zsdvzsdij

Global Finance Journal 25 (2014) 1–16

Contents lists available at ScienceDirect

Global Finance Journal

jou rnal h omep age: w ww . elsevier . com/l ocate/gfj

Quantitative easing in an open economy—Not a liquidity but a reserve trap

Anthony F. Herbst a, , Joseph S.K. Wu b, Chi Pui Ho b

aThe University of Texas at El Paso, El Paso, TX, United States

b Faculty of Business and Economics, The University of Hong Kong, Hong Kong

a r t i c l e i n f o a b s t r a c t

Available online 25 March 2014

JEL classification:E58F41

Keywords:Quantitative easingCredit expansionMacroeconomicsExport-led

Expansionary monetary policy is ineffective in a liquidity trap. In another case, which we call a “reserve trap,” money supply increase is trapped in bank reserves; there is no credit expansion through the banking system. In such case, quantitative easing (QE) will not boost credit to the real sector and revitalize the economy. To analyze a reserve trap, we modify the open economy model to include multiple interest rates. Trade is included since exports can be financed externally even during domestic credit constriction. We show the conditions under which QE can lead to currency depreciation and trigger an export-led recovery.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

To prevent deflation caused by a financial crisis or to revive economic growth,1 central banks can use interest rate policy to promote lending by cutting rates to raise the level of economic activity. However, sometimes such policy may not be sufficiently effective even

when rates are cut to near zero.2 Then central banks can also pump money directly into the economy to try to increase circulating money supply. Today this is known as quantitative easing (QE). (It may also recall the term “monetizing the debt” that is associated with inflationary periods historically). However, if QE results only in boosting banks reserve balances, and not in expanding bank credit to the real sector, the consequences will be very different from those of conventional credit expansion policy. This in fact is what happened

in 20083 after the Global Financial

Corresponding author at: Herbst 625 La Chapa Ave. El Paso, TX 79912-2321 USA. Tel.: +1 915 833 6129. E-mail address: [email protected] (A.F. Herbst).

1 For example, after the 2008 financial crisis, the consumer price index in the U.S. dropped drastically (see Appendix A Chart 1); and real GDP declined (see Appendix A Chart 2). (Sources: Federal Reserve Bank of St. Louis, 2014).

2 Liquidity trap can occur when interest rate is near zero.

3 For instance, Appendix A Charts 3, 4 and 5 show after the 2008 financial crisis, U.S. money stock M1 increased steadily while banks reserve balances rose sharply and bank loans declined. (Sources: Federal Reserve Bank of St. Louis, 2014)

http://dx.doi.org/10.1016/j.gfj.2014.03.004 1044-0283/© 2014 Elsevier Inc. All rights reserved.

2 A.F. Herbst et al. / Global Finance Journal 25 (2014) 1–16

Crisis. We call this type of situation a “reserve trap” because the increase in money supply is trapped in banks reserves; there is no credit expansion through the fractional reserve banking process. A reserve trap can occur by central bank design, by bank unwillingness to lend, or by unintended consequences of policies. Since this kind of phenomenon does not involve technological/productivity shocks or credit expansion to the real sector, real business cycle (RBC) analysis may not be an appropriate tool. Furthermore, a reserve trap will cause tiered financial markets so that the traditional open macro model with a single composite interest rate will not be adequate. To remedy this, we introduce a modi fied

open economy macro model with multiple interest rates, which we will call the Requil − Mequil model. This is to distinguish it from the traditional IS-LM model with a single interest rate. Our analysis will reveal the importance of the oft-neglected trade factor in any economic recovery associated with QE. We will show that if QE induces a divergence of various interest rates, domestic currency can depreciate and thus trigger an economic recovery through export growth. In fact, given the possible benefit of such depreciation on competitiveness of domestic industries and export growth, governments might not hesitate to engage in QE to indirectly manage their balance of trade.

Prof. R. Werner introduced the expression “quantity easing” in his article in the Japanese newspaper Nikkei on 2nd

September 1995.4 He was discussing the Bank of Japan (BOJ) policy, and he meant QE to be a policy of “expansion in broad credit creation” and not just increasing banks reserve accounts as practiced by BOJ.

