2-08)

14,097,571 734,626,089 748,536,130 629,181,275

Negative Forward Risk Premium:

Oct.07 to Mar.08 (Q4-07 to Q1-08)

7,677,216 387,684,679 391,360,662 436,397,291

Total 21,774,787 1,122,310,768 1,139,896,792 1,065,578,566

Costs Assessment of energy purchased in OMIP Auctions by Spanish Distribution Companies

Source: OMIP, OMIE

From Figure 4.1 and Table 4.1 it can be observed that the average futures price for all the quotation period (Fall) provides smaller economic values both for the positive and negative premia periods than the official recognised price from the compulsory purchase auctions for the Spanish electricity distribution companies and the Portuguese last resort supplier (Feq). Therefore, the total economic costs do differ depending on which future price is considered as reference. For the considered data set, the total economic cost of the difference between the reference future price and the spot price Feq-Spot (74,318,226 €) is 31% bigger than Fall-Spot (56,732,202 €). Additionally, from Figure 4.1 it can be seen that Forward Risk Premium with Fall is smaller than Forward Risk Premium with Feq along year 2008.


63

Another analysis is performed for the total economic costs by distinguishing per contract type (monthly versus quarterly), as shown in Table 4.2. From that table, it can be seen that the costs of monthly contracts are 2.8% higher considering the official recognised price (Feq) instead of the average quotation price (Fall). Nonetheless, for the quarterly contracts, the costs with Feq are less (-0.7%). Such a difference in the costs could show somehow that there may be more competitive pressure in OMIP call auctions for the quarterly contracts than for the monthly ones. This could be provoked by the interaction with the other coexisting MIBEL market mechanisms (EPE and CESUR auctions) where quarterly contracts are also traded. However, this claim is not so strong and should be cautiously considered as Feq has resulted bigger than Fall for the two considered quarterly contracts of year 2008 (Q1-08 and Q2-08). The results for the subsequent quarters would reinforce or reject this hypothesis about the competitive nature of OMIP quarterly call auctions and therefore, further research is suggested with a larger data set. However, the analysis provided in Section 4.3, considering a larger data set, sheds light on the magnitudes of the forward risk premium in the Iberian electricity market distinguishing between maturity (monthly, quarterly and annual futures contracts) and between diverse forward contracting market mechanisms.

Table 4.2. Costs assessment of Energy purchased in OMIP Call Auctions by Spanish Distribution Companies. Distinction per contract type.

Contract Type MWh € Fall € Feq € Spot

Monthly 13,896,691 722,027,227 742,311,612 680,792,606

Quarterly 7,878,096 400,283,541 397,585,180 384,785,960

Total 21,774,787 1,122,310,768 1,139,896,792 1,065,578,566

Cost Assessment of energy purchased in OMIP Auctions by Spanish Distribution Companies

Source: OMIP, OMIE

On the other hand, t-Student tests, separately done for monthly and quarterly futures, considering two tails and heteroscedasticity, show no evidence (test result provides a probability value of 62.6% for monthly futures and 84.3% for quarterly futures) for assuming same average values for Fall and Feq.

Figure 4.2 shows the evolution of Feq and Fall for the quarterly contracts. Additionally, it shows the evolution of weighted averages of the underlying monthly average spot prices (“OMIE”, also known as “OMEL”) for each quarterly delivery period. The conclusions obtained from Figure 4.1, Table 4.1 and Table 4.2 are also found in this Figure. Furthermore, comparison with the weighted averages of the monthly contracts is provided (Feq* and Fall*). It can be appreciated that Feq* and Fall* are smaller than Feq and Fall respectively until the end of year 2007, but during year 2008 the situation is reversed. The spread between Feq and Feq* is smaller than between Fall and Fall*. Additionally, the spread between Feq* and Fall* is smaller than between Feq and Fall.


64

Figure 4.2. Comparison of OMIP Settlement Prices: Quarterly Contracts versus Weighted Average Monthly Contracts (*) and underlying spot prices (OMEL).

35

40

45

50

55

60

65

FTB

Q4-

06

FTB

Q1-

07

FTB

Q2-

07

FTB

Q3-

07

FTB

Q4-

07

FTB

Q1-

08

FTB

Q2-

08

€/M

Wh

Fall Fall* Feq Feq* OMEL Source: OMIP, OMIE

From the results obtained, the following conclusions can be drawn:

• In the short term, until enough liquidity is reached in OMIP continuous market, it would seem reasonable to continue arranging compulsory call auctions for the Spanish Distribution Companies. Nonetheless, the equilibrium price (Feq), used for the settlement of the purchased contracts, is resulting slightly higher compared to the average of settlement prices along the trading period of the contract (Fall). According to OMIP trading limits for mitigating the members’ credit risk – as theoretically defined in Section B2.12 Daily Price Variation Limits of OMIP Operational Guide and practically specified in OMIP Notice 04/2006 regarding Maximum Price Variation Limits – accepted bids and offers must be contained within an interval centered on the Trading Session Reference Price, i.e. the resulting Settlement Price of the previous session. In the case of yearly and quarterly contracts, the interval spans from the reference price ±6% (±9% for monthly contracts; ±15% for weekly contracts) (OMIP-OMIClear, 2008a; OMIP, 2011b). Spanish Distribution Companies and Portuguese Last Resort Supplier submit their compulsory bids in OMIP call auctions at the price given by the upper limit of the interval, in order to ensure that their bids are matched. Therefore, competition only arises from the sales side. If they submitted their bids at a maximum price which is somehow


65

smaller, the resulting equilibrium price might diminish, as desired in terms of economic costs to the regulated supplies. Therefore, it could make sense that as soon as the settlement price of the previous session is determined, OMIP and MIBEL Regulators’ Committee may agree upon a cap price for the compulsory call auction of the following day. Such a cap price would be carefully calculated per auction, taking into account the historical evolution of the forward prices and the fundamental data (i.e situation of the demand and the available generation in the electricity spot market as well as the gas demand context and gas prices) in order to get the desired effect on the auction equilibrium price without preventing competition on the sales curve (in other words, maximizing the social welfare through lower end-user prices not stifling the trading and retailing competition). The cap price would only apply for the compulsory call auction, not for the continuous market. For the sake of transparency, the auction cap price should be published in OMIP bulletin together with the results of the trading session previous to the compulsory call auction. This daily bulletin shows, per negotiated futures contract, all the traded volumes in OMIP continuous market, compulsory call auctions, and OTC settled by OMIP clearing house (OMIClear). Furthermore, last traded price, open price, daily high and low prices, closing bid-ask spread, aggregated traded volumes distinguishing between financial and physical settled contracts (excluding OTC settled by OMIClear), and open interest are also shown (OMIP-OMIClear, 2012).

• It might be reasonable to continue offering compulsory quantities via OMIP call auctions to distribution companies (substituted by last resort suppliers since July 2009) until desired liquidity levels are reached in the continuous market. At that stage, the Settlement Price published by OMIP should accurately reflect market prices and could be better utilised for the calculation of last resort supply costs. For instance, the average value of the reference price of the prompt quarter contract in its last 3 months of quotation could be considered. In case the MIBEL Regulators’ Committee and the Market Operator would detect any suspicious behavior during the trading sessions of such a three month period, the reference price could be timely corrected through an approved methodology by those entities, with the aim of mitigating any excessive speculation (i.e. the forward price formation would not respond to fundamental factors related to the demand and offer equilibrium). Such a new methodology would facilitate that the last resort suppliers would be able to cover their forward energy needs through OMIP continuous market. Therefore, further compulsory OMIP call auctions would no longer be necessary. Additionally, the competitive nature of the continuous market would theoretically provide lower prices than compulsory auctions, making the supply costs more affordable to last resort suppliers.


66


Test 2 analyses the main statistics of the Spanish electricity futures and spot prices (as published by OMIP and OMIE respectively) compared with other relevant international energy markets, namely the French electricity prices (published by Powernext), the Nordic electricity prices (published by Nord Pool, note that the Nordic electricity derivatives prices are currently published by Nasdaq OMX Commodities, formerly Nord Pool AS), the British gas prices at the British virtual point or National Balancing Point (NBP) published by Platts and ICE, the Brent prices published by Platts and ICE, and the coal prices delivered in Northwest Europe (ARA) published by Argus Mc Closkey and EEX. In particular, Tables 4.3 and 4.4 show basic statistics for the monthly and quarterly futures prices respectively:

Table 4.3. Basic Statistics of Fall & Underlying Spot Prices of Monthly Futures Contracts during period Aug.06-Jul.08.

Futures Spot Futures Spot Futures Spot Futures Spot Futures Spot Futures SpotAverage 51.74 46.82 54.61 47.97 43.45 35.94 47.76 37.99 64.81 82.20 87.02 103.87Median 53.02 45.06 54.77 42.55 44.81 34.91 43.75 33.86 63.25 74.99 71.86 82.34Max 67.16 70.22 82.89 88.33 63.17 66.48 80.08 62.18 77.64 133.18 154.32 209.73Min 38.35 29.68 27.87 27.02 23.77 16.53 26.48 16.24 51.02 53.91 62.63 65.70Std.Dev. 8.29 12.36 15.95 18.61 12.42 13.73 15.69 15.35 6.16 23.76 26.66 41.59Asymmetry -0.21 0.56 0.03 0.50 -0.01 0.63 0.71 0.27 0.12 0.92 1.15 1.01Kurtosis -1.01 -0.96 -0.93 -1.08 -1.15 -0.23 -0.11 -1.45 0.31 -0.08 0.21 0.19

Basic Statistics of Average Reference Prices of Monthly Contracts & Underlying Spot Prices. Period: Aug.06-Jul.08OMIP (€/MWh) Powernext (€/MWh) Nord Pool (€/MWh) NBP (GB p/therm) Brent (US $/Bbl) EEX ARA (US $/t)

Source: OMIP, OMIE, Powernext, Nord Pool, EEX, ICE, Platts, Argus McCloskey

Table 4.4. Basic Statistics of Fall & Underlying Spot Prices of Quarterly Futures Contracts during period Q4.06-Q2.08.

Futures Spot Futures Spot Futures Spot Futures Spot Futures SpotAverage 51.86 45.58 54.69 47.87 40.26 32.73 49.41 37.64 73.01 102.32Median 50.31 39.00 51.09 41.72 38.80 34.61 41.46 30.53 69.51 86.02Max 58.57 65.86 70.06 72.71 48.72 44.60 76,11 60.58 86.85 158.63Min 46.57 35.70 41.08 29.35 33.34 19.74 30.36 20.20 66.14 67.39Std.Dev. 4.36 11.83 11.56 19.05 5.48 9.91 16.86 15.79 7.65 37.01Asymmetry 0.55 1.02 0.25 0.30 0.51 -0.14 0.86 0.40 1.21 0.59Kurtosis -1.05 -0.42 -1.97 -2.38 -0.78 -1.87 -0.72 -1.70 0.40 -1.63

Basic Stats. Avge. Ref. Prices Quarterly Contracts & Underlying Spot Prices. Period: Q4 06 - Q2-08OMIP (€/MWh) Powernext (€/MWh) Nord Pool (€/MWh) NBP (GB p/therm) EEX ARA (US $/t)

Source: OMIP, OMIE, Powernext, Nord Pool, EEX, ICE, Platts, Argus McCloskey

From the information reflected in Table 4.3 and Table 4.4., the following conclusions can be drawn:

• Within each market and comparing Futures with Spot values, the same behaviour is detected for monthly and quarterly contracts, except for Asymmetry and Kurtosis. Nevertheless, as the data set is quite limited (especially for quarterly contracts), such differences are not relevant.

• The Average Risk Premia are positive in Power and Gas Markets, but negative in Oil and Coal Markets. To be more precise, in Power and Gas


67

Markets the average Risk Premia of positive values tend to be bigger in absolute value than the average Risk Premia of negative values. In Oil and Coal Markets, the average Risk Premia of negative values tend to be bigger in absolute value than the average Risk Premia of positive values.

• According to standard deviation values, spot markets show bigger volatility than their related futures markets, except for gas market. Due to that bigger volatility, more extreme values are presented in spot markets, with the exceptions of maximum values in gas forward market and minimum values in oil and coal futures markets. The biggest spreads (Futures versus Spot) regarding maximum values are produced for the oil and coal markets.

• In general, Asymmetry tends to be positive and Kurtosis tends to be negative.

• Although not reflected in the tables, similar results (except for Asymmetry of quarterly contracts) are obtained from OMIP Feq as those shown for OMIP Fall.


Test 3 analyses the magnitudes of the ex-post forward risk premium of the considered energy markets. The premium is expressed in percentage values due to the different energy and monetary units employed in those markets. Diverse tests are run and their results are described in the following subsections.

4.2.3.1 Test 3.1 results

Test 3.1 assesses the forward risk premium existence. The t-Student test detects, for each market, if the positive and negative forward risk premium have the same average value. The null hypothesis is a premium with zero value (i.e. if the risk premium tends to 0, there would not be evidence of the existence of such a premium).

The t-Student tests performed for each market (except for EEX ARA Coal where all the forward risk premia are negative for every month in both monthly and quarterly contracts), and distinguishing between monthly and quarterly contracts, render extremely low probability values rejecting the null hypothesis of no existence (or “zero value”) for the forward risk premium. Therefore the results of the test justify the existence of positive and negative forward risk premia. Table 4.5 shows the insignificant probability values.


68

Table 4.5. t-Student test regarding null hypothesis of no existence (“zero value”) for the Forward Risk Premium.

OMIP Powernext Nord Pool NBP Brent

pMonthly 0.00068026 0.00082985 0.00000032 0.00000002 0.00000014

pQuarterly 0.00216445 0.00068821 0.00078747 0.00248197 n.a.

Source: OMIP, OMIE, Powernext, Nord Pool, Platts

4.2.3.2 Test 3.2 results

Test 3.2 compares the behaviour of the futures contracts towards maturity (i.e how the price evolves from the first months of quotation towards the last trading sessions of a given contract previous to its final settlement). The final settlement could be pure financial (obtained as the daily calculation of the difference between the derivatives price and the underlying spot price) or physical (involving the physical delivery of the commodity at the derivatives price). The settlement of the Iberian electricity derivatives is mainly financial and the settlement of the OTC electricity and gas forward trading in North West Europe is mainly physical.

Figure 4.3 shows the evolution of the Forward Risk Premia (in percentage) for both monthly and quarterly OMIP futures contracts, considering the following four series of futures prices: Fall, FM-3, FM-2, and FM-1. The idea behind those different series is to compare the behaviour through different distances to the final settlement of the derivatives instruments, namely: all quotation period (“Fall”), third last month of quotation (“FM-3”), second last month of quotation (“FM-2”), and last month of quotation (“FM-1”). Figures 4.4-4.8 respectively show the equivalent information for the rest of considered energy markets in this research, namely: Powernext, Nord Pool, NBP Gas, Brent, and EEX ARA Coal.


69

Figure 4.3. OMIP Forward Risk Premia distinguishing Reference Prices per approach to

maturity.

-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-M

Jun

-07

M J

ul-0

7M

Aug

-07

M S

ep-0

7M

Oct

-07

M N

ov-0

7M

Dec

-07

M J

an-0

8M

Feb

-08

M M

ar-0

8M

Apr

-08

M M

ay-

M J

un-0

8M

Jul

-08

F_All F_M-3 F_M-2 F_M-1

-25%

-15%

-5%

5%

15%

25%

35%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08

F_All F_M-3 F_M-2 F_M-1-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-M

Jun

-07

M J

ul-0

7M

Aug

-07

M S

ep-0

7M

Oct

-07

M N

ov-0

7M

Dec

-07

M J

an-0

8M

Feb

-08

M M

ar-0

8M

Apr

-08

M M

ay-

M J

un-0

8M

Jul

-08


-25%

-15%

-5%

5%

15%

25%

35%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08


Source: OMIP, OMIE

Figure 4.4. Powernext Forward Risk Premia distinguishing Reference Prices per approach to maturity.

-85%

-65%

-45%

-25%

-5%

15%

35%

55%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-07

M J

un-0

7M

Jul

-07

M A

ug-0

7M

Sep

-07

M O

ct-0

7M

Nov

-07

M D

ec-0

7M

Jan

-08

M F

eb-0

8M

Mar

-08

M A

pr-0

8M

May

-08

M J

un-0

8M

Jul

-08

F_All F_M-3 F_M-2 F_M-1-70%

-50%

-30%

-10%

10%

30%

50%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08


-85%

-65%

-45%

-25%

-5%

15%

35%

55%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-07

M J

un-0

7M

Jul

-07

M A

ug-0

7M

Sep

-07

M O

ct-0

7M

Nov

-07

M D

ec-0

7M

Jan

-08

M F

eb-0

8M

Mar

-08

M A

pr-0

8M

May

-08

M J

un-0

8M

Jul

-08

F_All F_M-3 F_M-2 F_M-1-70%

-50%

-30%

-10%

10%

30%

50%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08

F_All F_M-3 F_M-2 F_M-1 Source: Powernext


70

Figure 4.5. Nord Pool Forward Risk Premia distinguishing Reference Prices per approach

to maturity.

-50%

-30%

-10%

10%

30%

50%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-07

M J

un-0

7M

Jul

-07

M A

ug-0

7M

Sep

-07

M O

ct-0

7M

Nov

-07

M D

ec-0

7M

Jan

-08

M F

eb-0

8M

Mar

-08

M A

pr-0

8M

May

-08

M J

un-0

8M

Jul

-08


-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

60%

70%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08

F_All F_M-3 F_M-2 F_M-1-50%

-30%

-10%

10%

30%

50%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-07

M J

un-0

7M

Jul

-07

M A

ug-0

7M

Sep

-07

M O

ct-0

7M

Nov

-07

M D

ec-0

7M

Jan

-08

M F

eb-0

8M

Mar

-08

M A

pr-0

8M

May

-08

M J

un-0

8M

Jul

-08


-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

60%

70%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08


Source: Nord Pool

Figure 4.6. NBP Gas Forward Risk Premia distinguishing Reference Prices per approach to maturity.

-85%

-65%

-45%

-25%

-5%

15%

35%

55%

75%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-M

Jun

-07

M J

ul-0

7M

Aug

-07

M S

ep-0

7M

Oct

-07

M N

ov-0

7M

Dec

-07

M J

an-0

8M

Feb

-08

M M

ar-0

8M

Apr

-08

M M

ay-

M J

un-0

8M

Jul

-08

F_All F_M-3 F_M-2 F_M-1-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08


-85%

-65%

-45%

-25%

-5%

15%

35%

55%

75%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-M

Jun

-07

M J

ul-0

7M

Aug

-07

M S

ep-0

7M

Oct

-07

M N

ov-0

7M

Dec

-07

M J

an-0

8M

Feb

-08

M M

ar-0

8M

Apr

-08

M M

ay-

M J

un-0

8M

Jul

-08

F_All F_M-3 F_M-2 F_M-1-80%

-60%

-40%

-20%

0%

20%

40%

60%

80%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08


Source: Platts


71

Figure 4.7. Brent Forward Risk Premia distinguishing Reference Prices per approach to

maturity.

-80%

-60%

-40%

-20%

0%

20%

M A

ug-0

6

M S

ep-0

6

M O

ct-0

6

M N

ov-0

6

M D

ec-0

6

M J

an-0

7

M F

eb-0

7

M M

ar-0

7

M A

pr-0

7

M M

ay-0

7

M J

un-0

7

M J

ul-0

7

M A

ug-0

7

M S

ep-0

7

M O

ct-0

7

M N

ov-0

7

M D

ec-0

7

M J

an-0

8

M F

eb-0

8

M M

ar-0

8

M A

pr-0

8

M M

ay-0

8

M J

un-0

8

M J

ul-0

8

F_All F_M-3 F_M-2 F_M-1 Source: Platts

Figure 4.8. EEX ARA Coal Forward Risk Premia distinguishing Reference Prices per approach to maturity.

-50%

-40%

-30%

-20%

-10%

0%

10%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-07

M J

un-0

7M

Jul

-07

M A

ug-0

7M

Sep

-07

M O

ct-0

7M

Nov

-07

M D

ec-0

7M

Jan

-08

M F

eb-0

8M

Mar

-08

M A

pr-0

8M

May

-08

M J

un-0

8M

Jul

-08

F_All F_M-3 F_M-2 F_M-1-85%

-75%

-65%

-55%

-45%

-35%

-25%

-15%

-5%

5%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08


-50%

-40%

-30%

-20%

-10%

0%

10%

M A

ug-0

6M

Sep

-06

M O

ct-0

6M

Nov

-06

M D

ec-0

6M

Jan

-07

M F

eb-0

7M

Mar

-07

M A

pr-0

7M

May

-07

M J

un-0

7M

Jul

-07

M A

ug-0

7M

Sep

-07

M O

ct-0

7M

Nov

-07

M D

ec-0

7M

Jan

-08

M F

eb-0

8M

Mar

-08

M A

pr-0

8M

May

-08

M J

un-0

8M

Jul

-08

F_All F_M-3 F_M-2 F_M-1-85%

-75%

-65%

-55%

-45%

-35%

-25%

-15%

-5%

5%

Q4-

06

Q1-

07

Q2-

07

Q3-

07

Q4-

07

Q1-

08

Q2-

08


Source: EEX, Argus McCloskey


72

By analysing all these charts, various trends are detected. The following conclusions can be drawn from Figures 4.3-4.8:

• Monthly and Quarterly contracts have similar Forward Risk Premium variation trends coinciding with alternant periods of positive Δex-post % or negative Δex-post %. In the case of power markets, a general trend change (“positive to negative”) is produced in autumn 2007.

• Quantitative variations of Δex-post % are similar for monthly and quarterly contracts. Whereas monthly average values tend to be slightly smaller than quarterly ones, extreme monthly values tend to be somewhat bigger than quarterly ones, explained by slightly bigger volatility of the monthly values (measured in terms of standard deviation). Regarding extreme variations, the smallest ones occur for OMIP (around ±40%), and the biggest ones for NBP (around ±80%).

• Whereas Δex-post % positive is dominant in power and gas markets, Δex-post

% negative is dominant in oil and coal markets, supposing different hedging strategies within each market. In absolute value, positive Δex-post % tends to be slightly bigger than negative Δex-post %.

• Correlation between Futures Series (Fall with each of the 3 series FM-3, FM-2, or FM-1) – analysing separately monthly and quarterly futures contracts – is only significant in EEX ARA Coal (correlation coefficient around 0.99). In the case of Power Markets, for the monthly contracts the correlation coefficients are around 0.90 and for the quarterly contracts, the correlation coefficients are around 0.60. Smaller coefficients for quarterly contracts can be caused by the limited data set (7 values) compared to wider monthly data set (24 values). Comparing Power Markets, Powernext presents the biggest correlation coefficients and Nord Pool the smallest ones. For NBP Gas, correlation coefficients are around 0.70 for monthly and quarterly contracts. For Brent, correlation coefficients are around 0.85 (monthly contracts). For all the markets considered, the smallest correlation is produced between Fall and FM-1 (i.e. correlation tends to diminish as futures contracts approach maturity).

• Samuelson’s maturity effect (increasing volatility when maturity approaches) is only noticeable in OMIP, Powernext, EEX ARA Coal, and Brent.

• Increasing convergence to spot price with maturity is fulfilled by all time series (trend towards smallest Δex-post % in absolute value, when comparing, in this sequence, FM-3, FM-2, and FM-1). Monthly and quarterly contracts are analysed separately, where comparison of Δex-post % in absolute value is separately done for positive Δex-post % and negative Δex-

post %. The convergence with maturity can be caused due to lack of accuracy in oldest quoted futures prices (in other words, the closer to the delivery, the more precise its price forecast).


73


Test 4 analyses the compliance of the Bessembinder´s & Lemmon’s (2002) hypothesis for the diverse European electricity markets considered. They asserted that the Forward Risk Premium decreases in the variance of spot prices and increases in the skewness of wholesale prices. Linear regression is applied, in which α is a constant, β is the coefficient related to the variance of the spot prices, γ is the coefficient related to the non-standardised Asymmetry Coefficient (“skewness”) of spot prices obtained as the Asymmetry Coefficient multiplied by cubed Standard Deviation of Spot Prices, and εT is an error term.

Table 4.6 summarises the results of applying multifactor linear regression:

Table 4.6. Regression results regarding compliance with Bessembinder-Lemmon's Hypothesis.

Bessembinder Lemmon hypothesis compliance for ex-post Forward Risk Premium in Power Markets

OMIP M Contracts Quot.Period α β γ R2 t tα tβ tγ

All 4.75 0.0177 -0.0034 0.07% 2.08 1.34 0.10 -0.12 M-3 5.38 -0.0056 0.0005 0.00% 2.09 1.23 -0.02 0.01 M-2 5.84 -0.0386 0.0009 0.48% 2.09 1.66 -0.21 0.03 M-1 3.63 0.0358 -0.0060 0.47% 2.08 1.49 0.28 -0.31

OMIP Q Contracts Quot.Period α β γ R2 t tα tβ tγ

All 5.83 -0.0373 0.0105 1.95% 2.78 0.48 -0.08 0.17 M-3 7.28 -0.0943 0.0116 1.00% 2.78 0.62 -0.20 0.19 M-2 6.50 0.0316 -0.0098 2.58% 2.78 0.62 0.08 -0.18 M-1 4.51 0.1223 -0.0268 14.48% 2.78 0.51 0.34 -0.58

Powernext M Contracts Quot.Period α β γ R2 t tα tβ tγ

All 18.32 -0.0637 0.0003 60.13% 2.08 5.28 -3.67 2.54 M-3 22.50 -0.0890 0.0004 64.79% 2.08 5.83 -4.61 3.47 M-2 19.53 -0.0669 0.0003 59.39% 2.08 5.13 -3.51 2.37 M-1 12.60 -0.0337 0.0001 44.99% 2.08 3.81 -2.04 1.11

Powernext Q Contracts Quot.Period Α β γ R2 t tα tβ tγ

All 29.66 -0.1547 0.0010 75.36% 2.78 3.39 -2.62 2.18 M-3 27.34 -0.1211 0.0007 69.98% 2.78 2.74 -1.80 1.35 M-2 27.74 -0.1131 0.0006 69.48% 2.78 2.66 -1.61 1.15 M-1 20.48 -0.0796 0.0004 68.58% 2.78 2.24 -1.29 0.82

Nord Pool M Contracts Quot.Period Α β γ R2 t tα tβ tγ

All 13.24 -0.2662 0.0092 15.02% 2.08 3.35 -1.93 0.28 M-3 12,99 -0.2712 0.0153 10.41% 2.08 2.61 -1.56 0.37 M-2 10.90 -0.2520 0.0022 15.68% 2.08 2.96 -1.96 0.07 M-1 6.04 -0.1224 -0.0091 12.65% 2.08 2.69 -1.56 -0.49

Nord Pool Q Contracts Quot.Period α β γ R2 t tα tβ tγ

All 16.82 -0.2115 -0.0078 43.96% 2.78 2.67 -1.41 -0.28 M-3 13.61 0.0424 0.0283 9.51% 2.78 1.14 0.15 0.54 M-2 4.78 0.2855 0.0642 45.35% 2.78 0.59 1.47 1.81 M-1 -0.18 0.3524 0.0783 82.26% 2.78 -0.04 3.52 4.26

Source: OMIP, OMIE, Powernext, and Nord Pool


74

The following conclusions can be drawn from the results in Table 4.6:

• In general, there is relatively poor compliance for the three power markets. No significant differences are obtained between the four Futures series considered within each market. In the case of OMIP, similar results are also obtained by using Feq instead of Fall. Although the quarterly contracts are composed of a limited data set per serie (7 values versus 24 of the monthly contracts), R-squared statistic is larger for the quarterly contracts.

• OMIP is the least compliant market, as for both monthly and quarterly contracts, coefficient signs for β and γ are not right as expected from the testable hypothesis, R-squared statistic results too low, and t-Student tests (significant values for the coefficients) are not satisfactory.

• Powernext is the best compliant market, as coefficient signs are right, R-squared statistic renders reasonable level, and t-Student tests are partly satisfactory.

• Nord Pool has a medium compliance, as the coefficient signs tend to be right, and t-Student tests are partly satisfactory. Reasonable values for R-squared statistic are only found for the quarterly contracts.

As a caveat to the poor compliance of Bessembinder’s and Lemmon’s hypothesis

for the considered electricity markets, it seems that the results are quite sensitive to the structure of the data set, e.g. consideration of monthly average values against daily values, different contract maturities (prompt month forward prices against day-ahead prices (as employed in this research) or short term (e.g. day-ahead prices or daily futures contracts) against intraday prices (as used by other researchers)), et cetera. In this sense, there are more studies finding poor compliance as well. For instance, Pietz (2009) rejects the relation between the risk premia and the spot price variance and skewness. He uses three spot markets segments (intraday, block and day-ahead markets) from the German electricity market (EEX) with a data set ranging from August 2002 to May 2009. Regarding the block contract market, specific blocks of consecutive hours are traded whereas single hour contracts are traded in the day-ahead and intraday market. He finds that the contracts vary in magnitude and in sign throughout the day, and that higher risk premia exist in the summer months. He finds positive risk premia in the block contract (significant on Mondays) and in the day-ahead market (extremely volatile). Finally, he detects that risk premia seem to be higher in contracts with a longer time-to-delivery.


75

4.3 Comparison of the ex-post forward risk premia in the Iberian power forward contracting mechanisms

4.3.1 Some introductory facts

The ex-post forward risk premia is analysed for diverse forward contracting mechanisms within the Iberian Electricity Market, namely: the call auctions organised by OMIP and the OMIP continuous market, the auctions catering for the last resort supplies (CESUR auctions) and the Spanish Virtual Power Plant auctions (VPP, also known as EPE). The data embraces forward and spot prices from July 3, 2006 (start of OMIP futures market) until the end of February 2010.

Table 4.7 summarises the dates and the products traded in all these auctions related to the data set considered in the analysis.

Table 4.7. Spanish Regulated Forward Contracting Mechanisms within the MIBEL Framework complementing the OMIP call auctions.

Celebration Products Celebration Products1st June 13, 2007 1st June 19, 20072nd Sept. 13, 2007 2nd Sept. 18, 2007

3rd Dec. 11, 2007 3rd Dec. 18, 2007

4th March 11, 2008 4th March 13, 20085th June 10, 2008 5th June 17, 20086th Sept. 23, 2008 6th Sept. 25, 2008

7th Dec. 16, 20087th March 24, 2009 8th March 26, 2009

9th June 25, 200910th Dec. 15, 2009

Baseload and peak:quarter "Q+1"

Base & peak quarters: "Q+1"; "Q+2"

Base & peak: 6 month "(Q+1)+(Q+2)"; year "(Q+1)+(...)+(Q+4)"

Baseload & peak: quarter "Q+1"; 6 month

"(Q+1)+(Q+2)"; year "(Q+1)+(...)+(Q+4)"

Baseload: quarter "Q+1"

Spanish VPP ("EPE") Auction CESUR Auction

Baseload:quarter "Q+1";

6 month "(Q+1)+(Q+2)"

Source: Capitán Herráiz and Rodríguez Monroy (2010b)

In order to understand the magnitude of each forward contracting mechanism, Fig. 4.9 shows the delivered energy per month resulting from OMIP call auctions, OMIP continuous market, EPE auctions, and CESUR auctions. For the sake of simplicity, only baseload products are considered in EPE and CESUR auctions, and the energy delivered in EPE auctions would correspond to the theoretical case of full exercise of the call options. Note that such delivered energy do not correspond to actual physical delivery due to the predominance of cash settlement. On average, during each month, around 3.9 TWh are delivered in CESUR auctions, 1.1 TWh in OMIP call auctions, 0.9 TWh in EPE auctions, and 0.4 TWh in OMIP continuous auctions.


76

Figure 4.9. Delivered Energy (MWh) per MIBEL Forward Contracting Mechanism.


Figure 4.10 and Figure 4.11 help to understand how the trading activity is produced in OMIP. Figure 4.10 shows the volumes traded in call auctions and in continuous for month contracts. The horizontal axis corresponds to the delivery month of each contract. Figure 4.11 does the same for quarter and year contracts. The horizontal axis corresponds to the delivery quarter or year of each contract, except for the quarter and year contracts with delivery in 2010, as only their delivery during January and February 2010 is considered. It can be observed that volumes tend to increase in continuous market, for all maturities, even beginning to surpass the auction volumes during years 2009 and 2010. Auctions volumes increased until the second half of year 2008 and decrease afterwards. Note that since July 2009 only the Portuguese Last Resort Supplier is obliged to purchase in OMIP Call Auctions. Regarding the aggregated delivery volumes during each year, in the auction trading mode, month contracts deliver more energy than quarter and even more than year ones, except for year 2009, with dominance of quarter contracts. In the continuous trading mode, quarter contracts deliver more than year contracts and even more than month contracts. In 2009, remarkable quarter volumes are delivered in continuous, maybe due to its immediate equivalence with contracts from CESUR and EPE auctions, facilitating arbitrages amongst them. No year volumes for 2007 were traded in continuous, as liquidity in the first months of OMIP (second half of year 2006) was very low.


77

Figure 4.10. Evolution of OMIP Auction and Continuous Volumes of Month Contracts.

Evolution of OMIP Auction & Continuos Volumes per Contract Maturity: Month Contracts

0

200.000

400.000

600.000

800.000

1.000.000

1.200.000FT

B M

Aug

-06

FTB

M S

ep-0

6FT

B M

Oct

-06

FTB

M N

ov-0

6FT

B M

Dec

-06

FTB

M J

an-0

7FT

B M

Feb

-07

FTB

M M

ar-0

7FT

B M

Apr

-07

FTB

M M

ay-

FTB

M J

un-0

7FT

B M

Jul

-07

FTB

M A

ug-0

7FT

B M

Sep

-07

FTB

M O

ct-0

7FT

B M

Nov

-07

FTB

M D

ec-0

7FT

B M

Jan

-08

FTB

M F

eb-0

8FT

B M

Mar

-08

FTB

M A

pr-0

8FT

B M

May

-FT

B M

Jun

-08

FTB

M J

ul-0

8FT

B M

Aug

-08

FTB

M S

ep-0

8FT

B M

Oct

-08

FTB

M N

ov-0

8FT

B M

Dec

-08

FTB

M J

an-0

9FT

B M

Feb

-09

FTB

M M

ar-0

9FT

B M

Apr

-09

FTB

M M

ay-

FTB

M J

un-0

9FT

B M

Jul

-09

FTB

M A

ug-0

9FT

B M

Sep

-09

FTB

M O

ct-0

9FT

B M

Nov

-09

FTB

M D

ec-0

9FT

B M

Jan

-10

FTB

M F

eb-1

0

Trad

ed V

olum

es (M

Wh)

Auction Continuous Source: OMIP-OMIClear (2012)

Figure 4.11. Evolution of OMIP Auction and Continuous Volumes of Quarter and Year

Contracts.

Evolution of OMIP Volumes per Contract Maturity: Quarter and Year Contracts

0

1.000.000

2.000.000

3.000.000

4.000.000

5.000.000

6.000.000

7.000.000

FTB

Q4-

06

FTB

Q1-

07

FTB

Q2-

07

FTB

Q3-

07

FTB

Q4-

07

FTB

Q1-

08

FTB

Q2-

08

FTB

Q3-

08

FTB

Q4-

08

FTB

Q1-

09

FTB

Q2-

09

FTB

Q3-

09

FTB

Q4-

09

FTB

Q1-

10*

FTB

YR-0

7

FTB

YR-0

8

FTB

YR-0

9

FTB

YR-1

0*

Trad

ed V

olum

es (M

Wh)

Auction Continuous Source: OMIP-OMIClear (2012)


78

4.3.2 Analysis of the forward risk premium

First of all, evolution of the premium is analysed for OMIP Month Contracts (the so-called “FTB M”). Different futures prices are considered: (i) “PRNall” refers to the arithmetical mean of OMIP Settlement Prices during the whole quotation life of the contract; (ii) “PRNa” refers to the weighted average of the equilibrium prices of OMIP call auctions, considering the matched volumes in each auction; (iii) “PRNc” refers to the weighted average of OMIP Settlement Prices, considering the traded volumes in each continuos market session; (iv) “PRNa&c” refers to the weighted average of OMIP Settlement Prices, considering the sum of the auction and continuos traded volumes in each OMIP market session.

Figure 4.12 shows the results of this analysis, that can be summarised as follows: PRNall is a rough estimation, and therefore biggest extreme values and standard deviation are observed (note that, according to OMIP (2009a), when no trades occur, the settlement price can be obtained by means of the closing bid-ask spread, or through theoretical arbitrage with other quoted contracts, or even through ad-hoc Price Committee, and all these estimations may reflect speculative quotations); PRNa&c and PRNc provide the best centered values, as it reflects the trading sessions with effective activity (existence of trades); PRNall and PRNa provide the most positive values, somehow upward biased, due to buying pressure in OMIP call auctions, and maybe some speculative activity for arbitrage gains; PRNall and PRNc provide the most negative values, maybe due to selling pressure in the continuous market and speculative quotations aforementioned.

