economic bulletin 44 - bankofgreece.gr · contents the economic behaviour of households in greece:...

BA

NK

OF

GR

EEC

E

EUR

OSY

STE

ME

CO

NO

MIC

BU

LL

ET

IN N

o 4

4D

EC

EM

BE

R 2

01

6

DECEMBER2016

ECONOMICBULLETIN

No 44

DECEMBER2016

ECONOMICBULLETIN

No 44

BANK OF GREECE21, E. Venizelos AvenueGR-102 50 Athens

www.bankofgreece.gr

Economic Analysis and Research Department - SecretariatTel. +30 210 320 2393Fax +30 210 323 3025

Printed in Athens, Greeceat the Bank of Greece Printing Works

ISSN 1105 - 9729

CONT ENT S THE ECONOMIC BEHAVIOUR OF HOUSEHOLDS IN GREECE: RECENT DEVELOPMENTS AND PROSPECTS 7Konstantina ManouEvangelia Papapetrou

EUROPE AS AN OPTIMUM CURRENCY AREA: THE EXPERIENCE OF THE BALTIC COUNTRIES 35Christos Papazoglou Dimitrios Sideris Orestis Tsagklas

THE EVOLUTION OF FINANCIAL TECHNOLOGY (FINTECH) 47Sofia Anyfantaki

INTRODUCTION TO SOLVENCY II FOR (RE)INSURANCE UNDERTAKINGS 63Ioannis Chatzivasiloglou

WORKING PAPERS(July – December 2016) 91

ARTICLES PUBLISHED IN PREVIOUS ISSUES OF THE ECONOMIC BULLETIN 99

44Economic BulletinDecember 20166

1 INTRODUCTION

Economic theory suggests that it is importantto study and analyse the consumption and sav-ing behaviour of households.1 Consumption,the largest component of aggregate demand,has an impact on total demand; thus, changesin consumption cause fluctuations in economicactivity. Moreover, the saving rate determinesthe economy’s capital stock, which in turninfluences the transition to a steady state andthe future growth potential of an economy.Household saving rates differ across EuropeanUnion (EU) countries, and various studiespoint out the heterogeneity observed in thesaving behaviour among EU households (seeRocher and Stierle 2015 and ECB 2016).Greece is among the countries with a negativesaving rate.

Households’ consumption and saving behav-iour has been studied according to several eco-nomic theories, such as the permanent incomehypothesis, the life cycle hypothesis, the socio-demographic hypotheses, uncertainty, etc.Research uses macroeconomic and microeco-nomic data to study and analyse households’saving/consumption behaviour and employseconometric models to estimate the economicrelationships among the variables.

Economists examine the underlying factors ofhousehold consumption, such as disposableincome, pointing out that its increase leads toan increase in consumption and vice-versa.Moreover in the economic literature, wealth isidentified as another factor that affects house-holds’ consumption behaviour. Householdspossess financial assets (currency in circula-tion, deposits, shares, etc.) and non-financialassets (such as houses, equipment, etc.), which

together make up their total wealth. When thevalue of these assets increases, households per-ceive themselves as wealthier and tend to con-sume more. Therefore, disposable income andwealth are the key determinants of householdconsumption.

Empirical findings regarding wealth effects onconsumption vary across countries and acrossperiods. However, it is recognised that aninvestigation of such effects needs to distin-guish between financial and non-financialwealth (see Cussen and Phelan 2010). Empir-ical evidence implies that financial wealth hasan impact on consumption in the euro area,while non-financial assets may not play a sig-nificant role (see Skudelny 2009). Other papersshow that non-financial (real) wealth affectshousehold consumption in the United Statesand the United Kingdom, probably reflectingdeeper housing markets. A recent study oneuro area countries has shown that both finan-cial and real wealth have a positive effect onconsumption, but the impact of financial assetsis stronger than that of real assets (see DeBonis and Silvestrini 2012).

Recent contributions to international litera-ture underline the role of deleveraging inhousehold saving and consumption. Specifi-cally, it is argued that the increased household

44Economic BulletinDecember 2016 7

THE E CONOM I C B EHAV I OUR O F HOU S EHOLD S I N G R E E C E : R E C EN T D E V E LOPMENT S AND P RO S P E C T S *

Konstantina Manou Bank of Greece, Economic Analysis and Research Department

Evangelia Papapetrou Bank of Greece, Economic Analysis and Research Department,and National and Kapodistrian University of Athens, Department of Economics

* The views expressed in this article are personal and do notnecessarily reflect those of the Bank of Greece. The authors wouldlike to thank Heather Gibson, the participants in the internalworkshop of the Economic Analysis and Research Department ofthe Bank of Greece, Hiona Balfoussia, Nikos Kamberoglou,Evangelia Georgiou, Giannis Asimakopoulos, Maria Albani, SteliosKorres and Olga Lymberopoulou for their useful remarks andcomments, as well as Konstantinos Kallias for his help with theprocessing of data. Warm thanks are also extended to ELSTAT,in particular Anastasios Nikolaidis (National AccountsDirectorate), for the supply of statistical data and useful remarks.

1 Throughout this article, reference is made to households only,although the household sector also includes non-profit institutionsserving households (NPISHs).

deleveraging observed in the recent period hasbeen associated with higher saving and lowerconsumption (see McCarthy and McQuinn2014, Glick and Lansing 2010). However, morerecent empirical research suggests that thedeleveraging effects on household consump-tion and saving are strongly heterogeneousacross countries and are closely linked with thefinancial conditions prevailing in each country(see Bouis 2015).

While disposable income and wealth and theirrelationship with consumption and saving havebeen extensively analysed in international lit-erature, studies focusing on the case of Greeceare scarce, and a comprehensive analysis ofhousehold financial wealth is not available asyet. The econometric investigation of theabove relationship is beyond the scope of thisarticle, which analyses household financialwealth and examines its evolution over time,with particular emphasis on the recent period.In addition, an in-depth analysis of the com-ponents of net financial wealth, i.e. assets andliabilities, as well as their composition, is per-formed. In the context of this analysis, thechanges in the size of household financialassets and liabilities are decomposed into thosestemming from: (a) transactions (e.g. changein investment choices in the case of assets ordebt increase/decrease in the case of liabili-ties); (b) changes in the valuation of assets/lia-bilities (valuation gains/losses); or (c) a com-bination of the above.

The sections below analyse Greek households’saving behaviour, initially through the inter-action between investment and debt and sub-sequently through the interaction between con-sumption and disposable income. To this end,saving is measured by two different methods,each based on a different statistical source.The first method is based on financialaccounts, whereby the financial definition ofsaving is derived. The second method uses datafrom non-financial (national) accounts, on thebasis of which the non-financial (traditional)definition of saving is derived. Finally, a briefreview of the evolution of disposable income

and its components, as well as of the evolutionof household real final consumption expendi-ture by functional purpose is performed, focus-ing mainly on the recent period.

The data used in this article are drawn fromthe Statistical Data Warehouse of the Euro-pean Central Bank (ECB), specifically thequarterly accounts of the euro area (Euro AreaAccounts, EAA), which provide detailed infor-mation on income, expenditure, financing andinvestment for the institutional sectors of eachcountry (households, non-financial corpora-tions, general government, financial corpora-tions and the external sector). These EAA areproduced by integrating the quarterly non-financial (national) accounts with the quarterlyfinancial accounts (central bank data) by insti-tutional sector and for all EU countries, whiletheir reliability, consistency and comparabilityare ensured by the European System ofNational and Regional Accounts (ESA 2010).2

It is the first time that these data are used inGreece for a study and analysis of householdfinancial wealth and saving behaviour throughthe interaction of investment and debt.

The remainder of the article is structured as fol-lows: Section 2 presents the evolution over timeof Greek households’ net financial wealth andhow it has been affected by the recent crisis.Section 3 looks at the composition of the house-hold asset portfolio and examines the relativecontributions of transactions and asset valua-tions to changes in household assets. Section 4analyses the components of household liabili-ties and examines the relative contributions oftransactions and valuations to changes in house-hold liabilities. Section 5 provides the financialand non-financial definition of saving. Section6 examines disposable income and its compo-


2 Non-financial accounts describe the value of goods and servicesresulting from the production activity, the income flows generatedby this process and their uses within the economy. On the otherhand, financial accounts describe the stocks of financial assets andliabilities. The two sets of accounts are linked with one another,as the deficit (or surplus) created by income and expenditure innon-financial accounts appears on the financial side as a decrease(increase) in financial assets and/or an increase (decrease) inliabilities (https://www.ecb.europa.eu/stats/prices/acc/html/index.en.html).

nents and Section 7 provides an overview of theevolution of households’ real consumptionexpenditure by functional purpose. Finally, Sec-tion 8 summarises and concludes.

2 EVOLUTION OF HOUSEHOLD NET FINANCIALWEALTH

Chart 1 shows the evolution of Greek house-holds’ net financial wealth, which is defined,according to ESA 2010, as the differencebetween total assets and liabilities, during theperiod from the first quarter of 2002 to the firstquarter of 2016.

The period under review can be divided intothree subperiods:

(a) The first subperiod is between the firstquarter of 2002 and the fourth quarter of 2007,when Greek households’ net financial wealthrose by 30.4% (average annual rate of changefor that period: 2.6%). This rise occurred as

the combined result of increases in total house-hold assets (up by 69.0%, or 7.3% in averageannual terms) and of total liabilities (up by254.1%,3 or 24.8% in average annual terms).Thus, in the fourth quarter of 2007, net finan-cial wealth came to €217.5 billion.

(b) The second subperiod is between the firstquarter of 2008 and the second quarter of2012, when household total liabilities contin-ued to increase (up by 21.7%, or 5.9% in aver-age annual terms), while at the same time theirtotal assets decreased (-33.8%, or -7.6% onaverage annually), as the upward trend in theprices of their financial assets had beenreversed in the context of the financial crisisthat evolved into a debt crisis. As a result, bythe second quarter of 2012 net financial wealthhad declined by 65.3% (18.4% on averageannually) compared with the first quarter of2008, to stand at €75.6 billion.


3 The substantial increase in liabilities is partly due to a strong baseeffect from the low stock of liabilities in the first quarter of 2002.

(c) The third subperiod lasts from the thirdquarter of 2012 to the first quarter of 2016,when household total liabilities decreased (-20.9%, or -5.2% in average annual terms),while their total assets increased (+12.9%, or3.4% on average annually). The combined out-come of these divergent developments was apositive impact on household net wealth, whichincreased by 79.8% (16.4% in average annualterms) relative to the third quarter of 2012 andcame to €135.9 billion in the first quarter of2016. Overall, between the first quarter of 2008and the first quarter of 2016, household netfinancial wealth declined by 37.5%.

Sections 3 and 4 below discuss in more detailthe components of net financial wealth, i.e.assets and liabilities.

3 ANALYSIS OF HOUSEHOLD ASSETS

Economic theory argues that the value of thehousehold portfolio, whether it refers to realassets (houses) or financial assets (shares,bonds, deposits, etc.), is influenced by a num-ber of factors, including the macroeconomicenvironment, changes in real estate prices, theinternational economic environment, etc. Inparticular, declines in real estate prices, thefinancial crisis, an uncertain political envi-ronment, as well as the recessionary phase ofthe business cycle are expected to have anadverse effect on the value of the householdportfolio. This section analyses householdfinancial wealth.4

According to international literature, theinteraction of the aforementioned factorsmotivates households to restructure their port-folios in an effort to preserve the value of theirassets and limit, to the extent possible, impair-ments (see Arrondel et al. 2014, Cussen andPhelan 2010, Cussen et al. 2012 and Cooper2013). This restructuring occurs either directlyas a result of transactions, i.e. acquisition ofnew assets or sale of assets already possessedby households (shares, bonds, deposits, etc.) ora targeted shift from high-risk financial assets

to less risky ones, or indirectly throughchanges in the value of financial assets, i.e.asset and exchange rate revaluations/devalu-ations, reclassifications, etc.

The analysis of the changing composition ofhousehold portfolios over time is important, asit reflects what part of these changes is due totransactions (investment choices) and whatpart is due to valuation adjustments. In the for-mer case, information is derived on the house-hold investment pattern, namely the risk, yieldand liquidity preferences of households, and itschanges. In the latter case, the focus is on theimpact of the macroeconomic environmentand stock exchange fluctuations on changes inthe value of the household portfolio, hence inits composition.

A recent study by Cussen et al. (2012) exam-ines, among other things, the behaviour ofhouseholds from 24 EU countries during therecent financial crisis. The authors find that in2008 almost all households in the sample5

shifted towards safer assets, such as currencyand deposits, away from shares and otherequity, thus showing a clear preference for lessrisky/lower-yield and more liquid assets.

3A COMPOSITION AND RESTRUCTURING OF THEHOUSEHOLD ASSET PORTFOLIO

Charts 2 and 3 illustrate the impact of macro-economic conditions and the recent financialcrisis on the restructuring of Greek house-holds’ asset portfolio.

As shown in Chart 2, prior to the financial cri-sis (average of the period from the first quarter2002 to the fourth quarter 2007), the composi-tion of the household financial asset portfoliowas as follows: shares and other equity: 21.5%;debt securities: 12.3%; investment fund shares:8.8%; and deposits: 48.5%, demonstratingGreek households’ preference for conservativeinvestment options. The role of the remaining


4 Data on the housing wealth of Greek households are not available,preventing an assessment of their total wealth.

5 Excluding Estonia, Lithuania and Luxembourg.

categories, such as currency, insurance techni-cal reserves and other accounts receivable/payable, was insignificant (see Zarco 2009).

After the onset of the financial crisis, risk aver-sion and a restructuring of the household port-folio can be observed, with a strong shift tosafer and more liquid financial assets, such asdeposits, and a marked decline in shares andother equity, debt securities and mutual fundshares/units as a percentage of the stock ofhousehold financial assets. Furthermore, inperiods of heightened uncertainty in Greece,even deposits lost their attractiveness and cur-rency holdings tended to increase.

In more detail, Chart 3 depicts the compositionof the household asset portfolio at four distinctpoints in time.

(a) Before the financial crisis (specifically inthe fourth quarter of 2007), shares and otherequity, debt securities and investment fund

shares together accounted for 40.2% of totalassets. Deposits represented a percentage of49.0%, which is particularly high consideringthe favourable macroeconomic environment atthe time, indicating risk aversion on the part ofGreek households, which favoured safer andmore liquid asset holdings. Currencyaccounted for a negligible percentage (3.2%).

(b) At the peak of the financial crisis in the sec-ond quarter of 2012, economic uncertaintyincreased households’ propensity to hoard,resulting in a substantial increase in the rela-tive weight of currency, from 3.2% of totalassets before the crisis to 17.6%. The relativeweight of deposits rose from 49.0% to 59.7%,mainly at the expense of shares and otherequity, which fell to 3.9% from 26.9%, anddebt securities, which fell to 6.3% from 8.3%of total assets.

(c) In the first quarter of 2014, while sharesand other equity and deposits remained


broadly unchanged as a percentage of totalassets relative to the pre-crisis period (at26.7% and 49.3%, respectively), the percent-age of currency increased to 10.3% from 3.2%)and that of debt securities fell to 1.4% from8.3%.

(d) The recent picture of the household port-folio (first quarter of 2016) reflects a newincrease in the percentage of currency hold-ings, to 17.1% from 10.3% in the first quarterof 2014, and a decline in the percentage ofshares and other equity to 18.1% from 26.7%.Deposits remained almost unchanged as a per-centage of total assets compared with the firstquarter of 2014.

3B ANALYSIS OF HOUSEHOLDS’ INVESTMENTCHOICES (TRANSACTIONS)

As mentioned above, the analysis of transac-tions focuses on the factors that can influence

households’ investment decisions and overallbehaviour under the prevailing economic cir-cumstances, especially in crisis periods. In par-ticular, the financial crisis motivates house-holds to invest in more liquid and less riskyassets, in an effort to reduce their exposure tofuture stock exchange fluctuations and miti-gate any further financial wealth losses. More-over, households’ reduced risk tolerance isreflected in their shift to safer assets, which canalso be used to hedge risky assets in their port-folio. The changes in the value of assets due totransactions are shown in Charts 4 and 5.

Specifically, Chart 46 shows household invest-ment in fixed assets (mainly houses),7 repre-senting a major share of total investment


6 The four-quarter moving average is used in the analysis and thecharts, in order to smooth out the seasonality of the series.

7 Houses account for the bulk of household gross capital formation,which also includes purchases of equipment by sole proprietors, aswell as cultivated assets (ESA 2010).

before the crisis, as well as in financial assets,between the first quarter of 2006 and the firstquarter of 2016.

Households seem to have reduced their resi-dential investment starting from the first quar-ter of 2008, i.e. half a year before real estateprices began to decline. Subsequently, in linewith the continued downward trend of realestate prices, residential investment fell dra-matically, to €1.2 billion in the first quarter of2016 from €8.1 billion in the first quarter of2008, in terms of four-quarter moving sums.8

Besides, the decline in household financialinvestment started from the third quarter of2007 and intensified as from the first quarterof 2009, on the backdrop of rising unemploy-ment and shrinking household disposableincome.

Furthermore, Chart 5 shows the compositionof household investment flows (transactions inassets), in four-quarter moving averageterms.9

In the period before 2009, deposits were thepreferred investment instrument for house-holds, suggesting their caution towards riskyassets. Similar findings are reported in a Euro-stat study on the financial assets and liabilitiesof households in the EU, according to which,in 2007, Greece had the second highest share,after Slovakia, of currency and deposits in thetotal financial portfolio of households (seeZarco 2009).

Next in households’ preferences, and reflect-ing a search of higher yield associated withhigher corresponding risk, came investment inshares and other equity, followed by debt secu-rities. Insurance technical reserves were verylow, as were also households’ currency hold-ings. It is worth noting that, even the period ofstrong economic growth in Greece (2004-2007)


8 As mentioned in the Monetary Policy 2015-2016 report of the Bankof Greece (June 2016), the cumulative decline in the average levelof house prices between the onset of the economic crisis in 2008and the first quarter of 2016 was close to 41.3% in nominal terms.

9 It should be noted that investment in deposits, debt securities,shares and investment fund shares also includes those held abroad.

saw household portfolio shifts away frominvestment fund shares, whose yields were seenas less attractive.

This picture changed markedly following theonset of the economic crisis, when house-holds, more manifestly as from early 2010,tended to withdraw deposits and increasetheir currency holdings amid an adverse eco-nomic environment and heightened economicuncertainty. At the same time, they drasticallyreduced their net investment in shares anddebt securities10 and virtually eliminated theirnet investment in insurance products. House-holds’ strong shift to safer assets may alsoreflect, apart from their increased risk aver-sion under conditions of high uncertainty, theneed to hedge risky assets already held in theirportfolio.

In the first half of 2015, amid mounting eco-nomic uncertainty as a result of protracted

negotiations with Greece’s creditors, house-holds proceeded to mass deposit withdrawals,while at the same time increasing their cur-rency holdings (hoarding) and investment inforeign investment fund shares (mainly euromoney market funds). The imposition of cap-ital controls on 28 June 2015 reined in hoard-ing and deposit withdrawals by households.

3C ANALYSIS OF VALUATION ADJUSTMENTS OFHOUSEHOLD ASSETS

The changes in household total financial assetsas a result of valuation changes or statisticalreclassifications are shown in Chart 6.

The conclusions of the analysis are the fol-lowing:


10 Transactions in debt securities in 2011-2012 were related to theexchange of Greek government bonds under the private sectorinvolvement (PSI) programme.

(a) The most part of household asset devalu-ation took place in the early phase of the cri-sis (2008-2009) and continued at a weaker pacein the next three years (2010-2012). By con-trast, in the period from the fourth quarter of2012 to the third quarter of 2014, the value ofthe household financial portfolio rebounded,in line with the improving overall economicenvironment. This was followed by a newdevaluation in the first half of 2015, in the con-text of heightened uncertainty.

(b) Revaluations of household total assets weredriven primarily by shares and, secondarily,debt securities. The role of investment fundshares, insurance, pension schemes and otheraccounts receivable was negligible.

(c) In the period from the first quarter of 2008to the second quarter of 2012, shares and otherequity held in the household portfolio weresubject to massive valuation losses, which were

―to a large extent― recouped by the firstquarter of 2014, before recording new lossesuntil the first half of 2016.

3D THE RELATIVE CONTRIBUTIONS OFTRANSACTIONS AND VALUATION ADJUSTMENTS TO THE CHANGE IN HOUSEHOLDTOTAL ASSETS

The relative contributions of transactions andvaluation effects to the annual percentagechanges in household total assets are shown inChart 7.

Prior to 2008, the change (increase) in thevalue of household total assets was almostexclusively due to transactions, with the excep-tion of the period from the fourth quarter of2004 to the fourth quarter of 2005, when val-uation effects also had a positive contribution.A different picture can be seen later on, par-ticularly in the first phase of the financial cri-


sis (2008-2009), when the impairments ofhousehold assets could not be offset by thepositive flows (transactions) that were also tak-ing place at the time.11 As a result, the value ofhousehold total assets shrank.

In 2010-2012, household assets continued toshow sharp declines, this time attributable alsoto negative transactions (disinvestment),besides adverse valuation effects. By contrast,in the period 2013-2014, household assetsrebounded strongly, as a result of valuationgains that more than offset the comparativelylow negative transactions. From 2015 onwards,extensive devaluations and, less importantly,disinvestment drove household assets down.

4 ANALYSIS OF HOUSEHOLD LIABILITIES

4A HOUSEHOLD LIABILITIES

The years that preceded the financial crisis,especially the 2004-2007 period of strong eco-

nomic growth in Greece, saw a continuous andsharp increase in households’ debt liabilities,12

mainly associated with the financing of resi-dential investment, as shown in Chart 8.

Similar findings for various industrial countriesare reported by Bȇ Duc and Le Breton (2009)and Glick and Lansing (2010). Glick and Lans-ing (2010) analyse the relationship betweenhousehold leverage, house prices and con-sumption for various industrial countries,including the United States, focusing on thedecade preceding the 2007 financial crisis.They argue that the larger the expansion in theuse of borrowed money (leverage), the morerapid the rises in house prices in the countries


11 Cussen and Phelan (2010), using data on Irish households, showthat, in the years preceding the financial crisis (2003-2007), changesin the value of household assets were almost equally driven bytransactions and valuation effects. Subsequently, during the crisisof 2007-2009, the pattern of the decline in total assets of Irishhouseholds is similar to that observed for Greek households, withthe contribution of valuation changes exceeding that oftransactions.

12 As mentioned in footnote 3, household liabilities did increase, butfrom a low level in comparison with the other euro area countries.

examined. Also, when economic conditionsdeteriorated and house prices started falling,the negative impact on consumption was largerfor countries with high household leverage.

Greek households’ total liabilities increasedsteadily from the first quarter of 2002 to thethird quarter of 2010,13 when they peaked, hav-ing increased by 347% during that period. It isworth noting that the increases, albeitweaker, continued beyond the first quarter of2008 that marks the start of the decline inhouseholds’ real property values. However,during the deterioration of the crisis and withthe implementation of economic adjustmentprogrammes for Greece from 2010 onwards,households gradually began to reduce their netdebt incurrence.

In more detail, some of the factors that con-tributed in this direction were the adoption bybanks of tighter credit standards, as suggestedby the results of Bank Lending Surveys, alongwith interest rate increases for the main loan

categories. At the same time, as pointed out inthe Annual Report 2010 of the Bank of Greece(April 2011), the deterioration in macroeco-nomic conditions (rising unemployment,falling disposable income) also played a majorrole, as it affected the financial condition ofhouseholds by reducing their capacity as wellas their willingness to take on new debt. Over-all, between the fourth quarter of 2010 and thefirst quarter of 2016, households reduced theirtotal debt liabilities by 23.8% relative to thepeak observed in the third quarter of 2010.

4B THE COMPOSITION OF HOUSEHOLDLIABILITIES

The main categories that make up householdliabilities are loans and other accounts payable,as shown in Chart 9. Total household liabilities


13 The sharp increase in household total liabilities observed since June2010 largely reflects statistical reclassifications. Specifically, fromthat month onwards, loans to sole proprietors, farmers andunincorporated enterprises were reclassified from corporate loansto household loans, and loans to religious institutions are includedin loans to private non-profit institutions.

were 21.7% lower in the first quarter of 2016relative to the peak of the third quarter of2010.

Within household debt liabilities, the mainsubcategories are housing loans, consumerloans, loans to sole proprietors (as from June2010) and other loans.14 As shown in Chart 10,household liabilities almost quadrupled, bothfor housing and consumer loans, from 2002 toJune 2010.15 Thereafter, the stock of theseloans started to decline, also in the context ofintensified loan restructuring by banks. Soleproprietors, farmers and unincorporatedenterprises began to reduce their debt liabili-ties as from January 2011.

Between January 2002 and March 2016, thecomposition of household loans remainedremarkably stable, with housing loans account-ing for 70% of total loans and consumer loansfor 30%.

Other accounts payable comprise various lia-bilities that have not fallen due, mainly tax andsocial security liabilities, liabilities vis-à-visnon-financial corporations such as the PublicPower Corporation (DEH), and other liabili-ties arising from trade transactions. Overall,other accounts payable peaked in the secondquarter of 2012, falling by 51.1% thereafteruntil the first quarter of 2016.

4C LEVERAGE RATIOS

The expansion in the use of borrowed moneyby households between the first quarter of 2006and the first quarter of 2016 can be measuredby two ratios, as illustrated in Chart 11:

The first is debt-to-income ratio, i.e. debt as apercentage of disposable income, and isdefined as follows:

Total liabilitiesDebt-to-income ratio= Disposable income

Thus, while in the fourth quarter of 2006households’ debt represented 68.5% of theirdisposable income, in the first quarter of 2010

it amounted to 81%. This development was thecombined result of countervailing effects, i.e.an increase in both household disposableincome and net borrowing (change in the nom-inal debt stock). As the increase in net bor-rowing more than offset the increase in dis-posable income, the ratio showed this strongrise. Subsequently, the ratio kept increasing,16

mainly due to a reversal of the upward trendof disposable income, and peaked at about109.8% in the second quarter of 2014. Sincethen, the ratio gradually declined (first quar-ter of 2016: 101.8%), as households reduced,although at a slow pace, their net borrowingand the decrease in disposable income wasweaker. Still, the ratio remains high comparedwith its pre-crisis levels, as for every €100 ofdisposable income, households have higher lia-bilities (€101.8).

The second leverage ratio is the debt-to-assetsratio, measuring the extent to which householdassets have been financed with debt, and isdefined as follows:

Total liabilitiesDebt-to-assets ratio=

Total assets

Between the first quarter of 2006 and the sec-ond quarter of 2012, this ratio rose substan-tially, more than doubling from about 30.6% to66.5%, thus indicating the high leverage ofhouseholds. This development reflected, apartfrom the increase in household debt, a paralleldecrease in the value of their assets as a resultof the debt crisis.17 Subsequently, the ratioimproved noticeably, falling to 45.3% by thesecond quarter of 2014, due to a recovery in thevalue of household assets and a decline in net


14 Total household loans also include some other categories, such asloans from other financial institutions (OFIs), insurancecorporations and occupational pension funds. As these involvesmall amounts, they are not reported separately.

15 Brissimis et al. (2012) examine the factors that determined theevolution of consumer credit in Greece in the recent past.Papapetrou and Lolos (2011) investigate the interdependencebetween housing credit and the labour market.

16 In addition to declining disposable income, this also partly reflectsstatistical reclassifications affecting total household liabilities.

17 As calculated over the same period, the debt-to-assets ratio forthe euro area also showed an increase (from 31.0% in the firstquarter of 2006 to 36.0% in the second quarter of 2012), whichwas far smaller than in the corresponding Greek ratio. In the firstquarter of 2016, the debt-to-assets ratio for the euro area stoodat 31.7%.

borrowing. In the first quarter of 2016, totalhousehold liabilities corresponded to 46.6% oftheir total assets. The small increase in leverageobserved more recently is mainly due to valu-ation losses on household assets, which werenot offset by the decrease in their debt.

International literature points out that theseratios should be interpreted with some caution,as it is important to examine, among otherthings, how debt is allocated among house-holds and how it is linked with the expectedpath of household income. Furthermore, it isargued that the debt-to-assets ratio does notfully reflect households’ debt servicing capac-ity, given that certain financial assets are, bytheir nature, illiquid and thus cannot be usedfor immediate debt reduction, particularly intimes of adverse economic conditions.

4D THE RELATIVE CONTRIBUTIONS OFTRANSACTIONS AND VALUATION ADJUSTMENTSTO CHANGES IN HOUSEHOLD LIABILITIES

Deleveraging, i.e. household debt reduction,can be achieved through repayment or

through changes in the outstanding amount ofdebt due to write-offs/write-downs.18 Chart 12shows the relative contributions of these twofactors (referred to as transactions and adjust-ments, respectively) to the annual percentagechange in household total liabilities.19 As seenfrom the chart, households increased their lia-bilities at high rates until the first quarter of2006, followed by a reversal of this trend, moremanifestly as from the third quarter of 2007with the onset of the financial crisis. The firstsigns of deleveraging are visible as from thefourth quarter of 2010, with the reduction intotal liabilities being almost exclusivelydriven by transactions (i.e. net debt repay-ment) until the beginning of 2012 and, lateron, particularly in the period from the second


18 The credit institution may decide to write off the entire debt, if allextrajudicial and judicial actions have been exhausted and nofurther recovery can be expected, or to write down part of the debtso that the remaining part is reduced to a level likely to be servicedwithout problems.

19 Adjustments include any statistical reclassifications and/or changesin the outstanding amount of loans due to exchange rate variations.The latter applies e.g. to many households that have borrowed inforeign currency, notably the Swiss franc: when this currencyappreciated, it had an upward effect on the debt liabilities of thehouseholds concerned.

quarter of 2012 to the third quarter of 2014,by adjustments as well.20

5 HOUSEHOLD SAVING: FINANCIAL AND NON-FINANCIAL DEFINITION

Households’ saving behaviour can be analysedeither through the interaction between theirinvestment and debt (financial definition ofsaving) or through the interaction betweenconsumption and disposable income (tradi-tional or non-financial definition of saving).

5A HOUSEHOLD SAVING: FINANCIAL DEFINITION

Households may choose to channel a part oftheir income, which they do not spend on con-sumption, into investment in financial assets(shares, deposits, bonds, etc.) or non-financialassets21 (houses, equipment, etc.). They alsohave the option to borrow, if their income isnot sufficient to finance such investments. The

financial definition of saving, according toBerry et al. (2009), Cussen and Phelan (2010)and Cussen et al. (2012), can be derived as fol-lows:

Funds raised ≡ Funds invested (1)

Relationship (1) can be rewritten as follows:

Saving+Net borrowing=Net investment in financial assets+

Net investment in non-financial assets2 (2)

Rearranging relationship (2), saving can beexpressed as follows:


20 To a large extent, the size and the relative contribution of the“adjustments” component reflect the impact of bank resolutionsand loan write-offs/write-downs during that period.