After the 2008 financial crisis, various central banks combined lowering of target reserve rates5 with QE policy. QE was mostly in the form of injection of reserves to banks which involved central banks purchasing risky assets such as mortgage-backed securities (MBS) and other debts from the banks. In the case of the European Central Bank (ECB), this was in the form of expanding the types of collateral that banks can use for drawing reserves. The banking sector is

thus strengthened by having sufficient reserves to shore up its capital base and by unloading risks.6 All these actions are basically asset swaps between the central banks and their banking sectors. While this has the beneficial effect of stabilizing the banking sector, it does not promote investment necessary for economic growth. In fact, banks’ knee-jerk reaction to financial crisis is to shift toward risk aversion and to restrict lending. Some of the monetary policies adopted

after the crisis may in fact amplify the unwillingness of banks to lend to businesses.7

2. Literature review

Bernanke (1983) found that financial disruptions of the 1930–1933 depression years resulted in higher costs and reduced availability of credit. Bernanke and Gertler (1995, p. 41) highlighted one channel in which an increase in Fed fund rate reduces credit supply:

Open-market sale by the Fed—which shrinks banks' core deposit base and forces them to rely more on managed liabilities—also increases banks' (relative) cost of funds. An increase in the cost of funds to banks should shift the supply of loans inward, squeezing out bank dependent borrowers and raising the external finance premium.

Theoretically, this type of operation also works in reverse so that when a central bank cuts target reserve rates, the credit supply would increase. A main caveat of this thesis is that the banks’ own borrowing cost will decrease along

with the central bank target rate. If this is de-coupled, credit supply may not increase.8 Ivashina and Scharfstein (2010) showed that new loans to large borrowers in the United States fell by 47% during the peak period of the financial crisis

(4th quarter of 2008). Puri, Rocholl, and Steffen (2011) discovered evidence that in Germany banks affected by financial crisis rejected substantially more loan applications than non-affected banks. The Bank of England (2012)

stated that, since the 2008 Financial

4 See Centre for Banking, Finance and Sustainable Development (2012). 5 In the U.S. this is the Federal funds rate and in the U.K. this is the official bank rate.

6 In the case of the United States after the 2008 financial crisis, MBS became very risky and illiquid stemming from a collapsing U.S. housing market.

7 For example, in the United States, Section 128 of the Emergency Economic Stabilization Act of 2008 allows the Federal Reserve (Fed) to pay interest on bank excess reserves at the Fed. This type of risk-free returns dovetails nicely with the risk aversion adopted by the banks and takes away incentive for the banks to extend credit to businesses.

8 According to The Economist (2013), quoting IMF and Bank of Italy studies, investors ’ concern about bank and sovereign risks pushed up the banks’ borrowing cost in the euro zone, negating the effect of ECB’s easing. As a consequence, the supply of loans contracted.

A.F. Herbst et al. / Global Finance Journal 25 (2014) 1–16 3

Crisis, BoE has cut interest rate to 0.5% and purchased UK$375 billion of asset in the QE programme. However the lending to UK resident households and business had been flat up to 2012 OECD (2013) found that after the 2008 financial crisis, small and medium-sized enterprises (SMEs) in the euro zone and UK faced higher loan rates, greater collateral requirements, and shorter loan maturity than large firms. In particular, outstanding SME loans in Italy,

Portugal, UK and US declined continuously from 2008 to 2011.9 This has vital implication on employment levels as

SMEs create the majority of jobs in most OECD countries.10

On the modeling issue, Romer (2013, p. 129) pointed out the importance of multiple interest rates when there is divergence in various benchmark rates:

Sometimes, however, the fact that there are multiple interest rates is critical to behavior of the economy. This is especially true in periods when there are large changes in how well financial markets are functioning. As U.S. and world financial markets started to undergo stress in 2007 and the first part of 2008, for example, interest rates for very safe borrowers, such as the U.S. government, fell, while interest rates for riskier borrowers, such as small firms, rose. And when the world entered a full-blown financial crisis in late 2008, safe interest rates plummeted and risky interest rates skyrocketed. Analyzing such developments in a model where all interest rates move together is almost impossible.

3. The model

The traditional IS-LM model has a single composite interest rate for the whole economy. This will work if the various rates move proportionally in tandem with some benchmark rate. In a reserve trap, the bank loan rate to the real sector moves out of synchrony with the central bank’s benchmark rate, while interest rates for savings remain in synchrony, then a single composite rate analysis will not be adequate. To accommodate this divergence of rates, we

modify an open economy model to consist of two different rates; namely, r I, which is the bank loan rate to the real

sector, and rS, the rate on savings; and these rates are connected by a margin spread μ:

rI −rS ¼ μ; μ N 0 ð1AÞ

In case QE results in a reserve trap, banks ’ preference to keep reserves, and reluctance to extend credit to the real sector, will be reflected by a higher rI. If this is accompanied by a policy to lower a target interest rate, this will drag down rS and the spread μ will widen.