Figure 4.13 shows the results of the same analysis, but for the quarter and year contracts (“FTB Q” and “FTB YR” respectively). As previously found, PRNa&c and PRNc provide more centered values and PRNall the most extreme ones. Although PRNc does not provide now minimum premia, the average values for the three contract maturities obtained with PRNc result always the lowest. In the case of year contracts, maximum value is obtained for PRNa, however their data sample is very limited to draw strong conclusions. Note that premia for FTB Q1-10* and FTB YR-10* are obtained with average spot values for the period January 2010 – February 2010, and therefore their values are not so accurate. It is interesting to see that Year contracts provide bigger premia averages and standard deviation that Quarter contracts and even than Month contracts. This means that larger uncertainty and inaccuracy is assumed when hedging through larger maturities contracts.


79

Figure 4.12. Forward Risk Premia of OMIP Month Futures Contracts.

Source: OMIP-OMIClear (2012)

Figure 4.13 Forward Risk Premia of OMIP Quarter and Year Futures Contracts.



80

Table 4.8 shows the average, minimum, maximum, and standard deviations for all the premia and spot prices previously analysed.

Table 4.8. Analysis of OMIP Forward Risk Premia: basic statistics.

Month Contracts PRNall-Sp PRNa&c-Sp PRNa-Sp PRNc-Sp SpotAvge 5.35 5.11 5.42 4.34 45.60Min -15.41 -13.84 -13.61 -14.92 27.68Max 21.15 19.65 20.15 17.86 73.03Std Dev 8.36 7.07 7.61 6.49 13.63Quarter Contracts PRNall-Sp PRNa&c-Sp PRNa-Sp PRNc-Sp SpotAvge 8.63 6.11 6.49 5.87 45.02Min -14.33 -9.97 -10.28 -8.02 28.41Max 23.75 20.88 21.06 20.56 70.41Std Dev 13.71 9.75 9.83 9.20 13.74Year Contracts PRNall-Sp PRNa&c-Sp PRNa-Sp PRNc-Sp SpotAvge 12.63 10.41 11.01 9.10 42.43Min -13.07 -12.49 -12.90 -11.24 28.41Max 25.25 25.78 27.26 24.39 64.43Std Dev 17.90 16.23 17.00 18.34 15.43

Analysis of OMIP Forward Risk Premia (€/MWh)


In order to have a global average value of OMIP Forward Risk Premia, weighted

average of PRNa&c is obtained considering the delivered energy in each month by the corresponding month, quarter and year contract. Such an average (“W_Avge”) is compared with PRNa&c for the month contracts (“FTB M”) in Figure 4.14. As average and standard deviation values are bigger for quarter and year contracts, as shown in Table 4.8, the resulting W_avge shows more oscillating values than FTB M PRNa&c. This is well appreciated in Figure 4.14 in the right axis, where the Forward Risk Premium is divided by the Spot Price, providing a percentage (the forward risk premia are denoted in this case in Figure 4.14 as “%SpotW_Avge” and “%SpotFTB M” respectively). The differences between both series reveal clear arbitrage opportunities amongst the different maturities of OMIP Futures contracts. This limits the desired price efficiency of this market.

On the other hand, the global premia with W_avge do not tend to diminish due to offset effects amongst the different maturities: the only offset occurs during the month contracts Apr-08 to Jun-08 and quarter Q2-08 (positive premia) versus year YR-08 (negative). Regarding the percentage values, positive premia are larger than negative ones, mainly due to the buying pressure of the compulsory purchases in OMIP call auctions. Whereas the negative premia tend to diminish – signalising some price efficiency gain – the positive premia during year 2009 and 2010 show the same high levels as those ones presented at the beginning of the futures market. Further research is encouraged to diagnose this latter undesired price inefficiency.

Correlation analysis shows that PRNa&c for month contracts keeps strong correlation with the spot prices (0.86). Thus the month futures contracts reflect better the spot price dynamics than the quarter and year futures contracts. Quarter and Year contracts do not present such correlation with the spot prices. Although W_Avge keeps


81

less correlation with the spot prices (0.74), it has tight correlation with the month PRNa&c (0.92). Table 4.9 summarises the average forward risk premia for each delivery year, distinguishing between the global premia (obtained with W_Avge) and the premia obtained with PRNa&c for the month, quarter and year contracts. The unique annual average offset between maturities for the PRNa&c series appears during year 2008.

Figure 4.14. Forward Risk Premia of OMIP Futures: Global weighted average versus

weighted average of month contracts.

Forward Risk Premia of OMIP Futures: Weighted Average versus Month Contracts

-20

-15

-10

-5

0

5

10

15

20

25

30

Aug-

06Se

p-06

Oct

-06

Nov

-D

ec-0

6Ja

n-07

Feb-

07M

ar-0

7Ap

r-07

May

-Ju

n-07

Jul-0

7Au

g-07

Sep-

07O

ct-0

7N

ov-

Dec

-07

Jan-

08Fe

b-08

Mar

-08

Apr-

08M

ay-

Jun-

08Ju

l-08

Aug-

08Se

p-08

Oct

-08

Nov

-D

ec-0

8Ja

n-09

Feb-

09M

ar-0

9Ap

r-09

May

-Ju

n-09

Jul-0

9Au

g-09

Sep-

09O

ct-0

9N

ov-

Dec

-09

Jan-

10Fe

b-10

Futu

res

Pric

e-Sp

ot P

rice

€/M

Wh

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

60%

70%

Futu

res-

Spot

/Spo

t %W_Avge Fwd Risk Pr FTB M Fwd Risk Pr %Spot W_Avge %Spot FTB M


Table 4.9. Average OMIP Forward Risk Premia per delivery year for global and maturity weighted average series.

Year W_Avge Ma&c Qa&c Ya&c2006 14.29 13.59 20.55 -2007 5.57 4.95 6.57 13.442008 -3.15 1.37 -2.56 -12.492009 13.21 4.99 9.32 25.78

Average Forward Risk Premia per delivery year


Figure 4.15 shows the quarterly evolution of the Forward Risk Premia for the

different MIBEL forward contracting mechanisms. Note that, for the sake of simplicity and direct comparison with the rest of prices, the VPP forward risk premium has been obtained considering the resulting VPP prices as futures prices instead of option prices (i.e. it has been assumed an exercise of those call options in all the hours, facilitating the calculation of its ex-post forward risk premium). The resulting VPP price is obtained as the sum of the strike price of the option (published prior to the VPP auction) and the


82

resulting premium price in the corresponding VPP auction (i.e. the premium price is the equilibrium price of the auction). Further research is suggested employing the precise exercise of those options, assuming that the option is only exercised when the spot price is bigger than the strike price of the option.

In the case of OMIP, three average values are considered: (i) rough PRNall; (ii) PRNa&c of quarter contracts; (iii) global quarter average, obtained from the monthly values of W_Avge weighted by the number of hours of each month. In the quarter Q3-06*, the average is composed weighing the monthly W_Avge values of Aug-06 and Sep-06, as no futures contract for Jul-06 quoted in OMIP. For Q1-10*, the average is composed weighing the monthly values of Jan-10 and Feb-10, as the data set ends at February 2010.

It can be noticed the existence of arbitrage opportunities due to price differences amongst all these mechanisms. The extreme values and the biggest standard deviation are obtained through PRNall (i.e. the rough estimation of OMIP values: note the opposite behaviour with other average prices in Q2-08 and Q4-08) and VPP. Whereas VPP provides the lowest average –revealing good purchasing opportunities–, OMIP provides the biggest average values –good selling opportunities–. Regarding percentage values, similar results are derived, except for lowest average (CESUR). Table 4.10 provides such basic statistics.

Figure 4.15. Quarterly Forward Risk Premia of MIBEL Forward Contracting Mechanisms.



83

Table 4.10. Basic statistics of the quarterly Forward Risk Premia for the MIBEL forward contracting mechanisms.

Wavge-S PRNa&c-S PRNall-S VPP-S CESUR-S %SWav ge %SPRNa&c %SPRNall %SVPP %SCESURAvge 6.81 6.11 8.63 4.54 5.05 19.89% 18.02% 27.59% 16.23% 14.44%Min -7.85 -9.97 -14.33 -14.39 -9.48 -14.07% -15.14% -20.36% -20.79% -19.78%Max 18.49 20.88 23.75 24.65 18.89 45.35% 52.69% 61.96% 57.24% 43.86%Std Dev 8.74 9.75 13.71 13.21 8.48 21.51% 23.33% 32.24% 30.02% 21.15%

Quarterly Forward Risk Premia of MIBEL Forward Contracting Mechanisms (F-S) & (F-S)/S


4.4 Results

Market efficiency is analysed for the Iberian Power Futures Markets and other European power markets (Powernext and Nord Pool) and fuel markets (Brent, NBP Natural Gas, and EEX ARA Coal) through evaluation of ex-post Forward Risk Premium, with data set spanning between July 2006 and July 2008. The equilibrium price in OMIP compulsory call auctions for distribution companies is not optimal for remuneration purposes as the purchasing costs for regulated supplies tend to be slightly higher than those from OMIP average settlement prices along the whole quotation period. A regulated cap price for each OMIP compulsory call auction could be transitorily applied in order to obtain a lower equilibrium price and therefore diminish regulated costs of supply. Once OMIP continuous market has acceptable liquidity, the settlement price itself would reflect more accurately the market prices and could be used for evaluating the cost of last resort supplies.

In the period considered (August 2006 to July 2008), monthly futures contracts have a similar behaviour as quarterly contracts and average risk premia have been positive in power markets (especially until Q4-07) and in gas markets but negative in oil and coal markets. In all the examined markets, the Forward Risk Premium for a futures contract tends to diminish as it approaches maturity. Samuelson’s maturity effect (increasing volatility with maturity) is only noticeable in OMIP, Powernext, EEX ARA Coal, and Brent. Compliance with Bessembinder’s & Lemmon’s testable hypothesis regarding Forward Risk Premium variations in Power Markets (decreasing with variance of spot prices, and increasing with skewness of spot prices) is relatively low.

Further research is proposed considering an enlarged data set (especially with quarterly contracts) to better test all the hypotheses and draw additional conclusions. Inclusion of longer contracts (synthetic joint of two quarters, and calendar contracts) may also be of interest. In general, it can be concluded that none of the markets analysed presents a noticeable level of market efficiency as remarkable Forward Risk Premia exist in all the markets. Regarding electricity markets, the behaviour of OMIP futures prices do not differ much in terms of efficiency compared to more mature markets (Powernext and Nord Pool). Although liquidity is still poor in the first years of existence of the Iberian Derivatives Market managed by OMIP, its price efficiency has improved along with OMIP market development and should further increase with upcoming integration of European Power Regional Markets as well as with the development of the coexisting forward contracting mechanisms in the Iberian Energy


84

Market, namely, OTC, EPE auctions, and CESUR auctions. Additionally, due to the relevance of natural gas as power generation fuel within the Iberian energy mix, especially when the wind production is scarce, further development of the Iberian Natural Gas Market (the so-called MIBGAS) will presumably contribute to indirectly improve the desired OMIP price efficiency. Further research is encouraged to analyse the efficiency gains of the Iberian Derivatives Market caused by the dynamic evolution of the Iberian energy markets.

On the other hand, the ex-post forward risk premium of the Iberian power futures market is compared with the resulting forward risk premium of other interrelated forward contractiong mechanisms within the Iberian electricity market, namely the EPE (or VPP) and CESUR auctions. The data set spans between July 2006 and February 2010. The analysis of OMIP forward risk premia shows differences per contract maturity, allowing arbitrage gains through combined trading of the month, quarter and year futures contracts. Global analysis of the MIBEL forward contracting mechanisms, i.e. OMIP call auctions and continuous market, EPE auctions and CESUR auctions reveals also arbitrage opportunities amongst all these mechanisms. OMIP presented large positive premia at its beginning, showing limited price efficiency, due to the learning phase of the traders in such a nascent market, the initial low liquidity of the continuous market and some upward price pressure due to the dominance of compulsory call auctions. The introduction of EPE and CESUR auctions coexisting with OMIP provoked price efficiency gains, as less positive and negative forward risk premia were alternated. Nevertheless, since year 2009 onwards, large positive premia dominate in all these market mechanisms, limiting the desired price efficiency. The reasons of such large positive premia may arise from rising forward prices and decreasing spot prices in the Iberian energy market. The former were pushed up by the uncertainty in global financial markets and tight credit conditions –introducing a risk premium component– as well as some positive expectations of future global recovery, causing price increases in energy commodity prices. The latter were decreased by the hard economic situation, with less industrial production and national economic activity – both decreasing the energy demand –, and low prices in the power pool, due to such reduced demand and strong penetration of renewables – especially wind power, displacing coal plants and gas combined cycles from the power generation merit order – and large hydro reservoirs due to heavy rainfalls, as indicated in Llewellyn (2010).

As the forward risk premium is bigger for contracts with larger maturity (year contracts followed by quarter contracts), the market participants can benefit in many cases by means of trading short positions (i.e. keeping a selling net position) in such power contracts. Conversely, market participants with a natural long position (e.g. suppliers purchasing derivatives to hedge for their retail sales) might be more interested in trading month contracts, especially when the perception of uncertainty and thus energy risk is bigger. Since year 2011, the premium is smaller due to the existence of a regulatorily recognized price for coal power plants as described in Chapter 3 (Section 3.2.1). However, smaller premia should be the natural result of a well developed and efficient power forward market, rather than the effect of artificial mechanisms distorting the price formation.


85

Further research with larger data samples, including Portuguese VPP auctions and the Contract for Differences (CfD) auctions (replaced by Financial Transmission Rights (FTR) auctions, in which such FTRs are issued in a coordinated manner by the Spanish and Portuguese Transmission System Operators since March 2014) in the Portuguese-Spanish interconnection is encouraged. Those CfD/FTR auctions are issued to facilitate the hedging in the cross-border trades due to the risk derived from the price spread (i.e. price differential) between the spot price of the Spanish and Portuguese market areas. Those spot prices are only equal in the absence of congestion in the interconnection.

Regarding OMIP forward risk premium, month contracts reflect better the spot price dynamics than the quarter and year ones. Therefore, month contracts can be considered as the most efficient maturity. Introduction of month contracts in EPE and CESUR auctions could then improve the price efficiency of all these mechanisms.


86


87

CHAPTER 5. EVALUATION OF THE LIQUIDITY DEVELOPMENT

CHAPTER SUMMARY

A comprehensive assessment of the liquidity development in the Iberian power futures

market managed by OMIP (“Operador do Mercado Ibérico de Energia, Pólo

Português”) in its first four years of existence is performed. This market started on July

2006. A regression model tracking the evolution of the traded volumes in the

continuous market is built as a function of twelve potential liquidity drivers. The only

significant drivers are the traded volumes in OMIP compulsory auctions, the traded

volumes in the “Over The Counter” (OTC) market, and the OTC cleared volumes in

OMIP clearing house (OMIClear). Furthermore, the enrolment of financial members

shows strong correlation with the traded volumes in the continuous market. OMIP

liquidity is still far from the levels reached by the most mature European markets (Nord

Pool and EEX). The market operator and its clearing house could develop efficient

marketing actions to attract new entrants active in the spot market (energy intensive

industries, suppliers, and small producers) as well as volumes from the opaque OTC

market, and to improve the performance of existing illiquid products. An active dialogue

with all the stakeholders (market participants, spot market operator, and supervisory

authorities) will help to implement such actions.

5.1 Introduction

This chapter analyzes the efficiency of the Iberian power futures market focused on another cornerstone: liquidity. The analysis described in this chapter is based on the research performed by Capitán Herráiz and Rodríguez Monroy (2011a; 2013a). Note that the previous chapter analyzed the efficiency of the Iberian power futures market through the price formation (in particular, through the ex-post forward risk premium).

The employed data set in the liquidity analysis is robust, as it covers the first four years of existence of this market (from July 3, 2006, to June 30, 2010). Such an ample data set facilitates the detection of the most significant traded volume drivers and on the other hand, the identification of the products that are showing poor performance (i.e. illiquidity). These findings allow the formulation of policy recommendations for streamlining the efficiency of this market. A regression model using Ordinary Least Square methodology is estimated to assess the effect of twelve selected drivers (the independent variables) for the following key liquidity measure (the dependent variable): the evolution of the energy traded in the continuous market. The research is also


88

reinforced by means of a correlation analysis of the independent variables with the dependent variable.

The chapter is structured as follows: (i) Section 5.2 provides the grounds to analyse the liquidity development in the Iberian power futures market; (ii) Section 5.3 shows the results of the analysis of the traded volume drivers in OMIP continuous market by means of a regression model and correlation analysis; (iv) Section 5.4 formulates recommendations for the proper development of this market; (v) Section 5.5 shows the main results and suggests further research.

5.2 The basics to assess the liquidity development of the Iberian Power Futures Market

Chapter 1 provides a good overview to understand the performance and development of any energy derivatives market. In particular, Section 1.4 describes the fundamentals of forward energy trading. Regarding the specific features of the Iberian electricity market, Section 1.5 describes the basics of the Iberian power futures market and the interrelated market mechanisms (i.e. the OTC market, the VPP auctions and the CESUR auctions). As the derivatives market interacts with the spot market (i.e the market segment in which the underlying asset of the derivative is traded), Section 1.5.1 provides a concise description of the main features of the spot market (i.e. the day-ahead market and the intraday market) and the adjustment markets managed by the System Operator (i.e. the market for the resolution of technical constraints and the ancillary services and deviation management) within the scope of the Iberian electricity market. On the other hand, Chapter 3 provides in Section 3.3 an evaluation of the trading efficiency in the Iberian energy derivatives market. Therefore, a comparison of the traded volumes in the diverse Iberian electricity forward market mechanisms is provided. Section 5.2 complements the aforementioned sections as follows: Section 5.2.1 indicates the types of derivatives listed in the Iberian power futures market for trading or clearing and settlement in its clearing house (OMIClear); Section 5.2.2 describes the main features of the market maker agreements in the Iberian power futures market, helping to develop the liquidity; and Section 5.2.3 shows the traded and cleared volumes in OMIP and OMIClear respectively as well as the national electricity demand, to compare those key figures with the main European power exchanges.

5.2.1 The derivatives listed in OMIP

Table 5.1 lists the contract types (futures, forwards and swaps) traded in OMIP and/or cleared by OMIClear in the first four years of this market. The liquidity level is indicated, according to their traded volumes. Their nominal is 1 MW per delivery hour. They have financial settlement and some of them physical delivery through the spot market managed by OMEL. The spot reference price used in the settlement is the daily arithmetical mean of the OMEL day-ahead price in the Spanish (“SPEL”) or Portuguese


89

(“PTEL”) zone. The maturities of the contracts in the first four years of the Iberian power futures market are Week (“W”), Month (“M”), Quarter (“Q”) and Year (“Y”). Whereas baseload refers to each hour of the delivery period, peak only embraces twelve hours between 8a.m-8p.m. from Monday to Friday. The call auctions with peak futures during January 2010–July 2010 were only compulsory for the Portuguese last resort supplier (OMIP-OMIClear, 2012).

As shown in Table 5.1, almost all the traded volumes in the Iberian power futures market correspond to base load futures contracts (known as “FTB”) with underlying asset the spot price of the Spanish zone (known as “SPEL” index, i.e., Spanish electricity price). Thre traded volumes in forwards and swaps are almost null. In the case of the peak futures contracts, some liquidity was developed exclusively due to compulsory call auctions as stated above. The trading with base load futures contracts (“FPB”) with underlying asset the spot price of the Portuguese zone (the so-called “PTEL” index, Portuguese electricity price), is growing remarkably since December 2011 due to the introduction of Special Regime auctions, as described in Section 6.3.2.2.

Table 5.1. Derivatives listed in OMIP: basic features and liquidity diagnosis.

Derivative OMIP Load Underlying Settlement Market modes Liquidity Level Available sinceFutures FTB Base SPEL Financial Auction Good July 3, 2006

Can have physical delivery ContinuousOTC registered

Forwards FWB Base SPEL Financial with physical delivery OTC registered Null March 2, 2009Swaps SWB Financial Scarce Futures FPB Base PTEL Financial Auction Scarce July 1, 2009

ContinuousOTC registered

Futures FTK Peak SPEL Financial Auction Mainly due to January 20, 2010Can have physical delivery Continuous compulsory call

OTC registered auctions Source: OMIP-OMIClear (2012) adapted by authors

As previously stated, in the first four years of the Iberian power futures market only week, month, quarter and year contracts were quoted. Afterwards, shorter maturities were introduced, namely: day and weekend. Those maturities were introduced on 20 May 2011 (OMIP, 2012a). The liquidity of those shorter maturities is still low in OMIP continuos market. Nonetheless, the trading of such short term contracts in the OTC market is growing fast. Therefore, close monitoring and supervision should be done by the MIBEL regulatory agencies to assess the right price formation of the spot market (e.g. to detect any influence in the spot price due to eventual excessive speculation in short term derivatives trading). The monthly supervision reports of the Spanish Energy Regulator provide a good snapshot of the liquidity levels for each maturity. For instance, in November 2013, the traded volumes in OMIP continuous market were split per maturity as follows: only around 2 % correspond to short-term contract with delivery period shorter than one month (i.e. day, weekend, and week); the rest of maturities (month, quarter, and year) reckon each one around 1/3 of the total volume. Interestingly, the split in the OTC market is quite different: around 17% corresponds to delivery shorter than one month; around 21% for the month contracts as well as the quarter contracts; and the remaining 40% to year


90

contracts (CNMC, 2014b). The information in those reports would be even richer if the split per maturity is not only shown in terms of volume but also in terms of number of operations, to easily track the growth of the trading intensity in those emerging short-term segments very close to the spot market time window.

5.2.2 OMIP market makers as liquidity boosters

OMIP can establish a market maker agreement with a trading member, according to OMIP Instruction 01/2007, Market Makers, in force since September 1, 2007. Such an agreement specifies the particular conditions for the derivative – so far only market maker agreements with future scontracts have been established – the market maker is obliged to quote during the validity period. The quoted offers are to be kept at least during an agreed period within the continuous trading phase in each session. Those offers cannot spread more than a given amount of ticks (each tick equals 0.01 €/MWh), and should cover at least a minimum agreed amount of contracts (one contract equals 1 MW baseload per each delivery hour). The agreement may indicate under which conditions the market maker is allowed interrupting its activity in a predetermined number of trading sessions. The so-called fast market occurs when OMIP dictates that there is a significant increase in volatility of market prices. In that situation, the obligations to quote prices may be different to the normal situation and such particularities should be set in the agreement. The market maker becomes exempt from fulfilling its obligations when there is some technical failure on its trading system or on OMIP information system. If the market maker has privileged information (insider trading), it cannot quote until such information becomes public. OMIP may suspend the activity of the market maker in case of existence of strange trading circumstances. In order to supervise the market making activity, OMIP evaluates the performance of the market maker each month. As incentives for the market making activity, the market maker obtains discounts in trading and clearing fees regarding traded volumes in the continuous market, as well as monthly compensations from OMIP for each contract under its scope. These incentives should be stated in the agreement (OMIP, 2008b).

Table 5.2 summarizes the agreements signed during the first four years of the market. “M+1” stands for the prompt month, “Q+2” for the second prompt quarter, etc. As previously mentioned, the number of market makers active in OMIP has been kept stable in the succeeding years, as well as the maturities and type of contracts subject to such agreements. Therefore, Table 5.2 provides a quite illustrative snapshot of these liquidity fostering mechanisms. In practice, those agreements usually have a duration of approximately one year (or even shorter in some cases, around half a year) and can be further extended in time.


91

Table 5.2. Market maker agreements within the Iberian Power Futures Market.

Market Maker FTB Contracts Start Date End DateRBS Sempra M+1, M+2, M+3 September 1, 2007 March 31, 2008

M+1, M+2 May 1, 2008 October 31, 2008EGL Energía Iberia Q+1, Q+2, Y+1 November 10, 2008 November 11, 2009

M+1, M+2, Q+1, Q+2, Y+1 November 12, 2009 May 31, 2010M+1, M+2 June 1, 2010 December 31, 2010

Deutsche Bank M+1, M+2, Y+1 May 14, 2009 November 13, 2009M+1, M+2, Q+1, Q+2, Y+1 November 16, 2009 May 31, 2010

Q+1, Q+2, Y+1 June 1, 2010 December 31, 2010Citigroup Global Markets M+1, M+2 November 2, 2009 May 31, 2010

Source: OMIP-OMIClear (2012) adapted by authors

5.2.2.1 The effects of the market maker agreements in the bid ask spread reduction

Capitán Herráiz and Rodríguez Monroy (2009a) study the effects of the first market maker agreements signed within the Iberian power futures market on the effective reduction of the bid-ask spread. Therefore, such agreements positively contribute to the trading development (i.e. liquidity improvement) as well as the price efficiency. They considered daily data spanning from July 3rd 2006 until November 20th 2008. Therefore, the market maker agreements analysed are those from RBS Sempra and EGL Energia Iberia. Spreads are calculated for the futures contracts that have been until that date exposed to market maker agreements, namely: “M+1”, “M+2”, “M+3”, “Q+1”, “Q+2”, and “Y+1”. Note that, as usual in financial markets, the most liquid contracts are those ones close to maturity, as it is the case of the considered contracts. In OMIP, negotiation of weekly contracts or other ones far from maturity was very rare yet. The spread is composed of the Best Offer (“BO”) and the Best Bid (“BB”), i.e. the cheapest sale offer and the most expensive buy offer respectively, as provided by OMIP Daily Market Bulletins. Equation 5.1 shows the ratio employed (“average spread”) in order to better compare all the spreads:

Average Spread (%) = (BO-BB)/((BO+BB)/2) (5.1)

Figure 5.1 and Figure 5.2 show the evolution of the spreads in the first year of existence of the Iberian Power Futures Market managed by OMIP, for the monthly contracts (Figure 5.1) and the quarterly and yearly ones (Figure 5.2). Figure 5.3 and Figure 5.4 do respectively the same from July 3rd 2007 until November 20th 2008. In these figures vertical dotted lines embrace the periods in which some market maker agreement is in force. Table 5.3 provides the average spreads (in percentage values) and their variability measured in terms of standard deviation distinguishing three periods: the whole period embraced by the data set; the period prior to the first market maker agreement; and a separate analysis of the three held agreements (two for RBS Sempra and one for EGL Energía Iberia) for the whole data set considered.


92

Figure 5.1. Evolution of OMIP Closing Spreads: Monthly Futures contracts quoting from July 3rd 2006 to July 2nd 2007.

Evolution of OMIP Closing Spreads: Monthly Futures contracts quoting from July 3rd 2006 to July 2nd 2007

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

03/0

7/20

0610

/07/

2006

17/0

7/20

0624

/07/

2006

31/0

7/20

0607

/08/

2006

14/0

8/20

0621

/08/

2006

28/0

8/20

0604

/09/

2006

11/0

9/20

0618

/09/

2006

25/0

9/20

0602

/10/

2006

09/1

0/20

0616

/10/

2006

23/1

0/20

0630

/10/

2006

06/1

1/20

0613

/11/

2006

20/1

1/20

0627

/11/

2006

04/1

2/20

0611

/12/

2006

18/1

2/20

0625

/12/

2006

01/0

1/20

0708

/01/

2007

15/0

1/20

0722

/01/

2007

29/0

1/20

0705

/02/

2007

12/0

2/20

0719

/02/

2007

26/0

2/20

0705

/03/

2007

12/0

3/20

0719

/03/

2007

26/0

3/20

0702

/04/

2007

09/0

4/20

0716

/04/

2007

23/0

4/20

0730

/04/

2007

07/0

5/20

0714

/05/

2007

21/0

5/20

0728

/05/

2007

04/0

6/20

0711

/06/

2007

18/0

6/20

0725

/06/

2007

02/0

7/20

07

(Bes

t Offe

r-B

est B

id)/A

vera

ge(B

est O

ffer&

Bes

t Bid

)

M+1 M+2 M+3 Source: OMIP-OMIClear (2012) adapted by authors

Figure 5.2. Evolution of OMIP Closing Spreads: Quarterly & Yearly Futures contracts quoting from July 3rd 2006 to July 2nd 2007.

Evolution of OMIP Closing Spreads: Quarterly & Yearly Futures Contracts quoting from July 3rd 2006 to July 2nd 2007

0%

2%

4%

6%

8%

10%

12%

03/0

7/20

0610

/07/

2006

17/0

7/20

0624

/07/

2006

31/0

7/20

0607

/08/

2006

14/0

8/20

0621

/08/

2006

28/0

8/20

0604

/09/

2006

11/0

9/20

0618

/09/

2006

25/0

9/20

0602

/10/

2006

09/1

0/20

0616

/10/

2006

23/1

0/20

0630

/10/

2006

06/1

1/20

0613

/11/

2006

20/1

1/20

0627

/11/

2006

04/1

2/20

0611

/12/

2006

18/1

2/20

0625

/12/

2006

01/0

1/20

0708

/01/

2007

15/0

1/20

0722

/01/

2007

29/0

1/20

0705

/02/

2007

12/0

2/20

0719

/02/

2007

26/0

2/20

0705

/03/

2007

12/0

3/20

0719

/03/

2007

26/0

3/20

0702

/04/

2007

09/0

4/20

0716

/04/

2007

23/0

4/20

0730

/04/

2007

07/0

5/20

0714

/05/

2007

21/0

5/20

0728

/05/

2007

04/0

6/20

0711

/06/

2007

18/0

6/20

0725

/06/

2007

02/0

7/20

07

(Bes

t Offe

r-B

est B

id)/A

vera

ge(B

est O

ffer&

Bes

t Bid

)

Q+1 Q+2 Y+1 Source: OMIP-OMIClear (2012) adapted by authors


93

Figure 5.3. Evolution of OMIP Closing Spreads: Monthly Futures contracts quoting from July 3rd 2007 to November 20th 2008.

Evolution of OMIP Closing Spreads: Monthly Futures contracts quoting from July 3rd 2007 to November 20th 2008

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

03/0

7/20

0712

/07/

2007

21/0

7/20

0730

/07/

2007

08/0

8/20

0717

/08/

2007

26/0

8/20

0704

/09/

2007

13/0

9/20

0722

/09/

2007

01/1

0/20

0710

/10/

2007

19/1

0/20

0728

/10/

2007

06/1

1/20

0715

/11/

2007

24/1

1/20

0703

/12/

2007

12/1

2/20

0721

/12/

2007

30/1

2/20

0708

/01/

2008

17/0

1/20

0826

/01/

2008

04/0

2/20

0813

/02/

2008

22/0

2/20

0802

/03/

2008

11/0

3/20

0820

/03/

2008

29/0

3/20

0807

/04/

2008

16/0

4/20

0825

/04/

2008

04/0

5/20

0813

/05/

2008

22/0

5/20

0831

/05/

2008

09/0

6/20

0818

/06/

2008

27/0

6/20

0806

/07/

2008

15/0

7/20

0824

/07/

2008

02/0

8/20

0811

/08/

2008

20/0

8/20

0829

/08/

2008

07/0

9/20

0816

/09/

2008

25/0

9/20

0804

/10/

2008

13/1

0/20

0822

/10/

2008

31/1

0/20

0809

/11/

2008

18/1

1/20

08

(Bes

t Offe

r-B

est B

id)/A

vera

ge(B

est O

ffer&

Bes

t Bid

)

M+1 M+2 M+3

1st Market Maker Agreement of RBS Sempra 2nd Market Maker Agreement of RBS Sempra


Figure 5.4. Evolution of OMIP Closing Spreads: Quarterly & Yearly Futures contracts quoting from July 3rd 2007 to November 20th 2008.

Evolution of OMIP Closing Spreads: Quarterly & Yearly Futures Contracts quoting from July 3rd 2007 to November 20th

2008

0%

1%

2%

3%

4%

5%

6%

7%

03/0

7/20

0712

/07/

2007

21/0

7/20

0730

/07/

2007

08/0

8/20

0717

/08/

2007

26/0

8/20

0704

/09/

2007

13/0

9/20

0722

/09/

2007

01/1

0/20

0710

/10/

2007

19/1

0/20

0728

/10/

2007

06/1

1/20

0715

/11/

2007

24/1

1/20

0703

/12/

2007

12/1

2/20

0721

/12/

2007

30/1

2/20

0708

/01/

2008

17/0

1/20

0826

/01/

2008

04/0

2/20

0813

/02/

2008

22/0

2/20

0802

/03/

2008

11/0

3/20

0820

/03/

2008

29/0

3/20

0807

/04/

2008

16/0

4/20

0825

/04/

2008

04/0

5/20

0813

/05/

2008

22/0

5/20

0831

/05/

2008

09/0

6/20

0818

/06/

2008

27/0

6/20

0806

/07/

2008

15/0

7/20

0824

/07/

2008

02/0

8/20

0811

/08/

2008

20/0

8/20

0829

/08/

2008

07/0

9/20

0816

/09/

2008

25/0

9/20

0804

/10/

2008

13/1

0/20

0822

/10/

2008

31/1

0/20

0809

/11/

2008

18/1

1/20

08

(Bes

t Offe

r-B

est B

id)/A

vera

ge(B

est O

ffer&

Bes

t Bid

)

Q+1 Q+2 Y+1

1st Market Maker Agreement of RBS Sempra 2nd Market Maker Agreement of RBS Sempra



94

Table 5.3. Evolution of OMIP Closing Spreads: Futures contracts quoting from July 3rd 2006 to November 20th 2008.

Period Identifier M+1 M+2 M+3 Q+1 Q+2 Y+1 M+1 M+2 M+3 Q+1 Q+2 Y+1Jul.3 2006-

Nov.20 2008Whole period 3.03 2.90 3.80 2.79 2.39 1.83 0.024 0.020 0.019 0.021 0.016 0.014

Jul.3 2006-Aug.31 2007

Prior to first Agreement

4.57 4.27 5.32 4.04 3.15 2.86 0.034 0.031 0.031 0.026 0.020 0.018

Sep.1 2007-Mar.31 2008

1st AgreementRBS Sempra

2.85 2.84 3.67 1.97 2.00 1.28 0.012 0.012 0.013 0.014 0.015 0.008

May.1 2008-Oct.31 2008

2nd AgreementRBS Sempra

1.98 2.11 2.22 1.88 1.78 1.28 0.009 0.009 0.008 0.012 0.012 0.007

Nov.10 2008-Apr.10 2009

1st Agreement EGL Energía Iberia

1.26 2.41 - 2.59 3.37 2.15 0.004 - - 0.006 0.004 0.007

Average Spread (%): (BO-BB)/((BO+BB)/2)

Standard Deviation of Average Spread


The following observed trends are drawn by analyising Figures 5.1 – 5.4 and Table 5.3:

• In general, the introduction of market maker agreements has improved the level of liquidity, as all the spreads have diminished in magnitude (ratio) and in variability (standard deviation).

• The periods between market maker agreements do not worsen the achievements of those agreements, therefore the effect of the agreements are robust and the learning capabilities of the trading members are good (their ability to trade is not much dependent on continued existence of these agreements).

• Though the two first market maker agreements did not apply for quarterly and yearly contracts, it seems that spreads of these contracts tend to behave slightly better than the monthly ones (the quarterly and yearly contracts have lower spread ratios and variability than the monthly ones). Thus it seems that the liquidity of the quarterly and yearly contracts is more developed than the liquidity of the monthly contracts in OMIP continuous market. Note that the smaller and less variable quarterly and yearly spreads are partly caused by OMIP market rules themselves and not totally by trading efficiencies: as previously cited (see Section 4.2.1), OMIP Notice 04/2006, Maximum Price Variation Limits, states that trades exceeding a given percentage from the Settlement Price of the previous trading session are rejected. The allowed trading spread is ±6% of previous settlement price for quarterly and yearly contracts, ±9% for monthly contracts, and ±15% for weekly contracts. Thus such limited intervals are affecting somehow the resulting trading spreads for each contract type.

• Further research is encouraged to measure the spread evolution for the different maturities, considering a larger data set and the market maker agreements signed during that larger data period. In case the liquidity differences grow between the monthly contracts and the quarterly and


95

yearly ones, it might be reasonable that OMIP strives to foster further market maker agreements of monthly and even weekly contracts, as both contracts might need some support for their proper development.

5.2.3 Comparison with the most mature European power futures markets

As previously indicated, regarding the amount of members and traded volumes, although holding a steady growing pace, OMIP liquidity is still quite reduced compared to other more mature European power futures markets, as Nord Pool (Nordic Countries), EEX (Germany) and Powernext (France). Table 5.4 compares those markets for year 2008 – prior to the merge of EEX and Powernext in September 2009 (OFGEM, 2009) – with data obtained from the market operators (OMIP-OMIClear, 2012; EEX, 2009b; Nord Pool, 2009; Powernext, 2009), the British Regulator (OFGEM, 2009), and the Spanish System Operator (REE, 2009). OMIP is close to the figures provided by its neighbour Powernext, but far from the more developed EEX and the outstanding Nord Pool. The churn is defined as the ratio of traded volume of a commodity to throughput or generated output, or some other measure denoting physical consumption (OFGEM, 2009). It is calculated here as the ratio between the total cleared volumes in the derivatives exchange during year 2008 and the demand for that year. In the case of OMIP, the demand is the Spanish one at busbar, and for Nord Pool, it refers to the aggregation from all the Nordic countries. It can be appreciated that the churn ratio in the Iberian power futures market is still quite smaller compared to the most developed European energy derivatives exchanges (Nord Pool and EEX), and in general, OMIP basic comparative statistics are closer to the neighbouring energy derivative exchange (the French Powernext).