21 Non-financial investment includes, in addition to gross fixed capitalformation (houses, equipment, etc.), also acquisitions less disposalsof non-produced assets (patents, intellectual property rights, leasesof land or buildings, etc.).

22 In the case of net lending/borrowing, “net” refers to the incurrenceof new debt minus repayment of existing debt. Similarly, in the caseof investment, “net” refers to the acquisition of new assets minusdisposal of existing assets.

Saving=Net investment in financial assets –

Net borrowing+Net investment in non-financial assets (3)

Therefore, from (3) it is deduced that:

Saving = Net lending/borrowing+

Net investment in non-financial assets (4)

Thus, an increase in saving results from a netincrease in some type of investment, a netdecrease in debt or a combination of both.

Chart 13 shows the prevailing trends in house-hold net lending/borrowing,23 as derived fromtheir transactions on the asset side minus trans-actions on the liability side. A positive sign sug-gests that households are net lenders, i.e.acquisition of financial assets exceeds netincurrence of debt, or net repayment of debtexceeds net disinvestment.

Specifically, in the period up to the first quar-ter of 2010, households were net lenders, astheir net financial investment more than offsetnet incurrence of new debt. Two distinct sub-periods can be identified:

(a) The first subperiod is between the first quar-ter of 2002 and the third quarter of 2007, whenhouseholds invested in financial assets, while atthe same time constantly and considerablyincreasing their net borrowing. As their risinginvestment more than offset their rising liabil-ities, households ended up as net lenders.

(b) In the second subperiod, from the fourthquarter of 2007 to the first quarter of 2010,households started to gradually reduce both


23 A net lending position of households in non-financial accountsimplies that households have a surplus of funds can thus financethe other institutional sectors of the economy (non-financialcorporations, financial corporations, general government and theexternal sector). Conversely, a net borrowing position implies thathouseholds are net borrowers, i.e. they face a deficit of funds andneed to obtain financing from the other sectors.

their net lending and their net financial invest-ment, thus remaining net lenders in this sub-period too.

Thereafter, households became net borrowers,as their net debt reduction, which started fromthe fourth quarter of 2010, fell short of theirnet disinvestment. Exceptions were the periodfrom the fourth quarter of 2011 to the secondquarter of 2012 and the more recent periodfrom the second quarter of 2015 to the firstquarter of 2016, when households were againnet lenders.

Between the third quarter of 2013 and the firstquarter of 2015, although households pro-ceeded to increased net debt reduction, theirnet disinvestment was even larger, makingthem net borrowers. From the second quarterof 2015 onwards, net debt reduction retainedits momentum, but was accompanied byweaker net disinvestment, which made house-holds net lenders.

From relationship (4) and under its financialdefinition, saving is the sum of net investmentin non-financial assets (mostly houses) and infinancial assets, minus net transactions in lia-bilities. Therefore, by introducing also non-financial investment (houses, etc.) in the analy-sis, we can obtain saving according to its finan-cial definition (see Chart 14).

As shown in Chart 14, household savingremained at high levels until the fourth quar-ter of 2008, supported by strong investment inreal estate, other non-financial assets andfinancial assets, which outweighed net incur-rence of new debt. The subsequent downwardtrend in saving is explained by the fact that thedecline in total household net investment out-paced the decline in their liabilities. Actually,saving turned negative in the first quarter of2011, although households had started,already from the fourth quarter of 2010, thenet repayment of their existing debt. This neg-ative outcome reflected both reduced invest-


ment mainly in houses and strong net disin-vestment of financial assets. Since then, savinghas alternated from positive to marginally neg-ative values, depending on which componentdominates each time.

From an analysis of data for the recent period,a number of significant conclusions can bedrawn:

(a) As from the second quarter of 2013, house-holds intensified their net reduction of liabil-ities, and this trend continued into the nextquarters. It is important to note that house-holds proceeded to a net reduction of their lia-bilities for the first time in late 2010.

(b) As from the second quarter of 2014, savingfollowed an upward trend, which was sup-ported mainly by deleveraging rather than non-financial investment, which had been drasti-cally curtailed by then. Whenever these twoexplanatory factors together exceeded net dis-investment of financial assets, saving was inpositive territory, a fact that is more manifestin more recent quarters.

(c) Saving is exceptionally low compared withits pre-crisis levels, because of a contraction ofinvestment in houses/equipment and large dis-investment of financial assets.

5B HOUSEHOLD SAVING: NON-FINANCIALDEFINITION

Economic theory posits that household dis-posable income is channelled either into con-sumption or into saving. Consequently,household saving, according to the analysis ofnon-financial accounts, is defined as house-hold disposable income, minus consumption,that is:

Saving = Disposable income – Consumption (5)

As shown in Chart 15, the household savingrate,24 in terms of four-quarter moving sums,followed a downward path as from the secondquarter of 2009, falling from 7.4% in that quar-

ter to -5.6% in the first quarter of 2016. We canidentify two subperiods: the first subperiod isup to the first quarter of 2012, when the house-hold saving rate was positive, and the secondsubperiod, when it was negative. In general,when the rate of increase (decrease) in dis-posable income falls short of the rate ofincrease (decrease) in consumption, the savingrate decreases (increases). It is pointed outthat the decline in household saving ―partic-ularly from end-2009 onwards― is linked withthe fall in disposable income. Specifically, thecontinuous downward trend in the householdsaving rate from the second quarter of 2012 tothe second quarter of 2014, i.e. when the ratiowas in negative territory, was due to the factthat the decrease in disposable income out-paced the decrease in consumption. As fromthe second quarter of 2014, the household sav-ing rate stabilised at low levels, standing at -5.6% in the first quarter of 2016 (comparedwith -5.3% one year earlier), as the rate ofdecrease in disposable income (-1.7%)exceeded the rate of decrease in consumption(-1.4%).

It should be noted that the two methods ofmeasuring saving, according to its financial andnon-financial definition, should yield the sameresult (see Lequiller and Blades 2014 andCussen et al. 2012). In practice, however, thisis not the case, due to the statistical discrep-ancy25 arising from the different statisticalsources used under the two approaches. Thismeans that, adding this statistical discrepancyto the financial definition of saving, the non-financial (traditional) definition of saving canbe derived as follows:

Saving: Non-financial accounts ≡Saving: Financial accounts+

statistical discrepancy (6)


24 The saving rate is defined as the ratio of saving to disposableincome, in terms of four-quarter moving sums.

25 Where statistical discrepancy is equal to net lending/borrowing innon-financial accounts minus net lending/borrowing in financialaccounts. It should be noted that in the case of Greece thestatistical discrepancy for the household sector is significant for theyears before 2010, while it has been declining in more recent years.Also, statistical discrepancies are acknowledged to exist in ECBstatistics.

Equation (6) is illustrated in Chart 16, whichdecomposes the evolution of household savinginto contributing factors, in terms of four-quar-ter moving sums. Until early 2010, saving wassupported by household financial and non-financial investment and a gradual decline inincurrence of new liabilities. Thereafter, thealmost continuous downward path of savingwas driven by falling investment inhouses/equipment, but also by strong net dis-investment from financial assets. Debt reduc-tion, which started from the fourth quarter of2010 and intensified from the second quarterof 2013, had a positive contribution to saving.

Sections 6 and 7 take a closer look at the fac-tors of equation (5), namely household dispos-able income and consumption. Specifically, Sec-tion 6 provides a brief overview of develop-ments in household disposable income and itscomponents, and Section 7 discusses the evo-lution of real final consumption expenditure ofhouseholds by purpose in the recent period.

6 HOUSEHOLD DISPOSABLE INCOME AND ITSCOMPONENTS

Chart 17 shows the evolution of householdgross disposable income and its components.

On the income side, the main components ofdisposable income are compensation ofemployees (wages and salaries and employers’social contributions), operating surplus andmixed income (in the case of sole proprietors,this refers to the mixed income of the propri-etor, whereas in the case of households it refersto the own-account production of housing serv-ices by owner-occupiers), social transfers otherthan in kind (pensions and various social ben-efits, such as unemployment/maternity/ familybenefits, etc.), income from property receivable(interest, dividends, rents received for land) andother current transfers receivable. On theexpenditure side, the main components aresocial security contributions, income and wealthtaxes, property income payable and other cur-rent transfers payable26 (see Chart 17).

According to economic theory, compensationof employees, operating surplus and mixedincome, and property income exhibit a pro-cyclical behaviour, i.e. they tend to deterioratein economic downturns, as wages fall, jobs arelost, sole proprietors’ business activity slumpsand dividends and land rents decline.

A breakdown of household disposable incomeinto components shows that the largest per-centage contributions come from compensa-tion of employees, and operating surplus (seeChart 17), followed by social benefits (mainlypensions) and net property income. Anincrease (decrease) in these components has apositive (negative) effect on disposableincome. By contrast, an increase in social con-tributions and taxes has a negative impact ondisposable income and vice-versa. Between thefirst quarter of 2010 and the first quarter of2016, households’ nominal disposable incomeshrank by 32.8% (from €173.5 billion to €116.5billion, in terms of four-quarter moving sums),largely due to sharp falls of 31.5% in com-pensation of employees, (from €86.8 billion to€59.4 billion), 29.1% in operating surplus(from €76.2 billion to €54.0 billion), 17.2% insocial benefits (mainly pensions, from €43.5billion to €36.1 billion) and 45.6% in net prop-erty income (from €9.5 billion to €3.5 billion).During the same period, real disposableincome declined by 32.5%.27

Consumption theories, such as the permanentincome hypothesis and the life cycle hypothe-sis, associated the evolution of householdincome with household consumption expendi-ture, identifying a positive relationshipbetween the two variables (see Friedman1957). On the back of falling disposableincome, households’ consumption expendituredeclined, as shown in Chart 15. Between thefirst quarter of 2010 and the first quarter of2016, this decline was in real terms 24.7%, i.e.from €169.4 billion to €127.5 billion.


26 These include non-life insurance premiums, contributions to non-profit institutions, etc.

27 Calculated as nominal disposable income divided by the deflatorof private consumption.

It should, however, be pointed out that theincrease in compensation of employees (interms of four-quarter moving sums) observedas from the third quarter of 2014 is associatedwith a rise in dependent employment, as con-firmed by data from the ERGANI informationsystem, and in total employment (ELSTAT,Labour Force Surveys). The maintenance ofthe rise in employment and the decline inunemployment, as a consequence of the grad-ual restoration of confidence and the return ofthe economy to positive growth rates, couldsignal an increase in household permanentincome, ceteris paribus, and thus lead to anincrease in their consumer spending.

The following section briefly reviews the evo-lution of households’ consumption expenditurein the recent period. In order to provide adeeper insight into households’ consumptionbehaviour, the discussion focuses on real domes-

tic consumption expenditure by functional pur-pose.28 It is noted that this analysis includeshousehold consumption expenditure by resi-dents and non-residents (tourists) incurred in


28 According to ELSTAT, the main source of data on household finalconsumption is the Household Budget Survey, whilecomplementary sources and methods are used where necessary.The data used to calculate household final consumptionexpenditure refer to the average monthly expenditure perhousehold based on the resident/non-resident distinction(including expenditure by resident households incurred abroad andexcluding expenditure by non-residents (tourists) incurred in theGreek territory) and are available broken down by manner ofacquisition of the goods and services concerned (purchased,produced and consumed by the same household, received in kindfrom employers, organisations or other households). Furthermore,the Household Budget Survey provides data on the average numberof members per household. These data, along with Greece’sestimated population figures, are used to compile the nationalaggregate of final consumption expenditure from annualised data.Complementary data sources most notably include the Survey ofPrivate Legal Building Activity of ELSTAT, receipts and paymentsfor travel services from the Balance of Payments of the Bank ofGreece, as well as administrative data sources. Finally, in certaincases, data on household consumption expenditure is derived as abalancing item of the supply and use tables compiled by theNational Accounts Directorate of ELSTAT. The analysis, which iscarried out on an annual basis, includes households only, excludingnon-profit institutions serving households.

the economic territory of Greece, excluding res-idents’ expenditure incurred abroad, and usesannual data for the period up to the end of 2015.

7 HOUSEHOLD FINAL CONSUMPTIONEXPENDITURE BY PURPOSE

In the period 2009-2015, household final con-sumption expenditure declined in real terms by19.3%, from €163.7 billion in 2009 to €132.0billion in 2015. As shown in Chart 18, thisdecline was broadly based across all the twelvecategories of goods and services29 and was sig-nificant for all categories with the exception ofhotels and restaurants, for which it was onlymarginal, due to the strong growth in tourism(in terms of both arrivals and receipts) from2013 onwards.

The period 2009-2015 saw a marked fall in theconsumption of durable goods, demand forwhich is more elastic and, as argued in inter-national literature, moves in line with house-holds’ perceptions of economic conditions.Specifically, it is argued that factors such ashigh unemployment, falling household dis-posable income and wealth, high indebtednessand economic uncertainty have a negativeeffect on the consumption behaviour of house-holds, making them unwilling to make pur-chases of big-ticket items (see ECB 2015). It isworth noting that consumer spending in the


29 The twelve categories are the following: (1) food and non-alcoholicbeverages; (2) alcoholic beverages and tobacco; (3) clothing andfootwear; (4) housing, water, electricity and gas; (5) furnishings,household equipment and routine household maintenance; (6)health; (7) transport; (8) communication; (9) recreation andculture; (10) education; (11) hotels and restaurants; and (12)miscellaneous goods and services.

categories of “furnishing, household equip-ment, etc.”30 and “clothing and footwear”declined by 54.5% and 46.8%, respectively, farmore strongly than total consumer expenditure(19.3%) in the same period.

Chart 19 shows the percentage allocation ofhousehold annual expenditure to the twelvemain categories of goods and services,enabling to identify the evolution of householdconsumption patterns between 1996 and 2015.Comparing the allocation for 2009 and 2015,the ranking of categories in decreasing orderof their relative shares in total consumptionexpenditure remained unchanged for the cat-egories with the largest shares, i.e. “housing”,“water supply, electricity and gas”, “food andnon-alcoholic beverages”, “transport”, “hotelsand restaurants”, as well as for “education”,which continued to rank last with the smallest

share. On the other hand, the categories of“furnishing, house equipment and routinehousehold maintenance” and “clothing andfootwear” fell to lower places in the ranking,while “recreation and culture” and “alcoholicbeverages, tobacco” moved to higher places.

Furthermore, it can be observed that the shareof expenditure on basic needs (food, housingand network services) in total consumptionincreased at the expense of non-basic expen-diture (furnishing etc., clothing and footwear).These changes provide indications on the evo-lution of consumption patterns.

Specifically, between 2009 and 2015, the cate-gories with the most significant increases in


30 The decline in this category of expenditure is associated with thedownturn in residential construction, as mentioned in previoussections.

their shares in total expenditure were “hotelsand restaurants” (from 11.8% to 14.4%), dueto the positive impact of tourism, and “hous-ing, water supply, electricity and gas” (from19.4% to 21.2%), followed by “transport”(from 13.2% to 14.4%). Only marginalincreases were recorded in the shares of “alco-holic beverages and tobacco” (from 4.5% to4.9%) and “food and non-alcoholic beverages”(from 15.7% to 15.9%). Interestingly, “trans-port” not only increased its share in totalexpenditure but also showed compositionalchanges within this category: the share of “pur-chase of vehicles” decreased (-49.5%) and thatof “transport services” (i.e. use of tram, buses,trains, etc.) increased (+41.2%), probably alsoreflecting the pick-up in tourism.

In the same period, the share of expenditureon “health services” declined from 4.6% to4.2%, accompanied by a reallocation withinthis category, as expenditure on outpatientservices fell (-58.1%) and that on medicalproducts, appliances and equipment more thandoubled (117%).

“Miscellaneous goods and services” alsorecorded a decline in their share in total expen-diture, from 8.4% to 7.5%, along with a real-location within this category: expenditure onsocial protection (elderly care at home, nursinghomes, recovery and rehabilitation centres pro-viding long-term support, schools for childrenwith special needs, etc.) fell by 66.1% and per-sonal care (grooming and beauty services) by33.4%, while expenditure on insurance servicesincreased by 45.3%, with an emphasis on healthand saving insurance products. This trendreflects a precautionary motive, on the part ofhouseholds, to guard against potential uncer-tainties, including uncertainty about their futureincome and access to healthcare services.

Finally, of particular note is the category of“education”, given its importance for humancapital formation, hence the future productivecapacity of the economy. Although educationaccounts for a very small share in households’total expenditure (2.1% in 2015), an important

reallocation within this category can beobserved, with a shift of expenditure away from“pre-primary and primary education” towards“secondary education”.

8 CONCLUDING REMARKS

Economic theory highlights household con-sumption and saving as important factorsbehind the fluctuations of economic activityand economic growth. The importance ofhousehold disposable income and wealth, bothfinancial (currency, deposits, shares, etc.) andnon-financial (houses, etc.), as determinants ofhouseholds’ consumption and saving behaviouris recognised in international literature.

The aim of this study was, first, to examinehouseholds’ (mainly financial) net wealth andits evolution over time, with a focus on themost recent period, by analysing its compo-nents, i.e. assets and liabilities, and their com-position. Next, the Greek households’ savingbehaviour was analysed, by exploring the inter-actions between investment and debt andbetween consumption and disposable income.To this end, two approaches were adopted,based on the financial and the non-financial(traditional) definition of saving, respectively.Finally, in the context of the non-financial def-inition of saving, the study briefly reviewed theevolution of disposable income and its com-ponents and the evolution of households’ realconsumption expenditure by purpose, focusingon the recent period.

Some key conclusions drawn from the analysisare the following:

The financial crisis affected significantly Greekhouseholds’ net financial wealth. Betweenearly 2008 and the first half of 2012, house-holds’ net financial wealth was reduced by65.3% (average annual rate of change duringthat period: 18.4%), reflecting both a decreasein assets and an increase in liabilities. From thethird quarter of 2012 to the first quarter of2016, net financial wealth increased by 79.8%


(average annual rate of change during thatperiod: 16.4%), as households’ total liabilitiesdeclined, while the value of their total assetsincreased. Overall, between early 2008 andearly 2016, net financial wealth fell by 37.5%.

During the deterioration of the crisis, risk aver-sion prevailed and household portfolio shiftsoccurred, away from shares and other equity,debt securities and investment fund sharestowards more liquid financial assets, such asdeposits. At times of heightened uncertainty inGreece, even deposits lost their attractiveness,and cash holdings (currency) increased con-siderably.

Before 2008, the increase in household totalassets stemmed almost exclusively from trans-actions. This pattern changed thereafter, par-ticularly in the early phase of the crisis (2008-2009), as household assets sustained large val-uation losses that could not be offset by thepositive flows (transactions) occurring in thatperiod. By contrast, in 2013-2014, householdtotal assets showed a marked increase, largelyreflecting upward revaluations that exceededthe comparatively low negative transactions.From 2015 onwards, sharp devaluations cou-pled with disinvestment contributed negativelyto the change in the value of household assets.

In the period that preceded the financial cri-sis, in particular during the Greek economy’sboom years from 2004 to 2007, a continuousand considerable increase in households’ debtliabilities was observed, mostly associated withresidential investment. Starting from early2008, residential investment shrank consider-ably. Moreover, households almost quadrupledtheir liabilities, both for housing and consumerloans, from 2002 to June 2010. The composi-tion of household loans remained remarkablyunchanged, with housing loans accounting forabout 70% of total loans and consumer loansfor 30%. A net reduction in household liabil-ities occurred for the first time in late 2010and, intensified as from the second quarter of2013, continued into the next quarters. In spiteof a decline in household net borrowing, total

liabilities as reflected in leverage ratiosremained high.

According to the financial definition of saving(interaction between household investmentand debt), up to the end of 2008 household sav-ing remained robust, supported by households’high investment in real estate, other non-finan-cial assets and financial assets, which out-weighed net new debt incurrence. Thereafter,savings followed a downward trend, as thedecline in total household net investment out-paced the reduction in net liabilities. Althoughhouseholds had started already from the fourthquarter of 2010 a net repayment of their exist-ing debt, saving became negative in the firstquarter of 2011, reflecting both a decline in(mainly residential) investment and strong netdisinvestment of financial assets. Since then,saving has alternated from positive to mar-ginally negative, depending on which compo-nent dominates each time. Household savingis exceptionally low compared with its pre-cri-sis levels, because of a contraction of invest-ment in houses/equipment and large disin-vestment of financial assets.

According to the non-financial accounts of thehousehold sector, the household saving rate, interms of four-quarter moving sums, followeda downward path as from the second quarterof 2009, falling from 7.4% to -5.6% by the firstquarter of 2016. Two subperiods can be iden-tified in this regard: one until the first quarterof 2012, when the household saving rate was inpositive territory, and a second when it becamenegative. It is pointed out that the decrease inhousehold saving ―in particular from end-2009 onwards― is linked with the decline indisposable income. Specifically, the continuousdownward trend in the household saving ratefrom the second quarter of 2012 through thesecond quarter of 2014, when this ratio was innegative territory, is attributable to the factthat the rate of decrease in disposable incomeexceeded the rate of decrease in consumption.Since the second quarter of 2014, the house-hold saving rate appears to have stabilised atvery low levels.


A decomposition of household disposableincome indicates that the largest percentagecontribution comes from compensation ofemployees, followed by operating surplus,social benefits and, finally, net propertyincome. In the period from the first quarter of2010 to the first quarter of 2016, all these fourcomponents fell sharply. In the same period,household disposable income shrank by32.8% in nominal terms and by 32.5% in realterms; as a result, household consumptionexpenditure declined, by 24.7% in real terms,from €169.4 billion in the first quarter of 2010to €127.5 billion in the first quarter of 2016.

Finally, from 2009 onwards, household finalconsumption expenditure by purpose declinedin real terms across all individual categories ofexpenditure. Expenditure on consumerdurables fell sharply, on the back of risingunemployment and shrinking disposableincome. In the same period, the share ofexpenditure on basic needs (food, housing andnetwork services) in total consumptionincreased, and the share of expenditure onnon-basic needs (furnishing etc., clothing andfootwear) decreased accordingly. The abovechanges provide indications on the evolutionof household consumption pattern.


Arrondel, L., L. Bartiloro, P. Fessler, P. Lindner, T.Y. Mathä, C. Rampazzi, F. Savignac, T.Schmidt, M. Schürz and P. Vermeulen (2014), “How do households allocate their assets? Stylisedfacts from the Eurosystem Household Finance and Consumption Survey”, European CentralBank (ECB), Working Paper Series, No. 1722.

Bank of Greece (2011), Annual Report 2010. Bank of Greece (2015), Annual Report 2014. Bank of Greece (2016), Monetary Policy – Interim Report 2016.Bê Duc, L. and G. Le Breton (2009), “Flow-of-funds analysis at the ECB – framework and appli-

cations”, European Central Bank (ECB), Occasional Paper Series, No. 105.Berry, S., R. Williams and M. Waldron (2009), “Household saving”, Bank of England, QuarterlyBulletin Articles, 49(3).

Bouis, R. (2015), “Household Deleveraging and Saving Rates: A Cross-Country Analysis”,mimeo, February.

Brissimis, S.N., E.N. Garganas and S.G. Hall (2014), “Consumer credit in an era of financial lib-eralisation: an overreaction to repressed demand?”, Applied Economics, 46(2), 139-152.

Cooper, D. (2013), “Changes in US household balance sheet behavior after the housing bust andGreat Recession: Evidence from panel data”, Federal Reserve Bank of Boston, Public PolicyDiscussion Papers, 13(6).

Cussen, M., B. O’Leary and D. Smith (2012), “The impact of the financial turmoil on households:a cross country comparison”, Central Bank of Ireland, Quarterly Bulletin Articles, 2(12), 78-98.

Cussen, M. and G. Phelan (2010), “Irish households: assessing the impact of the economic cri-sis”, Central Bank of Ireland, Quarterly Bulletin Articles, 4, 62-76.

De Bonis, R. and A. Silvestrini (2012), “The effects of financial and real wealth on consump-tion: New evidence from OECD countries”, Applied Financial Economics, 22(5), 409-425.

European Central Bank (ECB)(2015), Economic Bulletin, No. 3, April.European Central Bank (ECB)(2016), Household Sector Report – 2016 Q1, Quarterly report

of financial and non-financial accounts for the household sector in the euro area.Friedman, M. (1957), A theory of the consumption function, NBER, Princeton University Press,

20-37.Glick, R. and K. Lansing (2010), “Global Household Leverage, House Prices, and Consumption”,

Federal Reserve Bank of San Francisco, Economic Letters, 1(11).Lequiller, F. and D. Blades (2014), Understanding national accounts, OECD Publishing, Sec-

ond edition.Lolos, S.E. and E. Papapetrou (2011), “Housing credit and female labour supply: assessing the

evidence from Greece”, Bank of Greece, Working Papers, No. 141.McCarthy, Y. and K. McQuinn (2014), “Deleveraging in a highly indebted property market: who

does it and are there implications for household consumption?”, Central Bank of Ireland,Research Technical Papers, No. 5.

Rocher, S. and M. Stierle (2015), “Household saving rates in the EU: Why do they differ somuch?”, European Commission, Discussion Papers, 005.

Skudelny, F. (2009), “Euro area private consumption: is there a role for housing wealth effects?”,European Central Bank (ECB), Working Paper Series, No. 1507, May.

Zarco, I.A. (2009), “Financial assets and liabilities of households in the European Union”, Euro-stat, Statistics in Focus, 32.


R E F E R ENC E S

1 INTRODUCTORY REMARKS

The theory of Optimum Currency Areas(OCAs) is central to international macroeco-nomic analysis and to the broad debate onmonetary integration and has grown to be ofparticular interest to the European currencyarea in recent decades.

The key conclusions of the OCA theory arebased on the premise that, for a currency areato be successful and labelled as optimum, thebenefits of joining should outweigh the coststhat the loss of monetary policy tools entails forprospective members (Van Overtveldt 2011).The literature on this topic lists a number ofcriteria that need to be met for a monetaryunion to qualify as an OCA. These criteriaform the basis of the OCA theory and serve atwofold purpose: on the one hand, they seek toreduce the incidence of asymmetric shocks byrequiring that participating economies sharesimilar structural characteristics (e.g. labourmarket institutions, inflation rates and levelsof economic development); on the other hand,they aim to establish adequate adjustmentmechanisms (e.g. labour mobility and fiscalintegration), to lessen the impact of asym-metric shocks, should they occur.1

The endogeneity of the OCA criteria, a notiondeveloped in the context of discussions on theOCA theory, assumes that monetary integra-tion leads to a significant deepening of recip-rocal trade. This has led to the idea that coun-tries may satisfy the OCA criteria ex post, evenif they do not ex ante (Frankel and Rose 1998).

By arguing that non-qualifying currency areascould, over time, turn into OCAs, the endo-geneity hypothesis provided the theoreticalunderpinning for refuelling the debate on Eco-nomic and Monetary Union (EMU). Thus, theOCA criteria could be fulfilled ex post, as aresult of the expected higher trade integrationand income correlation (Mongelli 2008). Onthe other hand, endogeneity would mean thatthe fulfilment of the criteria is the result of adynamic process, potentially involving factorsthat hamper, rather than facilitate, the devel-opment of an OCA. Thus, even economies thatmeet the OCA criteria before entering a mon-etary union may stop doing so after they havejoined (see, inter alia, Giannakopoulos andDemopoulos 2011).

In the case of the euro area, although it wasgenerally accepted that the participating coun-tries did not initially satisfy the conditions foran OCA, the monetary union seemed to workwell from its inception in 1999 to the outbreakof the financial crisis in 2008. Thereafter, how-ever, developments across the euro area coun-tries brought to light the flaws of this union, aseconomic convergence among the participat-ing countries proved to have been inadequate,and appropriate mechanisms to absorb asym-metric shocks were not in place. This articleaims to empirically test the validity of this nar-rative. To this end, it will attempt to empiri-


EUROPE A S AN OP T IMUM CURR ENCY A R E A : T H E EXP ER I EN C E O F TH E B A L T I C COUNTR I E S *�

Christos Papazoglou Bank of Greece, Economic Analysis and Research Department, and Panteion University

Dimitrios Sideris Bank of Greece, Economic Analysis and Research Department, and Panteion University

Orestis Tsagklas Panteion University

* The authors would like to thank Heather Gibson and HionaBalfoussia for their useful remarks. The views expressed in thisarticle are those of the authors and do not necessarily reflect thoseof the Bank of Greece.

1 See Gibson, Palivos and Tavlas (2014).

cally confirm the validity of the hypothesesthat: (a) the monetary union functionedsmoothly, with the participating economies fol-lowing a path of convergence, thereby sup-porting the case for endogeneity; and (b) thischanged in the period after the outbreak of thecrisis. If confirmed, this would certainly sup-port the argument that the initial perception ofa smooth path towards a European OCA wasoverly optimistic.

The analysis focuses on the case of the threeBaltic economies, i.e. Estonia, Lithuania andLatvia. These countries seceded from the for-mer Soviet Union in the early 1990s and simul-taneously embarked on a transition to marketeconomy. Having joined the European Union(EU) in 2004, they all adopted the single cur-rency within a period of five years (2011-2015).2 Although their euro area entry is rela-tively recent, their economic integration intothe EU began immediately after their inde-pendence from the Soviet Union, as all threecountries had set EU and euro area member-ship as a long-term national goal. Attesting tothis is the fact that it took them only a shorttime to achieve a very high degree of economicintegration with the rest of the EU. Indeed, asshown in Chart 1, by the start of the pastdecade, these economies had already had quitehigh trade openness and, as seen from Chart2, a high degree of trade integration with theEU countries.

Furthermore, all three countries adoptedfixed exchange rates of their currencies vis-à-vis the euro, which meant that their monetarypolicies had to be closely coordinated withthat of the European Central Bank (ECB).Due to the structural similarities of theireconomies, the three countries can be exam-ined together as a bloc for the purposes of thisinvestigation.

The analysis that follows will address thequestion of whether the abovementioned nar-rative holds true in the case of the Balticeconomies. That is, it will check the validityof the argument that the pre-crisis conver-

gence of the Baltics with the EU countries wasconjunctural, which would mean that the con-ditions for an OCA were not satisfied overtime, and that this was largely revealed by theadverse impact of the 2008 financial crisis. Infact, the divergence of these countries fromthe euro area after the outbreak of the crisisis reflected in the strong recovery from thesharp real GDP contraction they had experi-enced in 2009, as opposed to the anaemicrebound of euro area economies. More specif-ically, as seen in Chart 3, although the 2008crisis caused real GDP in 2009 to fall moresharply in the Baltic countries (-14.5%) thanit did in the euro area (-4.5%), the formerrecovered very strongly and soon. On theother hand, the euro area economies took aquite different path, with a lacklustre initialrecovery followed by a relapse to recession.Therefore, the divergence between the twoblocs also seems to be linked with the betterperformance of the Baltic economies, com-pared with the euro area. For this reason, theperiods before and after the onset of the cri-sis will be examined separately in the analy-sis below.