[For notational brevity, in our model we adopt the shorthand way of indicating the signs of the partial derivatives with respect to each variable by writing (+) or (−) signs underneath the variables.]

We assume the spread μ, savings S, investment I, net export NX and net capital out flow CF to take the following functional forms:

μ ¼ μ

MS;S ¼ S r S ;

YI ¼ I ðr I Þ;

NX¼

NX e Y Y_

;CF ¼ CF rS ;r_

! ð1BÞ_

ðþÞ _

!; − ; −

;

_ð þÞ ðþÞ ð −Þ

_ð Þ ð Þ ð þÞ ð−Þ

ðþÞ

where MS represents domestic money supply and Y is the national output. Y and r are exogenous foreign national output and interest rate, respectively. Exchange rate e is expressed as the amount of foreign currency per unit of home currency.

We see that the wider the spread μ, the lower will be the I curve in the rs − I space in Fig. 1. Equilibrium in the real sector means:

! ! _ _

S rS ; Y ¼ I rI þ NX e ; Y ; Y _ ð2Þ

ðþÞ ðþÞ

ð −Þ ð −Þ ð−Þ ðþÞ

9 For instance, Appendix A Chart 6 shows the dramatic decline in lending to UK resident households and businesses after the 2008 financial crisis. Note that FLS refers to the “Funding for Lending Scheme” launched by the Bank of England (Bank of England, 2012).

10OECD (1996) found that for the OECD Member countries (including United States, Japan, Germany, UK), SMEs make up over 99% of all firms, are the source of 40–80% of all employment, and contribute 30–70% of GDP. Similar findings are found in OECD (2010).


Fig. 1. The modified savings and investment schedules.

Balance of payment equilibrium:

!_

NX e ; Y ; Y _ −CF rS ; r _ ¼ 0 ð3Þ

ð−Þ ð −Þ ð þÞ ð −Þ ðþÞ

➔ S rS ; Y I rS ; μ CF rS ; r_

!

3A

ð þÞ ðþÞ!

¼

ð−Þ ð−Þ!

þ

ð−Þ ð þÞ ð ÞEq. (3A) will generate the following real sector equilibrium curve Requil, which is downward sloping in the rs − Y space11:

r

S ¼

R Y ; μ ; r_

!

4

equil ð −Þ ð−Þ ðþÞ ð Þ

We note that a wider interest rate spread μ will cause a larger drop in the Requil curve.12

Money market equilibrium (L denotes the liquidity preference function and P the domestic price level. To simplify the analysis, we assume P to remain constant for the period under discussion):

!MS

P ¼ L rS ; Y ð5Þ

ð −Þ ðþÞ

This curve is upward sloping in the rS − Y space13 and implies the following equilibrium curve Mequil for the monetary sector:

_ rS

¼ M

equil Y ; MS; P ð6Þ

ð þÞ ð −Þ ðþÞ

We reach a general equilibrium when the following Eq. (7) is satisfied (see point A in Fig. 2.):

!_

Requil Y ; μ ; r _

¼ M

equil Y ; MS ; P ð7Þ

ð−Þ ð −Þ ðþÞ ð þÞ ð−Þ ðþÞ

11 To show the Requil curve is downward sloping, we take the total derivative of (3A) and note at equilibrium, SrS drS þ SY dY ¼

IrS drS þ C FrS drS ➔drS SY

b 0.¼dY I

rS þC F

−S

12 r S r SHolding Y constant, we take the total derivative of (3A) with respect to rS and μ, we have SrS drS ¼ IrS drS þ Iμ dμ þ C FrS drS ➔

drS

¼Iμ

b 0.dμ SrS −C FrS −IrS

13 To show this curve is upward sloping, we totally differentiate (5) with respect to rS and Y, and at equilibrium: 0 ¼ LrS drS þ LY dY ➔drS

¼ −

LY

N0.dY LrS


Fig. 2. General equilibrium.

➔Y

¼

Y μ ; r_ ; MS ; P

!