Table 5.4. Comparison of the main European Power Derivatives Exchanges with data of year 2008.

Market Number of Members

Power FuturesVolumes (TWh)

OTC clearedin exchange (TWh)

Total Cleared in Exchange (TWh) Demand 2008 (TWh) Churn:

Total Cleared/Demand

EEX 118 266 899 1165 540 2.16Nord Pool 391 1437 1140 2577 426 6.05OMIP 30 22 9 32 264 0.12Powernext 43 87 4 91 503 0.18

Source: OMIP-OMIClear (2012), OFGEM (2009), EEX (2009), Nord Pool (2009), Powernext (2009) and REE (2009), adapted by authors


96

5.3 Analysis of the drivers developing the continuous market managed by OMIP

5.3.1 Evolution of the traded volumes in the continuous market

The evolution of the traded volumes (GWh) in OMIP continuous market can be tracked in Figure 3.2. Bigger volumes of year contracts since the autumn of year 2008 are due to bigger registration of OTC trades cleared by OMIClear, as the OTC market is the favourite mechanism for negotiating large contracts. The gradual increase of the cleared volumes in OMIP is due to bigger trading of quarter and year contracts, motivated by the negotiation of similar contracts in the coexisting auction mechanisms –mainly CESUR auctions– and the OTC market. The traded volumes in the continuos market, as well as the cleared volumes in OMIP clearing house, have kept a steady growing pace in the years following the data set considered in this analysis. As indicated by OMIP-OMIClear (2014a), during year 2013 the trade volume in the continuos market and in the auction mode (auctions related to the Special Regime Production in Portugal and auction related to Financial Transmission Rights issued by the Portuguese Transmission System Operator, described in Section 6.3.2.2 and Section 5.4.1.3 respectively) have increased 23% (47 TWh in year 2013 against 38 TWh in year 2012). Due to that, the number of trades increased 42%, with a record of continuous trading in one single day of 1.1 TWh. The open interest also increased (monthly average of 15.5 TWh at the end of the year, 27% higher than at the end of year 2012). The total cleared volume by OMIClear in year 2013 was 86 TWh (38 TWh corresponds to OTC registration), increasing 29% compared to the previous year. The number of cleared trades rose 45% compared to the previous year. The last quarter of year 2013 presented record figures regarding clearing (27.3 TWh in the last quarter of year 2013, and 12 TWh in December 2013). The larger increase in the number of trading and clearing operations compared to the resulting traded/cleared volumes means that the market players are increasingly using futures contracts of shorter maturity (delivery period shorter than one month).

5.3.2 The enrollment of trading members

The increasing enrollment of trading members can be observed in Figure 5.5. This chart shows the registered members at the beginning (July 3, 2006) and at the end of each month. Four categories are employed: “Distribution” refers to Spanish distribution companies and their related last resort suppliers, as well as the Portuguese last resort supplier; “Integrated” refers to generation and trading companies belonging to the Iberian vertically integrated energy groups; “Non Integrated” refers to energy traders not belonging to Iberian vertically integrated energy groups; “Financial” refers to pure financial agents as the commodities branches of investment banks or even brokers.


97

Figure 5.5. Enrollment of OMIP trading members.


The first entities entering the market are the consolidated market players in the Iberian energy market, i.e. the Spanish distribution companies and the Portuguese last resort supplier, as well as their affiliated power generation and trading companies. The international energy traders mainly join prior to the first regulated auctions shown in Table 1.1 and Table 4.7. The financial entities entered later, with a bigger participation since the end of year 2007, as profitable opportunities might arise in power markets compared to spoilt financial and credit markets. During June 2009-September 2009, the Spanish last resort suppliers enrolled. All the trading members are specialized agents with previous experience in the Spanish spot electricity market or in other European energy markets. There is much room for new enrollments of energy traders and financial players active in the spot market. Nonetheless, in order to get a balanced market structure as in Nord Pool (STEM, 2006) – contributing to increase the price efficiency –, the market operator and its clearing house should strive to attract large industrial consumers, suppliers and small producers active in the spot market, hedging their exposure to the spot price volatility. The years following this analysis have also shown a steady increase in the enrolment of trading members. As indicated in OMIP-OMIClear (2014a), fifteen agents joined the Iberian power futures markert during year 2013, totalling 47 trading members.


98

5.3.3 The discounts in OMIP trading fees

Special discounts in trading fees for strenghtening the continuous market began to be applied in the last quarter of 2007 (OMIP, 2007). Further campaigns were applied respectively in the first half of 2008, during the third quarter of 2009, and during 6 months since October 15, 2009 (OMIP, 2008a; 2009c; 2009d).

5.3.4 The regression model for the continuous traded volumes

Section 1.6.2 describes the regression model using Ordinary Least Square methodology to assess the effect of twelve selected drivers (the independent variables) on the the monthly evolution of the energy traded in the continuous market (the dependent variable). The mathematical expression of the model is given by Equation (1.7).

5.3.4.1 The regression results

As shown in Table 5.5, the model renders high R-squared statistic (0.88) and the following findings: the only significant variables (t values bigger than 2.03), all with positive coefficients as predicted, are the OTC cleared volumes in the futures market, the OTC volumes themselves, and OMIP call auction volumes. Such variables are related to coefficients a6 (equal to 0.22), a5 (equal to 0.05) and a4 (equal to 0.24) respectively. The rest of actions directly performed by the market operator to develop the market – i.e. expansion of the continuous trading phase (related to coefficient a9), creation of market making agreements (related to coefficient a10), and discounts in the variable fees related to trading in the continuous market (related to coefficient a12) – do not present significant values (even only the market maker agreements, of those three variables, present a positive coefficient, with t-value 0.25). Therefore, no linear relation was found between the rest of actions directly performed by the market operator to develop the market, and the traded volumes in the continuous market.


99

Table 5.5. Regression model results of traded energy in OMIP continuous market.

a0 -2,498.10 t0 -1.51a1 190.89 t1 0.91a2 -44.28 t2 -1.15a3 90.20 t3 1.13a4 0.24 t4 2.07a5 0.05 t5 3.43a6 0.22 t6 3.53a7 47.58 t7 1.30a8 25.08 t8 1.69a9 -67.49 t9 -0.26a10 58.16 t10 0.25a11 114.36 t11 0.96a12 -94.58 t12 -0.65R2 0.88 t 2.03

Regression model results of traded energy in continuous market

Source: Authors based on data from OMIP-OMIClear (2012)

Further research with a larger data set, considering potentially new significant drivers in the forward trading within the Iberian electricity market (e.g the traded volumes in the auctions related to the Special Regime Production in Portugal, since December 2011; the auctions related to cross border trading in the Spanish-Portuguese interconnection; the clearing volumes in the Spanish clearing house MEFF Power, active since March 2011 with a remarkable growth especially since the beginning of year 2013; OMIClear cooperation agreements with other European clearing houses to allow the settlement of Iberian electricity OTC traded volumes by market participants abroad in those other Europan clearing houses, with potential OMIP liquidity gains; and even the potential enrollment of international cross-commodity traders (i.e. trading all kind of commodity derivatives related oil, gas, electricity, metals, et cetera in the international markets, namely European, American and Asian markets) and renewable energy companies searching for additional rents through derivatives trading due to the drastic reduction of the green subsidies previously commented in Chapter 3, and also hedging at a higher selling price compared to the resulting low price in the spot market when renewable production almost occupies the whole merit order curve in the matching of the generation and the demand).

Regarding the cooperation of OMIClear with other European clearing houses, as indicated in OMIP-OMIClear (2013a), the European Energy Exchange (EEX) and its subsidiary European Commodity Clearing (ECC) have agreed with OMIP and OMIClear to cooperate in trade registration of power derivatives and cross-list their products. EEX German and French Power Derivatives will then be listed at OMIP-OMIClear and OMIP Spanish Power Derivatives at EEX/ECC. Thus market participants


100

can register and clear several power derivatives in their preferred clearing house. These energy derivatives exchanges and their clearing houses will publish the same settlement prices for the same product. From year 2014, EEX intends to offer its trading participants the registration of financially settled Spanish (SPEL) power futures for clearing by ECC, and OMIP-OMIClear will ensure that their participants can register financially settled German power futures (Phelix Futures) and French power futures (French Financial Futures). These derivatives exchanges and clearing house will mutually support each other in the clearing and settlement of the registered transactions. In order to facilitate a comprehensive supervision, the entities composing the MIBEL Regulatory Council should have access to the Iberian derivatives transactions cleared in ECC. Additionally, the historical series of statistics published by OMIP-OMIClear would be richer for researchers if the clearing figures of Iberian derivatives from ECC appear separately (regarding cleared OTC volumes and open interest) and furthermore global figures (aggregated cleared volumes and open interest from both clearing houses regarding Iberian electricity derivatives) are also provided.

Capitán Herráiz and Rodríguez Monroy (2011a) expands the data set 6 months (i.e ending in December 2010, thus considering 54 observations) and finds the following results: the model renders as well high R2 (0.86) being the only significant variables (t-values bigger than 2.02), all with positive coefficients as predicted, the OTC cleared volumes in the futures market (coefficient a6 equal to 0.20 and t-value equal to 2.71), the OTC volumes themselves (coefficient a5 equal to 0.05 and t-value equal to 2.83), and the number of financial agents (coefficient a3 equal to 186.05 and t-value equal to 2.31). With such expanded data set, the OMIP call auction volumes do not become a significant variable due to the fact that the OMIP auction volumes during year 2010 are very scarce, as previously commented in Section 5.2.1 and shown in Figure 3.2.

The OTC volumes cleared by OMIClear have grown due to the fast growth of the OTC market and due to OMIClear commission discounts. In the long term, OMIP continuous volumes are sustained by themselves without the need of the compulsory call auctions. However OMIP marketing campaigns regarding expansion of the continuous phase, promotion of commission discounts and captation of market makers seem not influence notably the development of the continuous market.

A closer dialogue of OMIP with the stakeholders and with the regulatory authorities would let OMIP identify the most important needs of the agents. The regulated forward contracting mechanisms (VPP and CESUR auctions) neither influence much the development of the continuous market, due to the fact that the frequency of such auctions is not high (every three months in the case of CESUR auctions; the last VPP auction was celebrated in March 2009). Both CESUR and VPP auctions could be held more often to promote forward contracting and provide more price signals, but a thorough cost-benefit analysis should be done by the regulatory agencies, as the management of multi-round electronic auctions with a previous qualification process implies substantial administrative costs – around 150.000 euros in the case of CESUR auctions (for electricity) and similarly for gas– as stated in the regulatory pieces for such auctions (e.g. MITyC (2010b; 2010c)). Cost-benefit analysis of stronger regulation versus market failures is well described by Joskow (2010).


101

5.3.5 Correlation analysis

The correlation of each variable introduced in the regression model expressed in Equation (1.7) with the evolution of the monthly traded volumes in the continuous market is analyzed. In the same way as in the regression model, monthly values for the four first four years of the Iberian power futures market are considered (i.e 48 observations corresponding to the months spanning from July 2006 to June 2010). The only remarkable correlation coefficients are found for the OTC volumes (0.89), the OTC cleared volumes by OMIClear (0.76) – both parameters were already identified as significant in the regression analysis –, and the enrolment of financial agents (0.77). Therefore, marketing incentives to attract these players could increase the trading activity in the continuous market. Note that Table 3.3 in Section 3.3.3 shows the results of the correlation analysis, considering a larger data set (from July 2006 to December 2011, i.e. 66 observations). The results from that analysis are quite consistent with the correlation analysis described in the current section. The analysis with a larger data set renders two additional factors showing a high correlation factor with the trading volumes in the continuous market (the existence of market maker agreements in OMIP and the enrolment of vertically integrated companies in the Iberian power futures market). In particular, the analysis performed in Section 3.3.3 indicates a correlation factor of 0.88 for the OTC volumes, 0.82 for OMIP market makers, 0.80 for OMIP financial agents, 0.76 for OMIP vertically integrated generation companies and 0.70 for the OTC volumes registered at OMIClear.

5.4 Efficiency recommendations

This section synthesizes all the research findings. They allow to formulate policy recommendations and supervision actions for the Iberian power futures market operator and its clearing house in order to improve their performance and yield liquidity gains.

5.4.1 The three-layers liquidity pyramid

Figure 5.6 shows a pyramid in which the diverse liquidity factors are placed according to their theoretical contribution to liquidity improvements. Three layers are distinguished. A bottom-up approach is used, showing the most valuable factors at the top. The highest an energy exchange can be located in the pyramid, the largest its efficiency. Therefore, this chart can be used as a graphical efficiency benchmark tool. The youngest and most illiquid power exchanges would be located around the basic layer. Their liquidity growth would mainly be of a quantitative nature (traded volumes, number of agents, et cetera). The most mature and liquid power exchanges would be


102

around the top layer, showing qualitative liquidity improvements (e.g. common use of more sophisticated derivatives). The regression model identified only some of the trading drivers as significantly contributing to the traded volume development of the continuous market. The market operator and its clearing house should continuously monitor all these factors through quantitative and qualitative analyses to reach the desired liquidity goals. It would be worthy that the results of such analyses are shared with the supervisory agencies – in this case, represented by the MIBEL Regulatory Council, composed of the Spanish and Portuguese energy and securities regulatory bodies (CNMC, ERSE, CNMV and CMVM). Such agencies could provide feedback and monitor better all the interrelated MIBEL forward markets.

Figure 5.6. The three-layers liquidity pyramid.

Growing traded volumes Variety of trading members Variety of basic derivatives

Coexistence with strong OTC market

Efficient market maker agreements

Small bid-ask spreads

Robust exchange & clearing operation through innovative IT platforms

Coexistence with regulated forward mechanisms

Open Interest development reflecting real hedging and sound risk management needs

Market competition through agents’ balanced structure

Short maturitybalancing contracts

Wider portfolio Multi-commodity

Price efficiency

Pre-trade & Post-trade fundamental data transparency

Cross-border capacityproducts

Compliance of Ethic Codes

Incentive campaigns (e.g. discounts)

Stakeholders’ communication

Source: Authors

5.4.1.1 The basic layer

The basic layer is composed of the ground drivers. Liquidity develops when a critical mass of trading members and products exist. Coexistence with OTC market and with regulated forward contracting mechanisms should be positive, offering such markets attractive and complementary trading possibilities for market agents. The possibility of price arbitrages – stimulating the trading activity – between all the existing mechanisms was considered in the regression model. Marketing campaigns would be desirable to attract market makers and to increase trading volumes, for instance,


103

through commission discounts. Efficient market maker agreements, requesting permanent performance measurement, could diminish bid-ask spreads and contribute to the accurate formation of the reference prices, the so-called settlement prices (STEM, 2006; Batlle et al., 2007). Robust operation of the exchange and the clearing house are of utmost importance. OMIP and OMIClear have so far performed well with no major incidences. The clearing house will get a fundamental role in the development of this market. Such a key function of the clearing house is in line with the conclusions of the G20 summit on September 25, 2009, in Pittsburgh (USA), and the envisaged policy actions for the proper regulation of derivatives markets by the European Commission. According to those policy guidelines, it is expected to increasingly clear the standardized OTC derivative contracts through central clearing houses to mitigate counterparty risk (European Commission, 2009c).

Pre- and post-trade transparency are key to ensure fair and orderly trading (European Commission DG TREN, 2008a; 2009). OMIP transparency compares well with European benchmarks (EEX and Nord Pool) regarding public data availability, but improvements should be done in providing timely and more analytical market monitoring reports as the other exchanges usually do. Stakeholders’ communication through steady dialogue is key to innovate in the right way, responding to market participants’ real needs, and to properly measure the market development, through the proposed win-win feedback from supervisory agencies.

OMIP and OMIClear meet regularly with their members through the Trading & Products and Clearing & Settlement Committees’ Meeting in order to improve members’ satisfaction. For instance, the 9th meeting was celebrated in Oporto on May 29, 2009 (OMIP, 2009b). The transparency would be substantially improved if a briefing of those meetings summarizing the main ideas were published, even if no mention to the authors of the comments exposed in those meetings were revealed for the sake of sensitive issues (this practition is known as the Chatham House rule). According to Chatham House, home of the Royal Institute of International Affairs, a world-leading source of independent analysis located in London, the Chatham House Rule, established in 1927, reads as follows: “When a meeting, or part thereof, is held under the Chatham House Rule, participants are free to use the information received, but neither the identity nor the affiliation of the speaker(s), nor that of any other participant, may be revealed”. This rule may be invoked at meetings to encourage openness and the sharing of information. It allows people to speak as individuals, and to express views that may not be those of their organizations, and therefore it encourages free discussion. People usually feel more relaxed if they do not have to worry about their reputation or the implications if they are publicly quoted (Chatham House, 2014).

5.4.1.2 The intermediate layer

In the intermediate layer, the proper evolution of the open interest related to real hedging practices (i.e. growing evolution of the open interest according to increased cleared volumes in the clearing house), opposed to pure speculative ones is remarked.


104

As indicated by Lucia and Pardo (2008), stagnation of the open interest is caused by intraday trading for strong speculation purposes: the positions are opened and closed in a single trading session. An analysis of the development of the traded and cleared volumes compared to the evolution of the open interest in OMIP-OMIClear is provided in Chapter 7. Additionally, the evolution of the cleared volumes and the open interest in the first months of the Spanish clearing house MEFF Power (currently known as BME Clearing) is also provided in Chapter 7.

A balanced structure of the market according to the desired equilibrated nature of the agents is also pointed out in the pyramid (STEM, 2006). The traditional classification of market agents as in Martín Martínez (2008), composed of hedgers, speculators and arbitrageurs is well represented in this market, as shown in Figure 5.5. Nonetheless, the relatively reduced amount of trading members in OMIP – compared to more mature European energy derivatives exchanges –, in which half of them is related to Iberian incumbents (i.e. vertically integrated companies historically settled in Spain and Portugal), does not help to the proper development of the market. For instance, in the case of the United Kingdom, the dominance of the so-called “Big 6” also causes liquidity strains (OFGEM, 2009).

OMIP should attract other market players active in the mature spot market, especially for intensive energy users and suppliers with strong purchasing hedging needs, as the existing hedgers in the continuous market correspond to a more prone selling nature (generation companies from large vertically integrated utilities). Should new financial agents enter without experience in energy markets, closer oversight should be performed to detect if excessive speculation may arise, increasing the price volatility (Universia, 2011). Note that the nature of the players in North American commodity derivatives markets has dramatically changed in the last decade, favouring excessive speculation: whereas in 1998, physical hedgers (in other words, they perform teading for commercial purposes and thus they are also called commercial traders) represented 77% of the market, traditional speculators 16% and index speculators 7%, in 2008, physical hedgers were only 31%, while traditional speculators rose to 28% and index speculators to 41% of the total (IECA, 2010). Small producers, as in STEM (2006), would also be welcome to enrol. Some of them could be attracted by improvements in the regulation of distributed generation, as suggested in Frías et al. (2009).

On the other hand, according to the international agreement for the constitution of the Iberian electricity market, signed in Santiago de Compostela on October 1, 2004, the futures market operator (OMIP) and the spot market operator (OMEL) must exchange stocks and constitute a single Iberian Market Operator (OMI). This monopolistic structure, as those ones described in Meeus (2011), can generate useful synergies incrementing the efficiency of both markets. For instance, marketing campaigns from OMIP to attract market participants active in OMEL (582 producers, 94 suppliers, and 4 intensive industries, according to OMEL (2010)) could help to equilibrate OMIP agents’ structure, producing liquidity growth and increasing the market competition.

OMIP and OMIClear approved their respective Ethic Codes in July 6, 2009, amended on May 18, 2010, in order to strenghten their governance structures and


105

supervisory capabilities, promoting and guaranteeing the transparency and competition amongst all the participants. The first meeting was held in Lisbon on July 9, 2009 (OMIP, 2009c). Strict compliance of these codes would ensure market integrity, contributing to the compliance of one of the main goal of the European Regulation on Market Integrity and Transparency of the European Wholesale Energy Markets, known as REMIT, aiming at ensuring the market integrity by means of avoiding the market abuse (insider trading and market manipulation) and increasing the market transparency (European Union, 2011). Derived from the achievement of higher levels of market integrity, market participants would rely on the derivatives exchange and clearing houses to manage their energy price risk and thus some liquidity gains were produced in the Iberian power futures market.

5.4.1.3 The top layer

In the top layer, the development of short maturity products with physical settlement for balancing purposes is suggested. Specific market maker agreements, as in Batlle et al. (2007), could make such an implementation more effective. Prior to their deployment, OMIP should properly market the still illiquid week contracts, by means of efficient market maker agreements, compulsory call auctions, and attractive commission discounts for those contracts. These actions are also recommended for the peak futures, and the base load futures with underlying price the spot price of the Portuguese zone (i.e. the PTEL index) shown in Table 5.1. Good development of such PTEL futures could facilitate cross-border trading in the Spanish-Portuguese interconnection – the introduction of those contracts is justified by the existence of market coupling in MIBEL since July 2007, which has increased the liquidity of the spot market (European Commission DGTREN, 2008b) –, and would justify the introduction of more capacity products (e.g. financial transmission rights, as in Molzahn and Singletary (2011)), serving as an efficient hedging tool in the Iberian energy market.

In this way, the first auction of financial transmission rights (FTR) related to the cross-border trades in the Spanish-Portuguese interconnection (FTR options for both flows, i.e. Spain to Portugal and Portugal to Spain) was managed by OMIP-OMIClear on 19 December 2013 (OMIP-OMIClear, 2013b). Furthermore, auctions related to contracts for differences (i.e. FTR obligations) have been managed by the Spanish market operator (OMIE) since year 2009 on a biannual basis (one auction is held every June and a second auction every December). Therefore, in December 2013 the 10th auction was held (OMEL Mercados A.V., 2013). Further research is suggested to measure if such FTR auctions have produced liquidity gains in the continuous market managed by OMIP and if the equilibrium price from those auctions (i.e. the price differences between the Spanish and Portuguese prices) is equivalent to those derived from the daily settlement prices of the Spanish and Portuguese Futures or remarkable arbitrage differences exist between those derivatives trading mechanisms. Such an analysis would strengthen the assessment of the efficient price formation of the Iberian power futures market.


106

Liquidity may grow by offering an ampler portfolio of derivatives (e.g. call and put options), other energy related products (e.g. derivatives related to the European Emission Trading Scheme, as described in Kockar et al. (2009)), and even market mechanisms for establishing a forward price related to the integration of renewable energy sources in the Spanish and Portuguese energy systems. A description of the penetration of the renewable generation in the Iberian electricity market can be found in Matos et al. (2009). A very first assessment of the first forward contracting mechanisms related to renewable generation trading (case of the Special Regime Production auctions held in Portugal) and clearing (related to the contracts for differences based on CESUR and spot prices for the Special Regime Production in Spain), both introduced in year 2011, is provided in Section 6.3.2. Further research in this original topic and of great socioeconomic impact, due to the ambitious European Union national governments’ renewable energy policy goals in the mid-term, is strongly encouraged

The summit of the pyramid shows the desired goal of the price efficiency, in terms of resilience as defined in Newbery et al. (2003), convergence with the OTC market (CNE, 2011a), and reasonable evolution (neither inflated nor biased) of the forward risk premium related to the spot market as analyzed in Capitán Herráiz and Rodríguez Monroy (2008a; 2008b; 2009a; 2009b; 2010b; 2012b) and Furió and Meneu (2010). For the sake of a bigger amount of futures trades originated by the incumbents instead of their natural trend to establish opaque bilateral trades (OFGEM, 2010), OMIP-OMIClear could provide trading and/or clearing commission discounts for the incumbents in case they exceed a given threshold of traded and/or cleared volumes. Those thresholds could be defined by the national energy regulatorory authorities (the Spanish CNMC and the Portuguese ERSE). In the case that such incumbents are active participants in the opaque OTC market, this measure would bring, in general, price transparency and, in particular, a potential remarkable liquidity growth in the futures market, both increasing the efficiency of the Iberian forward market.

As a final remark, whereas the most mature European energy derivatives markets (Nord Pool (whose current denomination is Nasdaq OMX Commodities) and EEX) can be located within the top layer – the first European exchange in reaching such recognition is Nord Pool (Mork, 2001; STEM, 2006) –, OMIP still strives to excel its already consolidated reputation in the bottom layer and pass to the intermediate layer. The exposed policy actions, expressing efficiency recommendations, would help OMIP-OMIClear to evolve upwards in this illustrative pyramid.

5.5 Results

Since its beginning on July 3, 2006, the Iberian power futures market managed by OMIP has experienced a steady development, in terms of number of participants and traded volumes. Nevertheless, OMIP liquidity is still poor compared to other more mature European power futures markets: Nord Pool (whose current name is Nasdaq OMX Commodities) and EEX (OMIP-OMIClear, 2012; OFGEM, 2009; EEX, 2009; Nord Pool, 2009). Therefore, whereas Nord Pool and EEX can be located at the top of a


107

graphical liquidity pyramid composed of three layers in a bottom-up approach, OMIP still strives to enter the intermediate layer. In this sense, whereas OMIP transparency compares well with those exchanges regarding public data availability, contributing such data transparency to trading confidence and thus liquidity growth (European Commission DG TREN, 2008a; 2009), bigger efforts and resources should be employed by OMIP to provide timely and more analytical market monitoring reports as the other exchanges usually do. A regression model is built to identify the main factors behind the development of one key liquidity measure: the traded volumes in OMIP continuous market. The model is composed of twelve parameters (independent variables) theoretically contributing to the development of the traded volumes in the continuous market (the dependent variable). The data set considers the first four years of the Iberian energy derivatives exchange (i.e. from July 2006 to June 2010, reckoning 48 monthly observations). According to the regression results, the only significant parameters acting as effective liquidity drivers are the traded volumes in the dominant OTC market, the OTC cleared volumes by OMIClear, and the call auctions in which the Spanish distribution companies and the Portuguese last resort supplier are obliged to purchase energy according to national legislation. Additionally, correlation analysis of the twelve variables with the traded volumes in the continuous market also shows a significant correlation of the enrolment of financial agents. Therefore, marketing incentives to attract these players could increase the trading activity in the continuous market. Apart from that potential group of new entrants, there is still much room for new enrolments in OMIP, especially for large industrial users, suppliers active in the spot market, small producers (e.g. renewable energy generation companies), as well as international cross-commodity traders. The enrolment of these companies would create a more balanced structure in the market (STEM, 2006; Martín Martínez, 2008). Apart from increasing the trading activity, it would facilitate a less biased price formation in terms of forward risk premium (Capitán Herráiz and Rodríguez Monroy, 2008a; 2008b; 2009a; 2009b; 2010b; 2012b; Furió and Meneu, 2010). On the other hand, specific commission discounts for the large vertically integrated groups to trade in OMIP and clear OTC volumes in OMIClear would bring price transparency and liquidity growth (OFGEM, 2010), in case such vertically integrated utilities were active OTC participants in opaque bilateral trades. The market operator could establish specific market maker agreements (Batlle et al., 2007), commission discounts, and auctions complementary to the continuous market to boost the still illiquid products. The market operator and its clearing house should hold a permanent dialogue with all the stakeholders, both with the agents and the spot market operator (as attractive products can only be offered if they respond to real market needs) and with the supervisory agencies (as new developments in market rules should improve the market efficiency). Collaborative efforts between market operator and supervisory agencies create strong synergies: the market operator could invite the supervisory agencies to provide feedback about its analyses measuring the performance of new business developments. Such shared vision would benefit both entities due to the liquidity and efficiency gains. Social welfare would increase as well, positively contributing to a better price formation of end users’ electricity prices. Further research is encouraged by analyzing the evolution of other liquidity measures, as the bid-ask spreads with a larger data set and considering short-term maturity contracts (i.e. day-ahead, weekend and week contracts) (STEM, 2006);


108

the open interest registered in the Portuguese clearing house (OMIClear) and the Spanish clearing house (MEFF Power, whose current name is BME Clearing) with a larger data set and considering as well short-term maturity contracts (Lucia and Pardo, 2008); the volatility and sensitivity of prices to additional demand (Krishna, 2008); and the resilience (Newbery et al., 2003). Especially interesting, in a context of high volatility in financial and commodities markets (Universia, 2011), would be the research about the open interest in OMIP futures market: whereas the daily trading volume reflects movements in the speculative activity, the daily open interest captures hedging activities as it excludes intraday positions taken by day traders, mainly inspired by speculative reasons (Lucia and Pardo, 2008). An analysis related to the evolution of the open interest and the traded and cleared volumes for the Iberian electricity derivatives is provided in Chapter 7, and stronger conclusions could be provided considering the suggestions for further research with a longer data set embracing as well shorter term contracts. Additionally, the research can also be enriched by consulting literature regarding similar experiences in North American energy derivatives markets and by analyzing the impact of coming new financial legislation in European and North American markets raised after the global concerns of the financial crisis suffered in year 2008 (European Commission, 2009c; European Commission DG TREN, 2009).


109

CHAPTER 6. EVALUATION OF THE FORWARD PRICE FORMATION THROUGH THE GENERATION COST ASSESSMENT

CHAPTER SUMMARY

The price formation of the Iberian Energy Derivatives Market – the power futures

market –, starting in July 2006, is assessed until November 2011, through the

comparison with the forward generation costs from natural gas (“clean spark spread”).

The power futures are strongly correlated with European gas prices. The spreads built

with prompt contracts tend to be positive, as the ex-post forward risk premium. The

biggest ones are for the month contract, followed by the quarter contract and then by

the year contract. Therefore, gas fired generation companies can maximize profits

trading with contracts of shorter maturity. Introduction of Iberian renewables market

mechanisms could contribute to a more efficient price formation, e.g.renewable

auctions for the most mature technologies (i.e. wind). Such mechanisms diminish the

tariff deficit caused by the massive deployment of the feed-in-tariff scheme. Liquidity in

the forward markets will also increase as a result of the entry of renewable generation

companies intending to maximize their profits due to gradual suppression of feed-in-

tariff schemes.

6.1 Introduction

The research in Chapter 4 analyzed the efficiency of the price formation of the Iberian energy derivatives market, the power futures market managed by OMIP (Iberian Energy Market Operator, Portuguese Pool), through the ex-post forward risk premium. The research in the current chapter employs alternatively another key indicator: the clean spark spread, obtained as the difference between the power futures price and the forward generation cost with a gas fired combined cycle plant taking into account the CO2 emission rates (e.g. Abadie and Chamorro, 2009). The analysis on the Iberian clean spark spread is based on the research performed by Capitán Herráiz and Rodríguez Monroy (2013c).

The data set considers the first five years and five months of the Iberian power futures market (from July 3, 2006, until November 30, 2011). For power futures prices, OMIP settlement prices for base load contracts with underlying price the spot price of the Spanish zone are considered. Such contracts are identified as “FTB” (Futures Base Load). For power spot prices, OMIE (Iberian Energy Market Operator, Spanish Pool) daily prices are calculated as the arithmetical average of the hourly prices in the day-


110

ahead auction for each day. For gas prices, Platts’ assessments of transactions traded OTC (Over-the-Counter, i.e. out of organized markets) in the Dutch virtual trading point (Title Transfer Facility, “TTF”) are used. The Spanish gas market lacks of price transparency, as physical swaps for balancing purposes are arranged amongst participants without disclosing the price (Honoré, 2011). Therefore the most liquid reference in continental Europe, as indicated in Heather (2012), has been taken instead.

Note that the TTF price presents high correlation and (on average) reduced price differentials with other relevant European hubs (e.g. the British NBP (this was the gas price reference used in Chapter 4) and the Belgian Zeebrugge hub). According to Petrovich (2013), the European wholesale gas prices traded in hubs present an increasing correlation. She uses OTC and exchange data from years 2007 to 2012. Her analysis finds support for the hypothesis that the European hubs provide a reliable price reference. Where anomalies occur they can be related either to hub immaturity in early periods or to physical connectivity constraints.

For the CO2 emissions, EUA (European Union Allowances) futures settlement prices in the ICE (InterContinelExchange) derivatives exchange are used. Brent crude oil prices – both spot prices assessed by Platts and ICE futures – are considered to analyze its correlation with gas prices and cointegration relationships with power prices.

Apart from the strong relationship between fuel prices and electricity prices in the forward price formation, the influence of the massive integration of renewable generation in the wholesale price formation is a topic of high interest. EWEA (2010) provides a comprehensive literature review and assessment of studies of the impact of wind energy on electricity prices, built with case studies of Germany, Denmark and Belgium. An increased penetration of wind power reduces CO2 emissions as well as the wholesale spot and end-user prices. Wholesale electricity prices (spot prices) are reduced between 3 and 23 €/MWh depending on the amount of wind power. Wind replaces hard coal power plants during hours of low demand and gas-fired power plants during hours of high demand in the countries analyzed. The low marginal costs of wind generation push more expensive technologies, such as gas and thermal plants, out of the market (the “merit order effect” in the matching of the supply and demand curves). Gelabert et al. (2011) analyze ex-post the effects of special regime generation (i.e. renewable sources, waste heat and co-generation) on Spanish wholesale electricity prices, with data for years 2005-2009. A marginal increase of 1 GWh of special regime generation is associated with a reduction of almost 2 €/MWh (ca. 4%) in wholesale power prices. More accurate production forecasts for intermittent renewable sources (wind and solar) could produce a downward pressure in the short-term prompt forward prices, contributing to lessen their forward risk premia.

The chapter is structured based on these two relevant themes in the price formation (fuel prices and renewable generation). Therefore, the clean spark spreads are analysed in Section 6.2. In order to provide a sound analysis, correlation and cointegration analyses are done to understand the relationships between the different energy prices and their impact in the electricity price formation. An analysis of the very first Iberian market mechanisms based on renewable generation is provided in Section


111

6.3 compared to more mature mechanisms previously introduced in Latin American countries. The analysis regarding the existing Iberian renewable trading/clearing mechanisms is based on the research performed by Capitán Herráiz and Rodríguez Monroy (2012a). Reflections for the development of Iberian renewable auctions contributing to a more efficient forward price formation are provided in that section. Finally, the article concludes stating the main results and suggesting further research work in Section 6.4.

6.2 Evaluation of the Forward Generation Costs

The Combined Cycle Gas Turbine (CCGT) generation accounts in Spain in 2009 for 24% of the installed capacity and for 29% of the electricity gross production, being the main generation technology (Moreno and Martínez-Val, 2011). In order to understand the power forward price formation, correlation analyses – between power, gas and emission prices as well as between oil and gas prices –, cointegration analyses of power, gas and oil prices, and tracking of the clean spark spreads are performed.

The main fundamentals for the Spanish electricity forward price formation are the expected demand, the CO2 allowances prices, and the fuel forward prices (especially the natural gas) (CNE, 2010). The evolution of the spot price and the neighboring power prices (France and Germany) should also be taken into account (CNE, 2012a; 2012f). The Brent crude oil and natural gas forward prices play a prominent role in the Spanish electricity forward price formation process (Furió and Chuliá, 2012). That research considers prompt month forward contracts and uses the gas prices of the Belgian gas hub, located at Zeebrugge, due to the commented lack of local reference for the Spanish gas market. The power, oil and gas series are cointegrated (the three markets respond to common information). The series of Spanish electricity forward prices adjusts to past disequilibria by moving toward the trend values of oil and gas prices. There is a short-run bidirectional relation between oil and gas forward prices. There are volatility spillovers from the oil and gas markets to the Spanish electricity market as well as between the two fuels markets. Any shock or increase in volatility originated in the oil and gas forward markets should be taken into account by market participants in the Spanish electricity forward market in order to anticipate a volatility rise in this market.

6.2.1 Correlation Analysis

Table 6.1 shows the correlations between daily power prices (OMIE spot price, OMIP futures prices for the prompt month (“M+1”), quarter (“Q+1”) and year (“Y+1”) contracts), daily gas prices (TTF spot and forward M+1, Q+1 and Y+1 contracts), and daily CO2 emission allowances (futures prices for the most liquid contract: the prompt December contract “Dec+1”). Different sub-periods have been considered to take into account significant events affecting the price formation, namely: (Period 1) July 3,


112

2006–May 31, 2007 (as since June 2007, the application of a fixed price of 42.35 €/MWh in the power pool for intra-group bilateral transactions expired (BOE, 2006; MITyC, 2007a); (Period 2) June 1, 2007–November 30, 2007 (as the second phase of the European Union Emissions Trading System (EU ETS) (2008-2012) begins in December 2007 (CRE, 2011)); (Period 3) December 3, 2007–August 29, 2008 (as September 2008 represents the beginning of the effects of the global financial crisis on the real economy (Furió and Chuliá, 2012)); (Period 4) September 1, 2008–January 31, 2011 (as the Royal Decree 134/2010 entered in force in February 2011 (MITyC, 2010a, 2011a)); and (Period 5) February 1, 2011–November 30, 2011. Correlation results are also shown for the whole data set.

In general, the CO2 prices show the smallest correlation, even with negative correlation especially in Period 2 due to near to zero price values of the emission allowances at the end of the first phase of the EU ETS and in Period 5 as the emission rates present a steady decline from April 2011 (quoting around 18 €/tCO2) till November 2011 (quoting around 8 €/tCO2). High correlation is observed amongst all the gas and power forward prices of same maturity (especially Q+1 and Y+1). The electricity spot prices tend to show negative correlation with gas and CO2 prices in Period 3, in which record oil prices are registered.