2 Estonia in 2011, Latvia in 2014 and Lithuania in 2015.

Our methodology is based on the theory ofPurchasing Power Parity (PPP),3 in particular,on the simple assumption that, in an econom-ically well-integrated monetary union, the realexchange rates of its member countries areexpected to converge. In other words, if theeconomies meet the criteria for joining a mon-etary union, any shocks will be symmetric andtheir macroeconomic variables will co-move.

Specifically, according to the theory of Gen-eralised PPP (G-PPP), if the fundamental eco-nomic variables determining the real exchangerates of a group of economies are non-sta-tionary, then the real exchange rates of theseeconomies are also non-stationary. Neverthe-less, if these variables tend to share commontrends in the long run, they can still form acointegrating relationship (Enders and Hurn1994). In this case, the relevant economies arelikely to constitute an OCA, if they face simi-lar real disturbances (Mundell 1961).

Finally, it should be noted that the economet-ric model used, apart from examining whethera G-PPP relationship exists between the realexchange rates of the three Baltic countriesagainst the euro, is also used to determine

whether a similar relationship exists betweentheir real exchange rates against the US dollar.The purpose of this latter empirical investiga-tion is to cross-check the findings of the analy-sis, given that the Baltic countries used the USdollar as an anchor currency in an early stageof their transition process and before the emer-gence of the euro, although their trade rela-tions with the United States are limited, par-ticularly in comparison with their trade withthe EU.

This article is structured as follows: The nextsection reviews the theoretical framework. Sec-tion 3 presents the data and the econometricmethodology. Section 4 reports the empiricalresults. Finally, Section 5 concludes.

2 THE THEORY OF GENERALISED PURCHASINGPOWER PARITY

In the OCA literature, the theory of Gener-alised Purchasing Power Parity (G-PPP) is the


3 The choice of method was largely determined by the availability ofthe monthly price and exchange rate data needed to ensure asufficiently large number of observations, and was also based onreliability considerations related to the increased transparency ofexchange rate and price level data.

most commonly used theory for testing whethera group of countries form a currency area.

The G-PPP theory was introduced by Endersand Hurn (1994) and is based on the followingidea: It could be that the fundamental eco-nomic variables determining the real exchangerates of a group of countries are non-station-ary, and consequently the real exchange ratesof the countries are non-stationary; neverthe-less, if the fundamentals are sufficiently inte-grated, the real exchange rates will share com-mon trends and therefore will convergetowards a long-run equilibrium relationship(i.e. they will form a cointegrating relation-ship). If this holds true, the economies will con-stitute an optimal currency area in the sense ofMundell (1961), who argues that two or moreeconomies constitute a currency area if theyface similar real disturbances. The theory alsosuggests that, when economic interdependencein a group of economies is high, an economy’sbilateral real exchange rate is influenced by theexchange rates of the other economies in thegroup and the fundamentals of the othereconomies.4

Testing for G-PPP initially entails univariatestationarity analysis of the individual realexchange rate series. The real exchange rate(R) is calculated as:

R = S (P / P*) (1)

where S is the nominal exchange rate (thevalue of the domestic currency expressed interms of the foreign currency), P* is the generallevel of prices in the foreign country and P isthe general level of domestic prices. Anincrease (decrease) in the real exchange ratemeans depreciation (appreciation) of thedomestic currency.

A stationary real exchange rate implies thatPPP holds between a given pair of countries(i.e. changes in the ratio of their national pricelevels are mirrored by changes in the nominalexchange rate between the relevant curren-cies), which, in turn, indicates that these coun-

tries are connected by strong trade and financelinks and that their economies are convergingtowards each other. By contrast, a non-sta-tionary real exchange rate would prima faciesuggest an absence of strong finance linksbetween the two countries. Nevertheless, non-stationary real exchange rates can still sharecommon trends in the long run, which is evi-dence of economic convergence/integrationbetween the economies and the existence of acurrency area.

Specifically, following the notation of Endersand Hurn (1994), G-PPP can be described asfollows: given a n-country world, a m (m≤n)country currency union exists when G-PPPholds, such that a long-run equilibrium rela-tionship exists between the m–1 bilateralexchange rates, of the form:

r2jt= a+b3jtr3jt+b4jtr4jt+b5jtr5jt+.....+bmjtrmjt+et (2)

where rijt is the log of the bilateral realexchange rate in period t between Country jand Country i; a is the intercept term; bijs arethe parameters of the cointegrating vector,which represent the degree of comovement ofthe real exchange rates; and et is a stationarystochastic disturbance term. The bij parametersreflect the economic interdependencies withinthe group of economies. Enders and Hurn(1994) show that the estimated bijs are closelylinked to the aggregate demand functions of agoods market-clearing relationship. They alsoindicate that the more similar the aggregatedemand functions in each country of the group,the lower the bijs in magnitude.

Numerous empirical studies have used the G-PPP theory to test whether a group of coun-tries with common characteristics form anOCA (see e.g. Sarno (1997), who focuses onEMS countries; Antonucci and Girardi (2006)on EMU countries; Kawasaki and Ogawa(2006), Wilson and Choy (2007) and Nusair(2012) on Eastern Asian countries; Neves et al.


4 The idea that third-country effects should be taken into accountwhen testing for bilateral PPP within a system of countries is furtherdeveloped by e.g. Sideris (2006b).

(2007) on Mercosur countries; and Sideris(2011) on Central European countries in rela-tion to the euro area).

3 DATA AND ECONOMETRIC METHODOLOGY

In this empirical investigation, we usemonthly observations for the nominalexchange rates of the domestic currency ofeach Baltic country vis-à-vis the euro and theUS dollar, respectively. For the calculation ofreal exchange rates, we use the consumer priceindices (CPIs). The choice of CPIs isexplained by the fact that these measures arepublished for all countries, ensuring a largesample of data compiled by a broadly similarmethodology.5

The sample period is determined by CPI dataavailability. In particular, monthly CPI data areavailable for the period from February 1995 toNovember 2014 (258 monthly observations).The sources of data are the IMF InternationalFinancial Statistics (IFS) online database andEurostat.

The price indices (P) refer to monthly datawith 2005 as base year (2005=100). The nom-inal exchange rates (S) of the three Balticcountries are end-month.6 The real exchangerate (R) of each Baltic country is derived fromits nominal exchange rate adjusted for prices.The logs of the real exchange rate series aredenoted by rij, where the subscript i takes thevalues la, li and es for Latvia, Lithuania andEstonia, respectively, and the subscript j takesthe values € και $ for the euro and the US dol-lar, respectively. The nominal exchange rateseries are taken from the monthly database ofthe Vienna Institute for International Eco-nomic Studies (WIIW).

To explore the potential relevance of the 2008crisis, the analysis is carried out (i) for the pre-crisis or pre-onset period, from February 1995to September 2008 (164 observations), and (ii)for the post-onset period, from October 2008to November 2014 (74 observations).7

We first test for stationarity of the euro (andthe dollar) real exchange rate series of Latvia,Lithuania and Estonia, applying unit root tests.If non-stationarity is established, we testwhether a G-PPP relationship exists betweenthe Baltics and the euro area (the US econ-omy), using the Johansen (1995) cointegrationtechnique.

4 EMPIRICAL RESULTS

4.1 UNIT ROOT TESTS

In order to test for stationarity of the individ-ual data series, we apply the Elliott-Rothen-berg-Stock (ERS) test (see, inter alia, Neves etal. 2007). In the regressions of the series, weinclude a constant and a trend based on testsfor their statistical significance. The lag length(known to have an impact on the results of theunit root tests) is selected based on theSchwarz Information Criterion (SIC). Theregressions are estimated using spectral ordi-nary least squares (OLS). The test results,which are shown in Table 1, provide evidencethat all series are I(1).8

According to the results for the euro realexchange rate series of Lithuania (rli€), Latvia(rla€) and Estonia (res€), as shown in Table 1A,the ERS test statistics (P-stats) take the values:Pli€=110.06 (Lithuania), Pla€=78.90 (Latvia)and Pes€=282.60 (Estonia), which are higher inabsolute terms than the critical value of the test(5.65) at the 5% level of significance. Hence, thenull hypothesis (Η0) cannot be rejected, and weconclude that the variables rli€, rla€ and res€ arenon-stationary in levels. Non-stationarityimplies that a series has at least one unit root,


5 The Harmonised Index of Consumer Prices (HICP) would havebeen even more relevant, but is available for a smaller sample ofobservations.

6 For the period up to end-December 1998, the exchange rates of thecurrencies of the Baltic countries are expressed in relation to theEuropean currency unit (ECU). The nominal effective exchangerates of the Estonian kroon and the Latvian lats against the USdollar as from December 2010 and December 2013, respectively,are expressed by reference to the euro/dollar parity.

7 The financial crisis began in September 2008 with the collapse ofLehman Brothers.

8 The results are consistent with the findings of Sideris (2006a) andHsing (2008).

i.e. it is at least integrated of order one (I(1)),without precluding a higher order of integration.To determine the order of integration of rli€, rla€and res€, the test is repeated using the first dif-ferences of each variable, which we denote byΔrli€, Δrla€ and Δres€. The ERS test statistics nowtake the values Pli€=0.91 (Lithuania), Pla€=0.99(Latvia) and Pes€=1.40 (Estonia), which are farlower in absolute terms than the critical value(5.65) at the 5% level of significance. Hence, thenull hypothesis (Η0) is rejected, and we concludethat the variables Δrli€, Δrla€ and Δres€ are sta-tionary. It ensues that rli€, rla€ and res€ are inte-grated of order one (Ι(1)).

Table 1B reports the results of the unit roottests for the real exchange rate series ofLithuania (rli$), Latvia (rla$) and Estonia (res$)vis-à-vis the US dollar. All three series arefound to be integrated of order one (Ι(1)).

4.2 TESTING FOR G-PPP USING COINTEGRATIONANALYSIS

Cointegration ranks

In this section, we investigate whether a long-run equilibrium G-PPP relationship of the type

described in equation (2) exists between thereal exchange rates. The analysis tests for coin-tegration using the Johansen VAR methodol-ogy (Johansen 1995). The number of lagsincluded in the VAR systems is determined bythe Akaike Information Criterion (AIC).Under this approach, the number of cointe-grating relationships is identified using theJohansen’s trace and maximum eigenvalue(Max-Eigen) tests. For each set of realexchange rates (i.e. vis-à-vis the euro and theUS dollar, respectively), cointegration analy-sis is performed for two different periods, i.e.before and after the onset of the crisis.

The results on the existence of a G-PPP rela-tionship between the real euro exchange ratesare reported in Table 2. Both tests indicate thepresence of one cointegrating vector in the sys-tem for the pre-onset period, i.e. February 1995-September 2008 (see Table 2A).9 The resultsprovide support to the existence of an equilib-rium relationship for the period before the onset


rli€ c, t 110.06 5.65 0.91 5.65

rla€ c, t 78.90 5.65 0.99 5.65

res€ c, t 282.60 5.65 1.40 5.65

Α. Real exchange rates vis-à-vis the euro

Variable Intercept terms

Testing for unit root in:

Levels First differences

P-statistic Critical value (5%) P-statistic Critical value (5%)

rli$ c, t 19.17 5.65 0.79 5.65

rla$ c, t 14.05 5.65 1.08 5.65

res$ c, t 11.40 5.65 0.97 5.65

B. Real exchange rates vis-à-vis the US dollar

Variable Intercept terms

Testing for unit root in:

Levels First differences

P-statistic Critical value (5%) P-statistic Critical value (5%)

Table 1 ERS unit root tests(February 1995-November 2014)

9 The estimated trace statistics (34.82) is greater than the criticalvalue at the 0.05% confidence level (29.80), thus leading to therejection of the null hypothesis of no cointegration (Η0). Likewise,the maximum eigenvalue (23.17) exceeds the critical value at the0.05% confidence level (21.13).

of the crisis. In other words, the cointegrationanalysis shows that the Baltic countries meet theconditions for an OCA with the euro area dur-ing the pre-onset period: the real exchange ratesappear to be closely integrated and form a G-PPP relationship during this period.

By contrast, the results for the post-onsetperiod (October 2008-November 2014), asreported in Table 2B, do not suggest an equi-librium relationship: neither the trace test northe maximum eigenvalue test provide evidenceof cointegration10 or long-run interactionamong the exchange rates.

Overall, the test findings suggest that theBaltics did form an OCA with the euro areabefore the crisis, but not afterwards. As far asthe pre-crisis period is concerned, this wouldmean that the negative impact of the fact thatthese economies were at the time undergoingtransition to market economy was fully offsetby the positive impact of: (a) their high degreeof economic integration with the countries ofthe euro area; (b) the stability of their nomi-nal exchange rates vis-à-vis the euro undertheir respective national exchange rate poli-cies; (c) the considerable flexibility of theirinstitutional framework, in particular regard-ing the labour market; and (d) a favourableeconomic conjuncture characterised by the

absence of major symmetric shocks, especiallyafter 2000 and until the outbreak of the GreatCrisis. However, according to the empiricalresults, this situation changed with the out-break of the 2008 crisis, after which economicactivity developments diverged between theBaltic States and the euro area, as a result ofasymmetric shocks.

The cointegration test results for the realexchange rates vis-à-vis the US dollar, asreported in Table 3, do not point to a long-runrelationship between the exchange rates, andin this sense, they are in line with expectations.As can be seen, there is no cointegrationamong the exchange rates in either the pre-onset or post-onset period.11 The realexchange rates do not share common trends orconverge towards one another.

The long-run relationship: Long-run elasticities

Table 4 shows the estimated cointegrating vec-tor, which describes the G-PPP relationship


0 23.17* 21.13 34.82* 29.80

1 8.66 14.26 11.66 15.49

2 3.0 3.84 3.0 3.84

Α. February 1995-September 1998

RankMaximum eigenvalue Critical value (95%) Trace Critical value (95%)

0 14.60 21.13 21.46 29.80

1 6.20 14.26 7.01 15.49

2 0.38 3.84 0.38 3.84

Β. October 2008-November 2014


Table 2 Cointegration analysis: real exchange rates vis-à-vis the euro

* Denotes rejection of the hypothesis at 0.05 significance level.

10 The estimated trace statistic (21.46) is lower than the critical valueat the 0.05% confidence level (29.80). Likewise, the maximumeigenvalue statistic (14.46) falls short of the critical value (21.13).Thus, based on both tests, we accept the null hypothesis of nocointegration (Η0).

11 For the pre-onset period, the trace statistic (14.55) is lower thanthe critical value at the 0.05% confidence level (29.80). Likewise,the maximum eigenvalue statistic (8.74) is lower than the criticalvalue (21.13).

between the three real euro exchange rates forthe period February 1995-September 2008.This relationship can be normalised on theEstonian kroon/euro real exchange rate, inorder to reflect a simplified form of the inter-relationship among these rates.12 The esti-mated coefficients can be interpreted as long-run elasticities. All coefficients are statisticallysignificant at the 5% level.

The estimated coefficients are less than unity. Inthe long-run relationship, a 1% increase(decrease) in the Estonian kroon/euro realexchange rate is associated with a 0.09%decrease (increase) in the Latvian lats/euro realexchange rate and a 0.19% increase (decrease)in the Lithuanian litas/euro real exchange rate.The low values of the coefficients can be inter-preted as evidence of significant homogeneity ofthe relevant economies. According to Enders

and Hurn (1994), the low values of parametersmean that the common path of the exchangerates is the result of significant homogeneity ofthe relevant economies rather than of mutualinteractions between the exchange rates.Assuming that the exchange rates are only influ-enced by real output processes of the variousnations, the normalised vector coefficients willbe smaller the more similar are a country’s aggre-gate demand parameters. The results thereforeindicate that the Baltic economies share a com-mon structure of aggregate demand.

The adjustment coefficients

Johansen’s maximum likelihood technique(Johansen 1995) also estimates the adjustment


12 Normalisation to any of these rates would be possible, by changingaccordingly the parameters of the long-run relationship.

0 8.74 21.13 14.55 29.80

1 4.03 14.26 5.80 15.49

2 1.77 3.84 1.77 3.84

Α. February 1995-September 1998


0 9.78 21.13 16.89 29.80

1 6.54 14.26 7.11 15.49

2 0.57 3.84 0.57 3.84

Β. October 2008-November 2014


Table 3 Johansen tests for cointegration rank: real exchange rates vis-à-vis the US dollar

Coefficient 1 -0.09 0.19

t-stats -3.70 2.64

Standard deviation 0.024 0.073

Probability 0.0003 0.009

Real exchange rates vis-à-vis the euro(February 1995-September 2008)

res€� rla€� rli€�

Table 4 Estimated cointegrating relationship

coefficients of each variable in the long-runrelationship. The adjustment coefficients indi-cate the speed at which the variables (in thiscase, real exchange rates) adjust towards theirlong-run equilibrium. If a certain variableadjustment coefficient is insignificantly dif-ferent from zero, then the variable is known tobe weakly exogenous, as the dynamics of thisvariable is not influenced by the long-run equi-librium relationship.

Table 5 presents the estimated speed-of-adjustment coefficients for the G-PPP rela-tionship with respect to the euro for the periodFebruary 1995-September 2008. Adjustmentcoefficients are found to be statistically sig-nificant for the Lithuanian litas and the Lat-vian lats. The highest coefficient is found in thecase of the real exchange rate of the Lithuan-ian litas vis-à-vis the euro. At 0.169, this coef-ficient implies that the litas/euro real exchangerate adjusts by 16.9% per month towards thelong-run equilibrium.

The adjustment coefficient of the Estoniankroon is not found to be statistically significant,indicating the G-PPP relationship does notinfluence its short-run dynamics of res€, i.e. thelatter is weakly exogenous to the cointegrationsystem. Yet, this weak exogeneity of res€ may bedue to the frequent interventions in foreignexchange markets undertaken by the Estonianmonetary authorities in order to keep the realexchange rate of the national currency at a pre-determined level. Moreover, weak exogeneitymay also reflect the impact of the institutionaland regulatory framework governing prices inEstonia.13

5 CONCLUSIONS

The results of the econometric analysis for thethree Baltic economies lead to conclusions thatare consistent with the common narrative ofhow the monetary union has performed.Specifically, we find that the Baltic countriesdid form an OCA with the euro area before theGreat crisis of 2008, but not afterwards, whenthe occurrence of significant asymmetricshocks mostly triggered by the global crisis ledto an economic divergence between the twoblocs.

In the empirical work, cointegration analysis isemployed to investigate the convergencebetween the three real exchange rates vis-à-visthe euro against the OCA criteria. Cointegra-tion between the Baltics and the euro area isexamined for two periods, the pre-crisis or pre-onset period, from February 1995 to Septem-ber 2008, and the post-onset period, fromOctober 2008 to November 2014.

The empirical findings provide evidence infavour of G-PPP (and therefore an OCA) withthe euro area only for the pre-onset period.More specifically, they indicate that, duringthat period, the Baltic economies were in aprocess of convergence with the euro area, asreflected in the gradual convergence of theirbilateral real exchange rates vis-à-vis the euro.Relative exchange-rate stability was supportedby increased trade integration, a favourableglobal economic environment and the con-


13 These results are indicative of the dynamics, but fail to identify thewhole cointegrating system.

Δrli 0.169 0.04 4.39

Δrla 0.088 0.03 2.60

Δres -0.02 0.014 -1.59

Exchange rates vis-à-vis the euro(February 1995-September 2008)

Adjustment coefficient Standard deviation t-stats

Table 5 Estimated adjustment coefficients

straints that the fixed exchange rate regimeoperated by the Baltic countries imposed onmonetary policy. As a result, the Balticeconomies followed a path of real convergencewith the euro area.

However, with the outbreak of the Great crisisin 2008, it became clear not only that their con-vergence with the euro area had been overes-timated, but also that the risks from the emer-gence of severe imbalances in these economieshad been underestimated. The crisis gave riseto significant asymmetric shocks, whichrevealed the inadequate degree of convergenceof the economies, and this was also reflected inthe path of the real exchange rates. More gen-erally, the experience of the crisis showed thatthe pre-crisis convergence of these economies,which was largely responsible for the build-upof significant, mostly external imbalances, wasunsustainable (IMF 2014).14 Against this back-drop, the crisis plunged the Baltic countries intoa deep recession, from which, however, theymanaged to recover quite soon, thanks toprompt policy responses, including adjustmentmeasures.15 The euro area economies, on theother hand, followed a different path, as men-tioned in the introduction. The results of ourresearch support this narrative, as the method-ology applied shows that, in the post-onsetperiod, convergence with the euro areaeconomies was insufficient.

The results indicate that the process of con-vergence towards the euro area has been weak-ened in recent years by the impact of the Greatcrisis of 2008. More specifically, prior to 2008,convergence was promoted by a favourable

economic conjuncture, the absence of asym-metric shocks and accelerated economic inte-gration with the EU, largely as a result of therole of the euro in European markets. How-ever, with the outbreak of the financial crisisin 2008, it became clear that the degree of eco-nomic integration was lower than pre-crisisconvergence would suggest (IMF 2014), withthe Baltic economies recovering faster than theeuro area.

A similar analysis has also been carried outwith respect to the US economy. The resultsshow no alignment between the Baltics and theUS economy for either the pre-onset or thepost-onset period. This confirms our initialhypothesis that the G-PPP theory does nothold for the United States. The results for theUnited States largely reflect the weakening ofthe US dollar in European markets, but alsothe limited economic integration of the Balticcountries with the United States.

In general, the 2008 crisis demonstrated thatthe previously achieved convergence betweenthe Baltics and the euro area, though signifi-cant, was not sufficient to keep the flaws of themonetary union from coming to light. In ourview, the results for the Baltic countries arequite representative of the general situationthat prevailed in the euro area for some timeafter the crisis.


14 According to several analysts, the Balassa-Samuelson effectexplained much of the convergence path of transition economiesand was used as an alibi for the large imbalances built up beforethe crisis. The methodology used in this article does not enable todisentangle the significance of this particular effect.

15 One important policy response involved internal devaluation, whichled to the speedy recovery in the very sizeable export sectors ofthese economies.

Aggarwal, R. and M. Mougoue (1993), “Cointegration among Southeast Asian and Japanese cur-rencies”, Economic Letters, 41(2), 161-166.

Antonucci, D. and A. Girardi (2006), “Structural changes and deviations from the PurchasingPower Parity within the euro area”, Applied Financial Economics, 16, 185-198.

Enders, W. and S. Hurn (1994), “Theory and tests of generalized purchasing-power parity: com-mon trends and real exchange rates in the Pacific Rim”, Review of International Economics,2(2), 179-190.

Frankel, J. and A. Rose (1998), “The endogeneity of the optimum currency area criteria”, Eco-nomic Journal, 108(449), 1009-1025.

Giannakopoulos, N. and G. Demopoulos (2001), “Theory of monetary unions”, in: Demopou-los, G., N. Baltas and J. Hassid (eds), Introduction to European Studies – Vol. 2: EconomicIntegration and Policies, Athens: I. Sideris, 329-369. [In Greek]

Gibson, H.D., Th. Palivos and G.S. Tavlas (2014), “The crisis in the euro area: an analyticoverview”, Journal of Macroeconomics, 39, 233-239.

Hsing, Y. (2008), “On the functional form of PPP: the case of nine new EU countries”, AppliedFinancial Economics Letters, 4(6), 389-393.

IMF (2014), 25 Υears of Transition: Post-Communist Europe and the IMF. Regional EconomicIssues, Special Report. Eds: J. Roaf, R. Atoyan, B. Joshi, K. Krogulski and an IMF Staff Team.Washington: International Monetary Fund Publication Services.

Johansen, S. (1995), Likelihood-Based Inference in Cointegrated Vector Autoregressive Mod-els, Oxford: Oxford University Press.

Kawasaki, K. and E. Ogawa (2006), “What should the weights of the three major currencies bein a common currency basket in East Asia?”, Asian Economic Journal, 20(1), 75-94.

Mongelli, F.P. (2008), “European economic and monetary integration and the optimum currencyarea theory”, European Economy, Economic Papers, 302, Brussels: European Commission.

Mundell, R.A. (1961), “A theory of optimum currency areas”, The American Economic Review,12(4), 657-665.

Neves, J.A., L. Stocco and S. Da Silva (2007), “Is Mercosur an optimum currency area?”, MunichPersonal RePEc Archive, MPRA Paper No. 2758.

Nusair, S. (2012), “Is East Asia an optimum currency area? A test of generalized purchasing powerparity in the presence of structural breaks”, Journal of the Asia Pacific Economy, 17(3), 399-425.

Sarno, L. (1997), “Policy convergence, the exchange rate mechanism and the misalignment ofexchange rates. Some tests of purchasing power parity and generalized purchasing power par-ity”, Applied Economics, 29(5), 591-605.

Sideris, D. (2006a), “Purchasing Power Parity in economies in transition: evidence from Cen-tral and East European countries”, Applied Financial Economics, 16(1-2), 135-143.

Sideris, D. (2006b), “Testing for long-run PPP in a system context: evidence for the US, Ger-many and Japan”, Journal of International Financial Markets, Institutions and Money, 16(2),143-154.

Sideris, D. (2011), “Optimum currency areas, structural changes and the endogeneity of the OCAcriteria: evidence from six new EU member states”, Applied Financial Economics, 21(4), 195-206.

Van Overtfeldt, J. (2011), Τhe End of the Euro: Τhe Uneasy Future of the European Union,Chicago: Agate.

Wilson, P. and K.M. Choy (2007), “Prospects for enhanced exchange rate cooperation in East Asia:some preliminary findings from generalized PPP theory”, Applied Economics, 39(8), 981-995.


R E F E R ENC E S

1 INTRODUCTION

The term “financial technology” (or FinTech)refers to the application of technology for theprovision of financial services. As a sector, Fin-Tech refers to technology startups that areemerging to compete with traditional bankingand financial market players, offering a numberof services, from mobile payment solutions (seeSection 2.1) and crowdfunding platforms (seeSection 2.2) to online portfolio management andinternational money transfers. FinTech compa-nies are attracting the interest of both financialservices users and investment firms, which seethem as the future of the financial sector.

The term FinTech can be traced back to theearly 1990s and more specifically to a projectinitiated by Citigroup.1 However, it was only in2014 that the sector started to attract theincreased attention of regulators, industry andconsumers. Although FinTech is seen as arecent close cooperation of financial servicesand information technology, the linkage ofthese two sectors has a long history. In fact,financial and technological developments havebeen interconnected and mutually reinforcingover time.

The global financial crisis of 2008 was a turn-ing point and one of the reasons that made Fin-Tech a new norm. This change has broughtabout challenges both for regulators and mar-ket participants, mainly in terms of striking abalance between the potential benefits and risksof innovation. Increased activity raises ques-tions like: What will the financial landscape belike after digitisation? What will be the role oftraditional banks? Will FinTech companiesexpand in tandem with the banking sector ornot? What are the new risks posed by these newsynergies to financial services users?

Regulators are faced by a difficult task in find-ing the right balance that will at the same time

allow incumbent firms to survive and new-comers to innovate, as hindering the entry ofnew firms would distort the market in favourof established firms. Thus, given the above-mentioned challenges, regulators should followsome general principles in order to strike a bal-ance.2 First, they should keep a neutral stancetowards technological developments. Rulesshould promote healthy competition amongmarket players, irrespective of whether theyoffer traditional approaches or new techno-logical solutions. Second, a harmonised, non-discriminatory set of rules should apply estab-lishing a level playing field for all participants,with a view to averting market fragmentationand low competition. Finally, regulators shouldabove all ensure the protection of users, as wellof the financial system itself.

This article is structured as follows: Section 2examines the drivers of the evolution of theFinTech sector. These include supply-side fac-tors, related to the digital revolution, anddemand-side factors, related to the emergenceof new consumer patterns. A more detailedanalysis is provided regarding the developmentof FinTech in payment services, lending andfunding. Section 3 focuses on the role of bankswith respect to FinTech, i.e. how big traditionalplayers react and what alternative strategiesthey could adopt. Next, Section 4 explores theensuing challenges for regulators, discusses dif-ferent approaches to the protection of finan-cial services users and outlines the existing reg-ulatory framework of the European Union


THE E VO LU T I ON O F F I N ANC I A L T E CHNOLOGY( F I N T E CH )

Sofia Anyfantaki*Economic Analysis and Research Department

* The author is grateful to Heather Gibson and Hiona Balfoussia fortheir very valuable guidance, comments and corrections. Theauthor is also grateful to Konstantia Tzortsou, Panagiotis Droukasand Yannis Kanellopoulos for the fruitful discussion. The viewsexpressed in this article are those of the author and do notnecessarily reflect those of the Bank of Greece. The author isresponsible for any errors or omisions.

1 FinTech is the initial name for the Financial Services TechnologyConsortium, a project initiated by Citicorp, former Citigroup. TheFinancial Services Technology Consortium (FSTC) started in 1993to facilitate technological cooperation efforts in the sector offinancial services.

2 See Darolles (2016).

(EU). In the last section, a brief overview ofthe landscape in Greece is provided.

2 THE EMERGENCE OF FINTECH

Once the global economy exited the crisis, itbecame clear that many customers, especiallythe younger generation, had lost their trust inbanks. Apart from an increased mistrust ofbanks, young people have developed differentconsumer patterns from those of their elders.They have grown up being used to having accessto personalised, tailor-made solutions, in starkcontrast with the past “mass” approach of banksand other traditional financial institutions.Against this backdrop, if traditional players wishto attract profitable clients, they need to evolveand offer interactive solutions of the same levelas those of their FinTech competitors.

This trend has been fuelled by a steady growthin global investment in the FinTech sector (seeChart 1), mainly by venture capital and privateequity. Between 2013 and 2014, in only oneyear, FinTech investment almost tripled in the

United States. London, San Francisco/SiliconValley and New York have already emerged asmajor financial innovation hubs, while newhubs have followed suit around the world,namely Amsterdam, Paris, Berlin and Dublin,which are the main centres of the EuropeanFinTech ecosystem (see Chart 2).

These new opportunities have the strongestimpact on emerging market economies, espe-cially those with a rapidly increasing middle-income population. More specifically, there isnow growing demand for financial services bypeople who previously had no access to thebanking sector, as mobile device-based tech-nology enables access to financial solutionswithout the need of physical banking infra-structure.

In developing countries, FinTech includesamong other things the following features: (1)a young population with digital literacy andequipped with mobile devices; (2) a fast-grow-ing middle class, with 60% of the world’s mid-dle class being identified in Asia by 2030; (3)inefficient financial markets, which allow for


informal alternative solutions; (4) lack of phys-ical banking infrastructure (1.2 billion peoplehave no bank account); and (7) underregulatedframeworks for data protection and competi-tion. The above features are further fuelled bythe interplay between a dynamic private sectorthat seeks to expand to the provision of finan-cial services and a public sector that aspiresafter market reform in order to achieve eco-nomic growth.