8

ð−Þ ðþÞ ð þÞ ð−Þ ð Þ_

➔e ¼ e CF ; Y ; Y_ð9Þ

ð−Þ ð−Þ ð þÞ

In our model, a drop in rS will affect the exchange rate through both net capital outflow and national output. In case

the effect from the former source is larger than the latter source, 14 the domestic currency will depreciate as shown in Fig. 3A and B. Such depreciation, barring competitive depreciation from trading partners, will cause net exports to increase.

An increase in MS under the QE program will shift the Mequil curve to the right. When we have the reserve trap

scenario, this increase in MS will cause the spread μ to widen and shift the Requil curve to the left (see Fig. 3A and B.)

To see the net effect of QE on Y, we take the total derivative of Eq. (7):

RequilY

dY þ

R

equilμ dμ

¼ M

equilY dY

þ

M

equilMS dMS

➔

dY

¼

_MequilMS

þ R

equilμ μMS

ð10ÞdMS MequilY

_ R

equilY

The sign of Eq. (10) depends on the numerator of Eq. (10),15 namely,

_MequilMS

þ R

equilμ μ

MS

ð11ÞWe have the following two cases:

Case A −MequilMS þ Requilμ μMS N0 will occur if_M

equil MS 16N

_ R

equilμ . This means when the effect of anμMS

increase of MS on Mequil exceeds its indirect effect through μ on Requil, thendY

N 0, and QE will bedMS

effective in promoting Y0 to Y1 (point A to B) in Fig. 3A.

de ∂e ∂ CF ∂ e ∂Y14

The condition is equivalent to ¼ þ N0.drS ∂CF ∂rS ∂Y ∂rS

15

The condition is equivalent to

de

¼

∂e ∂ CF

þ

∂ e ∂Y

N 0.drS ∂CF ∂rS ∂Y ∂rS

16 fi _MequilMS

From Eq. (6), MequilMS b 0,from Eq. (4), Requilμ b 0, and from rst expression in Eq. (1B), μMS N 0; hence, N _ R

equilμ ⇔μMS

_MequilMS þ Requilμ μMS N0.


Fig. 3. (A) Case A, effect of QE with reserve trap when −Mequil MS þ Requil μ μMS N0. (B) Case B, effect of QE with reserve trap when −Mequil MS þR

equilμ μ

MS b

0.

Case B −MequilMS þ Requilμ μMS b 0.17 In this case, since the effect of expansionary policy on Mequil in terms of effect on μ is not greater than the effect of μ on Requil, then dM

dYS b 0, and QE will result in Y0 contracting to Y2 in

Fig. 3B (from point A to C). 18

To see the importance of the trade component in our model, we use a counter-factual approach of converting our model into a closed economy and comparing the result. We note that the absolute value of

the slope of the open economy Requil curve (¼ _drS _ ¼_ SY _) will be smaller than that of a closeddY IrS þC FrS _SrS

economy ( _drS _ ¼_S

Y ). _ _ _ _IrS _SrS

¼ dY _ _ _ _ __ _ _ _

_ _ _ _For case A, in a closed economy, the equilibrium after the QE will move from point A to B ’ in Fig. 4A. In the

corresponding open economy case, the flatter Requil curves means equilibrium will move from A to B and output will be at Y1, which will be higher than Y1′ in a closed economy, suggesting a stronger recovery led by NX. Similarly for case B, QE and reserve trap will lower the output from Y0 to Y2′ in a closed economy (see Fig. 4B) and the flatter Requil

slope in an open economy will dampen the adverse effect on Y and move it to Y2.The above shows that the final impact of trade on the economy depends very much on the various conditions

assumed in our model, and we wish to point out that an export-led recovery is empirically still

17 There is the trivial case of _Mequil MS þ Requilμ μMS ¼ 0 where QE will have no effect on Y.18

This phenomenon actually occurred in the United States where both domestic investment and residential investment droppedand continue to drop after the QE program. See Charts 7A and 7B (Federal Reserve Bank of St. Louis, 2014).