For the spot power price, the biggest correlation with the power futures prices is for the M+1 contract for all the sub-periods (0.91 for the whole data set) as the time proximity is the biggest one. Equally, for the spot gas price, the biggest correlation is for the M+1 forward gas price for all the sub-periods (0.94 for the whole data set). Although the correlation amongst all the futures power prices in the whole data set is bigger than 0.8, only Q+1 and Y+1 contracts present a correlation bigger than 0.75 in all the sub-periods. In the same way, the correlation amongst all the forward gas prices is bigger than 0.8, but in Period 5, characterized by much uncertainty in the global economic recovery, the M+1 contract shows small correlation values with the Q+1 (0.56) and Y+1 (0.27). The prompt month and quarter futures power contracts show correlation bigger than 0.8 with all the gas forward contracts in the whole data set, though falling to much smaller values when considering sub-periods. The biggest correlation between futures power and forward gas prices is between the year contracts (0.92 for the whole data set, falling to 0.64 in Period 1) as the long-term futures power prices are closely influenced by the evolution of commodity prices (oil, gas and coal) as long as the scope of renewable forecasts is for the very short term. Although the emission allowances are influenced by the evolution of power and fuel prices, no high correlation values are obtained due to the excess of offer in the EU ETS (the Member States grant for free a quota of emission allowances to the polluting industries), provoking transitorily decoupled low emission prices and eventually near to zero values (CRE, 2011).


113

Table 6.1. Correlation Matrix between Wholesale Energy Prices.

Spot

ElectricityM+1

ElectricityQ+1

ElectricityY+1

ElectricitySpotGas

M+1Gas

Q+1Gas

Y+1Gas

Dec+1CO2

Spot Electricity 1.00 0.73 0,62 0.65 0.47 0.44 0.58 0.65 0.68M+1 Electricity 0.73 1.00 0.84 0.75 0.61 0.66 0.74 0.67 0.79Q+1 Electricity 0,62 0.84 1.00 0.77 0.65 0.72 0.86 0.79 0.86Y+1 electricity 0.65 0.75 0.77 1.00 0.42 0.45 0.60 0.64 0.66

Spot Gas 0.47 0.61 0.65 0.42 1.00 0.81 0.79 0.67 0.77M+1 Gas 0.44 0.66 0.72 0.45 0.81 1.00 0.89 0.58 0.78Q+1 Gas 0.58 0.74 0.86 0.60 0.79 0.89 1.00 0.86 0.96Y+1 Gas 0.65 0.67 0.79 0.64 0.67 0.58 0.86 1.00 0.92

Dec+1 CO2 0.68 0.79 0.86 0.66 0.77 0.78 0.96 0.92 1.00

Spot Electricity 1.00 0.74 0.69 0.73 0.51 0.47 0.36 0.61 -0.18M+1 Electricity 0.74 1.00 0.74 0.84 0.31 0.31 0.12 0.54 0.02Q+1 Electricity 0.69 0.74 1.00 0.87 0.72 0.79 0.64 0.89 -0.31Y+1 electricity 0.73 0.84 0.87 1.00 0.54 0.56 0.42 0.80 -0.07

Spot Gas 0.51 0.31 0.72 0.54 1.00 0.97 0.92 0.88 -0.65M+1 Gas 0.47 0.31 0.79 0.56 0.97 1.00 0.94 0.90 -0.65Q+1 Gas 0.36 0.12 0.64 0.42 0.92 0.94 1.00 0.81 -0.71Y+1 Gas 0.61 0.54 0.89 0.80 0.88 0.90 0.81 1.00 -0.45

Dec+1 CO2 -0.18 0.02 -0.31 -0.07 -0.65 -0.65 -0.71 -0.45 1.00

Spot Electricity 1.00 0.54 0.04 0.08 -0.21 -0.20 -0.20 -0.03 -0.33M+1 Electricity 0.54 1.00 0.21 0.30 0.05 0.23 0.40 0.28 0.14Q+1 Electricity 0.04 0,21 1.00 0.87 0.39 0.60 0.90 0.83 0.63Y+1 electricity 0.08 0.30 0.87 1.00 0.44 0.68 0.89 0.98 0.73

Spot Gas -0.21 0.05 0.39 0.44 1.00 0.78 0.34 0.52 0.72M+1 Gas -0.20 0.23 0.60 0.68 0.78 1.00 0.62 0.77 0.91Q+1 Gas -0.20 0.40 0.90 0.89 0.34 0.62 1.00 0.86 0.60Y+1 Gas -0.03 0.28 0.83 0.98 0.52 0.77 0.86 1.00 0.82

Dec+1 CO2 -0.33 0.14 0.63 0.73 0.72 0.91 0.60 0.82 1.00

Spot Electricity 1.00 0.94 0.90 0.87 0.77 0.84 0.86 0.82 0.71M+1 Electricity 0.94 1.00 0.97 0.93 0.81 0.89 0.93 0.87 0.81Q+1 Electricity 0.90 0.97 1.00 0.94 0.74 0.84 0.93 0.87 0.89Y+1 electricity 0.87 0.93 0.94 1.00 0.76 0.83 0.88 0.93 0.84


Dec+1 CO2 0.71 0.81 0.89 0.84 0.58 0.69 0.83 0.78 1.00

Spot Electricity 1.00 0.63 0.45 0.30 -0.08 0.42 0.55 0.18 -0.57M+1 Electricity 0.63 1.00 0.83 0.63 0.05 0.50 0.58 0.27 -0.63Q+1 Electricity 0.45 0.83 1.00 0.90 0.21 0.56 0.62 0.56 -0.32Y+1 electricity 0.30 0.63 0.90 1.00 0.33 0.55 0.68 0.75 -0.21

Spot Gas -0.08 0.05 0.21 0.33 1.00 0.53 0.10 0.11 0.08M+1 Gas 0.42 0.50 0.56 0.55 0.53 1.00 0.56 0.27 -0.34Q+1 Gas 0.55 0.58 0.62 0.68 0.10 0.56 1.00 0.68 -0.53Y+1 Gas 0.18 0.27 0.56 0.75 0.11 0.27 0.68 1.00 0.09

Dec+1 CO2 -0.57 -0.63 -0.32 -0.21 0.08 -0.34 -0.53 0.09 1.00

Spot Electricity 1.00 0.91 0.83 0.78 0.75 0.78 0.75 0.76 0.56M+1 Electricity 0.91 1.00 0.93 0.87 0.77 0.83 0.81 0.83 0.59Q+1 Electricity 0.83 0.93 1.00 0.91 0.75 0.83 0.88 0.89 0.54Y+1 electricity 0.78 0.87 0.91 1.00 0.68 0.73 0.78 0.92 0.42


Dec+1 CO2 0.56 0.59 0.54 0.42 0.53 0.51 0.50 0.56 1.00

Period 3: 3 December 2007 - 29 August 2008

Period 4: 1 September 2008 - 31 January 2011

Period 5: 1 February 2011 - 30 November 2011

Total (data set from July 3 2006 until November 30 2011)

Period 1: 3 July 2006- 31 May 2007

Period 2: 1 Jun 2007 - 30 November 2007

Source: OMIP-OMIClear, Platts, and ICE

The gas prices are influenced by the evolution of fundamental variables, linked to supply and demand (e.g. cold snaps, availability of interconnectors and Liquefied Natural Gas (LNG) carriers, stock levels at the underground storages and regasification


114

plants), as well as the macroeconomic situation and the evolution of the exchange rates ($/€) and oil prices.

Regarding the factors influencing crude oil price formation, EIA (2014) provides an interesting snapshot (with charts frequently updated, usually on a monthly basis) built with seven interrelated factors (essentially, very similar to the gas price formation factors commented above). Such factors (and their main indicators) are:

1. Supply from non-OPEC countries (Production & WTI crude prices; Changes in production capacity & Gross Domestic Product (GDP), price of WTI crude; Projected supply, annual average); OPEC stands for the Organization of the Petroleum Exporting Countries.

2. Supply from OPEC countries (Changes in production targets & WTI crude prices; Spare production capacity & WTI crude prices; Changes in production capacity & GDP, price of WTI crude).

3. Balance (OECD inventories & WTI futures spread); OECD stands for Organisation for Economic Co-operation and Development. This organization was established in year 1961 in Paris (France) and the current number of members reckons 34 countries. The OECD mission is to promote policies that will improve the economic and social well-being of people around the world (OECD, 2014).

4. Spot Prices (World crude oil prices; U.S. retail gasoline price, refiner acquisition cost of crude oil; Crude price reaction to events).

5. Financial Markets (Average daily open interest in crude oil futures; Futures positions by producers, merchants, processors, & end users; Futures positions by money managers; Correlations between daily prices changes of crude & other commodities; Commodity index assets under management & Dow Jones UBS price index; Composition of the Dow Jones UBS commodity index; Correlations between daily returns on crude oil & financial investments); As indicated in S&P Dow Jones Indices (2014), the The Dow Jones-UBS Commodity Index DJ-UBSCI is a broadly diversified index that allows investors to track futures of physical commodities through a single measure. The index is designed to minimize concentration in any commodity or sector. It currently has 22 commodity futures in seven sectors. No one commodity can compose less than 2% or more than 15% of the index, and no sector can represent more than 33% of the index as of the annual weightings of the components. The weightings for each commodity included in DJ-UBSCI are calculated in accordance with rules that ensure that the relative proportion of each of the underlying individual commodities reflects its global economic significance and market liquidity. Annual rebalancing and reweighting ensure that diversity is maintained over time. UBS AG is a Swiss global financial services company.

6. Demand from non-OECD countries (Consumption & GDP; World oil consumption, world GDP & WTI crude oil prices; GDP growth in Asia);

7. Demand from OECD countries (Crude oil consumption & WTI crude oil price; World oil consumption, world GDP & WTI crude oil prices).


115

Interestingly, EIA (2014) indicates that during the world financial crisis in the latter half of 2008 and 2009, markets saw a dramatic increase in the correlation between crude oil and other commodities as demand decreased for raw materials. Additionally, both before and after the world economic slowdown, there were observable increases in the correlations between commodity prices. There was at the same time a rise in interest in general commodity exposure. A growing number of investors have gained exposure to commodities by investing in index funds-market instruments that provide exposure to baskets of commodities. The high correlations could be caused by, amongst others, economic growth expectations.

The oil prices are still the main component in the indexed formulas of the majority of long term gas contracts in Europe (CNE, 2012e; Stern and Rogers, 2011; Muñoz and Dickey, 2009; European Commission, 2011d). Table 6.2 shows the correlation of gas prices (at TTF) with crude oil prices (Brent spot and futures prices, assessed by Platts and quoted at ICE respectively).

Table 6.2. Correlation between Gas Prices (TTF, in €/MWh) and Oil Prices (Brent, $/Bbl).

BrentSpot

Brent M+1

Brent M+3

Brent M+6

Brent M+9

BrentM+12

TTF Gas Spot 0.80 0.80 0.79 0.78 0.78 0.77TTF Gas M+1 0.69 0.70 0.70 0.70 0.70 0.70TTF Gas Q+1 0.57 0.59 0.60 0.61 0.61 0.61TTF Gas Y+1 0.84 0.86 0.87 0.89 0.90 0.90




Period 3: 1 July 2009 - 30 November 2011

Total data set (from July 3 2006 to November 30 2011)

Period 1: 3 July 2006 - 29 August 2008

Period 2: 1 September 2008 - 30 June 2009

Source: Platts and ICE

Different sub-periods have also been considered in Table 6.2: (Period 1) 3 July 2006 – 29 August 2008 (as September 2008 represents the beginning of the effects of the global financial crisis on the real economy (Furió and Chuliá, 2012)); (Period 2) 1 September 2008 – 30 June 2009 (as TTF increases its liquidity since July 2009 due to introduction of “quality conversion”, permitting gas trading gas irrespective of its High or Low Calorific nature (Heather, 2012)); (Period 3) 1 July 2009 – 30 November 2011. The prompt year gas prices present the biggest correlation with oil prices (around 0.70


116

for the whole data set). In general, there is high correlation between oil and gas prices for all the sub-periods analyzed and the lowest values occur in Period 2.

The correlation between energy prices for the prompt month, quarter, and year contracts can be appreciated in Figure 6.1, Figure 6.2 and Figure 6.3 respectively. The Spanish energy regulatory authority (Comisión Nacional de Energía, CNE, currently CNMC) publishes a monthly index of the LNG import prices in Spain, based on data declared at customs (CNE, 2012d). A cointegration relationship (monthly log prices) between such an index and the series built with the last 6 month average of Brent spot prices is identified in CNE (2012e). This oil price rolling average, typically used in long-term gas contracts, is employed in the Spanish last resort tariff for natural gas. Figure 6.1 includes the LNG series.

Figure 6.1. Evolution of power (OMIP), gas (TTF), LNG import prices in Spain and emission (ICE EUA) forward prices (month maturity).

0

10

20

30

40

50

60

70

80

90

100

3-7-

2006

3-9-

2006

3-11

-200

6

3-1-

2007

3-3-

2007

3-5-

2007

3-7-

2007

3-9-

2007

3-11

-200

7

3-1-

2008

3-3-

2008

3-5-

2008

3-7-

2008

3-9-

2008

3-11

-200

8

3-1-

2009

3-3-

2009

3-5-

2009

3-7-

2009

3-9-

2009

3-11

-200

9

3-1-

2010

3-3-

2010

3-5-

2010

3-7-

2010

3-9-

2010

3-11

-201

0

3-1-

2011

3-3-

2011

3-5-

2011

3-7-

2011

3-9-

2011

3-11

-201

1

€/M

Wh

0

3

6

9

12

15

18

21

24

27

30

€/tC

O2

FTB M+1 (€/MWh) TTF M+1 (€/MWh) LNG Spain ICE EUA Dec+1 (€/tCO2) Source: OMIP-OMIClear, Platts, ICE and CNE


117

Figure 6.2. Evolution of power (OMIP), gas (TTF) and emission (ICE EUA) forward prices (quarter maturity).

0

3

6

9

12

15

18

21

24

27

30

0

10

20

30

40

50

60

70

80

90

100

3-7-

2006

3-9-

2006

3-11

-200

6

3-1-

2007

3-3-

2007

3-5-

2007

3-7-

2007

3-9-

2007

3-11

-200

7

3-1-

2008

3-3-

2008

3-5-

2008

3-7-

2008

3-9-

2008

3-11

-200

8

3-1-

2009

3-3-

2009

3-5-

2009

3-7-

2009

3-9-

2009

3-11

-200

9

3-1-

2010

3-3-

2010

3-5-

2010

3-7-

2010

3-9-

2010

3-11

-201

0

3-1-

2011

3-3-

2011

3-5-

2011

3-7-

2011

3-9-

2011

3-11

-201

1

€/tC

O2

€/M

Wh

FTB Q+1 (€/MWh) TTF Q+1 (€/MWh) ICE EUA Dec+1 (€/tCO2) Source: OMIP-OMIClear, Platts, and ICE

Figure 6.3. Evolution of power (OMIP), gas (TTF) and emission (ICE EUA) forward prices (year maturity).

0

4

8

12

16

20

24

28

32

0

10

20

30

40

50

60

70

80

3-7-

2006

3-9-

2006

3-11

-200

6

3-1-

2007

3-3-

2007

3-5-

2007

3-7-

2007

3-9-

2007

3-11

-200

7

3-1-

2008

3-3-

2008

3-5-

2008

3-7-

2008

3-9-

2008

3-11

-200

8

3-1-

2009

3-3-

2009

3-5-

2009

3-7-

2009

3-9-

2009

3-11

-200

9

3-1-

2010

3-3-

2010

3-5-

2010

3-7-

2010

3-9-

2010

3-11

-201

0

3-1-

2011

3-3-

2011

3-5-

2011

3-7-

2011

3-9-

2011

3-11

-201

1

€/tC

O2

€/M

Wh

FTB Y+1 (€/MWh) TTF Y+1 (€/MWh) ICE EUA Dec+2 (€/tCO2) Source: OMIP-OMIClear, Platts, and ICE

The jump in the emission prices during 2007 is due to the pass from the first phase of the EU ETS (years 2005-2007) to the second phase (years 2008-2012). The year and quarter gas prices are higher than the month gas prices during the summer of


118

2008, as the contract with larger maturity were more influenced by the record in oil prices (on 3 July 2008 the Brent crude oil price reached 144.22 $/Bbl for the spot price and 147.05 $/Bbl for the M+3 futures contract). The EUA Dec+2 contract is used for calculating the clean spark spreads – see Section 6.2.3 below – for the 4th quarter contracts and for all the year contracts. Therefore, Figure 6.3 shows the evolution of EUA Dec+2. For computational simplicity, the daily values for EUA Dec+2 are assumed to be equal to EUA Dec+1 during the trading sessions in December (i.e. the December month contract of the prompt year is considered in both cases). The EUA Dec+2 prices have always been bigger than EUA Dec+1 prices in the considered period (on average only 3.4% bigger and showing a correlation of 0.99, discarding the trading sessions in December, due to the assumption aforementioned, and during the period January 2, 2007 – November 30, 2007, in which the EUA Dec+1 was almost nil as previously stated).

6.2.2 Cointegration Analysis of Energy Prices

Cointegration analyses of energy logarithmic price series are performed as detailed in Section 1.6.1.3. Table 6.3 shows the results of the Augmented Dickey-Fuller’s test for the existence of unit root variables of times series for log daily prices of OMIP M+1, Brent M+1, and TTF M+1, and log monthly prices of average Brent spot prices in the last 6 months. Although this variable presents low p-values, the existence of unit root cannot be rejected. The same unit root test has been applied to the first difference of the variables in Table 6.3. In all those cases the null hypothesis of unit root existence is rejected. Therefore all the series in Table 6.3 are non-stationary series whose first difference is stationary (I(1)).

Table 6.3. Dickey-Fuller’s Test for Analysis of Unit Root Variables in Energy Log Price Series.

OMIP M+1 (daily)

Brent M+1(daily)

TTF M+1 (daily)

6 month Average Brent Spot (monthly)

Neither Intercept nor Trend -0.182 0.338 -0.14 -0.084p-values >>0.1 >>0.1 >>0.1 >0.1Only Intercept -2.65 -1.33 -2.026 -3.163p-values 0.08 0.62 0.28 0.02Trend -2.613 -1.893 -2.09 -3.369p-values 0.27 0.66 0.55 0.06Diagnosis I(1) I(1) I(1) I(1)


Table 6.4 shows the cointegration results based on unitary root analysis for the residue of the regression for OMIP M+1 as dependent variable. Three regressions are built in which the single independent variables are Brent M+1 (daily), TTF M+1 (daily) and Brent spot 6 month rolling average (monthly, thus OMIP M+1 monthly average is


119

used as a dependent variable in its regression). Cointegration relationships between daily OMIP M+1 and daily TTF M+1 and between monthly OMIP M+1 and monthly Brent spot 6 month rolling average are found.

Table 6.4. Unitary Root Analysis of the Residue in Regression OMIP M+1 versus Fuels in Columns.

Brent M+1(daily)

TTF M+1(daily)


Neither Intercept nor Trend -2.805 -4.466 -4.748p-values <0.01 <<0.01 <<0.01Only Intercept -2.804 -4.465 -4.707p-values 0.0576 0.0002 0.0001Trend -2.944 -4.708 -4.914p-values 0.1483 0.0007 0.0003Diagnosis for the Residue I(1) I(0) I(0)Diagnosis for the Variable I(1) I(1) I(1)Co-integration No Yes Yes


Table 6.5 shows the regression results for the 3 cases analyzed in Table 6.4. Only the regressions built with TTF M+1 prices and with 6 month rolling average Brent Spot prices show significant R2 statistics (0.64 and 0.93 respectively). The cointegration results explain that electricity forward prices, due to the key role of CCGT as marginal cost technology in the Spanish electricity spot price, are influenced by the gas forward prices of the most liquid reference in continental Europe, and by the oil indexed prices used in long-term gas contracts and in Spanish gas last resort tariffs.

Table 6.5. Regression Results OMIP M+1 versus Fuels shown in columns.

Brent M+1(daily)

TTF M+1(daily)


Beta 0.40 0.51 1.09Error of Beta 0.02 0.01 0.04R2 (adjusted) 0.19 0.64 0.93



120

6.2.3 Analysis of the Clean Spark Spreads with Forward Prices

The Clean Spark Spreads (CSS) serve the power sellers to analyze if the gas fired generation is profitable and help to determine the proper power and gas hedges. They also serve the energy regulators to monitor if the electricity forward prices are directly influenced by the gas prices, and in case of remarkable divergences, to check if some market anomaly has occurred – e.g. excessive derivatives speculation or market abuse – in the scope of the European Regulation on Energy Market Integrity and Transparency, known as “REMIT”, in force since December 28, 2011 (European Union, 2011). The CSS, for each maturity (M+1, Q+1 or Y+1) is obtained as the difference between the power futures price and the forward generation cost of a gas combined cycle, as expressed in Equation (1.6) and explained in Section 1.6.1.3.2.

Figure 6.4, Figure 6.5 and Figure 6.6 show the daily evolution of the resulting CSS for the prompt month, quarter and year power and gas contracts according to Equation (1.6). Figure 6.4 also shows the CSS for the Spanish LNG import price monthly index. This CSS is built assuming a price charge of 5% in terms of third party access rates (regasification and transportation to the Spanish virtual trading point) and related to OMIP M+1 series. Whereas the largest positive spreads are found for prompt month contracts, the largest – and more frequent – negative spreads occur for prompt year contracts. The influence of high oil prices on gas prices is bigger for the prompt year gas contracts, as these contracts show the largest correlation with oil prices (Table 6.2). In the 3rd quarter of 2008 and the 1st half of 2010, the CSSLNG are negative whereas the CSSM+1 are positive. The remarkable differences of both spreads indicate the strategic role of a balanced gas portfolio structure for utilities in their gas procurement through oil indexed contracts and in gas hubs. As indicated by ACER (2013b) although around half of natural gas supply in the EU is still indexed to oil, recent developments may indicate that oil indexation is on the way out. This means that apart from the spot trading purely based on gas hub pricing (traditionally known as “gas-to-gas pricing”), long term contracts tend to include a larger proportion of gas hub price index (mainly the British National Balancing Point (NBP) and the US Henry Hub (corresponding to the Louisiana area) than the oil price index, and even a proportion of electricity prices (in the case of contracts signed by power utilities with CCGT assets). Despite the increase in the spot trading (especially due to the LNG business development), long term gas contracts will survive due to their key hedging role. However, they will shift towards a more balanced fuel index structure, as oil pricing provides more certainty in cash-flow calculations for the infrastructure investors, and on the other hand gas suppliers strongly rely on gas hub pricing (Heather, 2012; Critchlow, 2014).


121

Figure 6.4. Daily evolution of CSS built with M+1 power and gas contracts and Spanish month LNG index.

-15

-10

-5

0

5

10

15

20

25

30

35

03/0

7/20

06

03/0

9/20

06

03/1

1/20

06

03/0

1/20

07

03/0

3/20

07

03/0

5/20

07

03/0

7/20

07

03/0

9/20

07

03/1

1/20

07

03/0

1/20

08

03/0

3/20

08

03/0

5/20

08

03/0

7/20

08

03/0

9/20

08

03/1

1/20

08

03/0

1/20

09

03/0

3/20

09

03/0

5/20

09

03/0

7/20

09

03/0

9/20

09

03/1

1/20

09

03/0

1/20

10

03/0

3/20

10

03/0

5/20

10

03/0

7/20

10

03/0

9/20

10

03/1

1/20

10

03/0

1/20

11

03/0

3/20

11

03/0

5/20

11

03/0

7/20

11

03/0

9/20

11

03/1

1/20

11

€/M

Wh

CSS M+1 €/MWh) CSS LNG Source: OMIP-OMIClear, Platts, ICE, and CNE

Figure 6.5. Daily evolution of the CSS built with the prompt quarter power and gas contracts.

-15

-10

-5

0

5

10

15

20

25

30

35

03/0

7/20

06

03/0

9/20

06

03/1

1/20

06

03/0

1/20

07

03/0

3/20

07

03/0

5/20

07

03/0

7/20

07

03/0

9/20

07

03/1

1/20

07

03/0

1/20

08

03/0

3/20

08

03/0

5/20

08

03/0

7/20

08

03/0

9/20

08

03/1

1/20

08

03/0

1/20

09

03/0

3/20

09

03/0

5/20

09

03/0

7/20

09

03/0

9/20

09

03/1

1/20

09

03/0

1/20

10

03/0

3/20

10

03/0

5/20

10

03/0

7/20

10

03/0

9/20

10

03/1

1/20

10

03/0

1/20

11

03/0

3/20

11

03/0

5/20

11

03/0

7/20

11

03/0

9/20

11

03/1

1/20

11

€/M

Wh

CSS Q+1 (€/MWh) Source: OMIP-OMIClear, Platts, and ICE


122

Figure 6.6. Daily evolution of the CSS built with the prompt year power and gas contracts.

-15

-10

-5

0

5

10

15

20

25

03/0

7/20

06

03/0

9/20

06

03/1

1/20

06

03/0

1/20

07

03/0

3/20

07

03/0

5/20

07

03/0

7/20

07

03/0

9/20

07

03/1

1/20

07

03/0

1/20

08

03/0

3/20

08

03/0

5/20

08

03/0

7/20

08

03/0

9/20

08

03/1

1/20

08

03/0

1/20

09

03/0

3/20

09

03/0

5/20

09

03/0

7/20

09

03/0

9/20

09

03/1

1/20

09

03/0

1/20

10

03/0

3/20

10

03/0

5/20

10

03/0

7/20

10

03/0

9/20

10

03/1

1/20

10

03/0

1/20

11

03/0

3/20

11

03/0

5/20

11

03/0

7/20

11

03/0

9/20

11

03/1

1/20

11

€/M

Wh

CSS Y+1 (€/MWh) Source: OMIP-OMIClear, Platts, and ICE

Table 6.6 shows a comparison of the clean spark spreads for the prompt month, quarter and year maturities, and for the spreads built with Spanish LNG monthly index, providing annual average values. The spreads tend to be positive, except for the year maturity, due to the existence of more frequent periods with negative spreads (Figure 6.6). The smallest spreads are presented for year contracts, followed by quarter and month contracts. Therefore gas fired generation companies can maximize profits in their forward trading with gas and power contracts of shorter maturity. The CSS built with LNG prices shows the amplest spread between minimum and maximum prices and become much smaller than CSS M+1 in 2009.

Table 6.6. Comparison of Annual Average CSS per Maturity (“M+1”, “Q+1”, “Y+1”) and CSS built with Spanish LNG Monthly Index. *Data for Year 2006 span from July 3, 2006, to December 31, 2006. **Data for Year 2011 span from Jan. 3, 2011, to Nov., 30, 2011.

Year CSS M+1 €/MWh CSS LNG €/MWh CSS Q+1 €/MWh CSS Y+1 €/MWh2006* 9.15 13.42 0.19 2.352007 14.91 11.55 13.79 6.652008 9.54 15.97 2.87 -1.962009 11.61 2.31 10.54 5.482010 4.71 0.89 5.66 3.54

2011** 4.80 5.19 3.08 -0.48Average 9.12 8.22 6.02 2.60Minimum 4.71 0.89 0.19 -1.96Maximum 14.91 15.97 13.79 6.65

Source: OMIP-OMIClear, Platts, ICE, and CNE


123

On the other hand, as found in Capitán Herráiz and Rodríguez Monroy (2010b) – and described in Section 4.3 – for OMIP power futures contracts, the forward risk premium of year contracts is bigger than for quarter contracts and even bigger than for month contracts, as the largest the maturity, more uncertainty is summed up to the forward price. In a situation of contango (forward prices bigger than spot prices), power trading with year contracts can produce larger profits for net sellers (i.e. companies whose open interest in forward markets tends to have a selling nature, e.g. generation companies).

The finding commented in the previous paragraph is reassured by the energy companies. As described by Monforte (2014), according to the largest electricity company in Spain (Endesa), the forward risk premium added to the final electricity price increases with the maturity length of the derivatives contract, as the market participants have to consider larger extra costs. Therefore Endesa suggests not employing in the CESUR auctions derivatives contracts exceeding the quarter maturity. Furthermore, this company indicates that the creation of an annual regulated end-user tariff (basing the energy cost on the prices of year derivatives contracts instead of quarter contracts) is against the recommendations of the European Energy Regulators (lobbied at the Council of European Energy Regulators, CEER) and would produce opportunity behavior (arbitrage) regarding the switching between the liberalised market and the regulated tariff amongst certain consumers. Endesa also rejects the suggestion of the Spanish energy regulatory authority (CNMC) regarding the introduction of overlapped derivatives products in the CESUR auctions (e.g a prompt month quarter and prompt half year contracts were auctioned simultaneously).

6.3 The first renewable trading and clearing mechanisms in the Iberian Electricity Forward Market

A description of the first renewable forward market trading/clearing mechanisms in the Iberian Electricity Market is provided, due to the increasing relevance of the renewable energy sources in the Iberian wholesale electricity market.

Furthermore, a review of pioneering renewable market mechanisms in Latin America is provided. The special regime is mainly composed of cogeneration, photovoltaics, concentrating solar power, wind, small hydro, biomass and solid waste. The first experience in Europe of similar mechanisms for the special regime is found in Italy, in which the public company GSE (Gestore dei servizi energetici Spa) aggregates all the special regime production (the so-called “CIP6” covering the nonconventional incentivised generation) and sell it in the market in the form of CIP6 rights as stated in Decrees of the Ministry for Economic Development. The sold amount corresponds to conservative forecasts. In year 2010, 17% of such rights are bought by the single wholesale buyer (“Acquirente Unico”) for the last resort market. The remaining 83% is offered in the liberalised market. Each quarter, a regulated price is fixed for the sold energy. The buyer establishes a contract for differences with GSE and commits himself to buy at least in the spot market a quantity equal to his CIP6 rights. If the market price is higher (lower) than the regulated forward price, the buyer receives from (must pay to)


124

GSE an amount equal to the price difference multiplied by the allocated quantity (AEEG, 2011).

In order to understand the rationale behind the existing trading/clearing mechanisms within the Iberian electricity forward market, Section 6.3.1 provides an overview of the pioneering experiences in Latin American countries. Section 6.3.2 describes the Spanish case and Section 5.3.3 the Portuguese case.

6.3.1 Renewable auctions in Latin America Maurer and Barroso (2011) describe the existing auctions for competitive

procurement of electricity from renewable generation in Latin America as well as in other parts of the world. Auctions can contribute to the development of renewable sources in a more cost-competitive and sustainable way. Auctions foster competition pushing prices down, thus reducing the end-users’ tariffs. Auctions can either consider all forms of renewables as eligible to participate in the same auction process or restrict to particular types of technologies and/or sites. Two countries are reviewed below: a brief description of the Peruvian case is provided and a more extensive description of the Brazilian case is done as it can be considered the most mature case.

Apart from the renewable auctions celebrated in Peru and Brazil, description of similar experiences in other Latin American countries –Argentina and Uruguay– can be found in Batlle and Barroso (2011). A compreheensive description of the renewable auctions in Brazil can also be found in Müller-Monteiro and Moutinho dos Santos (2010).

6.3.1.1 The Peruvian case

The implementation of technology-specific auctions has been tried twice. The first time was in 2008, for hydropower, without great success and with limited bidders. In February 2010, a similar mechanism was applied again in an auction to contract renewables under a specific law (Legislative Decree 1002) covering diverse technologies. About 150 MW of wind power were competitively contracted at prices around 80 US$/MWh. For the case of small hydro and solar, contracting of 160 MW and 90 MW respectively was possible with contract durations of 20 years and delivery for three years ahead. The average price for biomass was 63.35 US$/MWh, for wind 80.35 US$/MWh, for solar 221.10 US $/MWh, and for small hydro 59.90 US $/MWh.

6.3.1.2 The Brazilian case

A feed-in-tariff scheme was in force since year 2002. The first renewable auction took place in 2009 for wind technology. Since then, the feed-in-tariff was no longer used, but instead comparable sources should compete to achieve the ad hoc quotas for non-conventional renewables, periodically set by the government. Separate auctions were held distinguishing the renewable technology.


125

So far, auctions for biomass and wind energy have been carried out. The main advantage of conducting auctions differentiated by technology is the possibility of explicitly introducing energy policy concerns, such as the renewable penetration in the energy mix, the socio-economic development, the development of some forms of generation technology, and an easy comparison of bids. Its main disadvantages include the criteria from which the quotas for different technologies should be selected, and the fragmentation of the procurement process, with risk of reduced competition and increased costs for end-users.

The so-called “reserve energy auctions”, meeting the demand of energy distribution companies according to amounts fixed by the government and not responding to demand forecasts, increase the reserve margin, and/or foster the development of particular sources of energy, such as renewables. They are fully specified by the government, including the definition of the technology (or project) and the portion of the demand to be contracted.

Energy Vortex (2014) defines reserve margin (synonymous of reserve capacity) as a measure of available capacity over and above the capacity needed to meet normal peak demand levels. For an energy producer, it refers to its capacity to generate more energy than the system normally requires. For a transmission company, it refers to the capacity of its infrastructure to handle additional energy transport if demand levels rise beyond expected peak levels. Regulatory agencies usually require producers and transmission facilities to maintain a constant reserve margin of 10-20% of normal capacity as insurance against breakdowns in part of the system or sudden increases in energy demand.

There is no requirement for a Firm Energy Certificate (i.e. a commitment to produce a certain amount of energy, otherwise a penalty is applied) in a reserve energy auction model, and the product delivered is basically a 15-year energy contract (20 years for wind). The total cost of the energy contracted is paid by all consumers (both regulated and free) through a fixed charge. All energy produced by the plants is sold at the spot market on a merchant basis, and the revenue is used to offset the fixed payment by consumers.

The first technology-specific reserve auction for the regulated market was carried out in 2007 and only renewable energy could participate. The participation was limited and discriminatory: prospective developers preferred to sell the energy to large end-users as they were eligible for discounts on the use of the transmission and distribution system and could therefore pay higher prices (discriminatory).

On the other hand, the fact that generators do not need a Firm Energy Certificate mitigates several risks, making those auctions very attractive for generators, which are basically selling their production for a fixed price. There is a natural production synergy between hydroelectric and biomass-fired electricity generation: the energy produced by biomass power plants during the dry season is more “valuable” because wholesale market spot prices are higher during that season than the annual average. The same counter-seasonal production behaviour is observed for wind plants, whose production pattern is complementary to hydro storage levels in some parts. This counter-cyclical characteristic represents a significant competitive advantage to


126

renewable sources: the benefits of portfolio diversity partially offset the higher unit costs of those technologies.

Regarding site and technology-specific auctions, project specific auctions to supply the regulated market have been carried out to develop large hydro plants in the Amazon region. Three hydro plants – Santo Antonio (3,150 MW), Jirau (3,300 MW), and Belo Monte (11,233 MW) – were auctioned in specific procurement processes during the years 2007, 2008, and 2010, respectively. Special financial conditions were created (e.g. granting of 30-year energy contracts, incentives for the formation of multiple consortia, tax incentives, et cetera).

The site and technology-specific auctions present a trade-off when there are few bidders: the government has to strive either to attract one more bidder with the expectation of pushing prices down, or to reduce the reserve price (the maximum price (i.e. a price cap) permitted by the auction administrator). The reduction in the reserve price could happen when only one bidder participates in the tender. In the case of Santo Antonio, the government’s efforts to find an alternative bidder apart from a well-positioned consortium was succesful compared to control the price via reserve prices: the winning bidder price was 78.9 R$/MWh, significantly lower than the 130 R$/MWh original estimate by the consortium that had carried out the initial pre-feasibility studies. The reserve price set by the government, and corresponding to the best available cost estimate for the site in question, including a prudent return on capital, was 122 R$/MWh.

6.3.2 The first mechanisms in the Iberian electricity market

6.3.2.1 The Contract for Differences derived from CESUR auctions in Spain

As indicated in CNE (2011b), the Royal Decree 302/2011, of 4 March 2011 regulates the sale of products to be settled through price differences by the special regime facilities with a regulated tariff scheme and the purchase by the last resort suppliers. Therefore this mechanism aims to establish a compulsory purchase mechanism for the last resort suppliers and compulsory sale mechanism for those special regimes facilities of products with price differences settlement between CESUR prices and the spot prices.

The special regime facilities considered in the Royal Decree 302/2011 are those choosing option 1.a) in the article 24 (Mechanisms for the retribution of the electricity produced through the special regime) of Royal Decree 661/2007 (MITyC, 2007). This Royal Decree regulates the special regime generation in the Spanish electricity market. The article 24 considers two modes of feed-in-tariffs, i.e. tariffs for the remuneration of the energy generated by a special regime facility. If the generation facility chooses option 1.a) then it receives a regulated tariff (fixed amount) per kWh generated; with option 1.b) the recognised price is the spot price plus a fixed premium.


127

Of the 59,211 installations registered in September 2011 as special regime facilities at the Spanish Energy Commission (CNE, currently CNMC) – the National Regulatory Authority manages the settlement of the feed-in-tariff scheme – 57,928 followed the regulated tariff mode and 1,283 the premium mode. The recognised amounts for the sold energy that month reckoned 430 million € for the regulated tariff and 115 million € for the premium facilities (CNE, 2011d).