In Asia and Africa, the recent growth of Fin-Tech is primarily driven by economic develop-ment. Hong Kong and Singapore saw in lessthan a year the creation of three FinΤech accel-erators,3 thus featuring the greatest concentra-tion of FinTech accelerators worldwide. Theemergence of FinTech in Asia is not unprece-dented in the wake of the crisis, but it is ratherthe combination of a number of business andregulatory factors. More specifically, IT spend-ing by traditional banks in Asia and Africa islower than in Europe and the United States.This can be explained by a slightly less com-petitive market, which is still largely controlledand subject to distortions by state-owned banks.Public mistrust of the state-owned banking sys-tem (because of corruption and inefficiency)means that users are willing to adopt alterna-tive solutions offered by non-banks. As a result,

mobile financial services and mobile phoneproducts are comparably more attractive.

Although Africa shares many common featureswith Asia-Pacific in terms of financial innova-tion development, the nature and the directionof the determinants of this sector are quite dif-ferent in Africa. According to G20, almost 2.5billion adults in the African continent (almosthalf of the working age population) have noaccess to the formal financial sector (see Table1). In this context, telecommunication com-panies, instead of banks, have taken the leadin the development of FinTech in the region.Mobile money, which means basic paymentand saving services whereby money is trans-ferred electronically using a mobile device,although initiated in the Philippines, achievedits greatest success in Kenya and more recentlyin Tanzania.

2.1 PAYMENT SERVICES

The past five years have witnessed a numberof novelties, which, by making a wide use of


3 Business development programmes for innovative firms that act asinnovation “incubators”, providing space, support (tailored trainingprogrammes, mentorship, networking, etc.) and every possibleassistance to new researchers, entrepreneurs and startups, in orderto develop new ideas and technology-driven solutions in the areaof financial services.

mobile devices and the internet, resulted insimpler payment solutions. Innovation comesin different forms, depending on the paymentssector and the market. Such noveltiesinclude, for instance, digital wallets, mobilepayments, contactless payments4 and real-timepayments.

At the same time, over the past five years anincreasing number of FinTech startups andnon-bank payment providers have entered thepayments industry (see Chart 3), taking advan-tage of an array of new technology conditionsprevailing in the market and using alternativebusiness models that could both disrupt andcomplement conventional payment practices.This new paradigm of non-bank paymentprovider has led to the emergence of FinTechstartups (which seek to apply technologicaladvances in payment services) as well as ofincumbent firms in other non-payment indus-tries (like Facebook and Apple).

A tangible proof of the potential market powerof technology-driven financial service providersis PayPal. Today, the company has more than100 million active accounts and processes adaily average of USD 315 million in payments.The use of prepaid cards also follows an upwardtrend. A report released in 2012 by Master-Card5 projected that the market for the so-called e-money (cards pre-loaded with cash)would be worth around USD 822 billion by

2017. Furthermore, it should be noted thatalmost every bank account holder in the EU hasa debit card and 40% of them also have a creditcard. 34% of EU citizens already shop onlineand more than 50% have a smartphone, whichallows them to access mobile payment. Someeconomy sectors ―like the travel industry―perform most of their sales online. Finally,small and medium-sized enterprises (SMEs) areamong the main beneficiaries of FinTech start-ups. They are willing to experiment with newtools that will have a material impact on theirbusiness activities. In fact, SMEs are the back-bone of many economies (accounting for 80%of global economic activity).

First, what are the benefits of mobile pay-ments? Due to the fact that mobile phones areall the more powerful and connected, the inte-gration of payments into a mobile phone offersmany potential advantages. The customer,using the computing and communicationpower of a mobile phone, may perform severalother activities simultaneously. For instance, aconsumer can compare retail prices, store thepayment record using a financial managementsoftware, download a warranty or instructionalvideo on how to use a product, etc.


4 Contactless payment allows consumers to pay for small purchasesby simply tapping their card (or their mobile device) near the point-of-sale terminal, while the intervention of third parties or signatureor PIN verification are typically not required.

5 2012 Global Prepaid Sizing Study, commissioned by MasterCard:A look at the potential for global prepaid growth by 2017.

Euro area 33.6 28.0

East Asia and Pacific 13.9 10.8

Least developed countries 1.2 3.3

Sub-Saharan Africa 1.4 3.9

South Asia 7.2 8.9

Latin America and Caribbean 12.5 15.7

Middle East and North Africa 11.0 15.2

Area 2004 2014

Table 1 Commercial bank branches (per 100,000 adults)

Source: The World Bank, http://data.worldbank.org/indicator/FB.CBK.BRCH.P5.

The main common characteristics of paymentsinnovations are: (1) simplicity (they allow cus-tomers to make payments in a single tap); (2)interoperability of IT systems (they are notrestricted to a single payment method, as thedigital wallet is linked to credit/debit cards ora bank account); and (3) supply of value-addedservices (customers, merchants and financialinstitutions interact more closely, whichenables them to offer additional services suchas offers, rebates and reward points).

These innovations have led to a shift away fromcash towards electronic payments, as con-sumers now benefit from the use of paymentcards even in small value transactions. Giventhat innovative solutions make use of the exist-ing infrastructure, which has very low variablecosts, the cost of electronic transactions isexpected to fall as the volume of electronicpayments increases. On the other hand, as aresult of this shift towards electronic paymentsand hence the accumulation of more personaldata, financial institutions, service providers

and merchants will be better informed abouttheir customers.

Furthermore, as transactions become all themore virtual and automated, an increasingnumber of payment processes will becomeinvisible to end users, changing in this way boththeir needs and their consuming behaviour. Inmore detail, the successful deployment of dig-ital wallets will free consumers from any lim-itations on the number of payment cards theycan hold and use for their transactions. Con-versely, customers may add multiple paymentcards to digital wallets and choose a differentcard each time with a few additional clicks orin just one click. On their part, to ensure faster,simpler and more efficient payments, a grow-ing number of merchants and payment solutionproviders will offer an automated or one-clickcheck-out in electronic payments, in whichconsumers will have set a default card for alltransactions unless a different paymentmethod is selected. As a result of the above,card issuers will have to differentiate in order


to compete for the default card, by providinge.g. rebates or loyalty points. Moreover, lever-aging data in specific customer segments willbecome a key component of financial institu-tions’ strategies to gain a dominant share indigital wallets.

With banks increasingly aware of the fact thatFinTech and developments in paymentsstrongly affect the future path of payment serv-ices, the payments industry is rapidly evolving,as traditional players and FinTech startupshave established collaborative partnerships tomake the best of both parties and to providecustomers with optimal solutions. New tech-nologies compromise the traditional role ofbanks in the payments landscape. Conven-tional payment solutions, such as credit anddebit cards, have been the main interestincome source for many banks. Such fee rev-enue is threatened by innovation, and banksmay see their share in the payment servicesmarket decline, since the adoption of a digitalwallet is as simple as the installation of anapplication in a smartphone. New, innovativebanking products and services are only avail-able digitally (e.g. digital wallets for mobiledevices), user-friendly, tailor-made and read-

ily available to users. Financial institutions’ability to partner with merchants will constitutea critical component of their strategies, eitherby offering merchants with special terms andconditions of use or by becoming the defaultcard for e-commerce platforms.

2.2 LENDING AND FUNDING

In the post-crisis period, lower risk appetiteamong retail banks has significantly limitedaccess to traditional bank lending. This mutualloss of trust created a lending gap, whichmeans that a considerable part of borrowingneeds is not adequately met by financial insti-tutions. Furthermore, customer preferences infinancial services are rapidly changing, whichcalls for increased transparency, effectivenessand control over savings and loans.

Over the same period, alternative peer-to-peer(P2P) lending platforms have emerged to fillgaps in the traditional lending model. Such plat-forms use alternative methods for assessing cus-tomers’ creditworthiness (for example, files ofsales history from eBay, social media data, etc.)and automated processes to offer loans to abroader base of customers, as well as a new class


of investment opportunities. P2P lending plat-forms are a new form of lending, without neces-sitating financial intermediation. Acting mainlyas online stock markets, lenders offer anamount, which is usually shared among bor-rowers to achieve risk diversification, while bor-rowers pick the lowest interest rate, i.e. the low-est return among those offered by lenders.

Emerging alternative lending models posecompetitive threats and create opportunitiesfor financial institutions, which highlights theimportance of close partnerships and syner-gies with a view to mutually sharing capabili-ties and learning from each other’s lessons. Inparticular, P2P lending processes are flexibleand automated, while P2P online platformscan process requests from investors and bor-rowers faster and more efficiently due tostate-of-the-art infrastructure and absence ofregulatory obligations. As a result, online plat-forms entail lower operating costs than tra-ditional financial institutions. In view of ris-ing customer demand for flexible, smart andtailored services, conventional financial insti-tutions are upgrading their financial products,focusing on sophisticated or highly person-alised products.

Typically, capital raising activities have beenfacilitated by specialised financial institutions,which on the back of their expertise are able toidentify and support investment opportunities.In view of growing interest in startups and digi-tisation, a number of alternative funding(crowdfunding) platforms have been launched,thereby increasing access to capital raisingactivities and providing funding to a greaternumber of companies and projects, wherepotential funders meet project developers viaan online platform.

Alternative funding platforms6 provide anopportunity for businesses to interact directlywith individual investors to widen their raisingcapital options. Crowdfunding serves as analternative model to funding for projects andbusinesses that lack access to capital invest-ment. It rests upon the active participation of

internet users, who are invited to financiallyback a new project or business often inexchange for some sort of “reward” (e.g.rebates and small gifts), without the need of afinancial intermediary. In recent years,“equity crowdfunding” has also emerged,under which backers receive equity shares ofthe company or buy part of the debt/lendmoney in return for a future premium.

Although these alternative funding platformsare not likely to replace the traditional fund-ing ecosystem in the short or medium term,their growth could change the role of incum-bent institutions. Against this backdrop, thepublic, investors and regulators have largelyfocused their attention on alternative financ-ing mechanisms. However, FinTech goesbeyond this narrow scope to include financingof technology itself (e.g. via venture capital,private equity, public offerings, etc.). In addi-tion to the continued development of alterna-tive financing mechanisms, FinTech is increas-ingly involved in areas such as robo-advisoryservices.7 Robo-advisors are just one exampleof the way incumbent firms are innovating inorder to recast their customer relationshipsand offer new banking approaches.

2.3 DIGITISATION

Digitisation is nothing new in the banking andfinancial sector. High-frequency trading andrelated arbitrage strategies are good exam-ples of the impact that new technologiesalready have. The increased use of mobilephone devices and smartphones (in 2014,active mobile devices outnumbered humanson the planet) has placed digital services inthe hands of consumers who previously couldnot be reached. Boasting access to cloud-based technology, smartphones enable digitalservices to be accessed by almost anyone, any-where and anytime.


6 The most popular platforms globally are kickstarter.com andindiegogo.com, while groopio.com is the first Greek crowdfundingplatform.

7 “Robo-advisors are a class of financial adviser that provide financialadvice or investment management online with moderate to minimalhuman intervention”, Wikipedia.

Besides, new technology has considerablyimproved storage of, access to and interpreta-tion of data, resulting in significant benefits,yet also the need for greater data protection.For the banking industry, perhaps the biggestpotential comes from “big data”.8 In recentyears, thanks to the development of new tech-nologies and applications ―such as the wide-spread use of social media and smartphones―the volume and the format of data havechanged drastically, and data analytical andmanagement capabilities are impressive.Technology advancements have made it pos-sible to effectively analyse and interpret vast,complex sets of data. This “smarter” data man-agement allows banks to create more effective,client centric solutions that are more in linewith customer needs.

For the largest part of the 20th century, pay-ments meant the exchange of banknotes orchecks. Even credit card transactions requiredthe submission of receipts and supporting doc-uments between banks. Nonetheless, digitisa-tion in payment services has taken place veryearly, making it hard to imagine the digitisationof other industries without a previous digiti-sation of the payments industry (PayPal forinstance). However, all these digital paymentsystems use a centralised network thatrequires users’ trust in a central counterparty.In 2009, a whitepaper proposed the creation ofa distributed ledger that facilitates transactionsbetween parties without the need of an inter-mediary via a cryptographic process. Such adistributed payment peer-to-peer protocol isthe Bitcoin network, with bitcoins as the digi-tal currency of the ledger. Digital currenciesbelong to the class of cryptocurrencies, usingcryptography to control the creation of addi-tional units and to secure transactions. Astransactions are made, changes in the owner-ship of cryptocurrencies are recorded in a pub-lic ledger which is known as the “blockchain”.9

Since 2009 a range of networks have beendeveloped, built on the same underlying prin-ciples and concepts but employing differentencryption technology or targetting on differ-ent usage.10

The outlook for digital currencies as a meansof payment is unclear. Some consider that thekey role of digital currencies will be cross-bor-der capital transfers, which are priced quitehighly by banks. Virtual currencies haveattracted great attention from the media andpolicymakers, while central banks are closelymonitoring this issue. Indeed, blockchain tech-nology poses a number of challenges that haveyet to be resolved. In particular, there is stillconsiderable uncertainty in many markets sur-rounding the future regulatory framework forbitcoins, with regulatory authorities puzzlingover whether digital currencies should betreated as a fiat currency (and thus as a foreigncurrency), as a commodity (and therefore as agood), as a form of money substitute (andtherefore not officially recognised by govern-ments) or as something completely new.

3 THE ROLE OF BANKS

The financial crisis of 2008 led to a series ofmajor upheavals in the banking and morebroadly the financial sector. First, it becameevident that the activities of large financialinstitutions generate systemic risk. This in turnled to the compilation of different metricsdesigned to quantify systemic risk. Bank finan-cial regulation tightened (Basel III) and manyfinancial institutions had to respond by, amongother things, adjusting their IT developmentmethods to the new regulatory framework. Atthe same time, banks had to confront not onlyan increasing number of competitors, but also


8 A 2012 definition by Gartner research company states that: “Bigdata is high-volume, high-velocity and/or high-variety informationassets that demand cost-effective and innovative forms ofinformation processing”. The three “Vs”, i.e. Volume, Velocity andVariety, are usually referred to in the literature as the key featuresof big data. Thus, the concept of big data is not merely about thesize, the type or the source of data but rather reflects a number ofprocesses that require enhanced insight.

9 In particular, when users spend digital currencies, the respectivenetwork records the transaction in a list called block. Every blockis linked to the previous and the next one and thus a chain iscreated, i.e. the blockchain. In this way, the blockchain is essentiallya public distributed ledger, since everyone on the network has acopy, allowing for security and transparency.

10 There are more than 500 different alternative digital currencies(altcoins). Most of them build up on the same framework providedby Bitcoin and die out shortly. Apart from Bitcoin, most populardistributed ledgers are Litecoin, Ripple and Νamecoin.

a new type of competitor, which is largely seenas better placed to respond to changing mar-ket regulations and customer needs. This isabout FinTech startups that can develop inno-vative products at a faster rate, showing a clearcompetitive advantage relative to the more tra-ditional methods employed by banks.11

2015 was the year when it became clear that theDigital Revolution had finally hit the bankingsector. Global investment in FinTech tripled toUSD 12 billion between 2013 and 2014,12 andthe British Bankers’ Association announcedthat mobile banking has become the preferredpayment method for customers. The bankingsector has a high IT spending-to-revenue ratio.However, between 2014 and 2015, total globalinvestment increased by only 4.6%, with thebulk referring to system maintenance.

Historically, banks have been responsible formost financial innovations. The launch ofcredit cards in the 1950s and of ATMs in the1970s revolutionised the way we access andpay for goods and services. The financial sec-tor has continued to witness many remarkableinnovations and technological advances, suchas contactless technology, digital wallets andcryptocurrencies. However, nowadays inno-vation rarely comes from banks, but fromsmall FinTechs. Despite their differences,both FinTechs and banks have a lot to gainfrom working together. FinTechs can benefitfrom the long history of banking operationsand banks’ institutional framework. On theirpart, banks can gain value added, eitherthrough synergies and partnerships with Fin-Techs or through the acquisition of theiradvanced technology offerings.

The European Banking Authority (ΕBA), theEuropean Securities and Markets Authority(ESMA) and the European Insurance andOccupational Pensions Authority (EIOPA), intheir joint report on risk and vulnerabilities inthe EU financial system that was published inMarch 2016,13 acknowledge that increasingproliferation of financial technology and ofFinTechs, digitisation and rapid innovation

characterise a changing financial sector, not-ing that technology advances may increasinglyaffect traditional providers of financial serv-ices and their revenues. The report states thatFinTech’s development can also provideopportunities, including wider access to finan-cial services for customers at lower costs,increasing competition and efficiency,reduced systemic risk, as well as access to bet-ter and more customer-friendly products. Onthe other hand though, concerns are expressedas to the impacts on the banking sector. Inmore detail, it is argued that FinTech couldaffect banks’ future profit generation capac-ity, promote the risk-taking behaviour of tra-ditional financial institutions and increaseoperating risk as a result of outsourcing toFinTech in an effort to reduce operating costs.At the same time, the report expresses con-cerns that FinTech may give rise to additionalrisks, such as money laundering and reputa-tional and integrity risks, as in the long runFinTech and digitisation could pose risks tofinancial stability and the orderly functioningof markets. Finally, the Board of Directors ofthe Euro Banking Association decided to cre-ate an open forum for banks, FinTechs andother stakeholders to exchange views andexperience on the various issues related to theimplementation of the New Payment ServicesDirective (PSD2) and the creation of an OpenBanking environment.

Banks maintain their reputation for reliableand secure transactions and invest in high reg-ulatory standards, but they also recognise theimportance of being at the forefront of inno-vation and seek to exploit the enormouspotential offered by FinTech. This has led alarge portion of banks to explore differentapproaches to leverage FinTech innovation(see Chart 5), including venture capital invest-ments, accelerator/incubator programmes and


11 In the financial sector, FinTech firms have a comparativeadvantage, due to the technical debt that was accumulated bytraditional players, notably banks. See Darolles (2016).

12 http://www.fintechinnovationlondon.net/media/730274/Accenture-The-Future-of-Fintech-and-Banking-digitallydisrupted-or-reima-.pdf.

13 Joint Committee Report on Risks and Vulnerabilities in the EUFinancial System, https://esas-joint-committee.europe.eu.

close collaboration with the FinTech com-munity. There are also several other initia-tives such as the “hackathon-type” innovationprogrammes, with a limited impact on theinternal performance of the sponsoringorganisations.

According to a survey conducted by Pricewa-terhouseCoopers (PwC 2016a) on 176 CEOsfrom the Banking and Capital Markets sectorin 62 countries, interviewed CEOs see cus-tomer relationship management systems(80% of respondents), data analytics (75%)and social media communication and engage-ment (56%) as the top three technologies thatwould generate the greatest returns in termsof customer engagement. The ability toanalyse a larger volume of data with higherspeed and more accurate predictions canensure a faster, targeted and forward-lookingresponse to customer demands and capitalmarket developments. Chart 6 also shows thatCEOs acknowledge the impact that customershave on their business strategy (for almost90% of respondents, meeting customers’expectations is a top priority).

As the rate of change is accelerating, banksengage with the FinTech community in orderto better understand future challenges andopportunities. Against this background,banks should continue (a) developing andpublishing an internal “road map” outlininghow to identify and respond to market threatsand opportunities; (b) conducting ongoingresearch to keep abreast of FinTech-drivenchanges; (c) ensuring that key staff are edu-cated on developments, threats and opportu-nities; (d) developing an innovation pro-gramme; and (e) optimising already availableinformation and data.14

Experience from several European FinTechshas shown that FinTechs and banks can worktogether at different levels. Notwithstandingthe level at which they engage in a partnership,both sides have potential valuable gains. Fin-Techs are technology-intensive companies thatseek to test new technologies and explore whatis technically feasible without being bound byrigorous legal frameworks. By implementing


14 See BNY Mellon (2015).

innovative approaches, they promote a largenumber of new ideas in a very flexible way. Onthe other hand, banks can add regulatory, legaland risk management expertise and can giveFinTechs access to global payment systems aswell as to their customer databases. Together,FinTechs and banks create an ecosystem thatallows them to better respond to customerneeds and bridge the gap between the servicesoffered by traditional banks and those actuallydemanded by customers.

4 THE ROLE OF REGULATORS

In recent years, retail payments have seen sig-nificant technical innovations with a rapidgrowth in the number of electronic and mobilepayments, as well as with the emergence of newtypes of payment services, which has chal-lenged the framework under the Payment Serv-ices Directive (PSD).15 Many innovative prod-ucts or services fall entirely or in large part out-side the scope of the PSD. At the same time,the EU market for card, internet and mobilepayments remains fragmented along nationalborders and faces serious challenges that hin-

der its further development and halt the EU’sgrowth momentum (different cost of paymentsfor consumers and merchants, different tech-nical infrastructures, lack of a common set oftechnical requirements among payment serv-ice providers, high interchange fees that trans-late into higher consumer prices).

In this context and in order to adapt the Euro-pean payments market to the opportunities ofthe single market and to support the growthof the EU economy, the European Commis-sion adopted a package of measures. Therevised Payment Services Directive (PSD2)16

introduces some new elements and significantimprovements in the EU payments market. Inparticular, it aims to facilitate and rendermore secure the use of internet payment serv-ices by including within its scope the new pay-ment initiation services. Such services operatebetween the merchant’s and the consumer’sbank, allowing for low-cost and efficient elec-tronic payments without the use of a creditcard. These service providers will now be sub-


15 Directive 2007/64/ΕC.16 These measures become effective on 31 January 2018. Directive

2015/2366/ΕU.

ject to the same high regulatory and supervi-sory standards as the other payment institu-tions. At the same time, banks and other pay-ment service providers should enhanceonline security by requesting strong customerauthentication.

Secure payment services are a prerequisite forthe smooth functioning of the payment servicemarket. Users should therefore be adequatelyprotected against potential risks.17 All onlinepayment services should be secured by adopt-ing technologies able to guarantee safe userauthentication and to mitigate the risk of fraud.In order to allow for user-friendly and easy toaccess means of payment for low-risk payments,such as low value contactless payments at thepoint of sale, mobile or not, security require-ment exemptions should be specified in regu-latory technical standards. In this respect, theuser should be able to rely on measures thatprotect the confidentiality and integrity of per-sonalised security credentials (for example, bySMS or email).

Moreover, in recent years technologicaladvances have given rise to the emergence ofa range of complementary services, such asaccount information services where the user isable to have an overall view of its financial sit-uation immediately and at any given moment.

The PSD2 also includes the above mentionedservices in order to provide consumers withadequate protection for their payment andaccount data. Finally, payment initiation serv-ices enable the payment initiation serviceprovider to provide comfort to a payee that thepayment has been initiated in order to providean incentive to the payee to release the goodsor to deliver the service without undue delay.Since payment initiation services were notincluded in the PSD, this raised a number oflegal issues, such as consumer protection, secu-rity and data protection. The new rules underthe PSD2 therefore address these issues.

From a regulatory point of view, a change ofattitude was warranted as to how FinTech prod-

ucts and services should be regulated. The digi-tisation of processes and services of financialinstitutions is a completely understandable mar-ket trend with regulatory implications and obli-gations related to the use of technology. On theother hand, tech startups enter the financialindustry with little or no past regulatory expe-rience. These companies tend to lack a cultureof compliance regarding their obligations forcustomer protection in the provision of finan-cial services. This is precisely where the currentdebate around FinTech regulation lies.

The objective of the new rules set by the PSD2is to close the regulatory gaps, as well as toprovide more legal clarity and ensure consis-tent application of the legislative frameworkacross the EU. In particular, a level playingfield is guaranteed for both incumbents andnew market participants, enabling new meansof payment to reach a broader market andensuring a high level of consumer protectionin the use of these payment services across theEU. This will improve the efficiency of thepayment system as a whole and lead to morechoice and more transparency of paymentservices, while strengthening the trust of con-sumers in a harmonised payments market.18

Moreover, the Directive states that the defi-nition of payment services should be “tech-nologically neutral” and should allow for thedevelopment of new types of payment services,while ensuring equivalent operating conditionsfor both existing and new payment serviceproviders.

The European Supervisory Authorities (com-prising the ESMA, the EBA and the EIOPA)are monitoring the growing number of insti-tutions offering automated services as well asthe use of Distributed Ledger Technology(DLT). The EBA also encourages regulators to


17 Directive 2015/2366/ΕU: “Consumers should be protected againstunfair and misleading practices...”.

18 It should also be noted that low value payment instruments are aninexpensive and easy-to-use alternative for goods and services oflow price and should thus not be overburdened. In order to enhanceconsumer confidence in a harmonised payments market, it isimportant that the payment services user is aware of the actual costsand charges.

closely monitor FinTech with a view to assess-ing potential risks to investor protection, e.g.information technology risk.

The banking and financial sector has under-gone profound changes and regulators need toavoid two pitfalls. The first is overprotectingincumbents by erecting barriers to entry fornewcomers. This would discourage financialinnovation and hinder competition in the sec-tor. Conversely, the second potential pitfall ischoosing to act in favour of newcomers, byimposing on them less strict regulating rulesthan on incumbents. Thus, they should providea level playing field to all participants, while atthe same time fostering an innovative, secureand competitive financial market. In additionto the rules per se, authorities look in generalat the incentives offered to market players andhow these could make them change theirbehaviour. In this vein, FinTech needs a frame-work that will be both harmonised anddynamic and from which market players (e.g.institutional or newcomers) and regulatorsalike benefit. Regulators’ objectives include:(a) financial stability, (b) prudential regulation,(c) fairness and (d) competition and marketdevelopment.19

5 THE GREEK FINTECH LANDSCAPE

In Greece, since the imposition of capital con-trols and the subsequent shock, an increasingnumber of individuals and businesses, follow-ing the global trend, have resorted to elec-tronic means of payment and banking.Restrictions on cash withdrawals, corporatetransactions and capital transfers abroad havepushed many consumers and firms towards abroader use of electronic payment methods.On 26 September 2016, the European CentralBank (ECB) published the 2015 statistics onnon-cash payments in the EU. On the basis ofthese statistics, in Greece the total number ofnon-cash payments, comprising cheques, cardpayments (excluding e-money payment trans-actions), credit transfers and direct debits,surged to 423 million in 2015.

In Greece, the most popular FinTech instru-ment is the e-wallet, which allows users tomake payments and transfer amounts to thirdparties or to bank accounts. Apart from wallet-to-wallet transfers (without transaction fees),the e-wallet also enables the payment of util-ity and other bills, the transfer of amounts toa mobile contact or a business VAT identifi-cation number, as well as payments at a phys-ical store without a POS terminal.

In the Greek market, banks as well offer elec-tronic banking services through different com-munication channels, such as e-banking andmobile banking. Furthermore, banks haveshown a keen interest in the provision of up-to-date electronic services and are active sup-porters of FinTech. A number of banks havelaunched the electronic wallet linked to a bankaccount or a debit/credit/prepaid card, whichsupports among other things contactless pay-ments, transactions history, transfers ofamounts to mobile or social media contactswithout the need to know the bank accountnumber of the recipient, one-click buying onelectronic stores, payments of utility and otherbills, rebates and reward points.

The competent authorities for the authorisa-tion and prudential supervision of credit insti-tutions, payment institutions and electronicmoney institutions20 are the Bank of Greeceand the ECB. In particular, the authorisationrequirements and the supervision rules gov-erning the operation of electronic money insti-tutions are laid down in Law 3862/2010, Law4021/2011,21 and Bank of Greece ExecutiveCommittee Acts 33/19.12.2013 and22/12.7.2013. Under Article 12(1) of Law


19 See Arner et al. (2015).20 Under Article 10 (1) of Law 4021/2011, “electronic money” means

“electronically, including magnetically, stored monetary value asrepresented by a claim on the issuer which is issued on receipt offunds for the purpose of making payment transactions .... and whichis accepted by a natural or legal person other than the electronicmoney issuer”.

21 Law 4021/2011 (Articles 9-30) transposes into Greek law theprovisions of Directive 2009/110/EC of the European Parliamentand of the Council of 16 September 2009 (OJ L267/10.10.2009) onthe taking up, pursuit and prudential supervision of the businessof electronic money institutions amending Directives 2005/60/ECand 2006/48/EC and repealing Directive 2000/46/EC.

4021/2011, the Bank of Greece is responsiblefor the authorisation and prudential supervi-sion of Electronic Money Institutions. In thiscapacity, the Bank of Greece issued ExecutiveCommittee Act 33/19.12.2013. In order to begranted authorisation, prospective electronicmoney institutions are required to hold initialcapital of not less than EUR 350,000 (threehundred and fifty thousand). The above Lawalso sets out the capital requirements and themethod of calculation of own funds require-ments for electronic money institutions, andspecifies the safeguarding requirements forfunds received, the optional exemptions andthe activities in which electronic money insti-tutions are entitled to engage in addition toissuing electronic money, either in Greece oron a cross-border basis. In accordance withArticle 25(1) of Law 4021/2011, issues con-cerning the provision of information, by elec-tronic money institutions, to electronic moneyholders on their rights, as well as the review ofcomplaints from electronic money holders donot fall under the competence of the Bank ofGreece, but under that of the Secretariat Gen-eral of Consumer Affairs.

In the context of providing information tointerested parties, the Bank of Greece pub-lishes online lists (register tables) of all creditor financial institutions authorised to providebanking services. According to the registertables, only one electronic money institutionand nine payment institutions are operating in2016. The tables also include those paymentand electronic money institutions that havenotified of their intention to provide servicesin Greece without establishment.22

6 CONCLUSIONS

As already mentioned in the introduction,financial sector and technology have longbeen intertwined concepts and these two sec-tors are mutually reinforcing. Although it isdifficult to determine how and where this shiftin financial services began, the global finan-cial crisis of 2008 was undeniably a turning

point, playing a key role in the emergence ofthe FinTech era.

FinTech covers digital innovations and tech-nology-enabled business innovation models inthe financial sector. Such innovations may dis-rupt existing structures in the industry, revo-lutionise the way existing businesses generateand distribute their products and services, andopen new avenues for entrepreneurship.Examples of innovations that are central toFinTech today include mobile payment sys-tems, new digital advisory and trading systems,peer-to-peer lending, equity crowdfunding,cryptocurrencies and blockchain.

Recent innovations might pose several chal-lenges for incumbent providers of financialservices. But they may also represent oppor-tunities for the financial sector, including wideraccess for consumers to financial advice andservices at a lower cost and increased compe-tition and efficiency. The growing diversity ofmarket participants and financial servicesoffered can also reduce systemic risk and leadto better and more customer-friendly products.The financial sector has benefited more com-pared to other industries from the improve-ments in the information technology sector.

In the banking sector, FinTech could impactbanks’ future revenue-generating capacity,with a negative impact on capital adequacy dueto loss of deposits. Financial institutions relymore heavily on non-interest income. FinTechcompanies are often able to offer more effi-cient services, while market entry barriers maybe lower than for traditional financial serviceproviders, as digitisation further facilitatestheir entry into the market.

Supervising authorities are closely monitoringthe evolution of financial technology in orderto fully understand developments in FinTechand innovation and be ready to respond effec-tively to a rapidly changing financial sector.


22 http://www.bankofgreece.gr/Pages/en/Supervision/SupervisedInsti-tutions/default.aspx.