Fig. 4. (A) Under reserve trap and if −MequilMS þ Requilμ μMS N0, open economy QE vs. closed economy QE. (B) Under reserve trap and if −MequilMS þ Requilμ μMS b0, open economy QE vs. closed economy QE.

an open question. However, a point that is often overlooked in the literature is that when domestic businesses are unable to expand due to lack of credit from the reserve trap, the export sector can still grow since it can be financed by foreign fund sources. There is evidence that a weaker domestic currency, barring competitive depreciation from trading

partners, will cause net exports to increase.19 This will be especially true if the domestic economy is experiencing chronic deflation. Then any increase in export-led economic activities will move the economy toward full employment, possibly pushing up labor costs. At the same time, the more expensive raw material imports due to the depreciation

will tend to increase the cost of manufacturing. Both these factors will help to re-inflate the economy.20

19Among the top four exporters, namely, China, Germany, United States, and Japan (World Fact Book 2012, CIA), China has a centrally controlled economy and government-determined exchange rates, which basically track the US$ and Euro. This makes any comparison based on our model meaningless as any currency depreciation or appreciation is not based on market adjustment. Hence, we will only examine the trade effect for the remaining three countries. See Appendix A Charts 8A, 8B and 8C for U.S./Euro, Japan/U.S. and Japan/Euro exchange rates, showing the relative weakness of the US$ and Euro against the Yen after the 2008 crisis. Appendix A Charts 9A and 9B show after an initial drop, exports from the US and Germany surged to an all time high and Chart 8C shows Japanese exports remained sluggish. (Sources for Chart 8A, 8B, 9A, 9B, 9C: Federal Reserve Bank of St. Louis, 2014; Source for Chart 8C: European Central Bank, 2014).

20Williams and Donnelly (2012) found that for the U.S., 2008 exports accounted for 9% of the rate of growth of GDP and in 2009 when GDP declined 3.5%, exports provided impetus for growth.


Fig. 5. (A) Liquidity trap and reserve trap in Japan. (B) A relative increase in non-Japanese interest rate r .

We have a peculiar case where both rI and rS are near zero, so that μ is practically zero. That is the liquidity trap scenario. This fits Japan’s “lost decade” state of affairs which can now be described by our model. We can see from Appendix A Charts 10A and 10B that the Japanese interest rates are virtually zero in the “lost decade.” Moreover, BOJ

policy led to a reserve trap situation, resulting in output Y0 dropping to Y1 in Fig. 5A (from point E to F) and caused an appreciation of the Japanese Yen. In addition, there is wide perception in the foreign exchange market that Japan was relatively unscathed in the 2008 financial crisis. Japan’s comparatively larger national debt was deemed to be a country-specific issue as most of the debt is financed domestically (inter-generational rather than international). All

these led to a relative appreciation of the Japanese Yen with respect to the US$ and Euro.21 This relative strength in the Yen was a drag on the economy. In case other economies taper off QE but Japan continues, or even expands her QE policy, then the differential between non-Japanese interest rates and the Japanese rate will increase and cause the yen to depreciate (see Fig. 5B, from point E to point G), until of course the Japanese domestic interest rate also increases and the differential narrows.

21 However, the Yen was relatively weaker against other unscathed currencies with fiscal discipline like the Singapore and Australia dollars.


4. Conclusion

The efficacy of QE is still being debated as its long-term effect is being played out, especially with respect to the possible inflation consequences. However, as our study shows, QE can definitely affect the real sector through the exchange rate and net export. Various stages of QE adopted by different countries are also time-staggered so that the relative exchange rate movements will depend on such timing, yielding different implications on various economies. In conclusion, we show that any discussion regarding QE should include the possibility of reserve trap and trade effects. We can explain how reserve traps can dampen the effectiveness of the QE on real sector recovery. We find that under certain conditions trade can theoretically promote recovery even when there is reserve trap. This suggests interesting topics for empirical studies and for further research in the field of monetary policy.

Acknowledgments

The authors wish to thank Dr. P. Swanson and Dr. C. Chittle for their helpful comments and suggestions that contributed tremendously to this study. All remaining errors are ours.

Appendix A

Chart 1. U.S. consumer price index for all urban consumers.

Chart 2. U.S. real gross domestic product.


Chart 3. U.S. M1 money stock.

Chart 4. U.S. reserve balances with federal reserve banks.

Chart 5. U.S. commercial and industrial loans at all commercial banks.


Chart 6. Lending to UK-resident households and businesses.

A

Chart 7A. U.S. real gross private domestic investment.


BReal Private Residential Fixed Investment, 3 Decimal (PRFIC96)

Chart 7B. U.S. real private residential fixed investment.