Leyton (2010) provides a review of the feed-in-tariff implementation schemes around the world in order to open a dialogue for the implementation of similar initiatives in the Chilean electricity market. Such a reflection exercise is motivated by the Law 20.257, published on April 1, 2008, obliging the power producers with an installed capacity bigger than 200 MW, to supply 10% of energy from renewable non-conventional or hydro power with capacity less than 40 MW, own or contracted, since January 1, 2010 (BCN, 2010).

Back to the Spanish case, the maximum compulsory volume to consider in the contract for differences is obtained through the difference between the sum of the quantities requested by the last resort suppliers during the period in force of the last resort rate (a natural quarter) and the quantities matched in the corresponding CESUR auction (the auction celebrated some days before the beginning of such a quarter). This mechanism reduces the last resorts suppliers’ risk, as it lets them purchase all the requested energy at the same cost (the equilibrium price of the CESUR auction). It is important to remark that the contract for differences are established with the real production of the special regime: in case that the production would be less than the difference between the last resort suppliers’ requested amounts and the matched amounts in the CESUR auction, a part of the purchases of the last resort suppliers would not be hedged neither through the CESUR auction nor through the contract for differences mechanism.

In other words, the contract for differences mechanism permits to transfer to the end-user the price differences between the equilibrium prices of the CESUR auctions and the spot prices for the energy requested by the last resort suppliers and not purchased in the CESUR auctions. Since the entry in force of the last resort supplies (July 2009), the ex-post forward risk premia between the CESUR equilibrium prices and the underlying spot prices is usually positive, see e.g. Villaplana and Cartea (2011; 2012) and Monforte Martín (2011). By means of the contract for difference mechanism the end-users can benefit from such a difference to mitigate the tariff deficit caused, among others, by the massive deployment of the special regime subject to attractive remuneration prices through the existing feed-in-tariffs. The solution is quite smart, as the affected special regime facilities will receive in any case the whole regulated tariff, and the end-users, in the cases when CESUR prices are higher than the underlying spot prices, although forced to face an increase in the last resort rate, will contain the tariff deficit due to the rents generated by the contract for differences mechanism. When the forward risk premium is negative (i.e. CESUR equilibrim price less than the spot price), the last resort suppliers will receive such a price difference (as a natural hedging at the CESUR equilibrium price) but in this case the end-users could as well benefit from smaller last resort rates due to lower CESUR prices. The spot market


128

operator (OMIE) acts as a clearing house facilitating the settlement of the contract for differences mechanism.

However, as the counterpart supporting the price differences for the special regime with regulated tariff is in fact the collective group of end-users, the design of the contract for differences mechanism could be improved (maximization of profits) considering:

• On the purchase side, not the requested demand by the last resort suppliers but the real demand (i.e. physically measured) when exceeding the requested demand (prior to each CESUR auction).

• On the sale side, not only the real production of the special regime with regulated tariff but the whole generated amount (by any technology, not necessaily renewable) to cover the difference between the total demand of the last resort suppliers and their purchases in the CESUR auction. In case of insufficient generation by the special regime, the rents generated by the other technologies selling that remaining part in the spot market could be settled equally and help to diminish the tariff deficit when the forward risk premium is positive. Note that the objective of this mechanism is to find this alternative way to finance the regulated tariff of the special regime, and any generation unit acting as a counterpart of the last resort suppliers is representing artificially the Spanish power system and thus the collective group of end-users, being the generated rent not received by any producer but entirely destined in an anonimous way to the financing of the special regime retributed with specific regulated tariffs for each technology.

6.3.2.2 The auctions for the sale of the special regime production in Portugal

On December 16, 2011, the first auction related to the Special Regime production was celebrated in Portugal. In these auctions, the Portuguese last resort supplier (EDP Serviço Universal, S.A.), who is in charge of managing all the purchase and sale of the Portuguese special regime energy, sells all the energy from the special regime. If the whole energy is not sold, it will automatically repurchase that remaining amount at the auction equilibrium price (ERSE, 2011b; 2011c).

These auctions contribute to develop the electricity market, as they facilitate the access to the energy via market mechanisms to the suppliers in the liberalised market as well as to the new entrants. They are a hedging tool for the price risk of such agents. They permit the market risk diverisification in the allocation of the special regime energy and mitigate the price volatility of its exclusive integration in the spot market. These auctions facilitate a level playing field for agents with or without own generation assets in Portugal. They auctions permit the stabilitiy of the last resort supplier’s cash-flow. The differences in the purchase price to the special regime producers by the last resort supplier and the sale price in these auctions are recognised as a regulated cost for the last resort supply. Therefore the end-users support the global cost of the special regime production (ERSE 2011a; ERSE 2011b).


129

Volumes on top of the issued ones cannot be matched in these auctions. ERSE, the Portuguese National Regulatory Authority, as the auction supervisor, has to publish not later than 2 days after the auction the equilibrium price and the matched amounts, the eventual amount repurchased by the last resort supplier, the amount of qualified agents, the buyers, and the number of rounds. OMIP is the auction administrator and OMIClear is the clearing house. The auctions are electronic, with multiple rounds and ascending price clock algorithm. Different products can be auctioned simultaneously without possibility to migrate bid blocks from one product to other products. It is mandatory to be a trading member in OMIP, and thus hold collateral in the futures market. ERSE communicates 10 days before the auction the quantities and the contracts (baseload and/or peak futures of the Portuguese zone with month, quarter and/or year maturity) (ERSE, 2011b; OMIP, 2011c).

In the first auction, 300 MW were sold (the total offered amount, 1,315,000 MWh in terms of energy, equivalent to 2.6% of the Portuguese mainland demand). The pre-qualification bids reckoned 7.15 times the offered amount. Eleven companies submitted initial bids, totalling 9,403,290 MWh (ca. 19% of the Portuguese mainland demand). The average equilibrium price was 53.12 €/MWh, aligned with the settlement prices of the previous OMIP trading session, and 2.00 €/MWh (year contract) and 1.85 €/MWh (quarter contract) above the reserve prices regulatorily fixed (i.e. the price floors). For the quarter contract (baseload with time horizon the second quarter of year 2012), 200 MW (436,600 MWh) were matched by 5 of a total of 10 bidders at 53.35 €/MWh in the fifth round. For the year contract (baseload with time horizon the year 2012), 100 MW (878,400 MWh) were matched by 5 of a total of 11 bidders at 53.00 €/MWh in the sixth round (ERSE, 2011d).

The last PRE auction held in year 2013 dated 12 December 2013 and was the 9th one, presenting, as in the previuos auctions, a good degree of participation. So far, the PRE auctions have been held in March, June, September, and December (i.e. the same frequency as in the case of CESUR auctions). The key regulation, as well as the specific terms, conditions, and results of each auction can be consulted in ERSE (2014).

In order to foster the liquidity of the futures products of the Portuguese zone, these auctions should be celebrated more regularly, at least once a month in order not to concentrate the negotiation in few points of time (case of quarterly frequency). As a next step in the integration of the special regime through forward market mechanisms, in order to provide a competitive price, especially when some specific technology reaches the grid parity (its price is competitive in the market without need of feed-in-tariff support schemes), it can make sense to directly arrange auctions in which these renewable producers sell the energy to the last resort supplier for the regulated market and to the suppliers for the liberalised market.


130

6.4 Results

The power futures are strongly correlated with the European gas prices, especially between the year contracts, as far as the renewable forecasts – a key signal for including renewable generation costs in the electricity price formation – are currently accurate just on the very short term. Results from cointegration analysis for the Iberian prompt month power futures contracts with gas and oil prices show that such electricity prices are influenced by gas forward prices of the most liquid reference in continental Europe (TTF), and by oil prices (6 month rolling average of Brent spot prices) used in long-term gas contracts and in Spanish gas last resort tariffs.

The clean spark spreads for prompt futures contracts, obtained as the difference between the power futures price and the forward generation cost of a combined cycle gas turbine, tend to be positive. The biggest spreads are for the month contract, followed by the quarter contract and then by the year contract. Therefore, gas fired generation companies can maximize profits trading more frequently with electricity contracts of shorter maturity.

Both the forward risk premium and the clean spark spreads are key indicators for market participants to optimize their trading strategies – they can exploit arbitrage gains through trading with contracts of different maturity and perform cross-market hedging operations – and for regulatory agencies to monitor the fair price formation in the wholesale energy markets, that ultimately affects the end-user prices.

Further research is encouraged taking into account the influence of the dark spreads (obtained as the difference between the power futures prices and the generation costs of a coal power plant, the latter calculated e.g. through coal futures contracts traded at the European Energy Exchange, EEX) to assess the influence of the coal generation – out of the regulatorily fixed prices – in the Iberian power forward price formation. A comparative analysis of both forward risk premia and spreads with other European markets (e.g. French, German or Nordic) could be used to measure their different efficiency levels. As the research performed is based on prompt base load contracts, the analysis could also be extended to further maturities (i.e. not immediately close to delivery) and for peak products.

There are currently two market mechanisms related to the forward integration of the special regime (i.e. renewable generation and cogeneration) in Spain and Portugal. In the Spanish case, a contract for differences mechanism is available in Spain since March 2011 between the last resort suppliers and the special regime facilities –those subjected to a regulated tariff for the retribution of their energy produced– settling the price differences between the equilibrium price of the forward regulated auctions for the last resort supply and the spot price. This mechanism helps to contain the tariff deficit by means of the rents generated when the CESUR equilibrium price is bigger than the underlying spot price. The design of this mechanism could be optimised (maximisation of profits) considering on the purchase side the real demand by the last resort suppliers when exceeding their requested demand in advance, and on the sale side the whole generated amount by any technology to cover the difference between the total demand of the last resort suppliers and their purchases in the CESUR auction. In the


131

Portuguese case, regulated auctions of baseload futures of the Portuguese zone in which the last resort supplier sells the special regime production exist since December 2011. In order to foster the still small liquidity of the futures products of the Portuguese zone, these auctions should be celebrated regularly, at least once a month in order not to concentrate the negotiation in few points of time.

Renewable auctions as those from the pioneering experiences in Brazil and Peru could be considered as the next step in the integration of the Iberian special regime through forward market mechanisms. Therefore, introduction of auctions at least for the most mature renewable technologies, providing a fair price for this type of generation, would help to diminish the tariff deficit caused by the massive deployment of the feed-in-tariff scheme. The introduction of renewable auctions for specific technologies would entail the gradual suppression of the existing feed-in-tariff scheme for such a technology (i.e. the new infrastructure would be built according to new auction programs substituting the former scheme). These auctions should be designed in a way that regulatory uncertainty is minimised. In this sense, successive modifications of the existing feed-in-tariff scheme have produced much uncertainty in the energy sector, damaging the investment climate, see e.g. García Breva (2012). As a natural consequence, liquidity in the forward markets will also increase as such renewable generation companies would try to maximize their profits by trading actively in the forward markets, as they could no longer benefit of the full hedge provided by the previously existing feed in tariff schemes.

Further research is encouraged to simulate the economic effects and the forward price formation in the Spanish power system due to the introduction of renewable auctions substituting the existing feed-in-tariff scheme. Regarding the economic effect, evaluation of the potential reduction of the accumulated tariff deficit will be very worthy. Regarding the price formation, analysis of the forward price formation taking into account reasonable prices of those potential renewable auctions as well as the generation costs (including the CO2 emission costs) of the thermal power plants – the gas and coal fired power plants acting as an effective back-up in the absence of windy days – would shed light about the most convenient assessment methodology for the energy costs in the last resort rates. As these rates are used as benchmark for the end-user rates in the liberalised market, social welfare would be obtained by improving the rate design including the penetration of renewable generation.


132


133

CHAPTER 7. EVALUATION OF THE HEDGING PERFORMANCE BASED ON OPEN INTEREST AND CLEARED VOLUMES

CHAPTER SUMMARY

An assessment of the hedging performance in the Iberian electricity forward market is

performed. Aggregated data from the Portuguese and Spanish clearing houses for

energy derivatives are considered. The hedging performance is measured through a

net position ratio obtained from the final open interest of a month derivatives contract

divided by its accumulated cleared volume. The base load futures in the Iberian energy

derivatives exchange show the lowest ratios due to good liquidity. The peak futures

show bigger ratios as their reduced liquidity is produced by auctions fixed by

Portuguese regulation. The base load swaps settled in the clearing house located in

Spain show initially large values due to low registered volumes, as this clearing house

is mainly used for short maturity (daily and weekly swaps). This ratio can be a powerful

oversight tool for energy regulators when accessing to all the derivatives transactions

as envisaged by European regulation.

7.1 Introduction

This chapter describes the evolution of the Iberian power futures market but focusing on another key efficiency parameter different to the forward risk premium and the traded volume: the open interest. The Chapter is based on the research performed by Capitán Herráiz and Rodríguez Monroy (2011b, 2012b, 2013b). It comprises the first five and a half years of this power futures market (i.e. the data set spans from the start of that market on 3 July 2006, until 31 January 2011). It aims to identify hedging patterns of the month contracts cleared and settled in OMIP clearing house (OMIClear) to shed light on how market participants manage their price risk and provide recommendations to the authorities in charge of the MIBEL supervision. Data cleared and settled in the energy derivatives clearing house located in Madrid, MEFF Power (currently BME Clearing), operating since March 21, 2011 are also considered. “MEFF” stands for Spanish Financial Futures Market.

The Iberian Electricity forward market is one of the most dynamic European energy markets, according to its growing traded volume rates. Due to its recent nature (the OTC market started in 1999 and the futures market in 2006) not much literature has been published analysing the evolution of the Iberian forward prices, the traded volumes and the open interest. This chapter provides a first approach to the relation


134

between the cleared volumes in the existing clearing houses and their open interest. One important feature of this market is its financial nature (in the rest of Europe the forward trading is essentially physical, with the exception of the Nordic countries). Therefore research about measurement of hedging and speculation trends in this emerging market is of paramount interest.

The hedging performance is tracked through the evolution of the net position ratio composed of the final open interest of each month contract divided by the accumulated cleared volumes in the clearing house. Month contracts are considered as their trading activity is good. The final open interest corresponds to the open positions at the end of the trading period of the contract. The accumulated cleared volume reflects the whole recorded volumes of the month contract (and the corresponding part of the quarter and year contracts containing that month) for clearing and settlement in the clearing house. The market participants can close positions during the trading period to limit their risk exposure or just for obtaining trading gains (i.e. arbitrage due to purchases at a lower price and sales at a higher price). The open positions will determine a hedge (i.e. a fixed price for the delivered energy) if the market participant performs the same operation in the spot market. The open position is cash-settled during the settlement period against the evolution of the daily spot price. This exposes the market participant to potential losses or gains in case he did not perform a similar operation in the spot market.

This analysis serves to identify how the Iberian electricity forward market is performing according to its original role as key hedging vehicle. Additionally, this research suggests further improvements to the design and application of this kind of ratios as a powerful market monitoring tool helping the supervisory agencies to detect and analyse suspicious trading behaviour bound to market abuse practices. The literature review – about stock and commodities markets, describing amongst other issues the traditional indicator in financial literature known as hedge ratio – and the description of the tests performed in this research can be consulted in Section 1.6.3.

The chapter is structured as follows: Section 7.2 presents the analysis of the net position ratio for the Iberian electricity derivatives; Section 7.3 provides reflections about the usefulness of the net position ratio in the prudential oversight of the systemic risk; Section 7.4 summarises all the insights of the research, suggests further lines of research and concludes.

7.2 Analysis of the net position ratio of the Spanish electricity derivatives

OMIP-OMIClear and MEFF Power final open interest (i.e. the value published for the last session in which a given contract is quoted) are studied for each month contract. Only Spanish month contracts are considered: OMIP-OMIClear base load and peak futures and MEFF Power base load swaps. The division of the final open interest and the accumulated cleared volume is used as a net position ratio to measure the potential interest of the traders in these contracts for risk management (i.e. hedging by means of final open positions, in case they trade the same amount with the same


135

nature afterwards in the spot market (e.g. spot purchases in case of long derivatives positions)).

Figure 7.1 shows the evolution of the final open interest for each month derivative divided by the total cleared volumes of such a derivative product (including the corresponding cleared volumes for that delivery month from the quarter and year products). The first month considered is August 2006, as OMIP started on 3 July 2006. Until July 2008, the net position ratio of the base load futures contracts fluctuates in a narrow spread (0.7-1.0) due to the fact that the final open interest is almost equal to the auction volumes and that both the continuous trading and the registration of OTC trades for central clearing and settlement were scarce. As practically all the auction volumes were due to compulsory purchases of the Spanish distribution companies and the Portuguese last resort supplier established by their respective national legislative pieces, no other hedges were established out of such compulsory trades. Afterwards, the volumes from compulsory auctions became smaller but the continuous volumes and the OTC registered volumes in OMIP-OMIClear grew, as shown in Figure 3.2. Due to that change in the nature of the trading activity, the net position ratio diminishes and oscillates since the beginning of year 2009 in a wider spread (0.1-0.5). The smallest ratio would indicate that the market participants tend to close positions for profit taking due to price fluctuations in the futures prices along the trading period of the contracts, minimising the portion of trades for hedging purposes.

Figure 7.1. Evolution of OMIP-OMIClear and MEFF Power net position ratio per

delivery month.

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0

Aug

.06

Sep

.06

Oct

.06

Nov

.06

Dec

.06

Jan.

07Fe

b.07

Mar

.07

Apr

.07

May

.07

Jun.

07Ju

l.07

Aug

.07

Sep

.07

Oct

.07

Nov

.07

Dec

.07

Jan.

08Fe

b.08

Mar

.08

Apr

.08

May

.08

Jun.

08Ju

l.08

Aug

.08

Sep

.08

Oct

.08

Nov

.08

Dec

.08

Jan.

09Fe

b.09

Mar

.09

Apr

.09

May

.09

Jun.

09Ju

l.09

Aug

.09

Sep

.09

Oct

.09

Nov

.09

Dec

.09

Jan.

10Fe

b.10

Mar

.10

Apr

.10

May

.10

Jun.

10Ju

l.10

Aug

.10

Sep

.10

Oct

.10

Nov

.10

Dec

.10

Jan.

11Fe

b.11

Mar

.11

Apr

.11

May

.11

Jun.

11Ju

l.11

Aug

.11

Sep

.11

Oct

.11

Nov

.11

Dec

.11

Jan.

12

Rat

io F

inal

Ope

n In

tere

st /

Acc

umul

ated

Cle

ared

Vol

umes

MWh_O I/MWh_Auction+C ontinuous+O TC cleared O MIC lear B aseload F uturesMWh_O I/MWh_Auction+C ontinuous+O TC cleared O MIC lear P eak F uturesMWh_O I/MWh_cleared ME F F P ower Month S waps

Source: OMIP-OMIClear (2012) and BME (2012) adapted by authors


136

In the case of OMIP peak futures, the series is composed of few values, as there are only cleared volumes for the months within the period February 2010–August 2010, and June 2011–July 2011. In the first period, the 97.6% of those volumes corresponded to auctions, 1.5% to continuous trading, and 0.8% to OTC registered volumes. The net position ratio of the peak futures fluctuated in a narrow spread (0.7-1.0) as almost all the transactions corresponded to hedges performed by the Portuguese last resort supplier in auctions for the purchase of regulatorily fixed amounts. The maximum ratios (1.0) in June 2011 and July 2011 correspond to single OTC registered transactions of those month contracts, due to the scarce liquidity of the peak contracts out of the compulsory auctions. An ampler data set is required to detect if these contracts experience a liquidity growth tending tower smaller net position ratios, as the more developed base load futures have shown due to the agents’ learning curve, which can trade more dynamically and maximize their portfolio returns.

Regarding the base load swaps cleared in MEFF Power, the ratio decreased sharply in January 2012 to 0.07 due to the fact that increasing volumes were registered, as shown in Figure 3.2. The series began in June 2011, as it was the first month swap with registered volumes. The months of July and August 2011 presented the maximum ratio (1.0) due to the minimal registration of contracts in the first months of the clearing house. There is only 1 transaction for the quarter swap with delivery period the third quarter of 2011, and 3 transactions for the month swap with delivery period August 2011. The accumulated volume for July-2011 only considers the corresponding volume of the quarter contract (Q3-11). The accumulated volume for August-2011 considers the volumes from the month contract and the fraction from the quarter contract (Q3-11). The accumulated cleared volumes coincide in both cases with the final open interest as no market participant closed some of these few open positions. Afterwards, the ratio presents a descending seesaw oscillation. A larger data set is required to draw stronger conclusions about the registration activity in this relatively new clearing house as the trend for the base load month swaps has been erratic so far (the liquidity of that clearing house at its start have been more concentrated on short term swaps (daily and weekly)).

7.3 The net position ratio and the prudential oversight of the systemic risk

Current financial reforms in both the United States and the European Union – the regulatory developments derived from Dodd-Frank Act and the Regulation on European Market Infrastructures (EMIR) respectively – consider the introduction of position limits in the commodities derivatives trading to counter disproportionate price movements or concentrations of speculative positions (U.S. Government, 2010; European Union, 2012). Such considerations stem from conclusions derived in the G-20 summit celebrated in Pittsburgh in September 2009, in order to prevent the systemic risk of the global financial market (European Commission, 2009c; IOSCO, 2011).

According to the current levels of trading volumes and prices in the Spanish OTC power market (CNE, 2011b, 2012b; CNMC, 2014b), assuming a yearly traded


137

volume of 300 TWh (300,000,000 MWh) at an average price of 50 €/MWh (i.e. the price level of the Spanish power futures market, very close to the price level of the spot market in the months with weakener renewable production and the price level of recognised prices for power plants burning indigenous coal), such a market would produce a turnover of 15 billion €. If the trading patterns observed in the clearing houses (i.e. net position ratios around 0.1 at the beginning of year 2012) were equally replicated in the dominant OTC market, only 1/10 of the turnover (1.5 billion €) would be exposed to the credit risk due to default of one counterpart (i.e. due to open positions assuming an OTC net position ratio built as the OTC open position divided by the total OTC traded volumes). Therefore, a steady low net position ratio in the OTC trading would mitigate the potential impact on the systemic risk. Were these figures right (so far only mere assumptions according to aggregated post-trade data published by the Spanish Energy Regulator (CNE, currently CNMC) in its monthly supervision reports of the Spanish electricity forward market and annual reports to the European Commission), the market participants would not be taking exaggerated open positions. In that sense, the probability of the feared systemic risk would be somehow moderate, even without the implementation of position limits.

7.4 Results A novel approach to measure the hedging performance in the Iberian electricity

forward market is employed, rather than the traditional method of the hedge ratio widely used in financial markets. A net position ratio is estimated with aggregated data from the Portuguese and Spanish clearing houses (OMIClear and MEFF Power (whose current name is BME Clearing) respectively) through a ratio obtained as the final open interest of a month derivatives contract divided by the accumulated cleared volume for that delivery month (i.e. the month contract and the corresponding part of the quarter and year contracts). The power futures base load contracts traded in the Iberian energy derivatives exchange show the lowest ratios due to their good liquidity. Since the beginning of year 2009, those ratios are less than 0.5 (around 0.1 at the beginning of year 2012), indicating that less than half of the cleared volumes can be used for hedging and the rest of cleared volumes correspond to positions being closed. The market participants closing such positions can benefit from price differences in the purchase and sale operations centrally cleared by OMIClear. They can also benefit through the portfolio trading based on the exploitation of the price differences (e.g. arbitrage) between the coexisting market mechanisms in the Iberian electricity forward market (namely, OMIP power futures market, CESUR auctions and OTC trading). The power futures peak contracts show bigger ratios as almost all the cleared amount correspond to hedges of the Portuguese last resort supplier in OMIP auctions in which this company has to purchase quantities regulatorily fixed. A larger data set is required to draw robust conclusions for the base load swaps registered in BME Clearing, as this clearing house started operations on 21 March 2011.

As the net position ratio for January 2012 is very low in both clearing houses (0.07 in BME Clearing and 0.13 in OMIClear), further research is suggested with a larger data set along year 2012 to detect if lower values respond to a change in the


138

trading behaviour. As the registration of short maturity contracts is becoming more frequent in both clearing houses – daily and weekly, and weekend in a less extent –, further research is encouraged to analyse if the net position ratios of those contracts are bigger, responding to hedging needs of the market participants close to the delivery period, or smaller as they could be used to close positions from the month contracts, either to adjust their hedging needs according to physical delivery commitments or to exploit existing price differences between the different maturities (arbitrage gains).

The ratio employed in this research is actually a straightforward static indicator, as it is built on the final open interest. A more complex and powerful dynamic indicator would consider the evolution of the open interest for all the trading sessions, not the last one. That daily ratio would help to better understand the trading patterns.

Further research comparing the daily evolution of the net position ratio with the daily evolution of the hedge ratio would provide powerful insights to grasp the hedging trends in this emerging market. Both ratios are not directly interrelated as the former depends exclusively on futures data and the latter depends as well on spot data. However, assuming for both cases that derivatives cleared volumes and spot volumes remained constant, the evolution of both ratios were led by the development of the net position (i.e. the open interest).

Although the global net position ratio shown in this research is a useful indicator of the trading development, it has some limitations. First, it only provides a partial snapshot of the trading, as the majority of the opaque OTC trading is still not centrally cleared and no aggregated data is published. Second, the hedge only exists if the agent finally performs the same operation in the spot market. Additionally, the net position ratio would be different per market participant type and the reasonable level of this indicator would only be useful per market participant type (not a global value as presented in this research). In the case of an energy company intending to hedge its generation or supply, the net position ratio would tend to present higher values. In the case of a financial entity without hedging goals and non-highly leveraged, its ratio would tend to present lower values (what minimizes its credit default risk). In case the net position ratio of any participant grew substantially affecting the price volatility, the supervisory authorities could monitor if the integrity of the whole system is in danger and take appropriate enforcement measures.

A more sophisticated design of the net position ratio (i.e. with daily values, distinguishing per market participant type and permitting screening for one agent) and the hedge ratio are suggested to be employed by energy regulators devoting resources to market oversight of wholesale (spot and derivatives) energy markets and by the Agency for the Cooperation of Energy Regulators (ACER). These authorities as well as other competent authorities – the national financial and competition authorities and the European Securities and Markets Authority (ESMA) – will have access to energy derivatives transaction details according to EU Regulation on Energy Market Integrity and Transparency (REMIT), in force since December 28, 2011 (European Union, 2011). The obligations for the market participants to submit their transaction records to ACER, that will grant data access to the energy regulators, will begin once the details of the transaction reporting are published in a European Commission Implementing Regulation (ACER, 2013). The use of daily net position and hedge ratios by the market


139

monitoring authorities, filtered per market participant type (utility, financial entity, dominant operator, new entrant, etc.) could help to detect any suspicious behaviour regarding market abuse. Publication of aggregated statistics as those shown in CFTC Commitment of Traders’ Reports would increase the ex-post trade transparency, enhancing the efficiency of the European wholesale energy markets.

The energy regulators can also access to OTC data by own initiative, if stated in their national law. For instance, the Spanish Energy Commission (CNE, currently CNMC) has access to limited information over OTC power transactions, namely volumes and anonymous (i.e. with no disclosure of the counterparts) transaction prices, through the information voluntarily submitted by the main brokers CNE (2011b). On 5 March 2011, the Law of Sustainable Economy was published in the Spanish Official Gazette (BOE). The fifth final disposition of this Law modifies the Securities Market Law, enabling the information exchange between CNMV (the Spanish Financial Services Authority) – CNMV is the authority empowered to request OTC power data in the Spanish financial market – and rest of entities belonging to the MIBEL Regulatory Council. As indicated in Section 3.5.2, this Council is composed of CNE, CNMV, ERSE (Portuguese National Regulatory Authority) and CMVM (Portuguese Financial Services Authority). Apart from the Multilateral Memorandum of Understanding (MoU) for the cooperation and efficient coordination in the MIBEL supervision, of 17 May 2011, permitting their coordinated OTC supervision, facilitating the data collection, CNE and CNMV have signed on July 3, 2012, a collaboration agreement in the framework of the supervision of energy forward markets, CESUR auctions and REMIT (MIBEL Regulatory Council, 2011; CNE and CNMV, 2012).

As a result of the enhanced cooperation between CNE and CNMV, the former published a report related to the supervision of the trading in OMIP and OTC and its effect on the 15th and 16th CESUR auctions supervised by CNE (CNE, 2012g). The main conclusion of such a pioneering report within the REMIT framework is the absence of suspicions regarding market manipulation in the months around the CESUR auctions held in June and September 2011. Further reports analysing the forward price formation in the Spanish electricity – and even gas – auctions would be very worthy supervisory works in the REMIT scope. The use of daily net position and hedge ratios per individual market participant would enrich substantially the analysis and insights of such reports and strengthen the daily market monitoring practices, helping to implement proper regulation in case market failures were detected.

Finally, a dynamic analysis of the net position and the hedging ratios could be used as well by policy makers to assess the convenience and estimate the proper magnitude of position limits. These limits are envisaged for the European and North American commodity derivatives markets, in order to avoid excessive speculation impacting on prices (Chilton, 2012b).


140


141

CHAPTER 8. RESULTS, CONCLUSIONS AND FUTURES LINES OF RESEARCH

8.1 Regulatory recommendations

The following conclusions summarize the main findings of the research and state regulatory proposals for the efficient development of the Iberian energy forward market:

1 - The Iberian power futures market managed by OMIP could be used for the last resort suppliers’ purchases together with the spot market. A scheme of auctions held every week with prompt quarter and prompt month contracts with tight oversight of the settlement price formation could be employed. The last resort rate (currently called “Voluntary Price for the Small Consumer”) energy price component could be composed as follows: 40% related to the weighted average price of futures prices and 60% related to the average spot price for the last three months period. New market players could be attracted to the futures market, increasing its liquidity and efficiency.

According to the experience gained from OMIP call auctions for distribution companies between years 2006 and 2009, the equilibrium price in such auctions was not optimal for remuneration purposes as the purchasing costs for regulated supplies tended to be slightly higher than those from OMIP average settlement prices along the whole quotation period.

A regulated cap price for each OMIP compulsory call auction could then be transitorily applied in order to obtain a lower equilibrium price and therefore diminish regulated costs of supply. Once OMIP continuous market had acceptable liquidity, the settlement price itself would reflect more accurately the market prices and could be used for evaluating the cost of last resort supplies. In that situation, each last resort supplier could trade directly in the continuous market.

However, as the price of the last resort supply should be unique for the 5 last resort suppliers (currently called reference suppliers, due to the electricity reform approved by the Spanish Government in December 2013), an auction scheme is deemed as more appropriate. OMIP auctions for the last resort supplies could substitute the auctions scheme for the last resort supplies held from July 2009 until December 2013 (such auctions were known as CESUR auctions).

Those CESUR auctions were held every three months. The transaction costs in the new substituting scheme, through OMIP platform, would be lower (the fixed membership fees and the variable trading and clearing fees charged by OMIP and its clearing house OMIClear could be more reduced than the costs charged by CESUR


142

auction administrator, OMEL Mercados, a subsidiary company of the Spanish market operator OMIE).

The optimal scheme design of such auctions in OMIP would consider prompt quarter contracts as in the CESUR scheme, as well as prompt month contracts (i.e. the three month contracts composing the prompt quarter), but held more frequently, e.g., once a week. This would avoid the concentration of liquidity in few trading sessions, and the potential threat of market abuse from excessive speculation in the spot and/or futures market in the weeks preceding a single last resort supply auction.

Market monitoring by the Iberian Electricity Market (MIBEL) national regulatory authorities coordinated with OMIP and OMIE (both futures and spot market operators acting through separate market monitoring units), could detect any anomaly in the daily price formation of OMIP settlement price and the spot price due to the potential of market manipulation in both interrelated market segments mentioned above.

In order to control such anomaly, ad-hoc price variation limits in the daily quotations in the futures market could be applied, once such a transitory measure is agreed between the national regulatory authorities and the futures market operator.

The coordination in the daily oversight of the price formation would be needed for both Spanish and Portuguese regulatory authorities, the former for the oversight of the last resort supply auctions and the latter for the oversight for the auctions for the sale of the Portuguese Special Regime generation (known as PRE auctions).

The Iberian power futures market would increase its liquidity from the last resort supply auctions, and more agents would enrol that market attracted by such trading opportunities. Hence, the efficiency of that market could increase. Nevertheless, such last resort supply auctions should have a limited life, as the end-users should not depend in the long term on regulated prices (such regulated tariffs should apply only to very particular cases and duly justified of end-users having problems to access to retail offers) but access to competitive supplies in the liberalised market instead.

Finally, as the Iberian electricity forward market tends to introduce a forward risk premium mainly due to both the regulatory and macroeconomic uncertainty, and the lack of accurate renewable forecasts for the medium and long term (i.e. not projected in the forward price forecasts except for the very short term, 24-48 hours), the Spanish Government should compel the last resort suppliers to purchase 40% of the forecasted energy through the OMIP auctions from prompt month and quarter contracts, and the remaining 60% through the day-ahead market.

The resulting energy cost to be considered in the last resort rate would be a weighted average of the energy bought in each market segment according to the ratio 40/60 above suggested. This 40/60 scheme would work well for potential situations of high renewable penetration (very low spot prices) and eventual supply shocks (see e.g. the “perfect storm” suffered in the Spanish power pool in the first half of December 2013 due to nuclear outages, thermal power plants unavailablity, and gas shortages from the largest gas import source due to haveries/maintenance works – note that more than half of the imported gas in Spain comes from Algeria –, analysed in detail in CNMC (2014a) producing a hike in the spot and forward prices).


143

The current mechanism applied by the Government relies 100% on the spot price and social alarm could be created in a situation of supply shocks (such a situation cannot be neglected considering e.g. the Ukrainian crisis). Therefore the 40/60 scheme could be a more robust solution in the long term, strengthening the price discovery function of the futures market and mitigating its forward risk premium through strict market monitoring by all the affected authorities and operators.

The ratio 40/60 is preferred to other alternatives (50/50, 30/70, 20/80, et cetera) as it results more equilibrated considering both the spot price volatility and the forward risk premium.

2 - The Iberian power futures market shows a limited market efficiency, as the majority of energy markets, due to the existence of remarkable forward risk premia. Strong oversight through integral access by the energy regulatory authorities to all the fundamental data, as well as transactional spot and derivatives data is key to assess inefficiencies and introduce proper remedies. The merging of the Spanish Energy Regulator and the Spanish Competition Authority (CNMC), as in the Netherlands (ACM, the Dutch Authority for Consumers and Markets), should streamline the cooperation and communication gateways between such merged Directions, improving the investigation and enforcement cases derived from market abuse suspects. Furthermore, active communication gateways between CNMC and the Spanish Financial Supervisory Authority (CNMV), grounded on the Collaboration Agreement of July 2012, would provide the former with quick access (not only on demand, i.e. case by case) to energy derivatives data (especially the electricity OTC trades) permitting a cross-check with the derivatives data provided by the Agency for the Cooperation of Energy Regulators (ACER) as defined in REMIT Implementing Acts related to data reporting. Cross-checked data would provide a more robust basis for investigation and thus facilitate the sanctioning of market abuse infringements. Quick wins could be obtained by attracting OTC trading to the futures market (OMIP) and the clearing houses (OMIClear and MEFF Power) and facilitate short term futures trading as the shorter the maturity (i.e. the time horizon or delivery period) the smaller the forward risk premium.

The ex-post forward risk premium in the Iberian power futures market, obtained as the difference of the average forward price during the trading period and the underlying spot price during the delivery period, is remarkable in its beginning as in the majority of energy markets and tends to be positive in the long term.

The same trend in the forward risk premium nature is appreciated for the rest of forward market mechanisms, namely, the dominant OTC trading and the CESUR auctions (the CESUR auction scheme ended in December 2013) in which the last resort suppliers purchased part of their regulated supplies.

The forward risk premium has diminished due to the agents’ learning curve and the effect since year 2011 of the regulatorily recognized price for coal fired generation burning indigenous coal for the sake of security of supply. As the forward risk premium


144

is bigger for contracts with larger maturity (year contracts followed by quarter contracts), the market participants can benefit in many cases by means of trading short positions (i.e. keeping a selling net position) in such power contracts.

Conversely, market participants with a natural long position (e.g. suppliers purchasing derivatives to hedge for their retail sales) might be more interested in trading month contracts, especially when the perception of uncertainty and thus energy risk is bigger. Smaller premia should be the natural result of a well developed and efficient power forward market, rather than the effect of artificial mechanisms distorting the price formation.

The Spanish and Portuguese national regulatory authorities should analyse the overall performance of the Iberian electricity market considering all the transactions from the spot and derivatives market and the fundamental data (e.g. available capacity, demand) to assess if the existing forward risk premia is justified or is augmented due to market abuse by some traders.

The European Union Regulation on Market Integrity and Transparency of Wholesale Energy Markets (REMIT), in force since 28 December 2011, empowers those authorities to access all the transactions data. Additionally, by means of the Multilateral Memorandum of Understanding between the members of the MIBEL Regulatory Council, the Spanish and Portuguese energy and financial regulators can exchange data for joint monitoring purposes. In case of market abuse suspicions, the energy regulatory authorities can perform investigations by means of REMIT and impose penalties.