This includes assessing potential risks toinvestor protection. Digitisation offers hugegrowth potential for the financial sector. How-ever, it is important that the necessary regu-latory changes do not stifle innovation and atthe same time ensure the stability that this sec-

tor needs in order to meet customer expecta-tions. The aim of the regulatory framework isto promote financial stability and access toservices, while regulators aim to continueexploring policies that promote innovation andthe entry of newcomers.


Accenture (2016), “FinTech and the Evolving Landscape: Landing Points for the Industry”,https://www.accenture.com/t20160427T053810__w__/us-en/_acnmedia/PDF-15/Accenture-Fin-tech-Evolving-Landscape.pdf#zoom=50.

Arner, D.W., J. Barberis and R.P. Buckley (2015), “The Evolution of FinTech: A New Post-Cri-sis Paradigm?”, University of Hong Kong Faculty of Law Research Paper No. 2015/047; UNSWLaw Research Paper No. 2016-62, http://ssrn.com/abstract=267655.

Bareisis, Z. and G. Lodge (2016), Banks, Retailers, and Fintech: Reimagining Payments Rela-tionships, Part One: The Bank Perspective, Celent Report, Oliver Wyman Group, January.

BNY Mellon (2015), “Innovation in Payments: The Future is Fintech”, https://www.bnymel-lon.com/_global-assets/pdf/our-thinking/innovation-in-payments-the-future-is-fintech.pdf.

Chishti, S. and J. Barberis (2016), The FINTECH Book: The Financial Technology Handbookfor Investors, Entrepreneurs and Visionaries, Wiley.

Darolles, S. (2016), “The rise of fintechs and their regulation”, Banque de France, Financial Sta-bility Review, 20, 85-92.

EY (2015), “Who will Disrupt the Disruptors?”, The Journal of Financial Perspectives: FinTech,Winter 2015.

EY (2016), EY FinTech Adoption Index.Kim, Y., J. Choi, Y.-J. Park and J. Yeon (2016), “The adoption of mobile payment services for

‘Fintech’ ”, International Journal of Applied Engineering Research, Vol. 11, Νο. 2, 1058-1061.McKinsey & Company (2016), “Cutting through the noise around financial technology”.Micu, I. and A. Micu (2016), “Financial technology (Fintech) and its implementation on the

Romanian non-banking capital market”, SEA-Practical Application in Finance, Vol. 4, No.2(11).

Philippon, T. (2016), “The FinTech opportunity”, NBER Working Paper Series, No. 22476:http://www.nber.org/papers/w22476.

PwC (2016a), 19th Annual Global CEO Survey – Key findings in the banking and capital mar-kets sector: Creating a platform for competitive regeneration, PricewaterhouseCoopers LLP,February.

PwC (2016b), Blurred lines: How FinTech is Shaping Financial Services, Global FinTech Report,PricewaterhouseCoopers LLP, March.

Rysman, M. and S. Schuh (2016), “New innovation in payments”, NBER Working Paper Series,No 22358: http://www.nber.org/papers/w22358.

Santander InnoVentures, Oliver Wyman and Anthemis Group (2015), “The Fintech 2.0 Paper:rebooting financial services”, http://santanderinnoventures.com/wp-content/uploads/2015/06/The-Fintech-2-0-Paper.pdf.

Walker, A. (2014), “Banking without banks: exploring the disruptive effects of converging tech-nologies that will shape the future of banking”, Journal of Securities Operations and Custody,Vol. 7, No. 1(12), 69-80.


R E F E R ENC E S

A INTRODUCTION

If we wish to draw a general picture of whatinsurance and reinsurance undertakings do forbusiness, we could say that their main objectiveis to provide (re)insurance products and cov-erages. In providing their services, (re)insurerspool funds by assuming insurance and financialrisks, while at the same time they channel thosefunds to the financial system, which makesthem major investors in the economy.

Insurance and reinsurance undertakingswhich are active in the European Union (EU)must meet specific solvency requirements.Those requirements (called Solvency I1 until 31December 2015) had been applicable for over30 years. Their main drawbacks were regardedas being: lack of sensitivity to risks, whichfailed to provide (re)insurers with the appro-priate incentives for adequate risk manage-ment; lack of transparency vis-à-vis customers,who tend to be quite sensitive to issues of trust;and reduced protection of policyholders. Thenew solvency requirements (now called Sol-vency II) became fully effective from 1 Janu-ary 2016 and cover a wide range of issues per-taining to almost all aspects of the prudentialsupervision of insurance and reinsuranceundertakings operating in the EU, thus shap-ing a so-called “solvency framework”.

Section A gives an introduction to the structureand the main concepts of Solvency II. SectionB presents the legal framework of Solvency II,placing emphasis on its European dimension.As the changes that were introduced impactednot only the requirements (for supervision,compliance, etc.) but most importantly therationale, relative to the previous framework,Section C outlines the main differences in theSolvency II approach, compared with SolvencyI. Next, Section D describes the components ofSolvency II and its pillars, and discusses severalissues of relevance. This section aspires to

address the issue without using legal, actuar-ial or other technical terms, but is rather tar-geted at anyone interested. Lastly, Section Eattempts to sum up the key points of the paperand concludes.

B THE LEGAL FRAMEWORK AND THEEUROPEAN DIMENSION OF SOLVENCY II

B.1 THE EUROPEAN SYSTEM OF FINANCIALSUPERVISION

The crisis of 2007-08 has brought to the foreserious weaknesses in the supervision of thefinancial system. Most importantly, it showedthat the EU supervisory system had almostreached its limits in its ability to effectivelyaddress EU-wide issues. As a result, the EUcould no longer afford to sustain a situation inwhich:

(a) there was no mechanism ensuring thatnational supervisory authorities make the bestpossible supervisory decisions on cross-borderfinancial institutions;

(b) there was no adequate cooperation andexchange of information among nationalsupervisors;

(c) the joint action of national authoritiescalled for complex rules based on fragmentedregulatory and supervisory requirements;

(d) national solutions were quite often the onlyfeasible option to deal with EU-wide problems;and, last but not least,

(e) the interpretation of the same legislativetexts varied across the EU.

44Economic Bulletin

December 2016 63

I N T RODUC T I ON TO S O L V ENCY I I F O R ( R E ) I N S URANC E UNDER T AK I NG S

Ioannis Chatzivasiloglou*

Private Insurance Supervision Department

1 The provisions of Solvency I were transposed into Greek law withDecree Law 400/1970, as amended.

In November 2008 the European Commissionassigned a high-level group headed by Jacques deLarosière to find and suggest ways of strength-ening European supervisory mechanisms, withthe ultimate goal of increasing the protection ofconsumers in financial services and rebuildingtrust in the European financial system.

In response to the aforementioned challengesand on the basis of the recommendations ofthe de Larosière report, the European Systemof Financial Supervision (ESFS) was establishedwith a view to overcoming those weaknessesand ensuring that the ESFS remains commit-ted to the objective of a stable and single Euro-pean financial market, bringing national super-visory authorities together in a strong, inte-grated EU network.

The ESFS rests upon two building blocks:

a) macroprudential supervision, consisting ofthe European Systemic Risk Board (ESRB).This building block is responsible for themacroprudential oversight of the EU financialsystem as a whole and is aimed at contributingto the prevention or reduction of systemic risksthat threaten financial stability in the EU, witha view to averting, to the extent possible, large-scale financial market turbulence, and

b) microprudential supervision, consisting ofthree European Supervisory Authorities, i.e.the European Banking Authority (EBA) for thebanking sector, the European Securities andMarkets Authority (ESMA) for the capital mar-ket and the European Insurance and Occupa-tional Pensions Authority (EIOPA) for theinsurance and occupational pensions sector.The second building block also includes allnational competent authorities (NCAs). Itsobjective is to increase the quality and consis-tency of national supervision, strengthen theoversight of cross-border groups and contributeto the drafting of a singe European rulebookper sector (insurance, banking, securities).

The key players under Solvency II are theEIOPA alongside the NCAs for insurance and

reinsurance undertakings (i.e. the Bank ofGreece and the respective supervisory author-ities of other Member States). The role of theESRB under Solvency II, albeit minimum, ispivotal, as it deals with cases which are criticalfor the recovery of insurance and reinsuranceundertakings.2

B.2 THE EUROPEAN INSURANCE ANDOCCUPATIONAL PENSIONS AUTHORITY (EIOPA)

The European Insurance and OccupationalPensions Authority (EIOPA) was establishedin accordance with Regulation (EU) 1094/2010of the European Parliament and the Council of24 November 2010 and, as discussed above, ispart of the European System of FinancialSupervision.

EIOPA’s mission is to work towards the betterfunctioning of the EU internal market, espe-cially by ensuring a high level of regulation andsupervision, taking account of the varyinginterests of all Member States and the differ-ent nature of financial institutions. In its capac-ity, EIOPA supports financial stability, trans-parency of markets and financial products, aswell as the protection of policyholders, pensionscheme members and beneficiaries.

Its tasks also include the promotion of super-visory harmonisation and the provision ofadvice to EU institutions in the areas of regu-lation and supervision of insurance and rein-surance undertakings and occupational pen-sions. ΕΙΟΡΑ replaced CEIOPS3 and isaccountable to the European Parliament andthe Council of the European Union.4


2 The ESRB is involved in those cases in which the recovery periodfor (re)insurance undertakings needs to be extended (under Article138 of Directive 2009/138/EC).

3 CEIOPS (Committee of European Insurance and OccupationalPensions Supervisors) ceased to operate on 31 December 2010.

4 The Council of the European Union is, along with the EuropeanParliament, the main decision-making body of the EU. It haslegislative and fiscal powers, in conjunction with the EuropeanParliament, and coordinates or develops Member States’ policies.Depending on the affairs under examination, the composition ofthe Council varies. Nine different configurations (or nine differentcouncils) have been established with the participation of allcompetent ministers in their respective areas. The Economic andFinancial Affairs Council configuration (ECOFIN), which is madeup of the economics and finance ministers from all Member States,is responsible for the area of private insurance.

EIOPA takes active part and assists in theeffective, efficient and consistent operation ofthe colleges of supervisors, which have been setup by supervisory authorities for each insur-ance group in the EU with cross-border activ-ities.5 Its assistance is for the time being lim-ited to the monitoring of the agenda items andannual work programmes, while in the nearfuture it is expected to further increase.

Another main task of EIOPA is to monitordevelopments in the financial markets and testthe resilience of (re)insurers as well as of theEuropean financial system as a whole to poten-tial adverse shocks. Monitoring comprises theconduct of studies and the publication of bi-annual reports, which present the developmentof risks to national markets. Testing theresilience of (re)insurers to adverse shocks isachieved by conducting periodic stress tests,with the participation of all EU Member States.

B.3 THE EUROPEAN DIMENSION OF SOLVENCY II

Solvency ΙΙ is a framework developed by theEU in line with the “Lamfalussy regulatoryprocess”6 and broadly transferred in December2009 in the Treaty on the Functioning of theEuropean Union (TFEU).

The Lamfalussy process involves four “levels”,each of which focuses on a specific stage of thelegislative process:

Level 1: The first level is based on a EuropeanCommission proposal for a directive or a reg-ulation7 following open consultation with allstakeholders. Level 1 measures are of a generalframework nature, set out the basic principlesof legislative acts and lay a clear foundation forLevel 2 implementing measures. The adoptionof Level 1 framework principles requires co-decision of the Council and the European Par-liament, and the final legislative text is calleda legislative act. Directive 2009/138/EC on thesolvency of insurance and reinsurance under-takings (Solvency II) constitutes a legislativeact (in the form of a Directive) and forms partof Level 1 measures in the Lamfalussy process.

Level 2: A legislative act (i.e. a Level 1 leg-islative text) may delegate to the EuropeanCommission the power to adopt non-legislativeacts (hence “delegated acts”)8 of general appli-cation to complement or amend certain non-essential elements of a legislative act. Theaforementioned delegation of power to theEuropean Commission may occur either underArticle 10 of Regulation (EU) 1094/2010 orunder Article 290 TFEU.

Delegated acts (DAs) under Article 10 ofRegulation (EU) 1094/2010 specifically indi-cate that they have been adopted as Regula-tory Technical Standards (RTS), supplementor amend certain non-essential elements ofthe legislative act (i.e. the Level 1 legislativetext), are rather technical and do not implystrategic decisions or policy choices. Thesestandards are endorsed by the EuropeanCommission, following submission of a draftby EIOPA.

Delegated acts under Article 290 TFEU(which are also DAs without any other indi-cation), although they are likewise limited tonon-essential elements of the legislative act,may imply strategic decisions or policy choicesand thus have a broader scope and grantgreater flexibility to the European Commis-sion. Their endorsement by the EuropeanCommission does not require EIOPA’sinvolvement.9

44Economic Bulletin

December 2016 65

5 In 2015 EIOPA participated in more than 100 colleges ofsupervisors.

6 This process was first introduced in March 2001, named after BaronAlexandre Lamfalussy, chairman of the Committee of Wise Menon the regulation of European securities markets, which wasentrusted with the task of assessing EU legislative practices in thearea of financial services and the submission of proposals forimproving the system.

7 A regulation has general application, is binding in its entirety onall parties and is directly applicable in all Member States. Adirective is binding, as to the result to be achieved, upon eachmember state to which it is addressed, but leaves to the nationalauthorities the choice of form and methods (under Article 288TFEU).

8 Delegated acts, although they are called non-legislative acts, areactually EU legislative texts. Their differentiation from the so-called legislative acts is the decision procedure: while legislativeacts require the full involvement and, subsequently the co-decision,of the Council and the European Parliament, this is not the casefor non-legislative acts, which are adopted by the EuropeanCommission and enter into force only if no objection has beenexpressed by the European Parliament or the Council within aperiod set by the legislative act.

9 Perhaps only in an advisory capacity.

In addition to delegated acts supplementing oramending non-essential elements of the Level1 legislative act, a uniform implementation ofthe legislative act may be needed. This need ismet through Implementing Technical Stan-dards (ITS), which are adopted by means ofimplementing acts.10

Implementing Technical Standards aim todetermine uniform conditions for implement-ing legislative and non-legislative acts. Thesestandards are purely technical and do not implyany strategic decisions or policy choices.EIOPA submits its draft standards to the Euro-pean Commission for endorsement.

Delegated acts (both DAs under Article 290TFEU and RTS) as well as ImplementingTechnical Standards are legally binding andtake the form of a regulation, directive or deci-sion11 (although their most common form isthat of a regulation, which has general appli-cation and is directly applicable in all MemberStates).

According to a strand of literature, only dele-gated acts under Article 290 TFEU are con-sidered to be Level 2 measures, whereas Reg-ulatory Technical Standards and ImplementingTechnical Standards are characterised as 2.5Level measures.

Level 3: The third level of the Lamfalussyprocess comprises those measures that havethe form of non-binding guidelines (GLs) orrecommendations, aimed at ensuring the com-mon, uniform and consistent application of EUlaw. Such measures are issued by EIOPA andare addressed to competent supervisoryauthorities and/or supervised insurance andreinsurance undertakings. Although thesemeasures are not binding, each national com-petent authority12 must notify whether itintends to comply with that guideline or rec-ommendation or not. If a national competentauthority does not comply or does not intendto comply, it must inform EIOPA, stating itsreasons.13 Furthermore, even though they arenon-binding on supervisory authorities, if

transposed into national law and depending onthe method of transposition, they becomebinding in their capacity as national law.

Level 4: The fourth level refers to the processundertaken by EIOPA, with a view to check-ing the application of EU rules by nationalcompetent authorities.

B.4 SOLVENCY II LEGISLATIVE TEXTS

Since, as already mentioned, Solvency II is thenatural outcome of the Lamfalussy process, inorder to gain insight into the specific provi-sions of this framework, one must consultmore than one legal texts, which cover all lev-els of that process. What further complicatesthings is that, apart from EU law, national (inour case, Greek) law should not be over-looked. All legislative (or non-legislative) actsby means of a directive must be transposedinto national legislation (in Greece, by law orpresidential decree) to have legal effect in aMember State. Furthermore, a similar trans-position into a Member State’s national lawis required for all (Level 3) guidelines issuedby EIOPA. Due to the fact that those guide-lines are not legally binding texts, each com-petent authority (the Bank of Greece forGreece) should choose which guidelines it willtranspose into national law, and how, andwhich not.14

The basic EU act under Solvency II is Direc-tive 2009/138/EC of the European Parliamentand of the Council of 25 November 2009 on thetaking-up and pursuit of the business of Insur-ance and Reinsurance (Solvency II). Thisdirective was subject to two major recasts.


10 Under Article 291 TFEU.11 A decision is binding in its entirety, while if specifying those to

whom it is addressed, it is binding only on them (Article 288TFEU).

12 Turning to Greece, the national competent authority for thesupervision of (re)insurance undertakings is the Bank of Greeceunder Article 3(10) of Law 4364/2016.

13 This process is also known as the “comply or explain process”.14 Of course, for those guidelines that it will choose not to transpose

into national law, a national supervisory authority must explain itsdecision to EIOPA, stating the reasons. For instance, a sufficient(and indeed quite typical) explanation is that a guideline providesfor matters concerning legal persons which do not exist in thatcountry.

The first recast took place in 2011 with Direc-tive 2011/89/EU of the European Parliamentand of the Council of 16 November 2011amending directives as regards the supple-mentary supervision of financial entities in afinancial conglomerate (Financial Conglom-erates Directive – FICOD). The aforemen-tioned directive amended specific articles ofthe Solvency II Directive with a view to ensur-ing the appropriate supplementary supervisionof insurance groups.

The second recast of the Solvency II Directivetook place in 2014 with Directive 2014/51/EUof the European Parliament and of the Coun-cil of 16 April 2014 amending directives andregulations in respect of the powers of theEuropean Insurance and Occupational Pen-sions Authority and the European Securitiesand Markets Authority (this directive is morecommonly known as Omnibus II Directive –OMD II). This second recast, apart from beingvery time-consuming (as it took more thanthree years to reach an agreement), was alsoquite extensive. Most important still, not onlydid it amend essential elements of the initialSolvency II Directive, but at some points it alsoamended even its original character.

Those three directives, recorded in a single(codified) legislative text, constitute Level 1 ofthe Lamfalussy process under the legal frame-work of Solvency II. All three directives havebeen transposed into Greek legislation by Law4364/2016.

With regard to Level 2 of the Lamfalussyprocess under Solvency II, the European Com-mission has so far issued a delegated regulationand four delegated decisions, under Article 290TFEU. The regulation supplements Directive2009/138/ΕC and contains all the modalities ofits implementation,15 while the decisions con-cern the equivalence of the solvency regimes inSwitzerland, Australia, Bermuda, Brazil,Canada, Mexico, the United States and Japan.

With respect to Level 2.5 of the Lamfalussyprocess, the European Commission is not

expected to adopt any delegated act by meansof Regulatory Technical Standards, at least inthe near future. Yet, things are quite the oppo-site with Implementing Technical Standards, asthe European Commission has already issued21 ITS so far.

Finally, turning to Level 3 of the Lamfalussyprocess under Solvency II, EIOPA has issueda series of guidelines, which have been trans-posed into Greek law by relevant Bank ofGreece decisions.16

C INSURANCE FROM THE NEW PERSPECTIVE OFSOLVENCY II

To fully grasp the new arrangements intro-duced by Solvency II, one must first under-stand the rationale behind them. Solvency IIrests upon a different approach to the functionof insurance undertakings and insurance con-tracts. In the following sub-sections, this newapproach is presented and its cornerstones areanalysed.

C.1 THE (RE)INSURANCE UNDERTAKING AS A FINANCIAL INSTITUTION

As mentioned in the introduction, (re)insur-ance undertakings raise funds by assuminginsurance and financial risks and channel thosefunds to the financial system.

The bulk of an insurance undertaking’s obli-gations consists of obligations vis-à-vis policy-holders and beneficiaries, which under Sol-vency II are called technical provisions.17 Thetype of technical provisions, as well as theunderlying need for provisioning (or reserving,as most commonly used), varies between lifeinsurance and non-life insurance. In life insur-ance, an insurance undertaking assumes a―usually long-term― commitment vis-à-vis

44Economic Bulletin

December 2016 67

15 Regulation (EU) 2015/35, as amended by Regulation (EU)2016/467.

16 A full list of all relevant legislative texts is available on the Bankof Greece website: http://www.bankofgreece.gr/Pages/en/deia/solvencyII.aspx.

17 In Greece, under Solvency I, these were called “technical reserves”.

policyholders to provide a predeterminedamount ―face amount― at an unknown pointin time. Even though this point in time is notgiven, it is associated with an ex ante deter-mined event, e.g. death, disability, retirement,etc. An insurance policy is offered for a fee,which is paid for in the form of, either peri-odical or one-off, insurance premiums. In thiscontext, technical provisions are the (mathe-matically calculated – hence “mathematicalreserve” under Solvency I) necessary amountof assets that an insurance undertaking musthold to be able to settle its contractual obli-gations vis-à-vis policyholders.

In non-life insurance, the insurance under-taking assumes risk for a short period of time– usually one year or less. In this context, tech-nical provisions are mainly specified as therequired amount of assets that an insuranceundertaking must hold to be in a position tocompensate policyholders and claimants in thefuture for a loss which is associated with aninsurance event that has already occurred.18

The fact that in non-life insurance both theamount to be paid and the time of payment arenot known at the time of the conclusion of thecontract makes risk assessment very challeng-ing. For a further clarification of the phrase“future compensation for a loss”, it should bepointed out that, although the insurance cov-erage spans a short period of time, the cover-age of a loss and therefore the obligation topay out the claim may extend to a time con-siderably later than the date of the event,sometimes even ten or twenty years after itsoccurrence, depending on the size of the lossas well as due to legal complications.19

In the light of the above, it becomes clear thatthe concept of technical provisions has threedimensions: the policyholder’s point of view,the insurer’s point of view and the supervisor’spoint of view. Policyholders see technical pro-visions as their own money, accumulated fromthe insurance premiums they have paid; there-fore, technical provisions are equal to theamount of their claims on the insurer. Insur-ers see technical provisions as the level of

funds available for investing and, lastly, super-visors view technical provisions as the neces-sary level of assets that insurers must hold inorder to fulfil, to the extent that it is reason-ably predictable, their obligations vis-à-vis pol-icyholders.

Certainly, insurers hold, apart from technicalprovisions, additional capital in the form ofown funds,20 but the focus from this point ofview, i.e. under Solvency I, is clearly on tech-nical provisions.

The above perspective is the most common,but not the only one. Another perspective,that of Solvency II, which could help us shedlight on certain other aspects of the insurancebusiness, views (re)insurers as financial insti-tutions. According to this perspective, share-holders-investors are at the centre, as theseinitially make available their funds in searchfor suitable investment opportunities, e.g.bonds, equity and other investment opportu-nities in the form of investment funds, loansto SMEs, construction mortgages, investmentin infrastructure, etc., with a view to investingthem to yield the highest possible profit. How-ever, as the larger the investment the higherthe profit, initial investors (shareholders) seekat the same time for additional funds comingfrom new investors to manage along with theirown funds and channel them to the identifiedinvestment opportunities. Making availablethose additional funds to the initial investorsmay materialise in various ways and take sev-eral forms: e.g. in the form of equity capital(i.e. by becoming shareholders) or in the formof loans. Such a loan may be a typical loan aswe know it or a sui generis loan. For example,a sui generis loan could imply that the newinvestor contributes to the undertaking a


18 The type of technical provisions that best fits this definition is the(outstanding) claims provision. Non-life insurers must also establishother types of technical provisions (such as premium provision),the amount of which however is usually lower relative to theaforementioned claims provision.

19 R.L. Brown and L.R. Gottlieb, “Introduction to Ratemaking andLoss Reserving for Property and Casualty Insurance”, 2nd ed.,Actex Publications Inc., 2001.

20 Under Solvency I, the additional capital, in excess of technicalprovisions, held by an insurer was called “available solvencymargin”.

determined amount of money in return foranother determined amount (which includeshis profit as creditor to the undertaking),which is payable only in the event of an acci-dental event, which is described in the con-tract of this sui generis loan, e.g. the occur-rence of material damages on a property dueto an earthquake). In this case, sui generislending is called “earthquake insurance”, thenew backer is called “policyholder” (and thatis how they are referred to hereinafter), thecontribution to the undertaking is called“insurance premium”, while the amountreceivable if that accidental event occurs iscalled “indemnity”.

From this perspective, the insurance under-taking initially makes available shareholders-investors’ funds. Subsequently, it collects pol-icyholders’ money and invests all of this (ini-tial capital and insurance premiums) in dif-ferent opportunities identified. If, for somereason, it is not possible to identify any appro-priate investment opportunities, then share-holders-investors, in search for yield, abandonthis business. They no longer want the fundsthat they have obtained from the otherinvestors (mainly the insurance portfolio, aswell as the investment that corresponds to themoney paid by policyholders) and transfer(sell) them to other shareholders-investors(buyers) who are more effective in identifyingsuitable investment opportunities.

On the basis of this perspective, investors’funds entail the risk of not yielding the antic-ipated profit to their holders or even of losingtheir value. Such risks mainly arise from thefact that the hypothetical buyer of the portfo-lio might ask for more money to assume it, ifneeded, and as a result, this gap should beclosed in their own expense (i.e. in the expenseof initial investors). From this point of view,insurers must maintain technical provisionswhich are tantamount to the sum of money thatthe new investor (buyer) would ask for in orderto take over the insurance portfolio. Besides,it becomes clear that the concept of technicalprovision adequacy is void and the focus is on

insurers’ own funds to finance such portfoliotransfers, where necessary.

C.2 INSURANCE CONTRACTS AS FINANCIALINSTRUMENTS

The new perspective, which was described inthe previous sub-section, is the most appro-priate stance that an insurer should in generalmaintain to achieve an effective managementunder Solvency II. In addition, it is also theactuaries who should change the way they seeinsurance policies. The traditional approachviews each insurance contract as a series offuture benefits, which can be perceived as theinsurer’s cash outflows to be paid out to thebeneficiary upon occurrence of a predeter-mined possible event (e.g. death, retirement,accident, etc.) According to the traditionalapproach, for policyholders to receive a ben-efit, they must either renounce other benefitsprovided for in the insurance policy (e.g. guar-anteed future cost of death charges or guar-anteed insurance premiums or guaranteedrenewals) or leave part of the insurance pol-icy’s value to the insurer (e.g. the surrender feein the event of an early termination of the con-tract or part of their accumulated insurancepremiums in the form of technical provisions,etc.) or have sustained a loss (e.g. occurrenceof the insured event, such as death, retirement,disability, earthquake, etc.).

On the other hand, insurance contracts may beconsidered as a pool of options which havebeen either offered, in return for a fee, by theinsurer to the policyholder21 (i.e. the insurerhas short position) or are held by the insurerin the policyholder’s knowledge (i.e. theinsurer has long position). With regard to theoptions that the insurer has offered the poli-cyholder, it should be noted that, as the exer-cise of such options does not always depend onthe policyholder’s will (e.g. earthquake, disas-ters, death, etc.), a distinction should be drawnbetween the options, the exercise of which is

44Economic Bulletin

December 2016 69

21 L.M. Tilman, “Asset/Liability Management of FinancialInstitutions”, Euromoney Institutional Investor, Ch. 16“Understanding options embedded in insurers’ balance sheets”.

left at the policyholder’s discretion, usuallydetermined on a financial basis , and thoseoptions, the exercise of which is not conditionalon the policyholder’s choice, but which remainembedded in insurance contracts, with insur-ers having nevertheless short position.

Of course, it is known that upon occurrence ofthe insured event beneficiaries do not take intoaccount the economic context to exercise theirrights (for example, in investment-linked insur-ance policies beneficiaries do not examinestock market indices and the insurer’s returnguarantees, in order to receive the death ben-efit, but the policyholder’s death is a sufficientcondition).

The most common options envisaged in aninsurance contract are the surrender option,the renewal privilege, the policy loan option,the settlement option and the over-depositingoption. Nevertheless, on top of the aboveoptions which explicitly stand out in insurancecontracts, there exist other options which arenot so explicit (or do not appear as such) butare still options of the insurer, such as insur-ers’ right to hold on their account the techni-cal provisions when policyholders discontinuethe payment of insurance premiums, the rightto levy mortality, expense and/or surrendercharges which are proportional to the level ofthe assets managed by the insurer, etc.

The above perspective has direct consequencesfor an undertaking’s risk valuation and riskmanagement methodologies. Risk valuation(in the form of either technical provisions orcapital requirements) under Solvency II mustemploy market-based economic theories(market consistent valuations), breaking withmore traditional, actuarial practices and meth-ods. The role of risk management in generaland of the risk manager in particular has nowbecome pivotal.

Finally, one could not help but wonder why theneed to view insurance contracts as a pool ofoptions has arisen now rather than earlier.What has changed today? To answer these

questions, we should start from the assumptionthat, in the past, insurance undertakings mainlyfocused on insurance risk and that the invest-ment risk assumed was not complex at all,while policyholders valued the long-term gainsand protection that they would enjoy from aninsurance coverage rather than the short-termprofits that would ensue from the (short-sighted) exerice of any rights or optionsoffered. In recent years, competition in theinsurance market has increased globally,mainly because of policyholders’ search foryield. In response to increased competition,insurers reduce their profit margins and offerpolicyholders a wider array of more complexrights and options, which are increasinglylinked to market returns (e.g. guaranteed inter-est rates, guaranteed returns, etc.). In tandemwith this trend in insurance products, capitalmarkets have become more efficient (in thesense that information is diffused very rapidlyand is available when investors plan theirinvestment choices) and uncertainty (which isobserved in markets as volatility) has height-ened, thereby considerably increasing the valueof rights and options that are embedded ininsurance contracts for policyholders.

Thus, it is clear that those insurers which willfail to recognise the rights and options offeredto policyholders and manage them in a waythat is consistent with capital markets jeopar-dise their financial soundness and ultimatelytheir viability, to the extent that market behav-iour as well as policyholders’ reactions arelargely included in the assumptions used byactuaries both in the pricing and reserving ofinsurance products. However, this is notallowed under Solvency II, as valuationmethodologies that are not market-basedignore the ever-changing reality. Against thisbackdrop, risk management and risk managersare a sine qua non for Solvency II.

C.3 RISK MANAGEMENT UNDER SOLVENCY II

If one discusses with insurance executives aboutcorporate risk management practices in aneffort to explain the new perspective that is


warranted under Solvency II, one will undoubt-edly be confronted by their prompt and justi-fiable answer: “But insurers are by default firmsthat have been dealing with risk for centuriesand, if they had not had effective risk manage-ment, the insurance market would not have sur-vived as a business industry”.

Undeniably, insurers do not expect a manualto highlight the necessity of risk managementor some innovative regulatory framework (likeSolvency II) to introduce such a requirement.They are already well aware of that necessityand have been managing risks for years. So,what’s all that fuss about a purportedly ground-breaking risk management practice introducedby Solvency II?

Let us then explain the true reasons for whichrisk management, as described in the interna-tional literature over recent years and envis-aged in Solvency II, is a novelty. Its originalityhas nothing to do with risk management actionsper se (e.g. that fact that reinsurance should beused as a tool for the management of mortal-ity risk is by no means seminal); it rather con-cerns the systemisation, the consistency and theeffectiveness of the methodology.