If the perception of stronger oversight were not enough to decrease somehow large forward risk premia partly caused by potential excessive speculation, other remedy actions should be properly designed by the MIBEL regulatory authorities. The most effective ones would be those attracting trading volumes from the opaque OTC market or from bilateral trades (e.g. intra-group transactions) towards the Iberian power futures market (e.g. through voluntary auctions with reduced variable fees) or, at least, clearing incentives to settle the out-of-exchange traded volumes in an Iberian clearing house (i.e. OMIClear and BME Clearing, formerly called MEFF Power) so that national regulatory authorities can properly analyse the market behaviour.

3 - The creation of an Iberian Gas Hub providing a transparent and reliable spot gas price at the Spanish Virtual Trading Point (known as “AOC”, the so-called “Centralized Operational Storage”) and a sound LNG spot price in the Spanish and Portuguese regasification terminals, as well as the subsequent introduction of Iberian gas derivatives would improve the price formation of the Iberian power futures and thus the efficiency of the forward market. The introduction of robust local gas price references would help large consumers to better negotiate the end users’ prices. Therefore, the efficiency of the Iberian energy market for both commodities, in all time frames (spot and derivatives) and segments (wholesale and retail) would increase.


145

The creation of an Iberian gas exchange is a must to overcome the chronical lack of price transparency in the Iberian Gas Market (known as MIBGAS). So far only OTC gas trades are concluded. They are registered in the electronic platform provided by the Spanish Gas TSO (Enagás acting as System Technical Manager (known in Spain as “GTS”) provides such an electronic platform called the Third Party Access – Secondary Market (MS-ATR)) in which no gas price is unveiled, despite of the large trading volumes and the key role of the Spanish gas market as European entry of LNG.

The working group for the creation of the Iberian gas hub, led by the Spanish energy regulator, and with monthly meetings with all the stakeholders is resulting very effective towards such a goal. The Spanish energy regulator has proposed regulatory changes in the third party access ground regulation (Royal Decree 949/2001) to accomodate all the requirements of a hub based market model, especially for the short term trades (balancing market).

Once the gas hub is properly working, further regulation would be needed to develop Iberian gas derivatives allowing gas hedging and cross-commodity hedging. The Iberian power futures prices are strongly correlated with the European gas prices, especially between the year contracts, as far as the renewable forecasts – a key signal for including renewable generation costs in the electricity price formation – are currently accurate just on the very short term.

Results from cointegration analysis for the Iberian prompt month power futures contracts with gas and oil prices show that such electricity prices are influenced by gas forward prices of the most liquid reference in continental Europe (TTF), and by oil prices (6 month rolling average of Brent spot prices) used in long-term gas contracts and in Spanish gas last resort tariffs. Introduction of Iberian gas derivatives would provide a more robust electricity price signal based on local gas-to-gas competition.

On the other hand, the clean spark spreads for prompt futures contracts, obtained as the difference between the Iberian power futures price and the forward generation cost of a combined cycle gas turbine (based on TTF gas prices and the month EUA futures emission allowances with December maturity), tend to be positive in the period July 2006 – November 2011.

The biggest spreads are for the month contract, followed by the quarter contract and then by the year contract. Therefore, gas fired generation companies can maximize profits trading more frequently with electricity contracts of shorter maturity. Both the forward risk premium and the clean spark spreads are key indicators for market participants to optimize their trading strategies – they can exploit arbitrage gains through trading with contracts of different maturity and perform cross-market hedging operations – and for regulatory agencies to monitor the fair price formation in the wholesale energy markets, ultimately affecting the end-user prices.

The Iberian power futures market seems the most appropriate venue to introduce such gas derivatives, due to the full reliability of this market during its first 8 years. The introduction of sound Iberian gas price references (both for the spot and derivatives trading) would facilitate stronger MIBEL oversight. Both the forward risk premia and the clean spark spreads based on an Iberian gas price would better reflect market fundamentals.


146

Such robust local price references would help large consumers to better negotiate the retail prices. Therefore, the efficiency of the Iberian energy market for both commodities, in all time frames (spot and derivatives) and segments (wholesale and retail) would benefit from the hub.

4 - The liquidity of the Iberian power futures market can substantially increase through marketing incentives (e.g. voluntary auctions, fee discounts) to attract financial players, large industrial users, small producers (e.g. renewable generation), and intra-group bilateral trades producing price efficiency gains. The market operator and its clearing house should hold a permanent dialogue with all the stakeholders and close cooperation with the supervision authorities to succeed in its business development improving the liquidity of all the listed products and the market efficiency. Any improvement in post-trade transparency and liquidity/centralised clearing growth due to cooperation actions of trading venues/clearing houses along Europe would facilitate wholesale market supervision and the retail price formation in the liberalised market based on robust wholesale price signals.

Though the Iberian power futures market shows steady development in terms of number of participants and traded and cleared volumes, as well as a good transparency level regarding public post-trade data availability, its liquidity is still poor compared to other more mature European power futures market (namely Nasdaq OMX Commodities located in Oslo and EEX located in Leipzig).

Bigger efforts and resources should be employed by OMIP to publish timely and more analytical market monitoring reports. The only significant parameters acting as effective liquidity drivers are the traded volumes in the dominant OTC market, the OTC cleared volumes by OMIClear, and the call auctions in which the Spanish distribution companies and the Portuguese last resort supplier were obliged to purchase energy according to national legislation.

The traded volumes in the continuous market shows a significant correlation to the OTC volumes, the amount of active market makers in OMIP, the enrolment in OMIP of financial agents and generation companies of vertically integrated groups, and the OTC cleared volumes by OMIClear. Marketing incentives to attract financial agents could increase the trading activity in the continuous market.

There is still much room for new enrolments in OMIP, as from such financial companies, energy intensive industries, international energy companies, small producers (as renewable generation companies) and other agents already active in the spot market. The enrolment of these companies would create a more balanced structure in the market and facilitate a less biased price formation in terms of forward risk premium.

Specific commission discounts for the large vertically integrated groups to trade in OMIP and clear OTC volumes in OMIClear would bring price transparency and liquidity growth, in case such vertically integrated utilities were active OTC participants. The


147

market operator could establish specific market maker agreements, commission discounts, and auctions complementary to the continuous market to boost the still illiquid products.

The market operator and its clearing house should hold a permanent dialogue with all the stakeholders, both with the agents and the spot market operator (as attractive products can only be offered if they respond to real market needs) and with the supervisory agencies (as new developments in market rules should improve the market efficiency).

Collaborative efforts between market operator and supervisory agencies create strong synergies: the market operator could invite the supervisory agencies to provide feedback about its analyses measuring the performance of new business developments. Such shared vision would benefit both entities due to the liquidity and efficiency gains. Social welfare would increase as well, due to a positive contribution towards a better price formation of end users’ electricity prices.

The launch by OMIP-OMIClear in the last week of May of year 2014 of European Options on base load Spanish power futures with two months ahead, two quarters ahead and one year ahead available maturities (with the first deal in the continuous trading mode (i.e. screen trading) on 29 May 2014) is a good initiative to attract OTC trading volumes and increase liquidity in the continuous trading mode as well as the cleared volumes, facilitating the market oversight to the energy regulators and increasing post-trade transparency (OMIP-OMIClear, 2014b).

This would help to build the retail prices in the liberalised market based on robust wholesale price signals. Finally, the effective implementation of the cooperation agreement signed between OMIP-OMIClear and Central Europe energy derivatives exchange (EEX) and clearing house (ECC) at the end of year 2013 would facilitate the registration of OTC trades from both the Iberian and Central European region in all those trading venues/clearing houses (OMIP-OMIClear, 2013a).

The affected regulatory authorities should ensure the proper data sharing agreements with those trading venues/clearing houses in order to access to the derivatives data whose underlying spot price is under their scope. In the Iberian case, shall remarkable Iberian OTC volumes traded along Europe would be now cleared and settled due to such trading venues/clearing houses agreements, a stronger picture of the derivatives trading would be easily accessed.

National Regulatory Authorities should not miss these “low hanging fruit” opportunities to improve their supervision input data spontaneously arisen from the market participants’ needs.

5 - The net position ratio obtained as the final open interest of a derivatives contract divided by the accumulated cleared volume for the delivery period specified in the contract maturity could be used by supervisory agencies as a trigger to detect eventual excessive speculation. This straightforward indicator, despite of conceptual limitations, as well as the traditionally used hedge ratio in financial literature, both applied with daily values, distinguishing per market


148

participant type and permitting single screening for one agent, can facilitate stronger market oversight to energy regulators when aggregating data from all the trading venues according to EU Regulation on Market Integrity and Transparency. A dynamic analysis with both ratios could be used as well by policy makers to assess the convenience and estimate the proper magnitude of position limits. Publication of aggregated statistics as those shown in US CFTC Commitment of Traders’ Reports would increase ex-post trade transparency and the efficiency of the European wholesale energy markets. The pro-active utilisation of such indicators and statistics reports by the MIBEL Council of Regulators could serve as a pilot experience producing efficiency gains in the Iberian electricity forward market that could be easily expanded to the rest of Europe.

The hedging performance in the Iberian Forward Electricity Market has been estimated with aggregated data from the Portuguese and Spanish clearing houses (OMIClear and MEFF Power (now BME Clearing) respectively) through the net position ratio obtained as the final open interest of a month derivatives contract divided by the accumulated cleared volume for that delivery month (i.e. the month contract and the corresponding part of the quarter and year contracts).

So far OMIP Spanish base load power futures contracts show the lowest net position ratio due to good liquidity. Since the beginning of year 2009, the net position ratios for these contracts are smaller than 0.5 (around 0.1 at the beginning of year 2012), indicating that less than half of the cleared volumes can be used for hedging and the rest of cleared volumes correspond to positions being closed.

The market participants closing such positions can benefit from price differences in the purchase and sale operations centrally cleared. They can also benefit through the portfolio trading based on the exploitation of the price differences (e.g. arbitrage) between the coexisting market mechanisms in the Iberian electricity forward market (namely, OMIP power futures market, CESUR auctions (until December 2013) and OTC trading).

The power futures peak contracts show bigger ratios as almost all the cleared amount correspond to hedges of the Portuguese last resort supplier in OMIP auctions in which this company has to purchase quantities regulatorily fixed. Both clearing houses (OMIClear and MEFF Power) behave complementarily, being OMIClear preferred for larger maturity contracts cleared as futures.

The net position ratio employed is actually a straightforward static indicator, as it is built on the final open interest. A more complex and powerful dynamic indicator would consider the evolution of the open interest for all the trading sessions, not the last one. That daily ratio would help to better understand the trading patterns.

However, its limitations when considering only centrally cleared volumes have to be taken into account: (i) it provides a trading partial snapshot, as the majority of the OTC trading is still not centrally cleared and no aggregated OTC data with enough granularity (except for the overall monthly CNMC’s electricity OTC statistics) are published; (ii) the hedge only exists if the trader does a similar operation in the spot


149

market. Additionally, the net position ratio would be different per market participant type and the reasonable level of this indicator would only be useful per market participant type (i.e. not a global value).

In the case of an energy company intending to hedge its generation or supply commitments, the net position ratio would tend to present higher values. In the case of a financial entity without hedging goals and non-highly leveraged, its ratio would tend to present lower values (what minimizes its credit default risk). In case the net position ratio of any participant grew substantially affecting the price volatility, the supervisory authorities could monitor if the integrity of the whole system is in danger and take appropriate enforcement measures.

The net position ratio can be very useful for energy regulators performing market oversight to detect excessive speculation, once they access to the details of all the transactions of energy derivatives (both from regulated exchanges and OTC) according to REMIT Implementing acts related to data reporting (its approval is foreseen for the second half of year 2014, after the adoption of the key European financial legislation (the so-called MiFIR/MiFID II and MAR) as the definition of financial instrument in MiFID II frames the scope of the rest of the mentioned legislation).

A more sophisticated design (i.e. with daily values, distinguishing per market participant type and permitting screening for one single agent) of the net position ratio and the hedge ratio – defined as the ratio of the position taken in the futures contracts that will exactly offset the size of the exposure in the spot market, traditionally used in financial literature – are suggested to be employed by energy regulators devoting resources to market oversight of wholesale (spot and derivatives) energy markets and by the Agency for the Cooperation of Energy Regulators (ACER).

These authorities as well as other competent authorities – the national financial and competition authorities and the European Securities and Markets Authority (ESMA) – will have access to energy derivatives transaction details according to REMIT. The obligations for the market participants to submit their transaction records to ACER, that will grant data access to the energy regulators, will commence six months after the details of the transaction reporting are published in a European Commission Implementing Regulation (the Implementing Acts related to data reporting mentioned above).

The use of daily net position and hedge ratios by the market monitoring authorities, filtered per market participant type (utility, financial entity, dominant operator, new entrant, etc.) could help to detect any suspicious behaviour regarding market abuse.

A dynamic analysis of the net position and the hedging ratios could be used as well by policy makers to assess the convenience and estimate the proper magnitude of position limits. These limits are envisaged for the European and North American commodity derivatives markets, in order to avoid excessive speculation impacting on prices.


150

6 – The publication of statistics through periodical market monitoring reports (e.g. weekly) regarding long and short positions for different types of market participant by the existing Iberian clearing houses would increase the post-trade transparency. Once the Spanish energy regulator (CNMC) has access to all the Iberian derivatives transactions from the centralised information system managed by the Agency for Cooperation of European Energy Regulators (ACER), it could publish such monitoring indicators considering both the centrally cleared transactions and the non cleared OTC transactions, providing a full picture of the derivatives market.

Neither the Iberian power futures market operator and Portuguese clearing house (OMIP-OMIClear) nor the Spanish clearing house for energy derivatives (MEFF Power, whose name was changed to BME Clearing on 9 September 2013) publish statistics regarding long and short positions associated to commercial traders (i.e. energy companies) or to financial entities.

The publication of such statistics via periodical market monitoring reports from each clearing house – or alternatively by the regulatory agencies having access to centrally cleared data – would provide more post-trade transparency, increasing the confidence in the market and thus its efficiency. Weekly publication as it is the case with the Commitments of Traders Reports by the US Commodities Futures Trading Commission (CFTC) is strongly suggested.

7 – Any detection of anomalous events in the electricity and natural gas markets affecting the price formation should be tackled in real time to avoid the exploitation of abusive practices. The development of agile communication gateways between the market and system operators, as well as energy brokers, with the energy regulators to inform the latter of suspects related to those undesired events is key to timely open investigation reports and succeed in such case handling by the competent authorities. In the case of the Iberian market, such communications gateways with the Spanish and Portuguese electricity and natural gas market and system operators and the main brokers operating in the Iberian market could be effectively implemented by the MIBEL Council of Regulators.

The Article 15 of the European Regulation on market integrity and transparency of the wholesale electricity and natural gas markets (REMIT), in force since 28 December 2011, establishes that the persons professionally arranging transactions should immediately inform the affected national energy regulators of any suspicious transaction related to trading based on inside information or (attempt of) market manipulation.

In order to do so, those persons should implement appropriate aarangements and procedures to identify such REMIT breaches. In this sense, the MIBEL Council of


151

Regulators could develop effective communication gateways with the Spanish and Portuguese market and system operators, namely: OMIE (electricity spot market), OMIP (electricity futures market), REE (Spanish electricity Transmission System Operator), REN (Portuguese electricity and natural gas Transmission System Operator), ENAGÁS GTS (Spanish natural gas Transmission System Operator), and the Iberian gas hub market operator (once it starts operations).

Additionally, the main brokers (for instance, the five brokers registered in the Iberian power futures market, namely: CIMD – Sociedad de Valores, ICAP Energy, Spectron Energy Services, Tradition Financial Services and Tullett Prebon (Europe)) should also develop such procedures to inform of any suspicious transaction regarding OTC trading.

Finally, the Spanish clearing house (BME Clearing) and the operators of LNG terminals and underground storages for natural gas should also be included to create a comprehensive compliance and supervisory culture minimising the incentives to market abuse practices. The effective collaboration between those operators who see everything in real time and the energy regulators supervising the wholesale markets would improve the market performance and increase the market confidence, providing more reliable price signals, considered in the price indexation of the retail contracts.

Such a collaboration – fully exploiting the synergies between the operators and supervisory agencies – is strongly needed in a context of budgetary cuts in the Public Administration, producing a lack of resources for regulatory and supervisory tasks (see e.g. ACER’s complaint regarding the scarcity of its currently available REMIT human resources in ACER (2014)).

8 – The creation of a centralised platform for the publication of inside information related to the Spanish electricity and natural gas markets is essential for creating a level playing field and favouring a proper wholesale price formation. It seems reasonable that both the Spanish electricity and gas transmission system operators develop such platforms. An active dialogue through workshops and public consultations managed by the Spanish energy regulator should be held between all the stakeholders (Spanish energy regulator, market participants, Transmission System Operators, market operators, brokers and infrastructure owners) to implement in a reasonable time such key platforms.

Article 4 of REMIT indicates that the market participants should timely publish any inside information regarding the business activities or the availability of the energy infrastructures related to their transactions in the wholesale electricity and natural gas markets.

The Agency for the Cooperation of Energy Regulators (ACER), suggests in its non binding guidance for the interpretation of REMIT that the most effective vehicle to publish inside information is a centralised platform rather than the individual market participants’ websites (ACER, 2013a).


152

ACER suggests the national regulatory authorities to consult with their affected market participants how to deploy national platforms regarding the publication of inside information.

The recently created REMIT page by the Spanish energy regulator (CNMC, in particular: www.cnmc.es/es-es/energ%C3%ADa/remit.aspx, officially launched in a first workshop with the Spanish market participants on 26 May 2014 focused on the creation of the Spanish register of market participants) is a powerful tool to manage efficiently all the REMIT implementing issues, especially the questions from the market participants, and the discussion of key points as the creation of the national platforms for the publication of inside information.

It seems reasonable that both the Spanish Transmission System Operators for electricity (REE) and natural gas (ENAGÁS GTS) develop such platforms, as they have vast experience in the management of fundamental data and publication of aggregated data in their respective websites for the sake of transparency.

Currently, only some market participants are publishing in their websites information about outages of their power plants, each one following its own criteria, not being an effective solution in the long term, as all the inside information should be disclosed by the market participants in an homogeneous way.

The centralised platforms are the appropriate solutions. Once the national platforms are working smoothly in the different countries, and based on the best practices (e.g. the Urgent Market Messages (UMM) in the Nordic Electricity Market have been successfully working since many years, even before the REMIT entry into force, and could thus be employed as a standard European format), it might make sense to provide a centralised European platform through the European networks of Transmission System Operators for electricity (ENTSO-E) and gas (ENTSO-G).

8.2 Further research

The research covered an emerging and quick developing topic. Therefore, the following suggestions for futures lines of research to expand the study and analysis of the Iberian energy derivatives trading are proposed:

1 - The research based the hedging efficiency analysis on the net position ratio. Further research is suggested based on the hedge ratio traditionally employed in the financial literature. Comparison of the daily evolution of both ratios would be useful.

Further research is encouraged with an ampler data set regarding the estimation of the optimal hedge ratio for the Iberian energy derivatives market, considering OMIP future prices and the underlying spot prices, to seize the magnitude of the resulting ratio compared to other commodities markets. The analysis of the evolution of that ratio would allow analysing any change in behaviour.

http://www.cnmc.es/es-es/energ%C3%ADa/remit.aspx


153

Furthermore, the analysis of the evolution of the net position ratio with a larger data set is strongly recommended to find results regarding the remarkable growth of the cleared volumes in Meff Power (BME Clearing) during years 2013 and 2014. Both ratios are not directly interrelated as the net position ratio depends exclusively on futures data and the hedge ratio depends as well on spot data.

However, assuming for both cases that derivatives cleared volumes and spot volumes remained constant, the evolution of both ratios were essentially led by the development of the net position (i.e. the open interest).

2 – In oil markets, an inverse relationship between the open interest and spot market volatility is found. Further research is encouraged with MIBEL spot and futures data to assess the impact of the development of the open interest on the spot market volatility.

When the open interest is greater, the volatility shock of the spot market associated with a given unexpected increase in futures trading is much smaller. Therefore the trading of futures contracts improves depth and liquidity in the underlying market. Further research is encouraged with MIBEL spot and futures data to assess the impact of the development of the open interest on the spot market volatility.

However, the main factors affecting the volatility of the Iberian spot market are the abundant wind generation and the regulatorily fixed prices for the remuneration of coal power plants burning indigenous coal. This subsidised price is applied since February 2011 by the Spanish government for the sake of security of supply.

Whereas the volatility is increased by the former due to the intermittent nature of wind generation, it is decreased by the latter as such fixed prices act as a price cap in the spot offers of thermal power plants limiting as well the price fluctuations in the futures market.

3 – An analysis of the impact of the contract for differences mechanism designed for the reduction of the tariff deficit caused by the massive deployment of feed-in-tariffs, based on the frequent positive forward risk premia (in average, 11%) from the electricity last resort supply auctions is suggested.

As a result of frequent positive forward risk premia, the electricity purchase by the Spanish last resort suppliers in CESUR auctions, with delivery between July 2009 and December 2011, resulted 587 million € more expensive than the electricity cost valued at the spot price.

That trend continue until the last CESUR auction held in December 2013, resulting the ex-post forward risk premium around 11% bigger than the related spot price in the period July 2009 – December 2013 (CNMC, 2014c). Therefore, an analysis of the economic impact of the contract for differences mechanism defined in Royal Decree 302/2011 is suggested for additional research.


154

This mechanism helps to mitigate the tariff deficit caused by regulated prices for special regime generation. The data set would cover until 31 December 2013, as the delivery period of the last validated CESUR auction (the one held in September 2013) spans until that date.

4 – Further research, based on literature review about regulatory risk, is suggested to quantify the increase in the Iberian forward risk premium due to regulatory risk provoked by intensive introduction of policy measures in Spain and Portugal with urgent nature.

Due to time pressure, many of the regulatory measures recently introduced by the Spanish Government with urgent nature for the reform of the electricity sector, whose main objective is to curb the tariff deficit, may lack a robust cost-benefit analysis.

The balance between the benefits and costs of imperfect regulation should overcome the costs of the related market failures. The outcomes of the suggested potential research to assess the impact of such reforms in the electricity forward price formation – in particular, the impact on the forward risk premium due to the uncertainty and derived effects of such frequent reforms – could be very useful for the Spanish energy regulator, as well as its peers from other countries, to streamline the existing regulation and supervise the performance of the wholesale energy markets.

5 – The analysis of the development of the Iberian electricity forward market can be enriched considering the hedging instruments for the cross-border trade, the auctions for selling the renewable production in Portugal, and the daily futures. Further research for those instruments is suggested based on the evolution of the traded and centrally cleared volumes, the forward risk premia, and the open interest. Regular monitoring of these parameters, especially for the fast developing short term derivatives, would be very useful to the entities supervising the Iberian electricity market.

The research has mainly considered the evolution of the electricity derivatives showing the highest degree of liquidity, i.e., the base load futures with month, quarter and year maturity whose underlying instrument is the spot price of the Spanish area.

Additionally, data from the CESUR and EPE auctions was also included, as well as aggregated OTC data. Further research is suggested taking into account the liquidity (and registration in the two Iberian clearing houses) growth in the short term base load futures (mainly daily and week, and weekend in a minor degree) with underlying instrument the Spanish spot price.

The daily and weekend contracts were introduced in the Iberian power futures market on 20 May 2011, and the main recent liquidity development of such instruments is related to OTC trading of daily contracts.


155

Analysis of the ex-post forward risk premia and open interest evolution in those short term futures could provide a richer picture of the price efficiency and the speculation levels (e.g. the spot price formation could be altered due to eventual excessive speculation in short term derivatives trading).

This would be very useful for the entities supervising the Iberian electricity market, to detect any signal of price distortion or eventual market manipulation and be able to react quicky with investigation and enforcement actions, as well as introduction of adequate regulatory measures to tackle any possible market failure.

Additionally, an analysis of the price formation of the financial transmission rights (options, formerly contract for differences (obligations)) currently traded in the Spanish-Portuguese interconnection, comparing its price evolution with that of the equivalent instruments on both sides of the interconnection (i.e. Spanish futures prices and Portuguese futures prices) would help to assess if the price formation of the less liquid Portuguese futures is evolving properly or even remarkable arbitrage differences exist between those derivatives trading mechanisms.

It is also interesting to measure if such financial transmission rights are providing any liquidity increase in OMIP continuous market. Finally, such an analysis can be strengthened taking into account the price formation of such Portuguese futures prices in the auctions in which the special regime generation is sold forward by the Portuguese last resort supplier, facilitating the energy access to new entrants in the Portuguese retail market.

6 – The research has considered the evolution of the forward generation costs for gas fired combined cycles compared to the electricity futures prices (clean spark spreads). Further research is suggested taking into account the forward generation costs for coal power plants (dark spark spreads). Additionally, the comparison of the forward risk premia and spreads with other European energy derivatives markets could be enriched with other maturities different to prompt base month, quarter and year contracts.

Further research is encouraged taking into account the influence of the dark spark spreads (obtained as the difference between the power futures prices and the generation costs of a coal power plant, the latter calculated e.g. through coal futures contracts traded at the European Energy Exchange, EEX) to assess the influence of the coal generation – out of the regulatorily fixed prices by the Spanish Government since year 2011 for the indigenous coal fired generation – in the Iberian power forward price formation.

A comparative analysis of both forward risk premia and clean (gas) & dark (coal) spark spreads with other European markets (e.g. French, German or Nordic) could be used to measure their different efficiency levels. As the research performed is based on prompt base load contracts, the analysis could also be extended to further maturities (i.e. not immediately close to delivery, and also for the very short term futures (i.e daily and weekly)) as well as for peak products.


156

7 – Further research is encouraged to simulate the economic effects and the forward price formation in the Spanish power system due to the introduction of renewable auctions substituting the existing feed-in-tariff scheme. For instance, the model proposed in the German energy derivatives market in which the renewable generation companies participate in an auction determining a premium over the spot price could make sense in the Iberian electricity market. Simulations based on that model are suggested.

Further research is encouraged to simulate the economic effects and the forward price formation in the Spanish power system due to the introduction of renewable auctions. It is of special interest to consider the auction model proposed by the German electricity derivatives market (EEX). In such auctions, the price of renewable generators is set for twenty years, as a result of adding to the spot market price the resulting premium from the proposed auctions.

This premium would be capacity based, with renewable power plants compensated for their capacity provided, rather than for the energy fed into the grid (Köhlner, 2014b). Regarding the economic effect, evaluation of the potential reduction of the accumulated tariff deficit would be very worthy.

Regarding the price formation, analysis of the forward price formation taking into account reasonable prices of those potential renewable auctions as well as the generation costs (including the CO2 emission costs) of the thermal power plants – the gas and coal fired power plants acting as an effective back-up in the absence of windy days – would shed light about the most convenient estimation methodology for the energy costs in the last resort rates, in case such rates were based on forward prices rather than spot prices.

The Royal Decree 216/2014, of 28 March 2014, recently introduced a new methodology based on spot prices (MINETUR, 2014), in which the last resort rates are termed as “Voluntary Price for the Small Consumer”. As the last resort rates are used as benchmark for the end-user rates in the liberalised market, social welfare would be obtained by improving the rate design including the penetration of renewable generation through more efficient competitive price mechanisms.

8 – There are two factors (the European Price Coupling for the day-ahead market, and the centralised data reporting through ACER IT system) that should theoretically provide a price convergence between the different European electricity prices. As the convergence should occur in the spot prices and on the other hand on the forward prices, the Iberian electricity forward risk premia should diminish, increasing the MIBEL price efficiency. Further research is suggested with an ampler data set to assess if such regulatory outcomes produce such efficiency gains or if additional regulatory measures should be introduced to tackle any shortcomings arisen when those two measures are fully in practice.


157

Further research is encouraged to assess the potential price efficiency gains arisen from the implementation of the Price Coupling Region (PCR) initiative for the majority of the European power exchanges introduced since February 2014 through the so-called EUPHEMIA algorithm for the day-ahead market.

In the case of the Spanish-French interconnection, the full implementation of the price coupling should be reached during year 2014 (so far it is performing in an asynchronous way).

On the other hand, after six months of the adoption of the REMIT implementing acts regarding data collection, the market participants will have to report all their electricity and natural gas (both spot and derivatives) transactions (and orders to trade) to the IT system of the Agency for the Cooperation of the European Energy Regulators (ACER).

The system is known as ARIS (ACER REMIT Information System). Due to such comprehensive European energy data reporting (its start is foreseen for the first half of year 2015), the supervision of the electricity and natural gas markets will be streamlined. As a result, the incentives for the market participants to introduce excessive speculation would be refrained, and less frequent atypical values for the spot and derivatives markets should be presented.

Therefore, the forward risk premium should diminish in general in all the European markets. In particular, the current forward risk premium in the Spanish electricity market tend to present more frequent positive values and more extreme positive and negative values than that of the French, German and Nordic electricity markets (see e.g. CNMC (2014b)).

A quantitative analysis with data from all these markets is suggested to see the impact of both regulatory outcomes in the efficient price formation of European energy markets, with special focus on the Iberian electricity market.

The ultimate consequence of the desired effects should be the formation of fair wholesale prices through competitive trading, improving as well the dynamics of the retail markets and thus more affordable end users’ prices.


158


159

CHAPTER 9. REFERENCES

Abadie, L.M., Chamorro, J.M., 2009. Income risk of EU coal-fired power plants after Kyoto. Energy Policy, 37, pp. 5304–5316.

ACER, 2011a. Questions and Answers on the third legislative package for an internal EU gas and electricity market, Agency for the Cooperation of Energy Regulators (ACER), 3 March 2011, Ljubljana. Last date accessed 6 June 2013, available from: www.acer.europa.eu/portal/page/portal/ACER_HOME/Communication/News/memo.doc

ACER, 2011b. Questions & Answers on REMIT, last updated 20 December 2011, Agency for the Cooperation of Energy Regulators, Ljubljana. Available from: www.acer.europa.eu/portal/page/portal/ACER_HOME/Activities/REMIT/ACER%20Guidance%20and%20Questions%20%20Answers%20on%20REMIT/QuestionsAndAnswers3%20ap_final%20vzu_cca.pdf Last date accessed: 8 February 2014.

ACER, 2013a. Guidance on the application of Regulation (EU) No 1227/2011 of the European Parliament and of the Council of 25 October 2011 on wholesale energy market integrity and transparency, 3rd Edition, published on 29 October 2013, by the Agency for the Cooperation of Energy Regulators (ACER), Ljubljana.

ACER, 2013b. ACER´s annual report on its activities under REMIT in 2012, issued on 7 November 2013, Agency for Cooperation of Energy Regulators, Ljubljana. Last date accessed 4 February 2014, available from:

www.acer.europa.eu/Official_documents/Acts_of_the_Agency/Publication/REMIT%20Annual%20Report%202013.pdf

ACER, 2014. ACER’s annual report on its activities under REMIT in 2013, published on 10 June 2014, Agency for Cooperation of Energy Regulators, Ljubljana.

AEEG, 2011. Relazione annuale sullo stato dei servizi e sull'attività svolta, Struttura, prezzi e qualità nel settore elettrico, 6 August 2011, Autorità per l'energia elettrica e il gas (AEEG), Milan.

Alba Ríos, J.J., Moreda Díaz, E., 2010. Competencia en el mercado mayorista de electricidad. Anuario de la competencia 2009, Fundación ICO- UAB, Ed. Marcial Pons, Madrid, pp. 215-293.

Allen, D.E., Cruickshank, S.N., 2002. Empirical Testing of the Samuelson Hypothesis: An Application to Futures Markets in Australia, Singapore and the UK. Working Paper, School of Finance and Business Economics, Edith Cowan University, Joondalup Campus, Western Australia.

Arciniegas, I., Barrett, C., Marathe, A., 2003. Assessing the efficiency of US electricity markets. Utilities Policy, 11, pp. 75-86.

Armstrong, M., Galli, A., 2005. Are day-ahead prices for electricity converging in Continental Europe? An exploratory data approach. CERNA Working Paper.

Avsar, S., Goss, B., 2001. Forecast Errors and Efficiency in the US Electricity Futures Market. Australian Economic Papers, 40, pp. 479-499.

Baillie, R.T., Myers, R.J., 1991. Bivariate GARCH estimation of the optimal commodity futures hedge, Journal of Applied Econometrics, Vol. 6(2), pp.109–124.

Ballester, J.M., Climent, F., Furió, D., 2012. Market efficiency and lead-lag relationships between spot, futures and forward prices: The case of the Iberian Electricity Market (MIBEL), Fundación de las Cajas de Ahorros, Documento de Trabajo Nº 695/2012, Madrid.

http://www.acer.europa.eu/portal/page/portal/ACER_HOME/Communication/News/memo.doc

http://www.acer.europa.eu/portal/page/portal/ACER_HOME/Communication/News/memo.doc

http://www.acer.europa.eu/portal/page/portal/ACER_HOME/Activities/REMIT/ACER%20Guidance%20and%20Questions%20%20Answers%20on%20REMIT/QuestionsAndAnswers3%20ap_final%20vzu_cca.pdf



http://www.acer.europa.eu/Official_documents/Acts_of_the_Agency/Publication/REMIT%20Annual%20Report%202013.pdf

http://www.acer.europa.eu/Official_documents/Acts_of_the_Agency/Publication/REMIT%20Annual%20Report%202013.pdf


160

Bartelj, L., Paravan, D., Gubina, A.F., Golob, R., 2010. Valuating risk from sales contract offer maturity in electricity market. International Journal of Electrical Power & Energy Systems, 32, pp. 147-155.

Batlle, C., 2007. Letters to the Editor: “Balancing Market Maker” Proposal for France Poses Several Key Questions. Carlos Batlle’s Reply. The Electricity Journal, 20, Nr. 10, pp. 5-6.

Batlle, C., Vázquez, C., Barquín, J., 2007. The Balancing Market Maker: A Tool to Enhance Liquidity in the French Balancing Mechanism. The Electricity Journal, 20, Nr. 6, pp. 76-86.

Batlle, C., Barroso, L.A., 2011. Support schemes for renewable energy sources in South America. MIT-CEEPR Working Paper 11-001.

BCN, 2010. Ley núm. 20.257, introduce modificaciones a la ley general de servicios eléctricos respecto de la generación de energía eléctrica con fuentes de energías renovables no convencionales, Ministerio de Economía; Fomento y Reconstrucción. Biblioteca del Congreso Nacional de Chile (BCN), Ley Chile, last date accessed 4 February 2014, available from:

www.leychile.cl/Navegar?idNorma=270212&tipoVersion=0

Bessembinder, H., Lemmon, M.L., 2002. Equilibrium Pricing and Optimal Hedging in Electricity Forward Markets. The Journal of Finance Vol. LVII, Nr. 3, pp. 1347-1382.

BME, 2011. MEFF Power comenzará a funcionar el 21 de marzo. MEFF press release, on March 2, 2011. Bolsas y Mercados Españoles (BME), Madrid. Available from:

www.bolsasymercados.es/aspx/RevOnLine/NPrensa_BME.aspx?id=17903 Last date accessed: 6 June 2013.

BME, 2012. MEFF Derivados Energía, Bolsas y Mercados Españoles (BME), Madrid. Historical market data, available from: www.meff.es/aspx/DerEnergia/Home.aspx Last date accessed: 6 June 2013.

BME, 2014a. January 2014 MEFF Power Memo, published on 10 January 2014, “MEFF Power cierra 2013 con 33.5 TWh, 4 veces el volumen de 2012”. Available at:

www.meff.es/docs/comunicados/2014/01/Balance_Actividad_Diciembre_2013_.pdf Last date accessed: 19 February 2014.

BME, 2014b. About us, accessed on 28 February 2014, available at: www.meff.es/aspx/Comun/Pagina.aspx?l1=QueEs&f=Home&id=ing

BOE, 2004. Convenio internacional relativo a la constitución de un mercado ibérico de la energía eléctrica entre el Reino de España y la República Portuguesa, hecho en Santiago de Compostela el 1 de octubre de 2004. Spanish Official Gazette (BOE), Madrid, 22 mayo 2006, pp. 19208-19212.

BOE, 2006. Real Decreto-Ley 3/2006, de 24 de febrero, por el que se modifica el mecanismo de casación de las ofertas de venta y adquisición de energía presentadas simultáneamente al mercado diario e intradiario de producción por sujetos del sector eléctrico pertenecientes al mismo grupo empresarial. Spanish Official Gazette (BOE), Madrid, 28 febrero 2006, pp. 8015-8016.

BOE, 2008. Acuerdo de modificación del Convenio Internacional relativo a la constitución de un mercado ibérico de la energía eléctrica entre el Reino de España y la República Portuguesa, de 1 de octubre de 2004, hecho en Braga el 18 de enero de 2008. Spanish Official Gazette (BOE), Madrid, 11 diciembre 2009, pp. 104682-104686.

http://www.leychile.cl/Navegar?idNorma=270212&tipoVersion=0

http://www.bolsasymercados.es/aspx/RevOnLine/NPrensa_BME.aspx?id=17903

http://www.meff.es/aspx/DerEnergia/Home.aspx

http://www.meff.es/docs/comunicados/2014/01/Balance_Actividad_Diciembre_2013_.pdf


161

BOE, 2011. Real Decreto 302/2011, de 4 de marzo, por el que se regula la venta de productos a liquidar por diferencia de precios por determinadas instalaciones de régimen especial y la adquisición por los comercializadores de último recurso del sector eléctrico. Spain’s Official Journal, BOE Number 55, March 5, 2011. Last date accessed: 19 February 2014, available from:

www.boe.es/boe/dias/2011/03/05/pdfs/BOE-A-2011-4119.pdf BOE, 2013. Ley 24/2013, de 26 de diciembre, del Sector Eléctrico. Spain’s Official

Journal, BOE Number 310, December 27, 2013. Last date accessed: 9 May 2014, available from: www.boe.es/boe/dias/2013/12/27/pdfs/BOE-A-2013-13645.pdf

Bolinger, M., Wiser, R., Golove, W., 2006. Accounting for fuel price risk when comparing renewable to gas-fired generation: the role of forward natural gas prices, Energy Policy, 34, pp. 706–720.