Regarding the systemisation of risk manage-ment, within an insurance firm it is acknowl-edged that, while in most cases all risks aredealt with, this is done in a non-systematic wayby several operational units or staff membersof the firm, with often overlapping or even con-flicting tasks, or that risks are managed in sucha way that the firm ignores the total level of riskassumed. In other words, management fails toaddress all risks individually and on aggregate,explore their possible interlinkages investigatetheir impact on the firm’s capital and assign tospecific operational units or employees clearresponsibilities and powers in order to takeconcrete action for the management of indi-vidual or aggregate risks. Such an endeavourmust be supported by the necessary culture, aswell as by a risk-oriented organisation at thefirm level or, as commonly known, by appro-priate risk governance. This is the only way to

reveal any inconsistencies as to how the samerisk is managed by different functions in a firm.

Turning to the effectiveness of risk manage-ment, it should be noted that, over time, crisismanagement has been the only viable alterna-tive to risk management – although it is muchcostlier, time-consuming and, in most cases,embarrassing.

In all business decisions there is an inherent riskelement. For instance, there are inherent risksin business decisions regarding the market onwhich a firm chooses to target, the productsoffered, their pricing, their reserving method,the management of investments made with themoney received by policyholders, the choice ofthe distribution channel for insurance products(e.g. online or via agents or brokers) etc.

As time passes, all individual business deci-sions and their ensuing risks make up the firm’stotal risk portfolio. This portfolio has its ownunique risk profile. The firm-specific risk pro-file determines the level as well the degree ofstability or variability of the profits dependingon the business and/or the economic cycle.From this perspective, it is acknowledged thatbusiness decisions are not disconnected fromone another and that a firm encompasses allthe decisions that it has made over time andwhich continue to have an impact. Some deci-sions have been made with a view to intro-ducing new risks into the firm (e.g. the decisionto develop a new insurance product or use anew distribution channel for those products,such as the Internet), others were aimed atreducing risks (e.g. the decision to purchasereinsurance or hedging, i.e. the introduction ofa market risk hedging system) and others atincreasing the risk that ensued from previousdecisions (e.g. the decision to further expanda business line). For a firm to effectively man-age its total risks, it is necessary to address andmanage underlying risks not only on an indi-vidual basis, but also their interdependencies.

The drawback of risk management is that it hasa different meaning to different people. Under

44Economic Bulletin

December 2016 71

Solvency II, the concept of risk managementtakes the broadest possible concept, at theenterprise level. In this vein, risk managementis all about acquiring a perspective which viewsthe firm as a single risk portfolio, with indi-vidual risks in constant interaction. From thispoint of view, management involves the fullspectrum of risks to which an insurer isexposed and subsequently calls for a singleapproach to addressing them. This perspectivehas the advantage of not overlooking the use-fulness of an appropriate risk-oriented atti-tude, while it should be noted that it does notrefer to the mitigation of risks but rather to theimprovement of profitability opportunities andprospects, with the ultimate goal of maximis-ing the value of a firm’s capital.

Without this approach to risks as a single port-folio, the Board of an insurer perceives risks asseparate pieces of a puzzle rather than gettingthe whole picture. In many firms, the risk man-agement function generates literally hundredsof pages in risk assessment reports on amonthly basis. Yet, even after all those reports,the Board is not adequately briefed to be ableto make informed decisions and, in many cases,to paint a full picture of its firm:

― What are the ten main risks assumed by thefirm?

― Is there a concise report that is delivered ona regular basis to the firm’s Board, showing thelevel of insurance risk, operational risk, coun-terparty risk and market risk, both currentlyand over time?

― What is the degree of the firm’s compliancewith internal procedures and existing legisla-tion?

― Were past actual losses or past actualadverse events inflicted on the firm associatedwith already identified risks?

― Is the firm managed with a view to max-imising risk-adjusted profits?

If a firm cannot answer any of the above ques-tions with certainty, then it is highly likely thatit will benefit from an integrated approach toaddressing all risk components, like theapproach introduced by Solvency II.

C.4 THE CORNERSTONES OF SOLVENCY II

The above analysis clearly demonstrates thatSolvency II calls for a different perspective onthe nature of an insurer as a financial institu-tion and on insurance contracts as financialinstruments. Against this background, it is eas-ily understandable that, as a result of theabove, the cornerstone of Solvency II shouldbe the systematic, effective and consistentmanagement of current or future risks facinginsurance and reinsurance undertakings.

Further lessons contributing to the formulationof Solvency II have been learnt from past insol-vency cases in Europe.

This is explained by the fact that private insur-ance is among the few business activities inwhich the consumer-policyholder pays inadvance to purchase the product (in the formof insurance premiums) in exchange for apromise of a future benefit. Besides, this prom-ise involves the payment of a much higheramount of money at an unknown yet criticaltime, as this is exactly when the consumer ismost financially vulnerable and in need of theinsurance payout. In this context, an eventualinsolvency of the insurer is crucial for the pol-icyholder.

Looking at insolvency cases in the broaderfinancial sector and despite any differences asto the manifestation of each case, a number ofcommon conclusions and lessons were drawn.Namely, the Board’s responsibility for thefunctioning of the firm, the implementation ofan effective governance system, the need offirms’ self-limitation in terms of risk taking andfinally the important role of cash.

One of the core elements, and ―according tomany― cornerstones, of Solvency II is the


Board’s responsibility for the smooth func-tioning of an insurance firm. If this appears tobe self-evident, it should be recalled thatunder Solvency I the Board of a firm bore nodirect responsibility, at least vis-à-vis thesupervisory authority, for its management,and accountability was sought ex post facto,after the firm had become insolvent. It wasbelieved ex ante that each Board had trans-ferred the two key responsibilities to others:on the one hand, the management responsi-bility to the “administration and managementofficer” and on the other hand, the responsi-bility for forming technical provisions (that is,the responsibility for checking almost all bal-ance sheet liabilities) to the appointed actu-ary, who was personally accountable to thesupervisory authority. Solvency II, besidesattributing to the Board full responsibility foradministration and management, takes onestep further, bringing to the heart of eacheffort the fact that all staff must be fully awareof their risk-roles and tasks, starting from theBoard members and ending to each and everyone of the employees, working even at themost remote branch. This knowledge must beaccompanied by a deep understanding of howtheir individual tasks can affect risks at theenterprise level, as well as how their opera-tions and responsibilities interact with therespective operations and responsibilities ofthe other functions of the firm.

A prerequisite for the effective risk manage-ment of an insurer is the existence of an effec-tive governance system, which shall ensure allnecessary safeguards to preventing anyone,either an individual or a group of people withinthe firm, from acquiring excessive power andtaking risks on behalf of the firm. This mayrefer, for instance, to a specific underwriterauthorised to assume excessive risks, or anemployee who may determine and amend atwill reinsurance contracts, or even a CEOwhose (mostly oral) orders are executed with-out any objections, even if they infringe uponpolicies already established by the Board. It isimportant to point out that an effective gov-ernance system, coupled with clearly distinct

responsibilities, is the only safeguard that afirm has against human errors and flaws in pro-cedures and systems. Solvency II, acknowl-edging this need, introduces the concept of keyfunctions, which are the actuarial function, therisk management function, the internal auditfunction and the compliance function. In addi-tion, it introduces requirements for specificprocedures and processes which are imple-mented by independent experts, such as theindependent overview of assets and liabilitiesvaluations and the independent validation ofrisk measurement models. Lastly, it establishesthe requirement that each firm recognises andaddresses all conflicts of interest that havearisen or may arise in the future.

The Solvency II framework can be charac-terised as quite liberal. There no longer existany restrictions which had been establishedunder Solvency I on the type and the level ofinvestments and insurance risks that a firmwas allowed to assume.22 From now on, eachfirm is free to determine its investments byitself, without any restrictions, and may alsoassume insurance risks unconditionally. Con-sidering that, each firm should exhibit self-dis-cipline, setting specific restrictions (tolerancelimits) on the nature of risks it proposes tocover. Furthermore, for each individual typeof risk falling within its tolerance limits, thefirm must set specific risk limits. Similar to thestrategic and business plans that determinewhere the firm plans to move to, restrictionsand limits determine when a firm must stop.Finally, on the basis of studies that have beenconducted from time to time worldwide, it isclear that all kinds of fraud, abuse, theft andinfringement in general within firms arealmost identical in one respect: they invariablytrack the firm’s cash flows. Besides, all kindsof errors or failures have multiple adverseeffects on an organisation if they affect cash.Under Solvency II, a firm’s cash outflows and

44Economic Bulletin

December 2016 73

22 Solvency I specified which investments an insurer was allowed toundertake within clear-cut limits. By way of illustration, an insurercould not invest its assets backing the technical reserves in financialderivatives. Moreover, there were currency matching requirements,in the event that investments and obligations were denominated indifferent currencies.

inflows are placed at the centre of attention.The actuarial function takes on a key role insafeguarding the flow of cash, as all valuationsthat it carries out and all procedures that itfollows for the validation of the models arebased on cash flows.

D THE COMPONENTS OF SOLVENCY II

D.1 THE PARTS OF THE LEGISLATIVE FRAMEWORKOF SOLVENCY II

Solvency II has been designed in such a way asto ensure a modern, innovative and liberalregime of prudential supervision. Its mainobjective is to strengthen the protection of pol-icyholders, whilst contributing to financial sta-bility, the creation of fair and stable marketsand the integration of the European insurancemarket. Solvency II rests upon three pillars,consisting of (i) quantitative requirements, (ii)qualitative requirements that are mainly asso-ciated with governance and supervisionrequirements, and (iii) supervisory reportingand public disclosure.

These three pillars, albeit core elements of Sol-vency II, are not expressly stated in legislativetexts, at least not explicitly. The provisionsregarding each pillar are scattered throughoutlegislative texts23 and the only way to pinpointthem is by making a reference to their content:anything associated with quantitative require-ments (calculations, methodologies, valua-tions) is part of Pillar I; anything relating toorganisational or supervisory requirements isincluded in Pillar II; and any arrangementsregarding public disclosure or reporting forsupervisory purposes belong to Pillar III.

The legislative framework of Solvency II isstructured in 6 parts. Those parts, as well astheir titles, can be easily identified in Law4364/2016.

― Part 1 is “General rules on the taking-upand pursuit of direct insurance and reinsuranceactivities”.

― Part 2 is “Specific provisions for insuranceand reinsurance”.

― Part 3 is “Supervision of insurance and rein-surance undertakings in a group”.

― Part 4 is “Reorganisation and winding-up ofinsurance undertakings”.

― Part 5 includes “Other provisions”.

― Finally, Part 6 comprises “Transitional andfinal provisions”.

Where in this section a reference is made to anarticle without specifying the law or regulationor decision thereof, it is assumed that the arti-cle refers to Law 4364/2016.

D.2 PART 1: “GENERAL RULES ON THE TAKING-UPAND PURSUIT OF DIRECT INSURANCE ANDREINSURANCE ACTIVITIES”

The first part of Solvency II includes all mainprovisions regarding the functioning of insur-ance and reinsurance undertakings on a solobasis (as opposed to groups, as laid down inPart 3). Part 1 comprises ten chapters, num-bered in Greek characters from A to J.

The pursuit of insurance and reinsurance inGreece, as is also the case in the rest of the EU,is a regulated activity, subject to the supervi-sion of the Bank of Greece, and is undertakenby firms that meet specific requirements andwhich have obtained an authorisation by theaforementioned supervisory authority. Thefirst three chapters A, B and C provide for suchmatters. In greater detail:

― Chapter A includes Articles 1-9. This chap-ter determines the scope of the law (i.e. towhich undertakings and activities Solvency IIapplies and which undertakings are excludedfrom its provisions). Furthermore, this chapterincludes a classification of insurance opera-


23 In some cases, it is not uncommon to find provisions regarding allthree pillars under the same article or even under the sameparagraph of an article.

tions according to classes of insurance (9classes for life insurance and 18 classes fornon-life insurance).

― Chapter Β includes Articles 10-18 and cov-ers the taking-up and pursuit of direct insur-ance or reinsurance. This chapter specifies thatthe pursuit of insurance and reinsurance oper-ations is a supervised activity and is subject toprior authorisation granted by the supervisoryauthority. In addition, it lays down the condi-tions for authorisation with regard to both theundertakings seeking authorisation and theirshareholders.

― Chapter C (Articles 19-28) comprises thegeneral rules of supervision and determines itsgeneral principles and its main objective, aswell as general supervisory powers. WhileChapters A and B are, as a rule, a copy ofexisting provisions from the previous super-visory framework (e.g. the authorisation ofinsurance undertakings is not a novelty intro-duced by Solvency II), Chapter C is an utterlynew chapter, which, as it regards supervision,is largely part of Pillar II, although there arealso provisions regarding reporting forsupervisory purposes (Article 24), which arepart of Pillar III.

Chapters D, F and G constitute the core of Sol-vency II. These chapters comprise the bulk ofarrangements that belong to all three Pillars,as they determine quantitative requirements,governance requirements and public disclosurerequirements. In more detail:

― Chapter D (Articles 29-47) comprises gov-ernance requirements, with its largest partbelonging to Pillar II, while Articles 38-42belong to Pillar III since they involve the pub-lic disclosure of information.

― Chapter F (Articles 50-106) contains quan-titative requirements, technical provisions andown funds.

― Finally, Chapter G (Articles 107-114)relates to cases of insurance and reinsurance

undertakings which are non-compliant withcapital requirements, or are in difficulty or inan irregular situation.

Chapter E, which was not mentioned above,consists of only two Articles (48 and 49)regarding composite insurance undertakingsthat pursue life and non-life insurance activi-ties simultaneously and the distinct separationof life and non-life insurance management.

Lastly, the last three chapters of Part A,namely H, I and J (Articles 115-144), relate tothe freedom of undertakings established inanother EU Member State to pursue cross-border activities, through a branch orremotely through the freedom to provide serv-ices, in Greece and vice versa, as well as theconditions for the taking-up of business byundertakings of third (non-EU) countriesthrough branch establishment.

D.2.1 Authorisation of insurance and reinsuranceundertakings

Insurance and reinsurance are regulated activ-ities. Therefore, their pursuit is subject to priorauthorisation by the supervisory authority(Article 10). Nevertheless, this does not implythat insurance undertakings established inGreece are not allowed to seek for and pur-chase reinsurance from reinsurance under-takings elsewhere in the world.

The operations which an insurance or rein-surance undertaking is allowed to carry out areclearly defined and distinctly separated intotwo large categories: life and non-life insur-ance. For reinsurance undertakings, this sep-aration is sufficient. The authorisation grantedby the supervisory authority (Article 11 para.4) may concern life reinsurance activity, non-life reinsurance activity, or all kinds of rein-surance activity (both life and non-life).

For insurance undertakings, the authorisationinvolves specific risks, classes and groups ofclasses of direct insurance and/or all activities,either in life or in non-life insurance. What

44Economic Bulletin

December 2016 75

insurance undertakings are not authorised is topursue life and non-life insurance activitiessimultaneously (in contrast with reinsuranceundertakings which, as already mentioned, areallowed to do so). At this point, there is a par-ticularity in terms of insurance undertakingsthat pursued simultaneously both life and non-life insurance activities before Greece joinedthe then European Economic Community(EEC) on 1 January 1981. Composite insur-ance undertakings which had received such anauthorisation in Greece before 1 January 1981may continue to pursue those activities simul-taneously (under Article 48 para. 5).

Insurance undertakings may pursue reinsur-ance activities only for the classes alreadyauthorised. Conversely, the pursuit of thedirect insurance business by reinsuranceundertakings is not authorised. The main rea-son for the above prohibition lies in the factthat the framework for the reorganisation andwinding-up of insurance undertakings with spe-cial privileges to policyholders and benefici-aries is not applicable to reinsurers, but exclu-sively to insurers.

There are two types of authorisations grantedby the supervisory authority to an insurance orreinsurance undertaking. Pursuant to the firstand most common type, the authorisation isvalid for the entire EU. In this way, all autho-rised insurance or reinsurance undertakingsare granted a “European passport” enablingthem to pursue business in any other EU Mem-ber State, either through branches (right ofestablishment) or remotely (freedom to pro-vide services). The second type is a differentversion of the above authorisation and isgranted to a particular category of insuranceundertakings (the so-called “excluded insur-ance undertakings”) which are excluded fromthe scope of the three Pillars of Solvency II.This special authorisation is valid only inGreek territory (Article 12).

Lastly, the aforementioned classification oflife and non-life insurance in classes or groupsof classes relates solely to the authorisation of

an insurance or reinsurance undertaking aswell as to the kind of activities it may pursue.For all necessary calculations and reporting tothe supervisory authority, in contrast with theprovisions of the previous framework, theabove classification is no longer used, but adifferent one is employed, which is based onhomogeneous risk groups and as a minimumby lines of business or other case-by-caseapproaches.24

D.2.2 Excluded insurance undertakings

Solvency II places emphasis on the protectionof all policyholders of all insurance undertak-ings. On that account, exclusion from the scopeof Solvency II [due to size] is quite limited, assmall and medium-sized insurance undertak-ings are subject to the principle of propor-tionality. However, in the event that the pro-visions of the framework, even after the appli-cation of the proportionality principle, are tooburdensome for such undertakings, then theseare excluded from the scope of Solvency II.

Exclusion of an undertaking from the scope ofSolvency II is made in such a way so as not tolead to distortions of competition by creatingdual-track insurance undertakings, or to avertany unreasonable or undue results, e.g. a sit-uation in which an undertaking falls within thescope of Solvency II at the stage of growth, butis excluded from it when it has accumulatedlarge obligations which are on a run-off situa-tion, or whether it uses reinsurance as an exclu-sion mechanism or not.

Before analysing what it means for an under-taking to be excluded from the scope of Sol-vency II, it is important to describe what it doesnot entail:

― When an undertaking is excluded from Sol-vency II, this does not mean that it is also


24 By way of illustration, portfolios consisting of obligations andcorresponding assets, in cases of transfer or when using specificadjustments (e.g. matching adjustment), or portfolios ofobligations for which an undertaking may change the insurancepremiums and/or the benefits under similar circumstances and withsimilar effects, when calculating the contract boundaries.

excluded from the requirement to obtainauthorisation by the competent supervisoryauthority; otherwise, we may end up with unau-thorised undertakings pursuing insuranceactivities in Greece.

― The fact that a concrete, uniform, EU-wideframework for the functioning of excludedundertakings is not envisaged does not implythat each Member State is not obliged to deter-mine its own framework. Member States arefree to determine on their own the frameworkfor the functioning of such undertakings. Ofcourse, it is not uncommon that national lawsin some countries regarding such undertakingsare consistent with the provisions of SolvencyII – in other words, in some countries theremay be no excluded undertakings.

― The right of establishment and freedom toprovide services, as specified in the Treaty onthe European Union, is valid, irrespective ofthe “passport” system that is provided for inthe framework. Yet, the “passport” con-tributes to an automated access, thereby lim-iting the supervisory responsibilities and pow-ers of the host State. Against this background,excluded undertakings are not prohibited topursue insurance or reinsurance activities inother EU Member States, but to do so, theyneed an approval (authorisation) by the hostMember State.

Undertakings which may be excluded from thescope of Solvency II must fulfil one of the fol-lowing conditions: they must either take theform of a mutual undertaking and at the sametime fully transfer their risk25 or be very small.26

In the first case, a full transfer of risks can taketwo alternative forms: either the conclusion ofan agreement which provides for the full rein-surance of the insurance policies issued or theconclusion of an agreement under which theaccepting undertaking is to meet the liabilitiesarising under such policies in the place of theceding undertaking. A prerequisite for rein-surance or accepting undertaking is that thecounterparty of the mutual undertaking is

another mutual undertaking.27 If all of the con-ditions are met, then the mutual undertakingis automatically excluded from the scope ofSolvency II without necessitating priorapproval by the supervisory authority.

In the second case (i.e. exclusion due to size),exclusion is subject to approval by the super-visory authority, while insurance undertakingsmust fulfil two general conditions: (a) theymust not have established a branch in anotherEU Member State, or must not pursue insur-ance activities in another EU Member State inaccordance with the provisions regarding thefreedom to provide services; and (b) their busi-ness must not include insurance or reinsuranceactivities covering liability, credit and surety-ship insurance risks.

In addition to the aforementioned general con-ditions, undertakings which may be excludedfrom the scope of Solvency II due to size mustmeet the following thresholds:

― their annual gross written premium incomedoes not exceed EUR 5 million;

― the total of technical provisions does notexceed EUR 25 million;

― the business of the undertaking does notinclude reinsurance operations or, if it does, itis very small relative to its direct insuranceoperations;28

― either they do not belong to an insurancegroup or where the undertaking belongs to agroup, the technical provisions of the group donot exceed EUR 25 million.

44Economic Bulletin

December 2016 77

25 Article 7(1) of Law 4364/2016.26 Article 7(2) 2 of Law 4364/2016.27 The rationale behind the counterparty being a mutual undertaking

is associated with the relatively stronger links that are developedin those cases, as for a mutual undertaking to provide reinsuranceto another mutual undertaking, the latter is required to becomemember of the first.

28 In this context, reinsurance operations should be not more than10% of gross written premium income or not more than 10% of thetotal of technical provisions – for the underlying insurance andreinsurance obligations. For reinsurance operations to becharacterised as small, on top of the above conditions, they shouldnot exceed EUR 0.5 million of gross written premium income orEUR 2.5 million of total reinsurance obligations.

The exclusion of an undertaking from SolvencyII requirements due to size is a dynamicprocess which is verified both during the initialauthorisation of the undertaking and for eachyear of operation. At the time of initial autho-risation, in order to be excluded from the scopeof Solvency II, the undertaking must establishin its scheme of operations that its activities arenot expected to exceed any of the thresholdswithin the following five years.

Undertakings may be excluded due to size notonly upon first authorisation but also through-out their operation. If an undertaking,although in the beginning it did not fall underthe scope of Solvency II, has followed a busi-ness downsizing practice with a view to meet-ing the conditions for exclusion for three con-secutive years and is expected to continue tomeet those conditions in the following fiveyears, it may request to be excluded due to size.

If an undertaking has been excluded due tosize either upon first authorisation or later, onaccount of reduced business, it must each yearconfirm that it has met all quantitative condi-tions for exclusion over the past three consec-utive years. If this fails to be verified, theundertaking ceases to be entitled to exclusionand falls within the scope of Solvency II fromthat year onwards.

Finally, it is important to note that the afore-mentioned amounts are adjusted for inflation(based on the Harmonised Index of ConsumerPrices of the EU). As those amounts mustreflect the actual size of each undertaking andshould not easily be affected by the manage-ment practices of each undertaking, technicalprovisions are gross of reinsurance, i.e. tech-nical provisions are not calculated net of theamounts recoverable from reinsurance con-tracts.

D.2.3 Head offices and legal form of the insuranceand reinsurance undertaking

While under the previous supervisory frame-work the head office of an insurance or rein-

surance undertaking should be located only inspecific Greek cities, such restriction is nowremoved by Solvency II. . Regarding the legalform of insurance undertakings, there arethree options: société anonyme, Europeancompany, or mutual association. In all threecases, the shares should not be anonymised.Mutual undertakings have an additionalrestriction: they may only be non-life insurers.All of the above, as well as the operations laiddown in their statute, any amendment theretoregarding the type of their activities and anyincrease or decrease in their share capital, aresubject to prior approval by the supervisoryauthority.

D.2.4 Pillar I: Quantitative requirements

“Pillar I” of Solvency II concerns the quanti-tative requirements on (re)insurance under-takings and comprises the following elements:

1. Valuation of assets and liabilities 2. Technical provisions 3. Own funds 4. Solvency capital requirement 5. Minimum capital requirement 6. Investments.

(Re)insurance undertakings value assets andliabilities as follows, unless stated otherwise:(a) assets are valued at the amount for whichthey could be exchanged between knowledge-able willing parties in an arm’s length trans-action; and (b) liabilities are valued at theamount for which they could be transferred, orsettled, between knowledgeable willing partiesin an arm’s length transaction.

When valuing liabilities, no adjustment is madeto take account of the own credit standing ofthe insurance or reinsurance undertaking.

Insurance and reinsurance undertakingsestablish technical provisions for all of theirinsurance and reinsurance obligations towardspolicyholders and beneficiaries of insurance orreinsurance contracts. The value of technicalprovisions corresponds to the current amount


that insurance and reinsurance undertakingswould have to pay if they were to transfer theirinsurance and reinsurance obligations imme-diately to another insurance or reinsuranceundertaking.

Apart from technical provisions and other obli-gations, undertakings hold an additionalamount of assets or third-party commitmentsthat can be called in specific circumstances.This additional amount takes the form of ownfunds, i.e. funds that can be called up to absorbany unexpected losses, if and when they occur,and is in principle calculated as the excess ofassets over liabilities, the so-called “economiccapital”. In this context, own funds are used asan extra level of security (protection of poli-cyholders). The amount of own funds is equalto the sum of on-balance sheet items and anyoff-balance sheet items. On-balance sheetitems are called basic own funds and are equalto the sum of economic capital and subordi-nated loans. Those loans, although they cannotbe treated as own funds in the narrow account-ing sense, since they are not “own” funds of theundertaking, exhibit some characteristics thatcould place them under own funds.29 Off-bal-ance sheet items are called ancillary own funds,usually comprise the unpaid share capital orinitial fund, letters of credit and guarantees, orany other commitments, such as any futuresupplementary contributions by members of amutual association, and are subject to priorsupervisory approval. Own funds are classifiedinto three tiers (Tier 1, 2 and 3) depending ontheir loss-absorbency and to the extent thatthey meet five main criteria:

― Permanent availability ― Subordination ― Absence of incentives to redeem ― Αbsence of mandatory servicing costs ― Absence of encumbrances.

Basic own funds can be classified in one of thethree tiers (Tier 1, 2 or 3), whereas ancillaryown funds can only be classified in Tier 2 or 3.30

Eligible own funds must be higher than the sol-vency capital requirement.

The solvency capital requirement correspondsto the economic capital that a (re)insuranceundertaking must hold to limit the annualprobability of default to 0.5%, that is theprobability of default should be 1 over 200years, and to reflect the actual risk profile ofthe undertaking, taking account of all quan-tifiable risks.

The solvency capital requirement is calculatedin two alternative ways. The first involves theuse of a standard formula (standardisedmethod), which aims to strike the right balancebetween sensitivity to risk and practicability.The standard formula enables the use ofparameters specific to each undertaking,depending on the case, as well as of standard-ised simplifications for small and medium-sized undertakings. The risk measure is theValue-at-Risk of basic own funds with a 99,5%confidence level over a one-year period, cov-ering at least the following risk modules: non-life underwriting risk, life underwriting risk,health underwriting risk, market risk, coun-terparty default risk, and operational risk. Thesecond method involves the use of internalmodels that have been developed by the(re)insurance undertakings themselves on thebasis of specific principles.

The minimum capital requirement representsan amount of capital below which policyhold-ers’ interests are exposed to an unacceptablelevel of risk, were insurance and reinsuranceundertakings allowed to continue their oper-ations. As a result, if the undertaking fails tocomply with the minimum capital requirement,the supervisory authority is allowed to activatea measure of last resort, i.e. a withdrawal ofauthorisation.

All investments by undertakings are managedand monitored in accordance with the prudentperson principle, which requires from

44Economic Bulletin

December 2016 79

29 The key characteristic of subordinated loans is the fact that thelender has agreed to not receive interest or principal, which willinstead be allocated in favour of policyholders in the event offinancial difficulties facing the undertaking or, otherwise, to receiveit only when financial difficulties are remedied.

30 They cannot be classified in Tier 1 as they are not readily available.

(re)insurance undertakings to give due con-sideration to the risks to which they areexposed. In this context, the legislation doesnot envisage any particular categories ofinvestments, and no prior supervisory approvalis required. Nevertheless, insurance and rein-surance undertakings should only invest inassets and instruments whose risks they canproperly manage, monitor and control.

Last but not least, a key chapter of Pillar I isthe long-term guarantee measures (LTG)which (re)insurance undertakings provide totheir policyholders. Those measures are aimedat addressing artificial volatility in the balancesheets of undertakings under Solvency II.31

Such measures are the symmetric adjustmentin the equity risk module (SA), the volatilityadjustment (VA), the matching adjustment(MA), transitional measures on risk-free rates(TMRFR), transitional measures on technicalprovisions (TMTP) and the extension of therecovery period (ERP).

D.2.5 Pillar II: Quantitative requirements –Supervision and Governance

The quantitative requirements and the super-visory rules for (re)insurance undertakings(Pillar II of the Solvency II framework) are laiddown in two sections, of which one comprisesthe Supervisory Authorities and General Rulesand applies to supervisory authorities, and theother comprises the Governance System andapplies to (re)insurance undertakings.

The main objective of supervision The main objective of supervision according toSolvency II is the protection of policyholdersand beneficiaries. The term beneficiary isintended to cover any natural or legal personwho is entitled to a right under an insurancecontract. Financial stability and fair and stablemarkets are other objectives of supervision.

Striking a balance between the main and otherobjectives of supervision calls for an ongoingeffort, as, although on a first reading theyappear to be consistent with one another,

under special circumstances the same objec-tives may dictate different decisions or actionson the part of a supervisory authority. On theone hand, the other objectives of supervisionmust be taken into account without under-mining the main objective, i.e. the protectionof policyholders and beneficiaries, as a sol-vency framework should give precedence to theprotection of those persons ―otherwise,what’s the point in undertakings being sol-vent?― while, on the other hand, supervisorsmust duly consider the potential impact oftheir decisions on the stability of the financialsystems concerned in the EU, in particular inemergency situations.

Solvency II is not limited to those objectives,but sets an additional mandate to the supervi-sory authorities, in the form of a limitation,requiring that, in times of exceptional move-ments in the financial markets, supervisoryauthorities should take into account the poten-tial pro-cyclical (recessionary) effects of theiractions.

Supervisory principles and general rules Under Solvency II, supervision rests upon arisk-oriented, forward-looking approach, whichis allowed for by the existence of a system thatcaptures the actual risk profile of undertakings,on the basis of economic principles, taking fulladvantage of the information provided by thefinancial markets. In addition, special care isgiven towards ensuring that the Solvency IIframework is not too burdensome for small andmedium-sized undertakings. Therefore, theproportionality principle is promoted, whichapplies to all Solvency II requirements.

The principle of proportionality The principle of proportionality is a new con-cept in private insurance law, based on an esca-lation of supervisory requirements on eachundertaking depending on the nature, scaleand complexity of the risks inherent in its busi-ness activities. Although the proportionality


31 The term “artificial volatility” refers to the volatility of own funds,as a result of a drop in interest rates with a concurrent increase incredit spreads.

principle governs the entire framework, it isnot clearly defined therein. Thus it is left to bedefined on a case-by-case basis and in a man-ner akin to the level and the type of risks ofeach undertaking (rather than the size of itsbusiness, unless expressly stated so by the law).It is clear that, on the basis of this principle,two seemingly identical undertakings but withdivergent risk management practices are sub-ject to different requirements.

Ladder of supervisory intervention Under the provisions regarding supervisoryauthorities, Solvency II grants to the supervi-sory authorities of all Member States the same,increased, powers, but is not limited to anexhaustive list. Thus, supervisory authoritiespossess an escalating ladder of intervention inthe undertakings depending on the degree ofany non-compliance.

First of all, supervision follows a principles-based approach without involving the imposi-tion of concrete measures and the interventionof the supervisory authority in the undertak-ing’s operation. In the event of significant non-compliance with the solvency capital require-ment, the undertaking submits a realisticrecovery plan for approval by the supervisoryauthority. In this case, supervision makes afirst-level intervention, requiring the under-taking to take concrete measures either to mit-igate the risk assumed or to raise additionalcapital with a view to remedying non-compli-ance within six months. In the event that thecondition of the undertaking deteriorates and,as a result of this deterioration, non-compli-ance with the minimum capital requirement isobserved, the undertaking submits a realisticshort-term finance scheme for approval by thesupervisory authority. In that case, supervisionmakes a second-level intervention in theundertaking, following a rules-based approachthis time for the remedy of non-compliancewithin three months. Finally, if the undertak-ing fails to comply with the rules for technicalprovisions, supervision shall not expect thesubmission of a recovery plan or financescheme; instead, it intervenes promptly and

may prohibit the free disposal of its assets oreven resort to more drastic measures. In anyevent, the measures adopted are proportion-ate, thus reflecting the level and duration ofthe deterioration of the solvency position ofthe undertaking concerned.

Stress testing A new power in the hands of the supervisoryauthorities is the possibility to require under-takings to assess their solvency standing underadverse scenarios (stress test). The supervisoryauthority may require that stress tests are con-ducted either by each undertaking on a solobasis, noting the risks specific to the under-taking concerned depending on its risk profile,or by the entire market.

Supervisory review process In order for the supervisory authorities to iden-tify those undertakings with financial, organi-sational or other characteristics that generatea high risk profile, the concept of the Super-visory Review Process (SRP) is introduced.This process is developed by the supervisoryauthorities and comprises the review andassessment of (a) each undertaking’s compli-ance with the Solvency II framework, (b) thesystem of governance, (c) the risks which theundertaking concerned faces or may face, and(d) its ability to address any possible events orfuture changes in economic conditions thatcould have adverse effects on its overall finan-cial standing.

Capital add-on under Pillar II Supervisory authorities may, only in excep-tional circumstances, impose a capital add-onon insurance or reinsurance undertakings as ameasure of last resort, in cases in which eithertheir risk profile deviates significantly from theassumptions underlying the Solvency CapitalRequirement or the system of governancedeviates significantly from the standards laiddown in the Solvency II framework.

Nevertheless, even in the aforementioned casesof deviations, the imposition of a capital add-on by the supervisory authority is exceptional,

44Economic Bulletin

December 2016 81

suggesting that there is a course of actions tobe followed by each supervisor in order to rem-edy those deficiencies, with the imposition ofa capital add-on being the very last step. On thebasis of the proportionality principle, thesupervisory authority is free to consider impos-ing a capital add-on and has the power toimpose it, but is by no means obliged to do so.32

Outsourced activities Another power that the supervisory authoritiespossess refers to the supervision of functionsthat have been outsourced by (re)insuranceundertakings, in particular given the possibil-ity of outsourcing key functions.

In this context and in order to ensure effectivesupervision of outsourced activities or func-tions, it is essential that the supervisory author-ities have access to all relevant data held by theoutsourcing service providers, as well as theright to conduct on-site inspections at the busi-ness premises of the outsourcing serviceprovider, regardless of whether the latter is aregulated or unregulated entity.

Governance requirements The governance requirements on (re)insuranceundertakings which are laid down by SolvencyII are aimed at achieving harmonisation withthe respective requirements on the banking andsecurities sectors. Under Solvency II, rigorousgovernance requirements are a prerequisite forthe existence of an effective solvency regime, ascertain risks can only be addressed through suchrequirements rather than with the establish-ment of quantitative requirements.

Against this background, undertakings havethe following four key functions:33 the actuar-ial function, the risk-management function, theinternal audit function and the compliancefunction. The system of governance alsoincludes compliance with fit and properrequirements for persons who effectively runthe undertaking or have other key functions.

To ensure the sound operation of the systemof governance, undertakings are required to

have written policies in relation to risk man-agement, internal control, internal audit and,where relevant, outsourcing.

Lastly, the Own Risk and Solvency Assessment(ORSA) is introduced, through a process ofinternal assessment conducted by the under-taking itself as part of its strategic decisions,which also serves as a supervisory tool.

D.2.6 Pillar III: Supervisory reporting and publicdisclosure

Supervisory reporting and public disclosuremake up Pillar III of the Solvency II frame-work and refer to the information to be pro-vided for supervisory purposes, as well as pub-lic disclosure.

Information to be disclosed34 is contained in asingle document entitled “Solvency andFinancial Condition Report”. In this report,which is released on annual basis, each under-taking includes information using specificstructure and, in some cases, according to aspecific format.35 The report comprises dataregarding the business and the performance ofthe undertaking in the previous year, a descrip-tion of the system of governance, and infor-mation on its risk profile. Furthermore, itincludes information on the methods used forthe valuation of assets, technical provisions andother liabilities as well as a description of itscapital management.

Information provided for supervisory pur-poses36 can be found under the general head-ing “regular supervisory reporting” and com-


32 If not (when capital add-ons are imposed in not so exceptionalcircumstances), there is a risk of indirectly admitting that theSolvency II framework fails to establish a new risks-based solvencyregime. Needless to say, the phrase “in exceptional circumstances”should in no way be interpreted as drawing the line in the numberof cases in which a supervisory authority may impose such capitaladd-ons.

33 A function is defined as an internal capacity to undertake practicaltasks of administration, management, representation or controlover specific operations of an undertaking, within a system ofgovernance (Article 3(29) of Law 4364/2016.

34 Articles 290-299 of the Regulation (EU) 2015/35.35 The procedures, formats and templates of the report on solvency

and financial condition are specified in Implementing Regulation(EU) 2015/2452.

36 Article 304 of Regulation (EU) 2015/35.

prises: (a) the (published) solvency and finan-cial condition report; (b) the regular supervi-sory report, which contains information spec-ifying and further elaborating on the publishedreport;37 (c) annual and quarterly quantitativetemplates specifying in greater detail and sup-plementing the information presented in theabove reports.

D.3 PART 2: “SPECIFIC PROVISIONS FORINSURANCE AND REINSURANCE”

The second part of Solvency II (Articles 145-169) comprises specific provisions to life insur-ance, non-life insurance, health insurance andlegal expenses insurance. These provisions areactually recasts of older provisions regarding,by way of illustration, the law applicable toinsurance contracts, issues specific to com-pulsory insurance, general information for pol-icyholders before the conclusion of insurancecontracts, cancellation periods, finite reinsur-ance and special purpose vehicles, while a setof provisions facilitating EU-wide co-insuranceoperations is contained at the end.

D.4 PART 3: “SUPERVISION OF INSURANCE ANDREINSURANCE UNDERTAKINGS IN A GROUP”

The third part of Solvency II provides for thesupervision of groups, in particular insurancegroups. For this kind of supervision, the defi-nition provided by the Solvency II frameworkis of great relevance. Insurance and reinsur-ance undertakings which are linked with oneanother in one of the following two ways areconsidered a group:38

― a group of undertakings that consists of aparticipating undertaking, its subsidiaries andthe entities in which the participating under-taking or its subsidiaries hold a participation,or

―a group of undertakings that is based on theestablishment, contractually or otherwise, ofstrong and sustainable financial relationshipsamong those undertakings, under certain con-ditions.

As for the understanding of Solvency IIrequirements on groups it is important to havepreviously acquired a full grasp of the rela-tionships which two or more undertakings canestablish, all relevant definitions are presentedin the Annex of the present study.

D.4.1 Cases of application of group supervision

Supervision at the group level applies to thefollowing three cases reflecting three differenttypes of group relationships. In the first case,the group relationship refers to either insur-ance and reinsurance undertakings, which area participating undertaking in at least oneinsurance or reinsurance undertaking, or insur-ance and reinsurance undertakings, the parentundertaking of which is an insurance holdingcompany or a mixed financial holding companywhich has its head office in the EU. In thiscase, the scope of supervision extends to theentire spectrum of the group’s solvency andfinancial condition.

In the second case, the group relationshipinvolves insurance and reinsurance undertak-ings, the parent undertaking of which is aninsurance or reinsurance undertaking or aninsurance holding company or a mixed finan-cial holding company that has its head officeoutside the EU. In this case, supervision restsupon an assessment of the equivalence ofthird-country supervisory regimes and variesaccordingly.

In the third case, the group relationship refersto insurance and reinsurance undertakings, theparent undertaking of which is a mixed-activ-ity insurance holding company. In this case, thescope of supervision focuses rather on thesupervision of intra-group transactions than onthe group’s solvency.

In the above analysis, it has been taken forgranted that the head of each group is clearly

44Economic Bulletin

December 2016 83

37 For this reason, both reports (the report on solvency and financialcondition and the regular supervisory report) have exactly the samestructure, with identical contents.

38 Article 170 of Law 4364/2016.

identifiable. However, in most cases things arenot so clear-cut. A group may operate in morethan one countries, either within or outside theEU. For this reason, Solvency ΙΙ (Article173)states that the ultimate parent undertaking atthe EU level is deemed head of the group.. Ifa group forms a subgroup which is active onlyin one country, the supervisory authority ofthat country may, without prejudice to nationallaw,39 deem the subgroup as a national sub-group which is subject to group supervision,irrespective of the remainder of the group.

D.4.2 European insurance group supervision

The supervision of EU insurance groups (i.e.the first case of group supervision) applies:

(a) to insurance or reinsurance undertakings,which are a participating undertaking in atleast one EU insurance undertaking, EU rein-surance undertaking, third-country insuranceundertaking or third-country reinsuranceundertaking; and

(b) to insurance or reinsurance undertakings,the parent undertaking of which is an insur-ance holding company or mixed financialholding company which has its head office inthe EU.

It should be noted that, in cases where thegroup also pursues activities other than insur-ance, Solvency II is interested in that part ofthe group which concerns insurance and rein-surance operations, addressing the remainingpart either in a simple (yet clear) manner orby way of reference to other legislative texts,for instance Law 3455/2006 regarding creditinstitutions, investment firms and financialinstitutions.

EU group supervision focuses on two aspects:(a) the financial position of the group and (b)the cooperation of supervisory authorities.

The assessment of a group’s financial positionincludes group solvency, the supervision of riskconcentration and of intra-group transactions,

and the supervision of the system of gover-nance. For the cooperation of the supervisoryauthorities that are associated with a group,the role of the colleges of supervisors is pivotal.For each group with cross-border operations acollege of supervisors is set up, headed by oneof the above supervisory authorities, namelythe group supervisor, and having as membersthe supervisory authorities of the other coun-tries in which the group is active, namely thesupervisory authorities concerned to groupsupervision. The college of supervisors meansa permanent but flexible structure for coop-eration and coordination among the supervi-sory authorities concerned to group supervi-sion, as well as for facilitation of the decision-making process.

D.4.3 European group solvency

The solvency of an EU group is calculated bytwo different ways which are called Method 1and 2, respectively. Method 1 is the defaultmethod and is carried out on the basis of theconsolidated accounts. Under the accountingconsolidation-based method, data from all(re)insurance undertakings of the group areconsolidated, and the group’s own funds andSolvency Capital Requirement are calculatedas if the group were a single undertaking. Inthis case, the resulting capital requirement isthe consolidated group Solvency CapitalRequirement. The consolidated group SolvencyCapital Requirement can be calculated eitherusing the standardised method, which alsoapplies to an undertaking on a solo basis, or onthe basis of an internal model for a group40 uponrequest of permission. The permission isgranted by the group supervisor, after con-sulting with the supervisors of all MemberStates in which the related undertakings fallingunder the scope of the internal model for thecalculation of the consolidated group SolvencyCapital Requirement have their head offices,


39 This leeway is provided by Solvency II Directive (Directive2009/138/EC) and is a national choice, left at the discretion of thenational legislator. In Greece, the national legislator chose to grantthis leeway to the Bank of Greece.

40 Article 188(2) of Law 4364/2016.

i.e. the supervisory authorities involved. In theevent that this internal model is used for thecalculation of the consolidated group SolvencyCapital Requirement and of the MinimumCapital Requirement of at least one insuranceor reinsurance undertaking of the group (on asolo basis), the model is called group internalmodel,41 the supervisory authorities of theMember States, in which the head office of theparticipating as well as of any related insuranceand reinsurance undertaking using the groupinternal model is located, are called supervisoryauthorities concerned, and all parties concernedmust jointly decide whether or not to grant thepermission sought.

Method 2 is called deduction and aggregationmethod, is an alternative method to Method1 and can be used only following permissionby the group supervisor. This method is usedto calculate the capital requirement of thegroup, the aggregated group Solvency capitalRequirement, employing the relevant succes-sive aggregations and deductions of individualcapital requirements and own funds of therelated insurance and reinsurance undertak-ings, as calculated for each undertaking on asolo basis.

D.4.4 Group solvency supervision for groups withcentralised risk management

A problem facing groups with insurance orreinsurance subsidiaries in several EU Mem-ber States is that each subsidiary is subject tothe supervision of the national competentauthority. As a result, if such a group aspiresto manage the risks assumed by all of its sub-sidiaries in a common (centralised) manner, itis faced by as many possible hindrances as thenumber of the national competent authoritiessupervising its subsidiaries.

Solvency II gives to such groups, provided thatthey have a proven centralised risk manage-ment, the option, with regard to specific issuesof their subsidiaries (e.g. the possible imposi-tion of a capital add-on), to be subject to thesupervision of a single supervisor, namely the

group supervisor, instead of the other super-visory authorities concerned or, with respect toother issues, to reduce any flexibility that thenational supervisory authorities possess inenforcing measures on the supervised sub-sidiary (e.g. in the event of non-compliancewith the Solvency Capital Requirement). Thisoption is given to a group following priorsupervisory approval.

D.4.5 Solvency of groups with parentundertakings outside the EU

The supervision of groups, the parent under-taking of which is an insurance or reinsuranceundertaking, an insurance holding company ora mixed financial holding company which hasits head office in a third, non-EU country (i.e.the second case of group supervision), is sub-ject to verification of equivalence of the super-visory regime of that third country. In theevent that the third-country supervisoryregime has been deemed (by the EuropeanCommission or the group supervisor) equiv-alent to the Solvency II regime, a college ofsupervisors is set up, with participation of thethird-country supervisory authority, with aview to ensuring a proper exchange of infor-mation, and the calculation of the group sol-vency is carried out in accordance with theregulation of the third country.

The assessment of the equivalence of third-country supervisory regimes is conducted byEIOPA, and the European Commissiondecides whether there is equivalence or not.The Commission’s equivalence decision maybe full or provisional. Their difference lies inthat a full recognition of equivalence has notime limit, unless of course changes are in themeantime made in the supervisory system ofthe third country, and in this case a re-assess-ment is warranted, whereas the provisionalrecognition of equivalence is valid over a pre-determined time horizon and specific condi-tions are determined which the third countryshould meet in order to achieve full recogni-

44Economic Bulletin

December 2016 85

41 Article 189 of Law 4364/2016.

tion. Furthermore, in the event of provisionalequivalence, for the favourable provisions ofequivalence to be enforceable,42 the parentundertaking of which the head office islocated in a third country must have a totalbalance sheet exceeding the balance sheet ofeach one of its EU subsidiaries. Finally, in theevent of lacking decisions (positive or nega-tive) by the European Commission, a verifi-cation of equivalence may be carried out anda full equivalence decision may be adopted bythe acting group supervisor,43 in cooperationwith ΕΙΟΡΑ.

If there is no equivalent supervision or if equiv-alence is provisional, but there is at least onesubsidiary of the group in the EU, the balancesheet of which exceeds that of the parentundertaking, then the calculation of group sol-vency is carried out in accordance with what isprovided for in Solvency II (and not accordingto the third-country methodology).

D.4.6 Mixed-activity insurance holding companies

Where the parent undertaking of one or moreinsurance or reinsurance undertakings is amixed-activity insurance holding company (i.e.the third case of group supervision), supervi-sion is general and exercised over transactionsbetween those insurance or reinsurance under-takings and the mixed-activity holding com-pany and its related undertakings. In this case,the group solvency is not calculated.

D.5 PART 4 “REORGANISATION AND WINDING-UPOF INSURANCE UNDERTAKINGS”

The fourth part of Solvency II (Articles 220-248) provides for matters relating to the reor-ganisation and winding-up of insurance44

undertakings with a head office in Greece,including their branches in other MemberStates, as well as of branches of third-countryinsurance undertakings situated in the territoryof Greece. This Part also applies to branchesof Greek insurance undertakings which areactive in third countries, unless stated other-wise in the third-country law.

Before resorting to the withdrawal of autho-risation of an insurance undertaking, andunder specific circumstances, the supervisoryauthority may take a series of measures,aimed at safeguarding or restoring the finan-cial condition of the insurance undertaking,the so-called reorganisation measures. Suchmeasures differ from administrative measuresagainst the undertaking or shareholders thatthe supervisory authority may adopt in itscapacity as supervisor, as reorganisation meas-ures may affect pre-existing rights ofclaimants, policyholders and beneficiaries, aswell as of other contracting parties and cred-itors. Such measures refer to mandatory sharecapital increase, mandatory transfer of port-folio, suspension of payments and reductionof claims.45 For the implementation of reor-ganisation measures, the supervisory author-ity may appoint an insurance administrator,either cooperating with the Board of theinsurance undertaking or acting as a replace-ment thereof.

The supervisory measure of last resort for theprotection of policyholders is the withdrawalof authorisation of an insurance undertakingand the opening of winding-up proceedings.46

In this case, the competent authority appointsa liquidator who is responsible for adminis-tering winding-up proceedings. The winding-up of an insurance undertaking involves therealisation of its assets and the distribution ofthe proceeds (indemnification) to policyhold-ers and beneficiaries. Winding-up proceedingsend either when the insurance portfolio hasbeen disposed of (i.e. the rights of all insuredparties have been satisfied) or when the assetsof the undertaking concerned have beenexhausted. In the first case, the phase of a typ-ical liquidation follows.


42 In other words, the calculation of group solvency is carried out inaccordance with the supervisory provisions of the third country.

43 As quasi group supervisor is defined that supervisory authoritywhich would hypothetically be the group supervisor if the groupwere European.

44 In contrast with reinsurance undertakings, to which specialprovisions on reorganisation and winding-up are not applicable, butto which only common provisions for their reorganisation apply,as is generally the case for all societes anonymes.

45 Articles 277-230 of Law 4364/2016.46 As opposed to a typical liquidation.

D.6 PART 5: “OTHER PROVISIONS”

The fifth part of Solvency II includes Articles249-268. This Part provides for a number ofissues, which are not directly connected withother Parts. By way of illustration, these pro-visions regard the right to apply to the courts,the exchange rates of the euro and the revisionof amounts expressed in euro, claims repre-sentatives, loss adjusters and legal represen-tatives. Moreover, provisions are includedregarding the operation of certain sectors orbusinesses, such as capital redemption, profitsharing mechanisms and surplus funds. Part 5concludes with administrative and criminalsanctions on breaches of private insurance leg-islation.

D.7 PART 6: “TRANSITIONAL AND FINALPROVISONS”

The sixth and last part of Solvency II comprisestransitional and final provisions. This Part isstructured in three large sections: the first sec-tion refers to the authorisations of existingundertakings and the rights acquired by exist-ing branches and reinsurance undertakings, thesecond section refers to a series of transitionalprovisions for the smooth transition from Sol-vency I to Solvency II, and the third sectioninvolves phase-in provisions. Phase-in provi-sions had initially been designed with a view toenabling the supervisory authority to grantand, subsequently the undertakings to obtain,any necessary permissions or approvalsregarding their operation on the first day ofapplication of the framework (1 January 2016)at a time prior to the first application, i.e. from1 April 2016 onwards. As the transposition ofthe Solvency II Directive in Greek law tookplace later than the first day of its application(on 5 February 2016) with retroactive effectfrom 1 January 2016, those phase-in provisionswere never implemented.

Special reference is made to the insurance sub-class IV.2 “Accident and sickness in non-lifeinsurance”, which was applicable under Sol-vency I. As by Law 4367/2017 this class no

longer exists, all insurance undertakings whichcomprised that sub-class received automati-cally an authorisation for both Solvency IIclasses 1 “Accident” and 2 “Sickness”. Thus, alllife insurance undertakings which included thissub-class were converted into special compos-ite insurance undertakings, in the sense that,although they are not treated as composites,they are not entitled to further expand theirauthorisation into other classes of non-lifeinsurance.

As already mentioned, Solvency II has intro-duced a series of transitional provisions with aview to ensuring the smoothest possible tran-sition from the former supervisory framework.

E CONCLUSIONS

As mentioned above, Solvency II provides asingle (and common) framework at the Euro-pean level (and not merely at the nationallevel). This has multiple and direct conse-quences:

a) The legal framework of Solvency II shall notvary across countries or per national legislator.As a result, any country-specific circumstances,which under the previous framework couldonly be addressed with special legislative meas-ures specific to the country concerned,47 cannow be dealt with only within the range of pos-sibilities provided by the common EU-widelegal framework.

b) The solvency of (re)insurance undertakingsin all EU Member States is comparable, as itis determined on the basis of uniform criteria,in contrast with the former Solvency I regimewhich was only seemingly common.48

44Economic Bulletin

December 2016 87

47 An illustrative example is the solutions provided by Greeklegislation in the past: mass losses from the drop in share prices in2000, which were likely to significantly impair the solvency ofinsurance undertakings, were capitalised by way of derogation andwere amortised in three years. Similarly, PSI-related losses werenot recognised immediately but gradually through a derogation forGreek government bonds.

48 As the previous directives that introduced Solvency I had minimumharmonisation, after more than 30 years of application, cross-country differences were so large that, arguably, there were 28different frameworks.

c) The operation of European groups hasbecome considerably simpler, as they are notfaced by divergent supervisory requirements ineach country in which they are active, but dealwith a single EU-wide regime, which has directpositive effects on their operating costs as theynow can benefit from economies of scale acrosscountries.

d) The supervisory authorities throughoutEurope no longer act on their own, at least tothe same degree as under the previous regime,or without any coordination between them, butthey make a concerted effort, developing com-mon supervisory practices or even commonsupervisory actions, both through the institu-tion of the college of supervisors with regardto groups with cross-border presence andthrough EIOPA, which now enjoys enhancedpowers relative to the past.

In Section C of this paper, the new perspectivedictated by Solvency II was presented. This per-spective calls for a different treatment of insur-ance undertakings and insurance contracts: itinvites us to view the insurance undertaking asa financial institution that seeks, through themaximisation of its shareholders’ profits, to pro-tect policyholders and insurance contracts asfinancial instruments. This new point of view iskey to the risk valuation and management of a(re)insurance undertaking. This inevitably leadsto the adoption of market consistent risk valu-ations, an integrated approach to risks and anenterprise risk management perspective. Thepractice of an integrated, rather than a solo-based approach to risks is corroborated by thefact that the main criterion for the solvency ofan undertaking is the level of aggregate risksassumed, either using the Solvency CapitalRequirement standard formula or on the basisof an internal model.49

As already noted, Solvency II rests upon threepillars (quantitative requirements ―gover-nance requirements― public disclosure ofinformation and supervisory reporting).Although the rationale behind Solvency II pil-lars is considerably different from the respec-tive arrangements of the previous regime,traces of all building blocks of the three pillarscould also be found in the former supervisoryframework, save one. Quantitative or gover-nance requirements as well as reporting forsupervisory purposes have always been present,albeit simpler and undeniably different. Nev-ertheless, there existed some. What was alto-gether lacking under the previous regime,given that it had not been envisaged at all, wasthe public disclosure by undertakings them-selves of detailed information on their solvencyand financial condition. This enhances trans-parency vis-à-vis policyholders and markets,i.e. any eventual future investors wishing toassume the insurance portfolio.

Of course, at the end of the day, the policy-holder cannot help but wonder (and this iswhat he is ultimately interested in) whetherthe application of a new supervisory frame-work, namely Solvency II, makes the provisionof benefits from the insurance undertakingmore certain. Even though time is the ulti-mate judge, it is a fact that (re)insuranceundertakings as well as supervisory authoritiesboast a wider array of reliable tools to bettercalculate and monitor the solvency of(re)insurance undertakings, which in turnleads to timely decision-making, while on theirpart policyholders are much better informedabout their insurer of choice, thanks to pub-lic disclosure.


49 As opposed to the former regime, under which solvency wascalculated as a fixed percentage of the technical provisions and/orinsurance premiums.

Relationships between two or more undertakings Two undertakings ―(re)insurance or not―may be linked to each other in the followingways:

a) Participating-related relationship:

This relationship can be described in two lev-els, depending on the degree of control orinfluence of the participating undertaking overthe related undertaking:

― At the minimum level of control, the firstundertaking (i.e. the participating undertak-ing) has a qualifying holding in the secondundertaking (the related undertaking),whereby a qualifying holding is defined as adirect or indirect holding of 10% or more ofthe shares or voting rights of the second under-taking, or any other means of providing sig-nificant influence over the management of thatundertaking.

― The next level of control is called partici-pation and means the direct or indirect,through a subsidiary undertaking, holding of atleast 20% of the shares or voting rights of thesecond undertaking.

(b) Parent-subsidiary relationship:

In this relationship, which is the highest levelof control, as parent undertaking is definedthat undertaking which meets one of the fol-lowing conditions:

― it is a parent undertaking within the mean-ing of Article 32(2) of Law 4308/2014, or

― it is a parent undertaking within the mean-ing of Article 1 of Directive 83/349/EEC, pro-vided that it has its head office in another EUMember State,

while a subsidiary undertaking is defined as anundertaking that fulfils one of the above con-ditions in the opposite sense, i.e. as a sub-sidiary.

The aforementioned conditions can besummed up to the fact that the first under-taking (the parent undertaking) holds themajority of the voting rights of the share-holders of the second undertaking (its sub-sidiary) or exercises a dominant influence overthe latter, having for instance the right toappoint or remove the majority of its man-agement board members.

Any subsidiary undertaking of a subsidiaryundertaking is also considered as a subsidiaryof the parent undertaking which is at the headof those undertakings.

It goes without saying that, if two undertak-ings are linked to each other by a parent-sub-sidiary relationship, then they are also relatedwith one another by a participating-relatedrelationship.

Undertakings in a group There are three kinds of groups which are ofinterest for Solvency II:

a) groups with undertakings which are activein areas of the financial sector other thaninsurance;

b) insurance groups; and

c) groups which are active in different businesssectors, but at the same time pursue insuranceactivities.

For understanding the first kind of groups, theconcept of the financial conglomerate is of rel-evance. This concept is defined in Law3455/2006 and refers to any group of under-takings, in which at least one of the entities iswithin the insurance sector and at least one iswithin the banking or investment services sec-tor. The undertaking which is at the head ofthe group may be either a regulated entity (i.e.is within the insurance or banking or invest-ment services sectors) or an unregulated entity.If the undertaking at the head of the group isan unregulated entity, it is called a mixed finan-cial holding company. If the undertaking at the

44Economic Bulletin

December 2016 89

ANNEXUNDER T AK I NG S I N A G ROUP

head of the group is a regulated entity, it isnamed after the business area in which it isactive, i.e. bank, insurance or reinsuranceundertaking or investment firm or asset man-agement company or alternative investmentfund manager.

Insurance groups are groups of undertakingswhich are mainly active in insurance and/orreinsurance. The undertaking at the head ofan insurance group may be either an insuranceor reinsurance undertaking itself or an under-taking, the main business of which is toacquire and hold participations in subsidiaryinsurance or reinsurance undertakings, beingneither an insurance or a reinsurance under-taking itself nor a mixed financial holdingcompany. In the last case, the undertaking atthe head of the group is called an insuranceholding company.

Finally, there are groups which are active indifferent business sectors other than the finan-cial sector, including at least one insurance orreinsurance undertaking among their sub-

sidiary undertakings. The undertaking at thehead of such groups, which is obviously neithera mixed financial holding company nor aninsurance holding company, is called a mixed-activity insurance holding company.

In the light of the above, it becomes clear thatone of the following undertakings can be at thehead of a group:

a) an insurance or reinsurance undertakingwhich has its head office within the EU;

b) an insurance or reinsurance undertakingwhich has its head office in a third (non-EU)country;

c) a mixed financial holding company;

d) an insurance holding company; or

e) a mixed-activity insurance holding company.

For each of the above cases, the form of groupsupervision exercised varies.



209. On the optimality of bank competitionpolicy Ioannis G. Samantas

210. The re-pricing of sovereign risks fol-lowing the Global Financial CrisisDimitris Malliaropulos and Petros M.Migiakis

211. Moral hazard and strategic default: evi-dence from Greek corporate loansIoannis Asimakopoulos, Panagiotis K.Avramidis, Dimitris Malliaropulos andNickolaos G. Travlos

212. New perspectives on the Great Depres-sion: a review essayGeorge S. Tavlas

213. Inter-industry wage differentials inGreece: rent-sharing and unobservedheterogeneity hypothesesEvangelia Papapetrou and PinelopiTsalaporta

214. Self-fulfilling dynamics: the interactionsof sovereign spreads, sovereign ratingsand bank ratings during the euro finan-cial crisisHeather D. Gibson, Stephen G. Halland George S. Tavlas

215. The slowdown in US productivitygrowth – what explains it and will it per-sist?Ursel Baumann and Melina Vasardani

216. Alternative Bayesian compression inVector Autoregressions and relatedmodelsMike G. Tsionas

217. Alternatives to large VAR, VARMAand multivariate stochastic volatilitymodels Mike G. Tsionas

218. Insights on the Greek economy fromthe 3D macro modelHiona Balfoussia and Dimitris Papa-georgiou

219. Non-performing loans in the euro area:are core-periphery banking marketsfragmented?Dimitrios Anastasiou, Helen Louri andMike G. Tsionas

This section contains the abstracts of Working Papers authored by Bank of Greece staff and/orexternal authors and published by the Bank of Greece. The unabridged version of these texts isavailable on the Bank of Greece’s website (www.bankofgreece.gr).

CONTENTS

WORK I NG P A P ER S(JULY – DECEMBER 2016)

This study examines whether the effect of mar-ket structure on financial stability is persistent,subject to current regulation and supervisionpolicies. Extreme Bounds Analysis (EBA) isemployed over a sample of 2,450 banks oper-ating within the EU-27 during the period2003-2010. The results show an inverse U-shaped association between market power and

soundness, as well as a stabilising tendency inmarkets of less concentration, where policieslean towards limited restrictions on non-inter-est income, official intervention in bank man-agement and book transparency. Regulationsignificantly contributes as a stability channelthrough which bank competition policy is opti-mally designed.