Bollerslev, T., 1986. Generalized Autoregressive Conditional Heteroskedasticity, Journal of Econometrics, 31, pp. 307-327.

Buchananan, W.K., Hodges, P., Theis, J., 2001. Which way the natural gas price: an attempt to predict the direction of natural gas spot price movements using trader positions, Energy Economics, 23, pp. 279-293.

Byström, H., 2003. The Hedging Performance of Electricity Futures on the Nordic Power Exchange Nord Pool. Applied Economics, 35, pp. 1-11.

Cabeza, L., 2011. Llegan los contadores inteligentes. Entrelíneas, Revista digital de Red Eléctrica de España, 20, abril-junio, 2011. Available from: www.revistaentrelineas.es/20/entretemas/reportajes/llegan-los-contadores-inteligentes?page=full Last date accessed: 6 June 2013.

Cabral, L.M., 2012. Faturas de luz e gás, Governo garante que haverá redução de custos. Diário de Notícias, 8 August 2012. Available from: www.dn.pt/inicio/economia/interior.aspx?content_id=2710269&page=-1 Last date accessed: 6 June 2013.

Capgemini, 2010. European Energy Markets Observatory, 2009 and Winter 2009/2010 Data Set, Twelfth Edition, November 2010.

Capitán Herráiz, A., Rodríguez Monroy, C., 2008a. Evaluation of the Efficiency of the Iberian Power Futures Market. Proceedings of the II International Conference on Industrial Engineering and Industrial Management (CIO 2008), XII Congreso de Ingeniería de Organización, September 3-5, 2008, Burgos, Spain, pp. 597-606.

Capitán Herráiz, A., Rodríguez Monroy, C., 2008b. Empirical evaluation of the efficiency of the Iberian power futures market. Journal of Industrial Engineering and Management, Vol 1, Nr 2, 2008, pp. 209-239.

Capitán Herráiz, A., Rodríguez Monroy, C., 2009a. Evaluation of the Liquidity in the Iberian Power Futures Market. IV Congress of Spanish Association of Energy Economists (AEEE), Seville (Spain), January 2009.

Capitán Herráiz, A., Rodríguez Monroy, C., 2009b. Analysis of the efficiency of the Iberian power futures market. Energy Policy, 37, pp. 3566–3579.

Capitán Herráiz, A., Rodríguez Monroy, C., 2010a. Evolution of the MIBEL Derivatives Market: a Model for Latin American Markets? Proceedings of the Eighth LACCEI Latin American and Caribbean Conference for Engineering and Technology (LACCEI’2010), “Innovation and Development for the Americas”, June 1-4, 2010, Arequipa, Perú.

Capitán Herráiz, A., Rodríguez Monroy, C., 2010b. Analysis of the price efficiency in the Iberian Power Forward Contracting Mechanisms. Proceedings of the 7th

http://www.boe.es/boe/dias/2011/03/05/pdfs/BOE-A-2011-4119.pdf


http://www.revistaentrelineas.es/20/entretemas/reportajes/llegan-los-contadores-inteligentes?page=full

http://www.revistaentrelineas.es/20/entretemas/reportajes/llegan-los-contadores-inteligentes?page=full

http://www.dn.pt/inicio/economia/interior.aspx?content_id=2710269&page=-1


162

International Conference on the European Energy Market, June 23-25, 2010, Madrid.

Capitán Herráiz, A., Rodríguez Monroy, C., 2011a. Assessment of the Operation of the Iberian Energy Forward Markets. Proceedings of the Ninth LACCEI Latin American and Caribbean Conference (LACCEI’2011), Engineering for a Smart Planet, Innovation, Information Technology and Computational Tools for Sustainable Development, August 3-5, 2011, Medellín, Colombia.

Capitán Herráiz, A., Rodríguez Monroy, C., 2011b. Analysis of the Iberian Power Futures Market Hedging Efficiency. 5th International Conference on Industrial Engineering and Industrial Management, XV Congreso de Ingeniería de Organización, Cartagena, 7 a 9 de Septiembre de 2011.

Capitán Herráiz, A., Rodríguez Monroy, C., 2012a. The first renewable forward market mechanisms in the Iberian Electricity Market. Proceedings of the Tenth LACCEI Latin American and Caribbean Conference (LACCEI’2012), Megaprojects: Building Infrastructure by fostering engineering collaboration, efficient and effective integration and innovative planning, July 23-27, 2012, Panama City, Panama.

Capitán Herráiz, A., Rodríguez Monroy, C., 2012b. Evaluation of the trading development in the Iberian Energy Derivatives Market. Energy Policy, 51, pp. 973–984.

Capitán Herráiz, A., Rodríguez Monroy, C., 2013a. Analysis of the traded volume drivers of the Iberian power futures market. International Journal of Electrical Power and Energy Systems, 44, pp. 431–440.

Capitán Herráiz, A., Rodríguez Monroy, C., 2013b. Analysis of the Iberian electricity forward market hedging efficiency. International Journal of Financial Engineering and Risk Management, Vol. 1, No. 1, 2013, pp. 20-34.

Capitán Herráiz, A., Rodríguez Monroy, C., 2013c. Analysis of the Iberian Power Forward Price Formation. IEEE Transactions on Power Systems. Vol. 28, Issue 3, August 2013, pp. 2942 – 2949.

Cartea, A., Villaplana, P., 2008. Spot Price Modelling and the Valuation of Electricity Forward Contracts: the Role of Demand and Capacity. Journal of Banking and Finance, 32, pp. 2502-2519.

CEER, 2011. CEER monitoring report on transaction reporting and detecting market misconduct in wholesale energy markets - Good practice examples from national regulatory authorities. Ref. C11-WMF-12-03, 1 December 2011. Council of European Energy Regulators (CEER), Brussels. Last date accessed 4 February 2014, available from:

www.energy-regulators.eu/portal/page/portal/EER_HOME/EER_PUBLICATIONS/CEER_PAPERS/Cross-Sectoral/2011/C11-WMF-12-03a_MR-MktMisconduct%2001122012.pdf

CESUR, 2012. Official website of the Spanish Bilateral Contracting Auctions for Last Resort Supplies (“CESUR Auctions”), last day accessed 13 February 2014, available from: www.subastascesur.omie.es/subastas-cesur/

CFTC, 2011. Disaggregated Commitments of Traders Report, Explanatory Notes, U.S. Commodity Futures Trading Commission (CFTC). Available from: www.cftc.gov/MarketReports/CommitmentsofTraders/DisaggregatedExplanatoryNotes/index.htm Last date accessed: 6 June 2013.

Chance, D.M., Brooks, R., 2010. Introduction to Derivatives and Risk Management, 8th Edition, South Western Cengage Learning, Ohio.

http://www.energy-regulators.eu/portal/page/portal/EER_HOME/EER_PUBLICATIONS/CEER_PAPERS/Cross-Sectoral/2011/C11-WMF-12-03a_MR-MktMisconduct%2001122012.pdf



http://www.subastascesur.omie.es/subastas-cesur/

http://www.cftc.gov/MarketReports/CommitmentsofTraders/DisaggregatedExplanatoryNotes/index.htm

http://www.cftc.gov/MarketReports/CommitmentsofTraders/DisaggregatedExplanatoryNotes/index.htm


163

Chatham House, 2014. Chatham House Rule, accessed on 28 January, 2014, available at: www.chathamhouse.org/about-us/chathamhouserule

Chen, S.-S., Lee, Ch.-f., Shrestha, K., 2003. Futures hedge ratios: a review, The Quarterly Review of Economics and Finance, Vol. 43, pp. 433–465.

Chilton, B., 2012a. Speculators and Commodity Prices – Redux, Speeches & Testimony, Commissioner Bart Chilton, February 24, 2012, Commodity Futures Trading Commssion (CTFC), Washington. Available from: www.cftc.gov/PressRoom/SpeechesTestimony/chiltonstatement022412 Last date accessed: 6 June 2013.

Chilton, B., 2012b. Sin-Orgy and Energy, Speeches & Testimony, Commissioner Bart Chilton, November 9, 2012, Commodity Futures Trading Commission (CTFC), Washington.

CNE, 2009. Comisión Nacional de Energía. Consulta pública sobre las emisiones primarias de energía y su impacto sobre el desarrollo de la contratación a plazo y la competencia. July 16, 2009, Madrid.

CNE, 2010. Informe sobre la relación de los precios del mercado spot y el de los mercados a plazo, Comisión Nacional de Energía (CNE), Madrid, 27 July 2010.

CNE, 2011a. Boletín semanal de futuros y OTC eléctricos. Semana del 18 al 22 de julio de 2011. Dirección de Mercados Derivados de Energía. Comisión Nacional de Energía, Madrid.

CNE, 2011b. Spanish Energy Regulator’s Annual Report to the European Commision, 22 July 2011, Comisión Nacional de Energía (CNE), Madrid. Last date accessed: 6 June 2013, available from:

www.energy-regulators.eu/portal/page/portal/EER_HOME/EER_PUBLICATIONS/NATIONAL_REPORTS/National%20Reporting%202011/NR_En/C11_NR_Spain-EN.pdf

CNE, 2011c. Informe sobre la evolución de la competencia en los mercados de gas y electricidad, Periodo 2007–2009, September 1, 2011, Comisión Nacional de Energía (CNE), Madrid. Available from: www.cne.es/cne/doc/publicaciones/PA-competencia-011.pdf Last date accessed: 6 June 2013.

CNE, 2011d. Liquidación de las primas equivalentes, primas, incentivos y complementos a las instalaciones de producción de energía eléctrica en régimen especial, mes de producción: 09/2011, 10 de noviembre de 2011,.Comisión Nacional de Energía (CNE), Madrid.

CNE, 2012a. Boletín semanal de Futuros y OTC Eléctricos, Dirección de Mercados Derivados de Energía, Comisión Nacional de Energía (CNE), Madrid, Semana del 2 al 6 de enero de 2012.

CNE, 2012b. Informe anual de supervisión del mercado eléctrico, año 2011. Comisión Nacional de Energía, 2 de febrero de 2012, Madrid. Available from: www.cne.es/cne/doc/publicaciones/PA_SUP-mercado-electrico11.pdf Last date accessed: 6 June 2013.

CNE, 2012c. Informe sobre el sector energético español. Comisión Nacional de Energía, 7 de marzo de 2012, Madrid. Available from: www.cne.es/cne/Publicaciones?accion=3&id=3026&id_nodo=32 Last date accessed: 6 June 2013.

CNE, 2012d. Informe Mensual de supervisión del mercado mayorista de gas, Dirección de Hidrocarburos, Comisión Nacional de Energía (CNE), Madrid, May 2012.

http://www.chathamhouse.org/about-us/chathamhouserule

http://www.cftc.gov/PressRoom/SpeechesTestimony/chiltonstatement022412

http://www.energy-regulators.eu/portal/page/portal/EER_HOME/EER_PUBLICATIONS/NATIONAL_REPORTS/National%20Reporting%202011/NR_En/C11_NR_Spain-EN.pdf



http://www.cne.es/cne/doc/publicaciones/PA-competencia-011.pdf

http://www.cne.es/cne/doc/publicaciones/PA-competencia-011.pdf

http://www.cne.es/cne/doc/publicaciones/PA_SUP-mercado-electrico11.pdf

http://www.cne.es/cne/Publicaciones?accion=3&id=3026&id_nodo=32


164

CNE, 2012e. Informe 15/2009 de la CNE sobre la propuesta de orden por la que se establece la metodología de cálculo de la tarifa de último recurso de gas natural. Anexo II: análisis econométrico de los precios de referencia de corto y largo plazo. Comisión Nacional de Energía (CNE), Madrid, 9 June 2012.

CNE, 2012f. Informe de Supervisión de Mercados a Plazo de Energía Eléctrica en España, Mayo 2012, 21 de junio de 2012, Dirección de Mercados Derivados de Energía, Comisión Nacional de Energía (CNE), Madrid.

CNE, 2012g. Informe CNE en relación a la supervisión de la negociación en el mercado OMIP y en el mercado OTC y su efecto en la 15ª y 16ª subastas CESUR supervisadas por la CNE, 12 de julio de 2012, Comisión Nacional de Energía (CNE), Madrid.

CNE, 2012h. Propuesta del mecanismo por el que se establece el servicio de garantía de suministro, Comisión Nacional de Energía (CNE), Madrid, 5 diciembre 2012. Available from: www.cne.es/cne/doc/publicaciones/cne51_12.pdf Last date accessed: 6 June 2013.

CNE and CNMV, 2012. Nota de prensa, La CNE y la CNMV suscriben un convenio para la supervisión de los mercados a plazo de energía y las subastas CESUR, 3 de julio de 2012, Comisión Nacional de Energía (CNE) y Comisión Nacional del Mercado de Valores (CNMV), Madrid.

CNMC, 2014a. Informe sobre el desarrollo de la 25ª subasta CESUR previsto en el artículo 14.3 de la orden itc/1659/2009, de 22 de junio, Comisión Nacional de los Mercados y la Competencia (CNMC), Madrid, 7 de enero de 2014.

CNMC, 2014b. Informe de supervisión de mercados a plazo de energía eléctrica en España (noviembre 2013), 9 de enero de 2014, Comisión Nacional de los Mercados y la Competencia (CNMC), Madrid. Last date accessed 10 May 2014, available from: www.cnmc.es/Portals/0/Ficheros/Energia/Informes/Inf%20mens%20superv%20mdos%20pzo_NOV_13.pdf

CNMC, 2014c. Informe sobre la propuesta de Real Decreto por el que se establece la metodología de cálculo de los precios voluntarios para el pequeño consumidor y su régimen jurídico de contratación, 25 de febrero de 2014, Madrid.

Conchado, A., Linares, P., 2010. Estimación de los beneficios de la gestión activa de la demanda. Revisión del estado del arte y propuestas, Cuadernos Económicos de ICE, 79, pp. 187-212.

CRE, 2011. Le fonctionnement des marchés de gros de l’électricité, du CO2 et du gaz naturel, Rapport 2010-2011, Novembre 2011, Commission de Régulation de L’Énergie, Paris. Available from: www.cre.fr/documents/publications/rapports-thematiques/rapport-de-surveillance-des-marches-2010-2011 Last date accessed: 6 June 2013.

CRE, 2012. Observatoire des marchés de l’électricité et du gaz, 4º trimestre 2011, March 2, 2012, Commission de Régulation de L’Énergie, Paris. Available from: www.cre.fr/marches/observatoire-et-indicateurs-des-marches#section1 Last date accessed: 6 June 2013.

Critchlow, A., 2014. US shale no 'game changer' for LNG, say Qatar energy minister, posted in The Telegraph on 2 February 2014, last date accessed 4 February 2014, available from:

www.telegraph.co.uk/finance/newsbysector/energy/oilandgas/10611960/US-shale-no-game-changer-for-LNG-say-Qatar-energy-minister.html

http://www.cne.es/cne/doc/publicaciones/cne51_12.pdf

http://www.cnmc.es/Portals/0/Ficheros/Energia/Informes/Inf%20mens%20superv%20mdos%20pzo_NOV_13.pdf

http://www.cnmc.es/Portals/0/Ficheros/Energia/Informes/Inf%20mens%20superv%20mdos%20pzo_NOV_13.pdf

http://www.cre.fr/documents/publications/rapports-thematiques/rapport-de-surveillance-des-marches-2010-2011

http://www.cre.fr/documents/publications/rapports-thematiques/rapport-de-surveillance-des-marches-2010-2011

http://www.cre.fr/marches/observatoire-et-indicateurs-des-marches#section1

http://www.telegraph.co.uk/finance/newsbysector/energy/oilandgas/10611960/US-shale-no-game-changer-for-LNG-say-Qatar-energy-minister.html

http://www.telegraph.co.uk/finance/newsbysector/energy/oilandgas/10611960/US-shale-no-game-changer-for-LNG-say-Qatar-energy-minister.html


165

EEX, 2009a. Market Monitor by the Market Surveillance of EEX, Q4/2008. European Energy Exchange AG, Leipzig. Last date accessed 6 June 2013, available from: http://s3.amazonaws.com/zanran_storage/www.eex.com/ContentPages/17765613.pdf

EEX, 2009b. Growth Through Security, 2008, Annual Report, EEX, European Energy Exchange AG, Leipzig.

EEX, 2010. Power Trading Results on EPEX Spot and EEX Power Derivatives in 2009, Press Release, EPEX Spot European Power Exchange and EEX Power Derivatives, Leipzig, Paris 13 January 2010.

EEX, 2012a. Press release: EEX Review of 2011 – EEX asserts itself in a difficult market environment – Boom in short-term trading – Natural Gas Market reaches record level. 12 January 2012. European Energy Exchange AG, Leipzig. Available from: http://cdn.eex.com/document/103018/20120112_EEX_Jahreszahlen_2011.pdf Last date accessed: 6 June 2013.

EEX, 2012b. EEX Company and Products, as of February 2012. European Energy Exchange AG, Leipzig. Last date accessed 6 June 2013, available from:

http://cdn.eex.com/document/103364/EEX_Company_Products_E.pdf

EIA, 2014. Energy & Financial Markets, What Drives Crude Oil Prices?, An analysis of 7 factors that influence oil markets, with chart data updated monthly and quarterly. U.S. Energy Information Administration (EIA), Independent Statistics & Analysis, Washington. Last date accessed: 4 February 2014. Available from: www.eia.gov/finance/markets/

Endesa, 2013. Smartcity Málaga: a sustainable management model for cities of the future. Relevant projects, Endesa. Last date accessed: 27 November 2013. Available from:

www.endesa.com/en/aboutendesa/businesslines/principalesproyectos/malaga_smartcity

Energy Market Price, 2012. Spain halts the award of new feed-in tariff contracts, 23 March 2012. Last date accessed 6 June 2013, available from:

www.energymarketprice.com/SitePage.asp?Command=ArticleView&ID=8342 Energy Vortex, 2014. Energy Dictionary, last date accessed: 7 February 2014,

available from:

www.energyvortex.com/frameset.cfm?source=/energydictionary/energyvortex.htm

Engel, C., 1996. The forward discount anomaly and the risk premium: A survey of recent evidence. Journal of Empirical Finance, 3(2), pp. 123-192.

ERSE, 2011a. Regulamento n.º 496/2011, Aprovação dos novos Regulamentos: Regulamento de Relações Comerciais, Regulamento Tarifário e Regulamento do Acesso às Redes e Interligações. Entidade Reguladora dos Serviços Energéticos (ERSE), Diário da República, 2.ª série — N.º 159 — 19 de Agosto de 2011, pp. 34035-34322, Lisbon.

ERSE, 2011b. Directiva n.º 5/2011, Termos e Condições de realização de Leilões de Colocação de PRE, Entidade Reguladora dos Serviços Energéticos (ERSE), Diário da República, 2.ª série, N.º 226, 24 de Novembro de 2011, pp. 46437-46444, Lisbon.

ERSE, 2011c. Termos e Condições específicas do 1.º Leilão PRE. Comunicado ERSE, 15 de Dezembro de 2011, Lisboa.

http://s3.amazonaws.com/zanran_storage/www.eex.com/ContentPages/17765613.pdf

http://s3.amazonaws.com/zanran_storage/www.eex.com/ContentPages/17765613.pdf

http://cdn.eex.com/document/103018/20120112_EEX_Jahreszahlen_2011.pdf

http://cdn.eex.com/document/103364/EEX_Company_Products_E.pdf

http://www.eia.gov/finance/markets/

http://www.endesa.com/en/aboutendesa/businesslines/principalesproyectos/malaga_smartcity

http://www.endesa.com/en/aboutendesa/businesslines/principalesproyectos/malaga_smartcity

http://www.energymarketprice.com/SitePage.asp?Command=ArticleView&ID=8342

http://www.energyvortex.com/frameset.cfm?source=/energydictionary/energyvortex.htm


166

ERSE, 2011d. Resultados do 1.º Leilão PRE. Comunicado ERSE, 16 de Dezembro de 2011, Lisboa.

ERSE, 2014. Offical website of the PRE auctions, last day accessed 4 February 2014, available from:

www.erse.pt/pt/supervisaodemercados/mercadodeelectricidade/contratacao/Paginas/leiloesdePRE.aspx

European Commission, 2009a. Directive 2009/72/EC of the European Parliament and of the Council of 13 July 2009 concerning common rules for the internal market in electricity and repealing Directive 2003/54/EC. Available from: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:211:0055:0093:EN:PDF Last date accessed: 6 June 2013.

European Commission, 2009b. Directive 2009/73/EC of the European Parliament and of the Council of 13 July 2009 concerning common rules for the internal market in natural gas and repealing Directive 2003/55/EC. Available from: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:211:0094:0136:en:PDF Last date accessed: 6 June 2013.

European Commission, 2009c. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee, the Committee of the Regions and the European Central Bank. Ensuring efficient, safe and sound derivatives markets: Future policy actions, DG MARKT, COM(2009) 563/4, Brussels.

European Commission, 2011a. Proposal for a Regulation of the European Parliament and of the Council on markets in financial instruments and amending Regulation [EMIR] on OTC derivatives, central counterparties and trade repositories, Brussels, 20 October 2011, COM(2011) 652 final, 2011/0296 (COD). Available from: http://ec.europa.eu/internal_market/securities/docs/isd/mifid/COM_2011_652_en.pdf Last date accessed: 19 February 2014.

European Commission, 2011b. Proposal for a Directive of the European Parliament and of the Council on markets in financial instruments repealing Directive 2004/39/EC of the European Parliament and of the Council, Brussels, 20 October 2011, COM(2011) 656 final, 2011/0298 (COD). Available from: http://ec.europa.eu/internal_market/securities/docs/isd/mifid/COM_2011_656_en.pdf Last date accessed: 19 February 2014.

European Commission, 2011c. Proposal for a Regulation of the European Parliament and of the Council on insider dealing and market manipulation (market abuse), Brussels, 20 October 2011, COM(2011) 651 final, 2011/0295 (COD). Available from: http://ec.europa.eu/internal_market/securities/docs/abuse/COM_2011_651_en.pdf Last date accessed: 19 February 2014.

European Commission, 2011d. Quarterly Report on European Gas Markets, DG Energy, Market Observatory for Energy, Volume 4, Issue 3: July 2011–September 2011, Brussels.

European Commission DG TREN, 2008a. Review and Analysis of EU Wholesale Energy Markets. Evaluation of Factors Impacting on Current and Future Market Liquidity and Efficiency. Research Findings and Conclusions, 2nd July 2008. Prepared by the Moffatt Associates Partnership, London.

European Commission DG TREN, 2008b. Review and Analysis of EU Wholesale Energy Markets. Historical and current data analysis of EU wholesale electricity, gas and CO2 markets. Final report, 9 September 2008. Prepared by ECORYS Nederland BV, Rotterdam.

http://www.erse.pt/pt/supervisaodemercados/mercadodeelectricidade/contratacao/Paginas/leiloesdePRE.aspx

http://www.erse.pt/pt/supervisaodemercados/mercadodeelectricidade/contratacao/Paginas/leiloesdePRE.aspx

http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:211:0055:0093:EN:PDF


http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:211:0094:0136:en:PDF

http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:211:0094:0136:en:PDF

http://ec.europa.eu/internal_market/securities/docs/isd/mifid/COM_2011_652_en.pdf

http://ec.europa.eu/internal_market/securities/docs/isd/mifid/COM_2011_656_en.pdf

http://ec.europa.eu/internal_market/securities/docs/abuse/COM_2011_651_en.pdf


167

European Commission DG TREN, 2009. Draft discussion paper by DG TREN on transparency and integrity of traded wholesale markets in electricity and gas, 09 December 2009.

European Parliament, 2014. MEPs vote laws to regulate financial markets and curb high-frequency trading, Press Release, Committees Committee on Economic and Monetary Affairs, Plenary sessions [15-04-2014 - 18:58], Brussels, last date accessed 12 May 2014, available at: www.europarl.europa.eu/news/en/news-room/content/20140411IPR43438/html/MEPs-vote-laws-to-regulate-financial-markets-and-curb-high-frequency-trading

European Union, 2011. Regulation (EU) No 1227/2011 of the European Parliament and of the Council of 25 October 2011 on wholesale energy market integrity and transparency. Official Journal of the European Union, December 8, 2011, Brussels. Last date accessed: 12 March 2013, available from: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2011:326:0001:0016:EN:PDF Last date accessed: 19 February 2014.

European Union, 2012. Regulation (EU) No 648/2012 of the European Parliament and of the Council of 4 July 2012 on OTC derivatives, central counterparties and trade repositories. Official Journal of the European Union, July 27, 2012, Brussels. Last date accessed 19 February 2014, available from: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2012:201:0001:0059:EN:PDF

EWEA, 2010. Wind Energy and Electricity Prices, Exploring the ‘merit order effect’, A literature review by Pöyry for the European Wind Energy Association (EWEA), Brussels, April 2010.

Fabra, N., 2009. Mesa Redonda: Temas Avanzados en Economía Energética, IV Congreso de la Asociación Española para la Economía Energética, IPTS, Seville, January 2009.

Farrar, D. E., Glauber, R. R., 1967. Multicollinearity in Regression Analysis: The Problem Revisited. Review of Economics and Statistics, 49 (1), pp. 92–107.

Fattouh, B., Mahadeva, L., 2012a. Assessing the Financialization Hypothesis, Oxford Institute for Energy Studies, Working Paper 49, November 2012, Oxford.

Fattouh, B., Mahadeva, L., 2012b. Financialization in Oil Markets: lessons for policy, Oxford Institute for Energy Studies, Oxford Energy Comment, December 2012, Oxford.

Federico, G., Vives, X., 2008. Competition and regulation in the Spanish gas and electricity Markets. Reports of the Public-Private Sector Research Center, 1, IESE Business School, University of Navarra, pp. 117-141.

FERC, 2012. Assessment of Demand Response & Advance Metering. Staff Report. Federal Energy Regulatory Commission (FERC), December 2012, Washington. Available from: www.ferc.gov/legal/staff-reports/12-20-12-demand-response.pdf Last date accessed: 6 June 2013.

Fernández Domínguez, E., Xiberta Bernat, J., 2007. Restructuring and generation of electrical energy in the Iberian Peninsula. Energy Policy, 35, pp. 5117-5129.

Fisher Box, J., 1987. Guinness, Gosset, Fisher, and Small Samples. Statistical Science 2 (1), pp. 45–52.

Fleming, J., Ostdiek, B., 1999. The impact of energy derivatives on the crude oil market, Energy Economics, 21, pp. 135-167.

http://www.europarl.europa.eu/news/en/news-room/content/20140411IPR43438/html/MEPs-vote-laws-to-regulate-financial-markets-and-curb-high-frequency-trading







http://www.ferc.gov/legal/staff-reports/12-20-12-demand-response.pdf


168

Frías, P., Gómez, T., Cossent, R., Rivier, J., 2009. Improvements in current European network regulation to facilitate the integration of distributed generation. International Journal of Electrical Power & Energy Systems, 31, pp. 445-451.

Fuller, W. A., 1976. Introduction to Statistical Time Series, Ed. John Wiley and Sons, New York.

Furió, D., Meneu, V., 2010. Expectations and forward risk premium in the Spanish deregulated power market. Energy Policy, 38, pp. 784-793.

Furió, D., Chuliá, H., 2012. Price and volatility dynamics between electricity and fuel costs: Some evidence for Spain, Energy Economics, 2012, vol. 34(6), pp. 2058-2065.

Fusaro, P.C., 2011. Energy Risk Management. Hedging Strategies and Instruments for the International Energy Markets. McGraw-Hill, New York.

García Breva, J., 2012. Renovables sin complejos. Posted in Energías Renovables on 30 March 2012. Last date accessed: 6 June 2013, available from:

www.energias-renovables.com/energias/renovables/index/pag/perfil/colleft//colright/perfil/tip/articulo/pagant/perfil/pagid/19840/botid/33/

Gelabert, L., Labandeira, X., Linares, P., 2011. An ex-post analysis of the effect of renewables and cogeneration on Spanish electricity prices, Energy Economics, 33, pp. 59–65.

Gensler, G., 2012. Remarks on the Dodd-Frank Act at the 6th Annual Capital Markets Summit Hosted by the U.S. Chamber of Commerce, “Transparent Markets – Key to a Growing Economy”, Speeches & Testimony, Chairman Gary Gensler, March 28, 2012, Commodity Futures Trading Commssion (CTFC), Washington. Available from: www.cftc.gov/PressRoom/SpeechesTestimony/opagensler-110 Last date accessed: 6 June 2013.

González-Páramo, J.M., 2011. Risk, return, resilience: The future financial system. European Central Bank. Speech at 3rd Annual Risk and Return Russia Conference, Moscow, 14 April 2011.

Grammatikos, T., Saunders, A., 1983. Stability and the hedging performance of foreign curency futures. The Journal of Futures Markets, Vol. 3, pp. 295-305.

Granger, C., 1969. Investigating Causal Relations by Econometric Models and Cross-spectral Methods. Econometrica, Vol. 37, No. 3, pp. 424–438.

Granger, C., Newbold, P., 1974. Spurious regressions in econometrics. Journal of Econometrics, Vol. 2, No. 2, pp. 111–120.

Greene, W.H., 1993. Econometric Analysis, 2nd Edition, New York: Macmillan Publishing Company.

Haites, E., Missfeldt, F., 2004. Liquidity implications of a commitment period reserve at national and global levels. Energy Economics, 26, pp. 845-868.

Heather, P., 2012. Continental European Gas Hubs: Are they fit for purpose? The Oxford Institute for Energy Studies, NG 63, June 2012, Oxford.

Hogg, R.V., Craig, A.T., 1995. Introduction to Mathematical Statistics, 5th ed. New York: Macmillan.

Honoré, A., 2011. The Spanish Gas Market: Demand Trends Post Recession and Consequences for the Industry. NG 55, July 2011, The Oxford Institute for Energy Studies. Last date accessed: 6 June 2013, available from:

http://www.energias-renovables.com/energias/renovables/index/pag/perfil/colleft/colright/perfil/tip/articulo/pagant/perfil/pagid/19840/botid/33/



http://www.cftc.gov/PressRoom/SpeechesTestimony/opagensler-110


169

www.oxfordenergy.org/wpcms/wp-content/uploads/2011/08/NG_55.pdf

Hull, J.C., 2005. Options, Futures and Other Derivatives, Prentice-Hall.

Iberdrola Distribución Eléctrica, 2013. First Smart Network: Castellón. Smart Grids, Iberdrola Distribución Eléctrica, last date accessed: 27 November 2013. Available from:

www.iberdroladistribucionelectrica.com/webibd/corporativa/iberdrola?IDPAG=ENSOCDISDISREDPRI

IECA, 2010. Industrial Energy Consumers of America Applaud Senator Lincoln for Derivatives Reform Legislation. News Release, April 20, 2010. Industrial Energy Consumers of America (IECA), Washington, D.C.

Intermoney, 2012. El Mercado Eléctrico en España, enero 2012, Madrid, 15 de febrero de 2012, im energía sa, consultoría, p. 57.

IOSCO, 2011. Principles for the Regulation and Supervision of Commodity Derivatives Markets, Task Force on Commodity Futures Markets, International Organization of Securities Commissions Organisation (IOSCO), Madrid, September 15, 2011.

Joskow, P.L., 2010. Market Imperfections versus Regulatory Imperfections. MIT Working Paper, draft of Jun 20, 2010. Alfred P. Sloan Foundation and MIT, Massachussets, USA.

Joskow, P.L., Wolfram, C. D., 2011. Dynamic Pricing of Electricity, Working Paper, December 2011, MIT Economics, Massachussets. Available from: http://economics.mit.edu/files/7428 Last date accessed: 6 June 2013.

Karakatsani, N.V., Bunn, D.W., 2005. Diurnal Reversals of Electricity Forward Premia. London Business School Working Paper.

Kockar, I., Conejo, A.J., McDonald, J.R., 2009. Influence of the Emissions Trading Scheme on generation scheduling. International Journal of Electrical Power & Energy Systems, 31, pp. 465-473.

Krishna, R., 2008. Liquidity in the Dutch Wholesale Electricity Markets. MSc Economics Thesis. Tilburg University.

Köhlner,S., 2014a. Being clear on clearing. Energy Risk, March 2014.

Köhlner,S., 2014b. What price is a fair price. Energy Risk, April 2014.

Leyton, S., 2010. Feed-In Tariff, posted on 13 July 2010. Last date accessed: 4 February 2014, available from: http://centralenergia.cl/2010/07/13/feed-in-tariff/

Linares, P., Conchado, A., 2010. The Economics of New Nuclear Power Plants in Liberalized Electricity Markets, WP 04/2010, Economics for Energy. Available from: www.eforenergy.org/docpublicaciones/documentos-de-trabajo/WP04-2010.pdf Last date accessed: 6 June 2013.

Llewellyn, H., 2010. Spanish thermal on its knees as wind booms, Analysis – Spain’s generation surplus, Platts Power in Europe, Issue 569, February 8, 2010.

Lomax, R.G., Hahs-Vaughn, D.L., 2013. Statistical Concepts: A Second Course, 4th Edition.

Longstaff, F.A, Wang, A.W., 2004. Electricity Forward Prices: A High-Frequency Empirical Analysis. Journal of Finance, 59(4), pp. 1877-1900.

Lucia, J., Pardo, Á, 2008. On measuring speculative and hedging activities in futures markets from volume and open interest data. Applied Economics, pp. 1-9.

http://www.oxfordenergy.org/wpcms/wp-content/uploads/2011/08/NG_55.pdf

http://www.iberdroladistribucionelectrica.com/webibd/corporativa/iberdrola?IDPAG=ENSOCDISDISREDPRI

http://www.iberdroladistribucionelectrica.com/webibd/corporativa/iberdrola?IDPAG=ENSOCDISDISREDPRI

http://economics.mit.edu/files/7428

http://centralenergia.cl/2010/07/13/feed-in-tariff/

http://www.eforenergy.org/docpublicaciones/documentos-de-trabajo/WP04-2010.pdf


170

Lucia, J., Schwartz, E., 2000. Electricity prices and power derivatives: Evidence from the Nordic Power Exchange. Working Paper 16_00, Anderson Graduate School of Management, Finance, University of California, Los Angeles, USA.

Mansanet-Bataller, M., Lucia, J., Pardo, Á, 2012. Evolution of the speculative activity in the European carbon market. Working Paper, preliminary version of June 2012, last date accessed 10 November 2013, available at: www.lameta.univ-montp1.fr/IMG/pdf/12-06-11_Speculation_and_hedging_in_carbon_markets_Maria_Mansanet-Bataller.pdf

Martín Martínez, M.J., 2008. Mecanismos de contratación a plazo de energía eléctrica. Energía: desarrollos regulatorios en Iberoamérica. ARIAE, 2008, XII Reunión anual iberoamericana de reguladores de la energía. Ed. Civitas, Navarra, Spain, pp.375-391.

Martín Martínez, M.J., Villaplana Conde, P., 2009. Los Mercados a Plazo en el Sector Eléctrico, in: Tratado de regulación del sector eléctrico, tomo II: aspectos económicos, Ed. Thomson-Aranzadi, Cizur Menor (Navarra), pp. 437-459.

Matos, M., Lopes, J.P., Rosa, M., Ferreira, R., Leite da Silva, A., Sales, W., Resende, L., Manso, L., Cabral, P., Ferreira, M., Martins, N., Artaiz, C., Soto, F., López, R., 2009. Probabilistic evaluation of reserve requirements of generating systems with renewable power sources: The Portuguese and Spanish cases. International Journal of Electrical Power & Energy Systems, 31, pp. 465-473.

Maurer, L.T.A., Barroso, L.A., 2011. Electricity Auctions: An Overview of Efficient Practices. The International Bank for Reconstruction and Development / The World Bank, Washington DC. Last date accessed 4 February 2014, available from: www.ifc.org/wps/wcm/connect/8a92fa004aabaa73977bd79e0dc67fc6/Electricity+and+Demand+Side+Auctions.pdf?MOD=AJPERES

Meeus, L., 2011. Why (and how) to regulate power exchanges in the EU market integration context? Energy Policy, 39, pp. 1470-1475.