The re-pricing of sovereign risks following the Global Financial Crisis

Working Paper No. 210Dimitris Malliaropulos and Petros M. Migiakis

How strong has been the effect of the GlobalFinancial Crisis (GFC) on systemic risk in sov-ereign bond markets? Was the increase incredit spreads relative to triple-A benchmarkswhich followed the GFC the result of highersovereign credit risk or the result of a re-pric-ing that reflected changes in broader marketconditions and risk aversion? This paper exam-ines these issues by specifying a sovereigncredit yield curve which relates sovereign yieldspreads to credit ratings and global variables.The model allows for time-variation in both theprice of credit risk and the average spreadacross all rating categories, which proxies theeffect of global risk factors on yield spreads.Daily data of ten-year bond yields and ratings

from a large database of 64 countries are used,covering both emerging markets and developedeconomies, for the period from 1.1.2000 to1.1.2015. Estimates suggest that sovereign riskpremia increased significantly after the GFC,with most of the increase due to a re-pricing ofbroader market risks rather than an increase inthe quantity or price of sovereign risk per se.This increase in global risk could be the resultof a flight-to-quality from lower-rated sover-eign bonds to AAA benchmark bonds. Inter-estingly, it is found that global risk in the sov-ereign bond market is driven by global vari-ables that relate to investor confidence, volatil-ity risk, central bank liquidity and the slope ofthe yield curve in the United States.

Moral hazard and strategic default: evidence from Greek corporate loans

Working Paper No. 211Ioannis Asimakopoulos, Panagiotis K. Avramidis, Dimitris Malliaropulos and Nickolaos G. Travlos

Using a unique dataset of corporate loans of13,070 Greek firms for the period 2008-2015and an identification strategy based on theinternal credit ratings of banks, it is shown thatone out of six firms with non-performing loans

are strategic defaulters. Furthermore, potentialdeterminants of firms’ behaviour were investi-gated by relating the probability of strategicdefault to a number of firm characteristics suchas size, age, liquidity, profitability and collateral


On the optimality of bank competition policy

Working Paper Νο. 209Ioannis G. Samantas

92

value. The results provide evidence of a posi-tive relationship of strategic default with out-standing debt and economic uncertainty and anegative relationship with the value of collat-eral. Also, profitability and collateral can beused to distinguish the strategic defaulters fromthe financially distressed defaulters. Finally, it

is found that the relationship of strategicdefault risk with firm size and age has aninverse U-shape, i.e. strategic default is morelikely among medium-sized firms compared tosmall and large firms and it is also more likelyamong middle-aged firms compared to new-founded and established firms.


New perspectives on the Great Depression: a review essay

Working Paper No. 212George S. Tavlas

Inter-industry wage differentials in Greece: rent-sharing and unobserved heterogeneity hypotheses

Working Paper No. 213Evangelia Papapetrou and Pinelopi Tsalaporta

The Great Depression of 1929 was the mostdevastating and destructive economic eventto afflict the global economy since the begin-ning of the twentieth century. What, then,were the origins of the Great Depression andwhat have we learned about the appropriatepolicy responses to economic depressionsfrom that episode? This essay reviews tworecently published books on the GreatDepression. Eric Rauchway’s The MoneyMakers: How Roosevelt and Keynes Endedthe Depression, Defeated Fascism, andSecured a Prosperous Peace (Basic Books,2015) tells the story of the ways Franklin D.

Roosevelt drew on the ideas of John MaynardKeynes to place monetary policy front-and-center to underpin the recovery from theGreat Depression and to underwrite the blue-print of the Bretton Woods System. BarryEichengreen’s Hall of Mirrors: The GreatDepression, the Great Recession, and theUses —and Misuses— of History (OxfordUniversity Press, 2015) shows the way the les-sons learned from analysis of the GreatDepression helped shape policy makers’response to the 2007-08 financial crisis, thushelping to avoid many of the mistakes madeby policy makers in the 1930s.

This paper examines the structure and deter-minants of inter-industry wage differentials inGreece, along with the role of the rent-shar-ing and unobserved heterogeneity hypotheses,employing restricted least squares and quan-tile regression techniques with cluster robuststandard errors at the firm level. To this end,a unique dataset, the European Union Struc-ture of Earnings Survey (SES), is utilised.Data refer to 2010 when the first elements ofthe economic adjustment programme to dealwith the chronic deficiencies of the Greekeconomy, restore sustainable public financesand competitiveness, and set the foundation

for long-term growth were beginning to beimplemented. Results point to high wage dis-persion across industries at the mean of theconditional wage distribution, even after con-trolling for personal and workplace charac-teristics. However, evidence for the unob-served heterogeneity hypothesis is ratherscant. Therefore, there is room for efficiencywage or rent-sharing theories in accounting fora large part of inter-industry wage differen-tials, tentatively implying that firm hetero-geneity in the ability-to-pay matters more thanemployee unobservable attributes in the wagedetermination process.

The slowdown in US productivity growth – what explains it and will it persist?

Working Paper No. 215Ursel Baumann and Melina Vasardani

Self-fulfilling dynamics: the interactions of sovereign spreads, sovereign ratings and bank ratings during the euro financial crisis

Working Paper No. 214Heather D. Gibson, Stephen G. Hall and George S. Tavlas

During the euro area financial crisis, interac-tions among sovereign spreads, sovereigncredit ratings, and bank credit ratingsappeared to have been characterised by self-generating feedback loops. To investigate theexistence of feedback loops, we consider apanel of five euro area stressed countrieswithin a three-equation simultaneous systemin which sovereign spreads, sovereign ratingsand bank ratings are endogenous. We estimate

the system using two approaches. First, weapply GMM estimation, which allows us to cal-culate persistence and multiplier effects. Sec-ond, we apply a new, system time-varying-parameter technique that provides bias-freeestimates. Our results show that sovereign rat-ings, sovereign spreads, and bank ratingsstrongly interacted with each other during theeuro crisis, confirming strong doom-loopeffects.


US productivity growth has been surprisinglyweak after the Great Recession, raising con-cerns among economists and policy makers.Shedding light on this weakness is crucial,given the importance of productivity in thelong-term growth prospects of the economy.

The literature relates the recent subdued paceof productivity growth to a number of factors,namely: (a) the fading effects of the informationand communication technology (ICT) revolu-tion, with the productivity slowdown merely sug-gesting a return to more “normal” rates fol-lowing the exceptional ICT boom in the mid-1990s; (b) capital mismeasurement, as the gainsfrom IT innovation for consumers’ prosperityare largely intangible; (c) a decline in businessdynamism and labour market fluidity, whichhinders the optimal allocation of resources aswell as technological diffusion or spilloversbetween firms and sectors; (d) a misallocationof resources due to the credit boom that pre-ceded the Great Recession, with a possible last-ing drag on productivity growth even after thedownturn; and (e) tighter credit standards dur-ing the crisis and in the first years of the recov-

ery, which restricted firms’ access to credit,thereby discouraging investment in innovationand the entry of new businesses into the market.

This paper contributes to the debate by empir-ically revising the main determinants of labourproductivity growth over the period 1999-2013for a panel of US states and explores howchanges in these determinants have affectedthe dynamics of labour productivity in recentyears. The results show that US productivitygrowth is determined by changes in capitaldeepening, the availability of credit and theextent to which credit growth is excessive, aswell as by the dynamism of the economy asmeasured by labour market churn and the firmentry rate. Besides, given the sectoral compo-sition of growth, it is important to control forthe impact of high-productivity sectors such asIT-producing and mining sectors.

The key findings of this paper suggest that morethan half of the slowdown in US productivitygrowth in the period 2011-2013 relative to itssample average is due to a decline in the rate ofcapital deepening. The other major factor

Alternative Bayesian compression in Vector Autoregressions and related models

Working Paper No. 216Mike G. Tsionas


explaining the recent weakness in productivitygrowth – more closely related to TFP – is theslowdown in business dynamism experienced bythe US economy. By contrast, while financialfactors had weighed on productivity growth dur-ing the Great Recession (i.e. during the 2008-2010 period), they now appear to have becomesupportive. Overall, part of the productivityslowdown appears to be cyclical, particularly inrelation to capital investment, but some otheraspects related to business dynamism, couldprove more persistent. Understanding the driv-ers and the evolution of business dynamism isthus crucial and a priority for future research.

The results of the study lead to a number ofpolicy implications. The authors suggest that

policies that support productivity growthshould focus primarily on measures facilitat-ing the diffusion of technology and enhancingthe effectiveness of resource allocation. Suchpolicies could include the promotion of cap-ital investment in physical and human capital,targeted fiscal measures and further tradeintegration, aimed at accelerating the devel-opment and wide adoption of commercial usesof new technologies. Moreover, measures thatmitigate uncertainty, address regulatory andcompetitive hurdles, and ensure the uninter-rupted provision of credit to small youngfirms/startups especially during downturns areestimated to re-invigorate business dynamism,ultimately yielding significant gains in TFPgrowth.

In this paper we reconsider large BayesianVector Autoregressions (BVAR) from thepoint of view of Bayesian CompressedRegression (BCR). First, we show that thereare substantial gains in terms of out-of-sam-ple forecasting by treating the problem as anerror-in-variables formulation and estimatingthe compression matrix instead of using ran-dom draws. As computations can be effi-

ciently organised around a standard Gibbssampler, timings and computational com-plexity are not affected severely. Second, weextend the Multivariate Autoregressive Indexmodel to the BCR context and show that wehave, again, gains in terms of out-of-sampleforecasting. The new techniques are used inUS data featuring medium-sized, large andhuge BVARs.

Alternatives to large VAR, VARMA and multivariate stochastic volatility models

Working Paper No. 217Mike G. Tsionas

In this paper, our proposal is to combine uni-variate ARMA models to produce a variant ofthe VARMA model that is much more easilyimplementable and does not involve certaincomplications. The original model is reducedto a series of univariate problems, and a cop-ula-like term (mixture-of-normals densities) isintroduced to handle dependence. Since theunivariate problems are easy to handle by

MCMC or other techniques, computations canbe parallelised easily, and only univariate dis-tribution functions are needed, which are quiteoften available in closed form. The results fromparallel MCMC or other posterior simulatorscan then be taken together and use simple sam-pling-resampling to obtain a draw from theexact posterior which includes the copula-liketerm. We avoid optimisation of the parameters

entering the copula mixture form, as its param-eters are optimised only once before MCMCbegins. We apply the new techniques in threetypes of challenging problems: large time-vary-ing parameter vector autoregressions (TVP-VAR) with nearly 100 macroeconomic vari-ables, multivariate ARMA models with 25

macroeconomic variables and multivariate sto-chastic volatility models with 100 stock returns.Finally, we perform impulse response analysisin the data of Giannone, Lenza, and Primiceri(2015) and compare, as they proposed withresults from a dynamic stochastic general equi-librium model.


Insights on the Greek economy from the 3D macro model

Working Paper No. 218Hiona Balfoussia and Dimitris Papageorgiou

Non-performing loans in the euro area: are core-periphery banking markets fragmented?

Working Paper No. 219Dimitrios Anastasiou, Helen Louri and Mike G. Tsionas

This paper examines the macroeconomic andwelfare implications of banking capitalrequirement policies and their interactionswith real and financial shocks for the Greekeconomy. The model employed is that ofClerc et al. (2015), a DSGE model featuringa detailed financial sector, banking capitalregulations and bank default in equilibrium.The key model implication is that capitalrequirements reduce bank leverage and thedefault risk of banks but their relationshipwith social welfare is hump-shaped, reflectinga trade-off. The model is calibrated to data onthe Greek economy and the dynamicresponses to a number of financial and realshocks which may have played a material role

in the unfolding of the Greek crisis areexplored. The results indicate inter alia thatan increase in depositors’ cost of bank defaultleads to a substantial increase in the depositrate, a decline in deposits and bank equity andan increase in bank fragility, while on the realside of the economy the decline in total creditprompts a deterioration of key macro vari-ables. Additionally, the results imply thatwhile recapitalisations increase bank networth and credit supply and boost economicactivity, this potential benefit is severely com-promised in a high financial distress scenario,as the positive real and financial implicationsof a recapitalisation become both smaller andmore short-lived.

The objective of this study is to examine thecauses of non-performing loans (NPLs) in thebanking system of the euro area for the period2003-2013 and distinguish between core andperiphery country determinants. The increasein NPLs post crisis has put into question therobustness of many European banks and thestability of the whole sector. It still remains aserious challenge, especially in peripheralcountries which are hardest hit by the finan-cial crisis. By employing both Fully ModifiedOLS and Panel Cointegrated VAR, we esti-

mate that NPLs are affected by the samemacroeconomic and bank-specific conditionsbut the responses are stronger in the periph-ery. Following the FMOLS estimations, NPLsin the euro area have performed an upward(much higher in the periphery) shift after2008 and are mostly related to worseningmacroeconomic conditions especially withrespect to unemployment, growth and taxes.Fiscal consolidation and interest rate marginsare significant for the periphery, while creditto GDP is significant only for the core. Qual-

ity of management and loans to deposits playan important role, while size is negatively sig-nificant only in the periphery. Most of thesefindings were confirmed by the panel Coin-tegrated VAR results. A chi-square test com-paring the estimated coefficients for the coreand periphery NPLs rejects the hypothesis of

equality, revealing another aspect of bankingfragmentation in the euro area. Such findingscan be helpful when designing macro-pru-dential as well as NPL resolution policies,which should be adjusted appropriately to thedifferent responses between core and periph-ery banks.


44Economic Bulletin

December 2016 99

Boutos Yannis, “Economic stabilisation and growth prospects”, No. 3, February 1994.Papademos Lucas, “Growth with stability: the role of monetary policy”, No. 5, March 1995.Voridis Hercules, “The special nature of banks and the transmission mechanism of monetary

policy: A review of recent literature”, No. 5, March 1995.Filippides Anastasios, Kyriakopoulos Panayotis and Moschos Dimitrios, “Bank of Greece mon-

etary policy instruments”, No. 6, November 1995.Haralabides Michael, Hardouvelis Gikas and Papageorgiou George, “Changeover to the single

currency: Prospects and challenges for credit institutions in Greece”, No. 6, November 1995.Karabalis Nikos, “Harmonisation of Consumer Price Indices in EU countries”, No. 7, March 1996.Public Sector Accounts Department, Studies, Planning and Operations Development Office,

“Government Securities Markets”, No. 7, March 1996.Saccomanni Fabrizio, “Opportunities and challenges in the process of European monetary inte-

gration”, No. 7, March 1996.Papademos Lucas, “Challenges facing monetary policy on the road to EMU”, No. 8, November

1996.Information Systems and Organisation Department, “Developments in EU Payment Systems:

TARGET and the Hellenic system Hermes”, No. 8, November 1996.Brissimis Sophocles and Gibson Heather, “Monetary policy, capital flows and Greek disinflation”,

No. 9, March 1997.Sabethai Isaac, “From contractual earnings to labour costs: incomes policy, collective bargain-

ing and inflation (1991-1996)”, No. 9, March 1997.Hall S.G. and Zonzilos Nicholas, “The output gap and inflation in Greece”, No. 9, March 1997.Pantelidis Evangelos, “The methodology of the new balance of payments compilation system of

the Bank of Greece”, No. 9, March 1997.Papademos Lucas, “The globalisation of financial markets and the conduct of economic and mon-

etary policy”, No. 10, December 1997.Brissimis Sophocles and Kastrisianakis Efstratios, “Is there a credit channel in the Greek econ-

omy?”, No. 10, December 1997.Gibson Heather, “The relationship between the financial system and the real economy”, No. 10,

December 1997.Hondroyiannis George and Papapetrou Evangelia, “The causal relationship between consumer

and wholesale prices in Greece”, No. 10, December 1997.Hardy Daniel and Simigiannis George, “Competition and efficiency of the Greek banking sys-

tem”, No. 11, June 1998.Brissimis Sophocles and Gibson Heather, “What can the yield curve tell us about inflation?”,

No. 11, June 1998.Pantazidis Stelios, “Inflation, investment and economic growth in Greece”, No. 11, June 1998.Mitrakos Theodoros, “The contribution of income sources to overall inequality”, No. 11, June 1998.Garganas Nicholas, “Greece and EMU: prospects and challenges”, No. 12, December 1998.Brissimis Sophocles, Sideris Dimitris and Voumvaki Fragiska, “Purchasing power parity as a long-

run relationship: an empirical investigation of the Greek case”, No. 12, December 1998.Manassaki Anna, “European Union transfers to Greece: historical background and prospects”,

No. 12, December 1998.Lazaretou Sophia, “The drachma on the metallic monetary standards: lessons from the past”,

No. 13, July 1999.Hondroyiannis George, “The causality between government spending and government revenue

in Greece”, No. 13, July 1999.Mitrakos Theodoros and Tsakloglou Panos, “The distributional impact of excise duties”, No. 13,

July 1999.

AR T I C L E S P UB L I S H ED I N P R E V I OU S I S S U E S O FTH E E CONOM I C BU L L E T I N


Dellas Harris, “ECB monetary policy strategy and macroeconomic volatility in Greece”, No. 13,July 1999.

Papazoglou Christos, “The real exchange rate and economic activity: is the hard-drachma pol-icy necessarily contractionary?”, No. 14, December 1999.

Gibson Heather and Lazaretou Sophia, “Leading inflation indicators for Greece: an evaluationof their predictive power”, No. 14, December 1999.

Karapappas Andreas and Milionis Alexandros, “Estimation and analysis of external debt in theprivate sector”, No. 14, December 1999.

Papademos Lucas, “From the drachma to the euro”, No. 15, July 2000.Zonzilos Nicholas, “The Greek Phillips curve and alternative measures of the NAIRU”, No. 15,

July 2000.Pantazidis Stelios, “The sustainability of the current account deficit”, No. 15, July 2000.Hondroyiannis George, “Investigating causality between prices and wages in Greece”, No. 15, July 2000.Sabethai Isaac, “The Greek labour market: features, problems and policies (with emphasis on

labour market flexibility and on combatting unemployment in the context of non-inflationarygrowth)”, No. 16, December 2000.

Brissimi Dimitra and Brissimis Sophocles, “The problem of unemployment in the EuropeanUnion”, No. 16, December 2000.

Bardakas Joanna, “Financial deregulation, liquidity constraints and private consumption inGreece”, No. 16, December 2000.

Lazaretou Sophia and Brissimis Sophocles, “Fiscal rules and stabilisation policy in the euro area”,No. 17, July 2001.

Lolos Sarantis-Evangelos, “The role of European structural resources in the development of theGreek economy”, No. 17, July 2001.

Papazoglou Christos, “Regional economic integration and inflows of foreign direct investment:the European experience”, No. 17, July 2001.

Garganas Nicholas, “The future of Europe and EMU”, No. 18, December 2001.Brissimis Sophocles and Papadopoulou Dafni-Marina, “The physical introduction of euro ban-

knotes and coins in Greece”, No. 18, December 2001.Zonzilos Nicholas, “The monetary transmission mechanism – the ECB experiment: results for

Greece”, No. 18, December 2001.Brissimis Sophocles, Kamberoglou Nicos and Simigiannis George, “Bank-specific characteris-

tics and their relevance to monetary policy transmission”, No. 18, December 2001.Tsaveas Nicholas, “Exchange rate regimes and exchange rate policy in South Eastern Europe”,

No. 18, December 2001.Demenagas Nicholas and Gibson Heather, “Competition in Greek banking: an empirical study

for the period 1993-1999”, No. 19, July 2002.Hondroyiannis George, “Economic and demographic determinants of private savings in Greece”,

No. 19, July 2002.Gatzonas Efthymios and Nonika Kalliopi, “Eligible assets and management of collateral in the

context of central bank monetary policy operations”, No. 19, July 2002.Voridis Hercules, Angelopoulou Eleni and Skotida Ifigeneia, “Monetary policy in Greece 1990-

2000 through the publications of the Bank of Greece”, No. 20, January 2003.Kaplanoglou Georgia and Newbery David, “The distributional impact of indirect taxation in

Greece”, No. 21, July 2003.Papapetrou Evangelia, “Wage differentials between the public and the private sector in Greece”,

No. 21, July 2003.Athanasoglou Panayiotis and Brissimis Sophocles, “The effect of mergers and acquisitions on

bank efficiency in Greece”, No. 22, January 2004.

Lazaretou Sophia, “Monetary system and macroeconomic policy in Greece: 1833-2003”, No. 22,January 2004.

Karabalis Nikos and Kondelis Euripides, “Alternative measures of inflation”, No. 22, January 2004.Papaspyrou Theodoros, “EMU strategies for new Member States: the role of Exchange Rate

Mechanism II”, No. 23, July 2004.Mitrakos Theodoros, “Education and economic inequalities”, No. 23, July 2004.Papapetrou Evangelia, “Gender wage differentials in Greece”, No. 23, July 2004.Gibson Heather, “Greek banking profitability: recent developments”, No. 24, January 2005.Athanasoglou Panayiotis, Asimakopoulos Ioannis and Georgiou Evangelia, “The effect of merger

and acquisition announcement on Greek bank stock returns”, No. 24, January 2005.Mitrakos Theodoros and Zografakis Stavros, “The redistributional impact of inflation in Greece”,

No. 24, January 2005.Theodossiou Ioannis and Pouliakas Konstantinos, “Socio-economic differences in the job sat-

isfaction of high-paid and low-paid workers in Greece”, No. 24, January 2005.Garganas Nicholas, “Adjusting to the single monetary policy of the ECB”, No. 25, August 2005.Mitrakos Theodoros, Simigiannis Georgios and Tzamourani Panagiota, “Indebtedness of Greek

households: evidence from a survey”, No. 25, August 2005.Nicolitsas Daphne, “Per capita income, productivity and labour market participation: recent devel-

opments in Greece”, No. 25, August 2005.Nicolitsas Daphne, “Female labour force participation in Greece: developments and determin-

ing factors”, No. 26, January 2006.Mitrakos Theodoros and Zonzilos Nicholas, “The impact of exogenous shocks on the dynamics

and persistence of inflation: a macroeconomic model-based approach for Greece”, No. 26, Jan-uary 2006.

Athanasoglou Panayiotis, Asimakopoulos Ioannis and Siriopoulos Konstantinos, “External financ-ing, growth and capital structure of the firms listed on the Athens Exchange”, No. 26, January2006.

Mitrakos Theodoros and Nicolitsas Daphne, “Long-term unemployment in Greece: developments,incidence and composition”, No. 27, July 2006.

Kalfaoglou Faidon, “Stress testing of the Greek banking system”, No. 27, July 2006.Pantelidis Evangelos and Kouvatseas Georgios, “Frontier survey on travel expenditure: method-

ology, presentation and output assessment (2003-2005)”, No. 27, July 2006.Brissimis Sophocles and Vlassopoulos Thomas, “Determinants of bank interest rates and com-

parisons between Greece and the euro area”, No. 28, February 2007.Simigiannis George and Tzamourani Panagiota, “Borrowing and socio-economic characteristics

of households: results of sample surveys carried out by the Bank of Greece”, No. 28, February2007.

Papapetrou Evangelia, “Education, labour market and wage differentials in Greece”, No. 28, Feb-ruary 2007.

Christodoulakis George, “The evolution of credit risk: phenomena, methods and management”,No. 28, February 2007.

Karabalis Nikos and Kondelis Euripides, “Inflation measurement in Greece”, No. 28, February2007.

Kapopoulos Panayotis and Lazaretou Sophia, “Foreign bank presence: the experience of South-East European countries during the transition process”, No. 29, October 2007.

Nicolitsas Daphne, “Youth participation in the Greek labour market: developments and obsta-cles”, No. 29, October 2007.

Albani Maria, Zonzilos Nicholas and Bragoudakis Zacharias, “An operational framework for theshort-term forecasting of inflation”, No. 29, October 2007.

44Economic Bulletin

December 2016 101

Asimakopoulos Ioannis, Brissimis Sophocles and Delis Manthos, “The efficiency of the Greekbanking system and its determinants”, No. 30, May 2008.

Balfoussia Hiona, “Stock market integration: the Athens Exchange in the European financialmarket”, No. 30, May 2008.

Mitrakos Theodoros, “Child poverty: recent developments and determinants”, No. 30, May 2008.Asimakopoulos Ioannis, Lalountas Dionysis and Siriopoulos Constantinos, “The determinants

of firm survival in the Athens Exchange”, No. 31, November 2008.Petroulas Pavlos, “Foreign direct investment in Greece: Productivity and Technology diffusion”,

No. 31, November 2008.Anastasatos Tasos and Manou Constantina, “Speculative attacks on the drachma and the

changeover to the euro”, No. 31, November 2008.Mitrakos Theodoros and Simigiannis George, “The determinants of Greek household indebt-

edness and financial stress”, No. 32, May 2009.Papazoglou Christos, “Is Greece’s export performance really low?”, No. 32, May 2009.Zonzilos Nicholas, Bragoudakis Zacharias and Pavlou Georgia, “An analysis of the revisions of

first (flash) quarterly national account data releases for Greece”, No. 32, May 2009.Rapanos Vassilis and Kaplanoglou Georgia, “Independent fiscal councils and their possible role

in Greece”, No. 33, May 2010.Mitrakos Theodoros, Tsakloglou Panos and Cholezas Ioannis, “Determinants of youth unem-

ployment in Greece with an emphasis on tertiary education graduates”, No. 33, May 2010.Gibson Heather and Balfoussia Hiona, “Inflation and nominal uncertainty: the case of Greece”,

No. 33, May 2010.Migiakis Petros, “Determinants of the Greek stock-bond correlation”, No. 33, May 2010.Mitrakos Theodoros, Tsakloglou Panos and Cholezas Ioannis, “Determinants of the wage rates

in Greece with an emphasis on the wages of tertiary education graduates”, No. 34, September2010.

Bragoudakis Zacharias and Panagiotou Stelios, “Determinants of the receipts from shippingservices: the case of Greece”, No. 34, September 2010.

Gibson Heather and Balfousia Hiona, “The impact of nominal and real uncertainty onmacroeconomic aggregates in Greece”, No. 34, September 2010.

Manesiotis Vasilios, “Numerical fiscal rules in practice”, No. 35, June 2011.Vasardani Melina, “Tax evasion in Greece: an overview”, No. 35, June 2011.Kalfaoglou Faidon, “The usefulness of stress testing exercises for assessing the soundness of the

banking system”, No. 35, June 2011.Nicolitsas Daphne, “On-the-job training in Greece: a brief overview”, No. 35, June 2011.Papaspyrou Theodoros, “EMU sustainability and the prospects of peripheral economies”, No.

36, April 2012.Kanellopoulos Costas Ν., “Employment and worker flows during the financial crisis”, No. 36,

April 2012.Kalfaoglou Faidon, “Bank capital adequacy framework”, No. 36, April 2012.Bragoudakis Zacharias and Sideris Dimitris, “Do retail gasoline prices adjust symmetrically to

crude oil price changes? The case of the Greek oil market”, No. 37, December 2012.Kanellopoulos Costas Ν., “The size and structure of uninsured employment”, No. 37, Decem-

ber 2012.Migiakis Petros M., “Reviewing the proposals for common bond issuances by the euro area sov-

ereign states under a long-term perspective”, No. 37, December 2012.Karabalis Nikos and Kondelis Euripides, “Indirect tax increases and their impact on inflation

over 2010-2012”, No. 38, November 2013.Vourvachaki Evangelia, “Structural reform of the road haulage sector”, No. 38, November 2013.


Kanellopoulos Nikolaos C., Mitrakos Theodoros, Tsakloglou Panos and Cholezas Ioannis, “Theimpact of the current crisis on private returns to education in Greece”, No. 38, November 2013.

Papapetrou Evangelia and Bakas Dimitrios, “The Greek labour market during the crisis: Unem-ployment, employment and labour force participation”, No. 38, November 2013.

Tzamourani Panagiota, “The Household Finance Survey: Description of the 2009 survey and mainresults on households’ income, wealth and debt in Greece”, No. 38, November 2013.

Dimitropoulou Dimitra, Koutsomanoli-Filippaki Anastasia, Charalambakis Evangelos and Agge-lis Georgios, “The financing of Greek enterprises before and during the crisis”, No. 39, July2014.

Kanellopoulos Costas, “The Greek private sector labour market during the crisis”, No. 39, July2014.

Kalfaoglou Faidon, “European Banking Union: “Europeanising” banks’ financial safety net”, No.39, July 2014.

Lazaretou Sofia, “The smart economy: cultural and creative industries in Greece. Can they bea way out of the crisis?”, No. 39, July 2014.

Chaireti Aikaterini, “The housing situation in Greece during the crisis years 2008-2012”, No. 39,July 2014.

Tagkalakis Athanasios, “Tax buoyancy in Greece”, No. 40, December 2014. Bragoudakis G. Zaharias, “An empirical investigation of the relationship between unit labour

cost and price developments: The case of Greece”, No. 40, December 2014. Migiakis Petros M., “The international financial markets as a source of funding for Greek non-

financial corporations”, No. 40, December 2014. Kalyvitis Sarantis, “Estimating the quality of Greek exports: A non-price based approach”, No.

40, December 2014. Kanellopoulos Costas, “The effects of minimum wages on wages and employment”, No. 41, July

2015. Georgiou Evangelia, “The use of cash and electronic payment instruments in the economy”, No.

41, July 2015. Kalfaoglou Faidon, “Alternative NPLs resolution regimes. A case studies approach”, No. 41, July

2015. Lazaretou Sophia, “How can the use of historical macroeconomic data series support economic

analysis and policy making? The new macro history database 1833-1949 as part of the large data-base of Southeast European Countries”, No. 41, July 2015.

Kasimati Evangelia and Sideris Dimitrios, “Towards a new growth model for tourism: structuralreforms and the tourism product in Greece during the crisis (2008-2014)”, No. 42, December2015.

Tagkalakis Athanasios, “The determinants of unemployment dynamics in Greece”, No. 42,December 2015.

Chaireti Aikaterini, “The Greek primary sector from 1995 to 2014”, No. 42, December 2015.Vettas Nikos, Giotopoulos Ioannis, Valavanioti Evangelia and Danchev Svetoslav, “The deter-

minants of new firms’ export performance”, No. 43, July 2016.Belli Styliani and Backinezos Constantina, “The transition to the new methodology for the com-

pilation of balance of payments statistics – BPM6”, No. 43, July 2016.Lazaretou Sophia, “The Greek brain drain: the new pattern of Greek emigration during the recent

crisis”, No. 43, July 2016.Kalfaoglou Faidon, “Bank recapitalisation: a necessary but not sufficient condition for resum-

ing lending”, No. 43, July 2016.

44Economic Bulletin

December 2016 103

BA

NK

OF

GR

EEC

E

EUR

OSY

STE

M

ISSN: 1105 - 9729

EC

ON

OM

IC B

UL

LE

TIN

No

44

DE

CE

MB

ER

20

16

economic bulletin 44 - bankofgreece.gr · contents the economic behaviour of households in greece:...

Documents