MIBEL Regulatory Council, 2009. Description of the operation of the MIBEL, November 2009, by CMVM, ERSE, CNMV, and CNE (Comissão do Mercado de Valores Mobiliários, Entidade Reguladora dos Serviços Energéticos, Comisión Nacional del Mercado de Valores y Comisión Nacional de Energía). Last date accessed: 11 September 2013, available from:

www.mibel.com/index.php?mod=pags&mem=detalle&relmenu=88&relcategoria=1057&idpag=97&lang=en

MIBEL Regulatory Council, 2011. Memorando de Entendimiento entre CMVM, CNE, CNMV y ERSE, para la cooperación y coordinación eficaz de la supervisión del Mercado Ibérico de Electricidad MIBEL, Madrid, 17 de Mayo de 2011. Last date accessed 4 February 2014, available from:

www.mibel.com/index.php?mod=pags&mem=detalle&relmenu=95&relcategoria=1069&idpag=104&lang=es

MINECO, 2013. Orden ECC/1796/2013, de 4 de octubre, por la que se determina la fecha de puesta en funcionamiento de la Comisión Nacional de los Mercados y la Competencia. Spanish Oficial Gazette (BOE) 239, 5 de octubre de 2013, p. 81839.

MINETUR, 2012a. Real Decreto-ley 1/2012, de 27 de enero, por el que se procede a la suspensión de los procedimientos de preasignación de retribución y a la supresión de los incentivos económicos para nuevas instalaciones de producción de energía eléctrica a partir de cogeneración, fuentes de energía renovables y residuos. MINETUR (Ministerio de Industria, Energía y Turismo). Spain’s Official Journal,

http://www.lameta.univ-montp1.fr/IMG/pdf/12-06-11_Speculation_and_hedging_in_carbon_markets_Maria_Mansanet-Bataller.pdf



http://www.ifc.org/wps/wcm/connect/8a92fa004aabaa73977bd79e0dc67fc6/Electricity+and+Demand+Side+Auctions.pdf?MOD=AJPERES

http://www.ifc.org/wps/wcm/connect/8a92fa004aabaa73977bd79e0dc67fc6/Electricity+and+Demand+Side+Auctions.pdf?MOD=AJPERES

http://www.mibel.com/index.php?mod=pags&mem=detalle&relmenu=88&relcategoria=1057&idpag=97&lang=en

http://www.mibel.com/index.php?mod=pags&mem=detalle&relmenu=88&relcategoria=1057&idpag=97&lang=en

http://www.mibel.com/index.php?mod=pags&mem=detalle&relmenu=95&relcategoria=1069&idpag=104&lang=es

http://www.mibel.com/index.php?mod=pags&mem=detalle&relmenu=95&relcategoria=1069&idpag=104&lang=es


171

BOE Number 24, January 28, 2012. Last date accessed: 6 June 2013, available from: www.boe.es/boe/dias/2012/01/28/pdfs/BOE-A-2012-1310.pdf

MINETUR, 2012b. Real Decreto-Ley 13/2012, de 30 de marzo, por el que se transponen directivas en materia de mercados interiores de electricidad y gas y en materia de comunicaciones electrónicas, y por el que se adoptan medidas para la corrección de las desviaciones por desajustes entre los costes e ingresos de los sectores eléctrico y gasista. Spain’s Official Journal, BOE Number 78, March 31, 2012. Available from: www.boe.es/boe/dias/2012/03/31/pdfs/BOE-A-2012-4442.pdf Last date accessed: 6 June 2013.

MINETUR, 2012c. El Gobierno reduce los costes del sistema eléctrico y limita la subida de la luz al 7% en abril, Press release, March 30, 2012. Available from: www.minetur.gob.es/es-es/gabineteprensa/notasprensa/documents/npmediasenergia300312.pdf Last date accessed: 6 June 2013.

MINETUR, 2014. Real Decreto 216/2014, de 28 de marzo, por el que se establece la metodología de cálculo de los precios voluntarios para el pequeño consumidor de energía eléctrica y su régimen jurídico de contratación. Spain’s Official Journal, BOE Number 77, March 29, 2014. Last date accessed: 9 May 2014, available from: www.boe.es/boe/dias/2014/03/29/pdfs/BOE-A-2014-3376.pdf

Ministério da Economia e da Inovação (Portugal), 2006. Portaria nº 643/2006, de 26 de Junho. Diário da República – I Série-B, nº 121 – 26 de Junho de 2006, pp. 4522-4523.

Ministério da Economia e da Inovação (Portugal), 2007a. Despacho nº 780/2007. Diário da República – II Série, nº 11 – 16 de Janeiro de 2007, pp. 1251-1252.

Ministério da Economia e da Inovação (Portugal), 2007b. Despacho nº /2007. Diário da República, 29 de Junho de 2007.

Ministério da Economia e da Inovação (Portugal), 2008. Despacho nº 19098/2008. Diário da República, Nº 137, Série II, 17 de Julho de 2008.

Ministério da Economia e da Inovação (Portugal), 2009. Despacho nº 125-A/2009. Diário da República, Nº 1, Série II, 2 de Janheiro de 2009.

MITyC, 2006a. Orden ITC/2129/2006, de 30 de junio, por la que se regula la contratación a plazo de energía eléctrica por los distribuidores en el segundo semestre de 2006. Ministerio de Industria, Turismo y Comercio (MITyC). Spanish Oficial Gazette (BOE) 158, 4 de julio de 2006, pp. 25020-25024.

MITyC, 2006b. Orden ITC/3990/2006, de 28 de diciembre, por la que se regula la contratación a plazo de energía eléctrica por los distribuidores en el primer semestre de 2007. BOE 312, 30 de diciembre de 2006, pp. 46680-46683.

MITyC, 2006c. Real Decreto 1634/2006, de 29 de diciembre, por el que se establece la tarifa eléctrica a partir de 1 de enero de 2007. BOE 312, 30 de diciembre de 2006, pp. 46656-46679.

MITyC, 2007a. Orden ITC/400/2007, de 26 de febrero, por la que se regulan los contratos bilaterales que firmen las empresas distribuidoras para el suministro a tarifa en el territorio peninsular. BOE 50, 27 de febrero de 2007, pp. 8205-8210.

MITyC, 2007b. Resolución de 19 de abril de 2007, de la Secretaría General de Energía, por la que se regulan las emisiones primarias de energía previstas en la disposición adicional vigésima del Real Decreto 1634/2006, de 29 de diciembre, por el que se revisa la tarifa eléctrica a partir del 1 de enero de 2007. BOE 98, 24 de abril de 2007, pp. 17933-17936.



http://www.minetur.gob.es/es-es/gabineteprensa/notasprensa/documents/npmediasenergia300312.pdf

http://www.minetur.gob.es/es-es/gabineteprensa/notasprensa/documents/npmediasenergia300312.pdf



172

MITyC, 2007c. Orden ITC/1865/2007, de 22 de junio, por la que se regula la contratación a plazo de energía eléctrica por los distribuidores en el segundo semestre de 2007 y en el primer semestre de 2008. BOE 152, 26 de junio de 2007, pp. 27314-27318.

MITyC, 2008a. Real Decreto 324/2008, de 29 de febrero, por el que se establecen las condiciones y el procedimiento de funcionamiento y participación en las emisiones primarias de energía eléctrica. BOE 69, 20 de marzo de 2008, pp. 16721-16723.

MITyC, 2008b. Resolución de 13 de mayo de 2008, de la Secretaría General de Energía, por la que se regulan las emisiones primarias de energía previstas en la disposición adicional única del Real Decreto 324/2008, de 29 de febrero, por el que se establecen las condiciones y el procedimiento de funcionamiento y participación en las emisiones primarias de energía eléctrica. BOE 129, 28 de mayo de 2008, pp. 24795-24798.

MITyC, 2008c. Orden ITC/1934/2008, de 3 de julio, por la que se regula la contratación a plazo de energía eléctrica por los distribuidores en el segundo semestre de 2008. BOE 162, 5 de julio de 2008, pp. 29517-29520.

MITyC, 2008d. Orden ITC/3789/2008, de 26 de diciembre, por la que se regula la contratación a plazo de energía eléctrica por los distribuidores en el primer semestre de 2009. BOE 314, 30 de diciembre de 2008, pp. 52430-52434.

MITyC, 2009a. Real Decreto 485/2009, de 3 de abril, por el que se regula la puesta en marcha del suministro de último recurso en el sector de la energía eléctrica. Spain’s Official Journal, BOE Number 82, April 4, 2009. Last date accessed: 5 June 2012, available from: www.boe.es/boe/dias/2009/04/04/pdfs/BOE-A-2009-5618.pdf

MITyC, 2009b. El Plan Contador ayudará a los consumidores a ahorrar electricidad y abaratar su factura, press release, June 8, 2009. Available from:

www.minetur.gob.es/es-es/gabineteprensa/notasprensa/documents/np%20envío%20cne%20plan%20contador%2008%2006%2009.pdf Last date accessed: 6 June 2013

MITyC, 2009c. Orden ITC/1659/2009, de 22 de junio, por la que se establece el mecanismo de traspaso de clientes del mercado a tarifa al suministro de último recurso de energía eléctrica y el procedimiento de cálculo y estructura de las tarifas de último recurso de energía eléctrica. BOE 151, 23 de junio de 2009, pp. 52252-52279.

MITyC, 2010a. Real Decreto 134/2010, de 12 de febrero, por el que se establece el procedimiento de resolución de restricciones por garantía de suministro y se modifica el Real Decreto 2019/1997, de 26 de diciembre, por el que se organiza y regula el mercado de producción de energía eléctrica. Spain’s Official Journal, BOE, Number 51, February 27, 2010. Last date accessed: 6 June 2013, available from: www.boe.es/boe/dias/2010/02/27/pdfs/BOE-A-2010-3158.pdf

MITyC, 2010b. Resolución de 11 de junio de 2010, de la Secretaría de Estado de Energía, por la que se aprueban las reglas de las subastas CESUR y se establecen las características de la undécima subasta CESUR. Spain’s Official Journal, BOE, Number 147, June 17, 2010. Last date accessed 6 June 2013, available from: www.subastascesur.omie.es/files/resolucion_see_reglas_11_cesur_con_correccion.pdf

MITyC, 2010c. Resolución de la Dirección General de Política Energética y Minas por la que se aprueban determinados parámetros de la subasta destinada a la adquisición de gas natural para la fijación de la tarifa de último recurso entre el 1 de julio y el 31 de diciembre de 2010. MITyC, 14 de junio de 2010, Madrid (España). Last date accessed 6 June 2013, available from:


http://www.minetur.gob.es/es-es/gabineteprensa/notasprensa/documents/np%20envío%20cne%20plan%20contador%2008%2006%2009.pdf




http://www.subastascesur.omie.es/files/resolucion_see_reglas_11_cesur_con_correccion.pdf

http://www.subastascesur.omie.es/files/resolucion_see_reglas_11_cesur_con_correccion.pdf


173

www.minetur.gob.es/energia/gas/Legislacion/2010/860_Resolucion_DGPEM_Precio_Salida_Subasta_TUR_Junio_2010_(sin%20Anexo).pdf

MITyC, 2011a. Resolución de 8 de febrero de 2011, de la Secretaría de Estado de Energía, por la que se fijan las cantidades de carbón, el volumen máximo de producción y los precios de retribución de la energía, para el año 2011 a aplicar en el proceso de resolución de restricciones por garantía de suministro. Spain’s Official Journal, BOE, Number 35, February 10, 2011, pp. 13927-13934. Available from: www.boe.es/boe/dias/2011/02/10/pdfs/BOE-A-2011-2542.pdf Last date accessed: 6 June 2013.

Molzahn, D., Singletary, C., 2011. An Empirical Investigation of Speculation in the MISO Financial Transmission Rights Auction Market. The Electricity Journal, Volume 24, Issue 5, pp. 57-68.

Monforte Martín, C., 2011. La especulación con la tarifa eléctrica ha costado a los españoles 726 millones, published in CincoDías, Madrid, on 21 March 2011, last date accessed: 4 February 2014, available from:

http://cincodias.com/cincodias/2011/03/21/empresas/1300718380_850215.html

Monforte, C., 2014. Endesa propone cambios en la Cesur como paso previo a eliminar la tarifa, published in Cinco Dias, Madrid, on 31 January 2014, last date accessed 4 February 2014, available from:


Moreno, F., 2011. Characterization of Peaks and Valleys of Electricity Demand. Application to the Spanish Mainland System in the Period 2000-2020. Energy and Power Engineering, 2011, 3, pp. 537-546.

Moreno, F., Martínez-Val, J.M., 2011. Collateral effects of renewable energies deployment in Spain: Impact on thermal power plants performance and management. Energy Policy, 39, pp. 6561–6574.

Mork, E., 2001. Emergence of financial markets for electricity: a European perspective. Energy Policy, 29, pp. 7-15.

Mundaca, L., 2008. Markets for energy efficiency: Exploring the implications of an EU-wide ‘Tradable White Certificate’ scheme. Energy Economics, 30, pp. 3016-3043.

Muñoz, M.P., Dickey, D.A., 2009. Are electricity prices affected by the US dollar to Euro exchange rate? The Spanish case, Energy Economics, 31, pp. 857–866.

Murray, M.P., 1994. A Drunk and Her Dog: An Illustration of Cointegration and Error Correction. The American Statistician, February 1994, Vol. 48, No. 1, pp. 37-39.

Müller-Monteiro, E., Moutinho dos Santos, E., 2010. Uso Político do Setor Elétrico Brasileiro: Uma matodologia de análise baseada na Teoria de Grupos de Pressão. Rio de Janeiro, Synergia Editora, FAPESP.

NASDAQ OMX Commodities, 2011. Market Report, December 2010. Nasdaq OMX Commodities and Nord Pool Spot, Lysaker. Available from: www.nasdaqomxcommodities.com/digitalAssets/72/72641_2010dec.pdf Last date accessed: 6 June 2013.

NASDAQ OMX Commodities, 2012. Market Report, December 2011. Nasdaq OMX Commodities and Nord Pool Spot, Lysaker. Available from: www.nasdaqomxcommodities.com/digitalAssets/77/77547_2011dec.pdf Last date accessed: 6 June 2013.

Newbery, D., von der Fehr, N.H., van Damme, E., 2003. Liquidity in the Dutch wholesale electricity market, Den Haag, May 2003.

http://www.minetur.gob.es/energia/gas/Legislacion/2010/860_Resolucion_DGPEM_Precio_Salida_Subasta_TUR_Junio_2010_(sin%20Anexo).pdf

http://www.minetur.gob.es/energia/gas/Legislacion/2010/860_Resolucion_DGPEM_Precio_Salida_Subasta_TUR_Junio_2010_(sin%20Anexo).pdf




http://www.nasdaqomxcommodities.com/digitalAssets/72/72641_2010dec.pdf

http://www.nasdaqomxcommodities.com/digitalAssets/77/77547_2011dec.pdf


174

Nord Pool, 2009. Nord Pool ASA, Annual Report, 2008, Lysaker.

NordREG, 2010. The Nordic financial electricity market, Report 8/2010, Nordic Energy Regulators (NordREG), Eskilstuna, November 2010.

OECD, 2014. About the OECD, The Organisation for Economic Co-operation and Development (OECD), Paris. Last date accessed: 4 February 2014, available from: http://www.oecd.org/about/

OFGEM, 2009. Liquidity in the GB wholesale energy markets. Discussion Paper, Ref. 62/09, 08 June 2009, Office for Electricity and Gas Markets (OFGEM), London. Last date accessed: 6 June 2013, available from:

www.ofgem.gov.uk/Markets/WhlMkts/CompandEff/Documents1/Liquidity%20in%20the%20GB%20wholesale%20energy%20markets.pdf

OFGEM, 2010. Liquidity Proposals for the GB wholesale electricity market. Consultation, 22 February 2010, London.

OMEL, 2010. El mercado de electricidad, octubre de 2010, Operador del Mercado Eléctrico (OMEL), Madrid.

OMEL, 2011. Resultados de la XVI subasta TUR electricidad, cuarto trimestre 2011, OMEL Mercados Agencia de Valores, S.A.U. Available from: www.subastascesur.omel.es/files/resultados_de_la_subasta_tur_electricidad_cuarto_trimestre_2011_2.pdf Last date accessed: 6 June 2013.

OMEL Mercados A.V., 2013. Resultados de la décima subasta de contratos financieros basados en la diferencia de precios entre el sistema eléctrico español y portugués, 17 de diciembre de 2013, OMEL MERCADOS AGENCIA DE VALORES S.A.U., Madrid. Last date accessed 15 May 2014, available at:

www.subastasdifprecios.omie.es/files/resultados_subasta_dif_precios_17_12_2013.pdf

OMIE, 2011. Resultados del mercado, OMI-POLO ESPAÑOL, S.A, (Iberian Market Operator, Spanish Pool). Available from:

www.omie.es/files/flash/ResultadosMercado.swf Last date accessed: 6 June 2013.

OMIE, 2013a. The Iberian Electricity Spot Market changes its gate closure time to 12:00h on October, 15th. Press release 14/10/2013, Iberian Market Operator, Spanish Pool (OMIE), Madrid. Last date accessed: 23 November 2013, available from: www.omie.es/en/home/market-information

OMIE, 2013b. Electricity market (Daily and intradaily). Iberian Market Operator, Spanish Pool (OMIE), Madrid. Last date accessed 25 November 2013, at: www.omie.es/en/home/markets-and-products

OMIE, 2014a. Power Exchanges agree on the European Cross-Border Intraday solution. Press release published in OMIE website on 10 February 2014, last date accessed: 25 February 2014, available from:

www.omie.es/files/20140210_press_release_agreement_on_delivery_of_european_intraday_market_0.pdf

OMIE, 2014b. Successful operation of PCR solution: Power Exchanges focus on Pan-European power market, PCR press release (Price Coupling of Regions). Published in OMIE website on 19 February 2014, last date accesed: 25 February 2014, available from:

www.omie.es/files/2014-02-19_PCR%20PR_Power%20Exchanges%20focus%20on%20Pan-European%20power%20market.pdf

http://www.oecd.org/about/

http://www.ofgem.gov.uk/Markets/WhlMkts/CompandEff/Documents1/Liquidity%20in%20the%20GB%20wholesale%20energy%20markets.pdf

http://www.ofgem.gov.uk/Markets/WhlMkts/CompandEff/Documents1/Liquidity%20in%20the%20GB%20wholesale%20energy%20markets.pdf

http://www.subastascesur.omel.es/files/resultados_de_la_subasta_tur_electricidad_cuarto_trimestre_2011_2.pdf

http://www.subastascesur.omel.es/files/resultados_de_la_subasta_tur_electricidad_cuarto_trimestre_2011_2.pdf

http://www.subastasdifprecios.omie.es/files/resultados_subasta_dif_precios_17_12_2013.pdf

http://www.subastasdifprecios.omie.es/files/resultados_subasta_dif_precios_17_12_2013.pdf

http://www.omie.es/files/flash/ResultadosMercado.swf

http://www.omie.es/en/home/market-information

http://www.omie.es/en/home/markets-and-products

http://www.omie.es/files/20140210_press_release_agreement_on_delivery_of_european_intraday_market_0.pdf

http://www.omie.es/files/20140210_press_release_agreement_on_delivery_of_european_intraday_market_0.pdf

http://www.omie.es/files/2014-02-19_PCR%20PR_Power%20Exchanges%20focus%20on%20Pan-European%20power%20market.pdf




175

OMIP, 2007. OMIP News, 4th Quarter 2007.

OMIP, 2008a. OMIP News, 1st Quarter 2008.

OMIP, 2008b. OMIP Instruction 01/2007, Market Makers, version of 18 November 2008.

OMIP, 2009a. Circular OMIP 01/2009, Preço de Referencia de Negociação, 02.março.2009.

OMIP, 2009b. OMIP News, 2nd Quarter 2009.

OMIP, 2009c. OMIP News, 3rd Quarter 2009.

OMIP, 2009d. OMIP News, 4th Quarter 2009. OMIP, 2011a. OMIP 2010 Results Show Strong Growth. Press release, 5 January

2011. Last date accessed 6 June 2013, available from:

http://omip.pt/Press/PressReleasestodos/tabid/100/nid/1183/language/en-GB/Default.aspx

OMIP, 2011b. OMIP Notice 04/2006. Price Variation Limits, version of 20 May 2011.

OMIP, 2011c. Circular OMIP 1/2011, Subasta de colocación de Producción en Régimen Especial, 14 diciembre 2011, OMIP, Lisboa.

OMIP, 2012a. Ficha Técnica, Contratos de Futuros MIBEL SPEL Base Financeiros, 24 Agosto 2012.

OMIP, 2012b. OMIP Instruction 04/2006, Auction Trading Mode, version of 7 September 2012.

OMIP, 2013. OMIP Notice 05/2007, Phases and Times of the Trading Session, version of 1 April 2013.

OMIP, 2014a. Derivatives Market Members. Last date accessed 9 January 2014, available at:

www.omip.pt/Members/DerivativesMarketMembers/tabid/88/language/en-GB/Default.aspx

OMIP, 2014b. OMIP Daily Market Bulletin, Nº 135 – VIII Série 2014/01/09. Available at: www.omip.pt/LinkClick.aspx?fileticket=bfUjnSYpaXk%3d&tabid=259&mid=739&language=es-ES Last date accessed: 19 February 2014.

OMIP-OMIClear, 2008a. MIBEL Derivatives Market Operational Guide, version of June 2008.

OMIP-OMIClear, 2012. Daily Market Bulletins of the MIBEL Derivatives Market, July 3, 2006–January 31, 2012. Last date accessed: 19 February 2014, available from:

www.omip.pt/Press/Bulletins/tabid/259/language/es-ES/Default.aspx OMIP-OMIClear, 2013a. Press release: EEX, ECC, OMIP and OMIClear to Cooperate

in Trade Registration of Power Derivatives, Leipzig and Lisbon, 12.December.2013, last date accessed 6 March 2014, available at:

www.omip.pt/LinkClick.aspx?fileticket=Da-G6_zzinE%3d&tabid=100&mid=781&language=es-ES

OMIP-OMIClear, 2013b. Press release: Successful first FTR exchange auction in Europe, Lisbon, 19.Dec.2013, last date accessed 15 May 2014, available at:

www.omip.pt/LinkClick.aspx?fileticket=_17k6NlCivQ%3d&tabid=100&mid=660&language=es-ES



http://www.omip.pt/Members/DerivativesMarketMembers/tabid/88/language/en-GB/Default.aspx

http://www.omip.pt/Members/DerivativesMarketMembers/tabid/88/language/en-GB/Default.aspx

http://www.omip.pt/LinkClick.aspx?fileticket=bfUjnSYpaXk%3d&tabid=259&mid=739&language=es-ES

http://www.omip.pt/LinkClick.aspx?fileticket=bfUjnSYpaXk%3d&tabid=259&mid=739&language=es-ES

http://www.omip.pt/Press/Bulletins/tabid/259/language/es-ES/Default.aspx

http://www.omip.pt/LinkClick.aspx?fileticket=Da-G6_zzinE%3d&tabid=100&mid=781&language=es-ES

http://www.omip.pt/LinkClick.aspx?fileticket=Da-G6_zzinE%3d&tabid=100&mid=781&language=es-ES

http://www.omip.pt/LinkClick.aspx?fileticket=_17k6NlCivQ%3d&tabid=100&mid=660&language=es-ES

http://www.omip.pt/LinkClick.aspx?fileticket=_17k6NlCivQ%3d&tabid=100&mid=660&language=es-ES


176

OMIP-OMIClear, 2014a. Press release, OMIP and OMIClear highlights 2013, Lisbon. 20 January 2014, last date accessed 15 May 2014, available at:

www.omip.pt/LinkClick.aspx?fileticket=fu_D3Ey2hGw%3d&tabid=100&mid=781&language=es-ES

OMIP-OMIClear, 2014b. Press release, First Trade on Spanish Power Options at OMIP/OMIClear, Lisbon, published on 30 May 2014.

Ordóñez, L., 2012. El gran reto energético es fijar un marco regulatorio serio donde cada tecnología sea rentable por sí misma. La nueva España, 22 March 2012. Available from: www.lne.es/asturama/2012/03/21/gran-reto-energetico-consiste-fijar-marco-regulatorio-serio-tecnologia-sea-rentable/1217467.html Last date accessed: 6 June 2013.

Pereira da Costa, A., Lage, S., 2006. MIBEL - mercado a prazo: caracterização e implicações fiscais, Cadernos do Mercado de Valores Mobiliários, N.º 25 - Dezembro 2006, Comissão do Mercado de Valores Mobiliários, Lisboa. Last date accessed: 7 November 2012, available from:

www.cmvm.pt/CMVM/Publicacoes/Cadernos/Documents/32d1f9ca47f04607b34f99ad40784051Artigo.pdf

Petrovich, B., 2013. European gas hubs: how strong is price correlation? The Oxford Institute for Energy Studies, NG 79, October 2013, Oxford. Available at: www.oxfordenergy.org/wpcms/wp-content/uploads/2013/10/NG-79.pdf Last day accessed: 7 April 2014.

Phillips, P.C.B., Perron, P., 1988. Testing for a Unit Root in Time Series Regression, Biometrika, 75(2), pp. 335–346.

Pietz, M., 2009. Risk premia in electricity wholesale spot markets: empirical evidence from Germany, CEFS working paper series, No. 2009-11.

Powernext, 2009. Market Activity 2008. Powernext SA, Paris.

Redl, C., Bunn, D., 2011. Determinants of the Premium in Forward Contracts, Working Paper, version March 2011, Vienna University of Technology & London Business School. Available from:

www.eeg.tuwien.ac.at/eeg.tuwien.ac.at_pages/publications/pdf/RED_PAP_2011_1.pdf Last day accessed: 10 February 2013

Redl, C., Haas, R., Huber, C., Böhm, B., 2009. Price formation in electricity forward markets and the relevance of systematic forecast errors. Energy Economics, 31, pp. 356–364.

REE, 2009. Red Eléctrica de España, Boletín Mensual, diciembre 2008, número 24. Fecha de edición: 27/01/09, Madrid.

REE, 2012. Informes mensuales, Red Eléctrica de España (REE), Madrid. Available from: www.ree.es/sistema_electrico/boletin_mensual_ree.asp Last date accessed: 6 June 2013.

Said, S.E., Dickey, D.A., 1984. Testing for Unit Roots in Autoregressive-Moving Average Models of Unknown Order, Biometrika, 71(3), pp. 599–607.

Samuelson, P., 1965. Proof that properly anticipated prices fluctuate randomly. Industrial Management Review, 6 (2), pp.41-49.

Sanders, D.R., Boris, K., Manfredo, M., 2004. Hedgers, funds, and small speculators in the energy futures markets: an analysis of the CFTC’s Commitments of Traders reports, Energy Economics, 26, pp. 425- 445.

http://www.omip.pt/LinkClick.aspx?fileticket=fu_D3Ey2hGw%3d&tabid=100&mid=781&language=es-ES

http://www.omip.pt/LinkClick.aspx?fileticket=fu_D3Ey2hGw%3d&tabid=100&mid=781&language=es-ES

http://www.lne.es/asturama/2012/03/21/gran-reto-energetico-consiste-fijar-marco-regulatorio-serio-tecnologia-sea-rentable/1217467.html

http://www.lne.es/asturama/2012/03/21/gran-reto-energetico-consiste-fijar-marco-regulatorio-serio-tecnologia-sea-rentable/1217467.html

http://www.cmvm.pt/CMVM/Publicacoes/Cadernos/Documents/32d1f9ca47f04607b34f99ad40784051Artigo.pdf

http://www.cmvm.pt/CMVM/Publicacoes/Cadernos/Documents/32d1f9ca47f04607b34f99ad40784051Artigo.pdf

http://www.oxfordenergy.org/wpcms/wp-content/uploads/2013/10/NG-79.pdf

http://www.eeg.tuwien.ac.at/eeg.tuwien.ac.at_pages/publications/pdf/RED_PAP_2011_1.pdf

http://www.eeg.tuwien.ac.at/eeg.tuwien.ac.at_pages/publications/pdf/RED_PAP_2011_1.pdf

http://www.ree.es/sistema_electrico/boletin_mensual_ree.asp


177

Sensfuß, F., Ragwitz, M., Genoese, M., 2008. The merit-order effect: a detailed analysis of the price effect of renewable electricity generation on spot market prices in Germany. Energy Policy, 36, pp. 3076–3084.

Serletis, A., 1992. Maturity effects in energy futures. Energy Economics, 14, pp.150-157.

Shawky, H.A., Marathe, A., Barrett, C.L., 2003. A First Look at the Empirical Relation between Spot and Futures Electricity Prices in the United States. The Journal of Futures Markets, Vol.23, Nr.10, pp. 931-955.

Shioshansi, F., 2013. FERC Survey: Millions Of Smart Meters, Only A Handful Smart Tariffs. Energy Policy Blog. Last date accessed 6 June 2013, available from: www.energypolicyblog.com/2013/02/02/ferc-survey-millions-of-smart-meters-only-a-handful-smart-tariffs/.

Steedman, I., 1987. Reservation price and reservation demand, The New Palgrave: A Dictionary of Economics, pp. 158-159.

Steel, R.G.D, Torrie, J.H., 1960. Principles and Procedures of Statistics with Special Reference to the Biological Sciences, Ed. McGraw Hill.

STEM, 2006. Finansiella elmarknaden. Statens energimyndigheten (STEM), Eskilstuna, Sweden.

Stern, J., Rogers, H., 2011. The Transition to Hub-Based Gas Pricing in Continental Europe, The Oxford Institute for Energy Studies, NG 49, March 2011, Oxford.

Student, 1908. The Probable Error of a Mean. Biometrika, Volume 6, Issue 1, Mar. 1908, pp. 1-25.

S&P Dow Jones Indices, 2014. Dow Jones UBS Commodity Indices. Last date accessed 7 February 2014, available from: www.djindexes.com/commodity/

Ullrich, C.J., 2007. Constrained Capacity and Equilibrium Forward Premia in Electricity Markets. Virginia Institute of Technology Working Paper.

Universia, 2011. Everything from Oil to Silver: Are Speculators Causing Too Much Volatility? Universia Knowledge Wharton, Finance and Investment, Publish Date: Jun 01, 2011.

U.S. Government, 2010. Public Law 111 - 203 – Dodd-Frank Wall Street Reform and Consumer Protection Act, July 21, 2010, Washington.

Villaplana, P., Cartea, A., 2011. Un análisis de la evolución de los precios a plazo de energía eléctrica en España. Los nuevos mercados energéticos, Fundación de Estudios Financieros, Colección Papeles de la Fundación nº 41, Madrid, pp. 193-242.

Villaplana, P., Cartea, A., 2012. Analysis of the main determinants of electricity forward prices and forward risk premia, in Quantitative energy finance: modelling, pricing and hedging in energy and commodity markets, Proceedings of the Wolfgang Pauli Institute special thematic year, Springer Verlag, pp. 1-29.

Walls, W., 1999. Volatility, Volume and Maturity in Electricity Futures. Applied Financial Economics, 9, pp. 283-287.

Zachmann, G., 2005. Convergence of Electricity Wholesale Prices in Europe? A Kalman Filter Approach, DIW Discussion Paper, 512.

http://www.energypolicyblog.com/2013/02/02/ferc-survey-millions-of-smart-meters-only-a-handful-smart-tariffs/

http://www.energypolicyblog.com/2013/02/02/ferc-survey-millions-of-smart-meters-only-a-handful-smart-tariffs/

http://www.djindexes.com/commodity/


178


179

ANNEX: LIST OF PUBLICATIONS

Publications in peer reviewed journals with Thomson Reuters Journal Citation Reports (JCR) impact factor:

Publications in Energy Policy (Elsevier), JCR 2012 impact factor 2.743 Journal official website: www.journals.elsevier.com/energy-policy/

Capitán Herráiz, A., Rodríguez Monroy, C., 2009b. Analysis of the efficiency of the Iberian power futures market. Energy Policy, 37, pp. 3566–3579.

Capitán Herráiz, A., Rodríguez Monroy, C., 2012b. Evaluation of the trading development in the Iberian Energy Derivatives Market. Energy Policy, 51, pp. 973–984.

Publication in International Journal of Electrical Power & Energy Systems

(Elsevier), JCR 2012 impact factor 3.432 Journal official website: www.journals.elsevier.com/international-journal-of-electrical-

power-and-energy-systems/

Capitán Herráiz, A., Rodríguez Monroy, C., 2013a. Analysis of the traded volume drivers of the Iberian power futures market. International Journal of Electrical Power and Energy Systems, 44, pp. 431–440.

Publication in IEEE Transactions on Power Systems (Institute of Electrical and

Electronics Engineers), JCR 2012 impact factor 2.921 Journal official website: www.ieee-pes.org/ieee-transactions-on-power-systems

Capitán Herráiz, A., Rodríguez Monroy, C., 2013c. Analysis of the Iberian Power Forward Price Formation. IEEE Transactions on Power Systems. Vol. 28, Issue 3, August 2013, pp. 2942 – 2949.

Publications in other peer reviewed journals:

Publication in Journal of industrial Engineering and Management (OmniaScience), indexed amongst others in Latindex and Scopus

Journal official website: www.jiem.org/index.php/jiem

Capitán Herráiz, A., Rodríguez Monroy, C., 2008b. Empirical evaluation of the efficiency of the Iberian power futures market.Journal of Industrial Engineering and Management, Vol 1, Nr 2, 2008, pp. 209-239.

Publication in International Journal of Financial Engineering and Risk

Management (Inderscience Publishers), indexed in Google Scholar Journal official website: www.inderscience.com/jhome.php?jcode=ijferm

http://www.journals.elsevier.com/energy-policy/

http://www.journals.elsevier.com/international-journal-of-electrical-power-and-energy-systems/

http://www.journals.elsevier.com/international-journal-of-electrical-power-and-energy-systems/

http://www.ieee-pes.org/ieee-transactions-on-power-systems

http://www.jiem.org/index.php/jiem

http://www.inderscience.com/jhome.php?jcode=ijferm


180

Capitán Herráiz, A., Rodríguez Monroy, C., 2013b. Analysis of the Iberian electricity forward market hedging efficiency. International Journal of Financial Engineering and Risk Management, Vol. 1, No. 1, 2013, pp. 20-34.

The article published in International Journal of Financial Engineering and Risk Management belongs to a Special Issue on Commodities Financial Management.

Publications in Congress Proceedings:

Publications in Industrial Organization Engineering Congress Proceedings of the Association of Spanish Industrial Organization Engineers (ADINGOR)

Capitán Herráiz, A., Rodríguez Monroy, C., 2008a. Evaluation of the Efficiency of the Iberian Power Futures Market. Proceedings of the II International Conference on Industrial Engineering and Industrial Management (CIO 2008), XII Congreso de Ingeniería de Organización, September 3-5, 2008, Burgos, Spain, pp. 597-606.

Capitán Herráiz, A., Rodríguez Monroy, C., 2011b. Analysis of the Iberian Power Futures Market Hedging Efficiency. 5th International Conference on Industrial Engineering and Industrial Management, XV Congreso de Ingeniería de Organización, Cartagena, 7 a 9 de Septiembre de 2011. Available from: http://oa.upm.es/9459/1/pag_259-265.pdf

Publications in Congress Proceedings of Latin American and Caribbean

Consortium of Engineering Institutions (LACCEI) Capitán Herráiz, A., Rodríguez Monroy, C., 2010a. Evolution of the MIBEL

Derivatives Market: a Model for Latin American Markets? Proceedings of the Eighth LACCEI Latin American and Caribbean Conference for Engineering and Technology (LACCEI’2010), “Innovation and Development for the Americas”, June 1-4, 2010, Arequipa, Perú.

The article presented in LACCEI 2010 Congress raised the interest of the Russian Non-Profit Partnership NP Market Council (Council for organizing Efficient System of Trading at Wholesale and Retail Electricity and Capacity Market) and was translated to Russian as a valuable resource for research purposes. More information about Market Council at: www.en.np-sr.ru/partnership/authority/supervisoryboard/

Capitán Herráiz, A., Rodríguez Monroy, C., 2011a. Assessment of the Operation of the Iberian Energy Forward Markets. Proceedings of the Ninth LACCEI Latin American and Caribbean Conference (LACCEI’2011), Engineering for a Smart Planet, Innovation, Information Technology and Computational Tools for Sustainable Development, August 3-5, 2011, Medellín, Colombia.

Capitán Herráiz, A., Rodríguez Monroy, C., 2012a. The first renewable forward market mechanisms in the Iberian Electricity Market. Proceedings of the Tenth LACCEI Latin American and Caribbean Conference (LACCEI’2012), Megaprojects: Building Infrastructure by fostering engineering collaboration, efficient and effective integration and innovative planning, July 23-27, 2012, Panama City, Panama.

http://oa.upm.es/9459/1/pag_259-265.pdf

http://www.en.np-sr.ru/partnership/authority/supervisoryboard/


181

Other publications: Publications in Congress of the Spanish Assocciation of Energy Economists

(AEEE, the Spanish section of the International Assocciation of Energy Economists, IAEE)

Capitán Herráiz, A., Rodríguez Monroy, C., 2009a. Evaluation of the Liquidity in the Iberian Power Futures Market. IV Congress of Spanish Association of Energy Economists (AEEE), Seville (Spain), January 2009.

Publications in Congress of the European Energy Markets (EEM, co-sponsored by the Power & Energy Society of the Institute of Electrical and Electronics Engineers, IEEE)

Capitán Herráiz, A., Rodríguez Monroy, C., 2010b. Analysis of the price efficiency in the Iberian Power Forward Contracting Mechanisms. Proceedings of the 7th International Conference on the European Energy Market, June 23-25, 2010, Madrid.