A

(U.S

. Dol

lars

to O

ne E

uro)

U.S. / Euro Foreign Exchange Rate (DEXUSEU)Source: Board of Governors of the Federal Reserve System

1.7

1.6

1.5

1.4

1.3

1.2

1.1

1.0

0.9

0.82000 2002 2004 2006 2008 2010 2012 20141998

Shaded areas indicate US recessions. 2013 research. stlouisfed.org

Chart 8A. US/Euro foreign exchange rate.


BJapan / U.S. Foreign Exchange Rate (DEXJPUS)

Source: Board of Governors of the Federal Reserve System

(Jap

anes

e Y

en to

One

U.S

. Dol

lar) 170

160

150

140

130

120

110

100

90

80

701995 2000 2005 2010 20151990


Chart 8B. Japan/U.S. foreign exchange rate.

C

(Jap

anes

e Y

en t

o O

ne

Eu

ro)

170

160

150

140

130

120

110

100

90

801999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Chart 8C. Japan/Euro foreign exchange rate.


A

(Bill

ions

of D

olla

rs)

B

(Mill

ions

of C

hain

ed 2

005

Eur

os)

Exports of Goods and Services (EXPGSA)Source: U.S. Department of Commerce: Bureau of

Economic Analysis2,200

2,000

1,800

1,600

1,400

1,200

1,000

800

600

4001995 2000 2005 2010 20151990


Chart 9A. U.S. exports of goods and services.

Exports of Goods and Services in Germany (DEUEXPORTQDSNAQ)

Source: Organisation for Economic Co-operation and Development

360,000

320,000

280,000

240,000

200,000

160,000

120,000

80,0001995 2000 2005 2010 20151990

Chart 9B. Exports of goods and services in Germany.


C

(Bill

ions

of C

hain

ed 2

005

Yen

)

A

Exports of Goods and Services in Japan (JPNEXPORTQDSNAQ)Source: Organisation for Economic Co-operation and Development

100,000

90,000

80,000

70,000

60,000

50,000

40,000

30,0001996 1998 2000 2002 2004 2006 2008 2010 20121994

2013 research. stlouisfed.org

Chart 9C. Exports of goods and services in Japan.

(Per

cent

per

Ann

um)

Interest Rates, Discount Rate for Japan (INTDSRJPM193N)

Source: International Monetary Fund7

6

5

4

3

2

1

01990 1995 2000 2005 2010

2013 research.stlouisfed.org

Chart 10A. Interest rates and discount rate for Japan.


B

(Per

cent

per

Ann

um)

Interest Rates, Government Securities, Government Bonds for Japan (INTGSBJPM193N)

Source: International Monetary Fund8

7

6

5

4

3

2

1

01990 1995 2000 2005 2010 2015

2013 research.stlouisfed.org

Chart 10B. Interest rates, government securities, government bonds for Japan.

References

Bank of England (2012). The Funding for Lending Scheme. viewed 23 January 2014. http://www.bankofengland.co.uk/publications/ Documents/quarterlybulletin/qb120401.pdf

Bernanke, B. S. (1983). Non-monetary effects of the financial crisis in the propagation of the Great Depression. NBER Working Paper, no. 1054.

Bernanke, B. S., & Gertler, M. (1995). Inside the black box: The credit channel of monetary policy transmission. Journal of Economic Perspectives, 9(4), 27–48.

Centre for Banking, Finance and Sustainable Development (2012). MP & Banking Expert Call For Change to BoE Policy. viewed 23 January 2014.

http://www.greennewdealgroup.org/wp-content/uploads/2012/03/Green-QE-report-CBFSD-Policy-News-2012-No-1.pdf European Central Bank (2014). Data Tools. viewed 23 January 2014.

http://www.ecb.europa.eu/stats/exchange/eurofxref/html/eurofxref-graph-jpy.en.html

Federal Reserve Bank of St. Louis (2014). Statistics. viewed 23 January 2014. http://research.stlouisfed.org/datatools.html Ivashina, V., & Scharfstein, D. (2010). Bank lending during the financial crisis of 2008. Journal of Financial Economics, 97(3), 319–338. OECD (1996). Financing SMEs: Employment, innovation and growth. viewed 23 January 2014. http://www.oecd.org/cfe/smes/

2090756.pdfOECD (2010). Innovative SMEs and entrepreneurship for job creation and growth. viewed 23 January

2014. http://www.oecd.org/ cfe/smes/46404350.pdfOECD (2013). Financing SMEs and entrepreneurs 2013: An OECD scoreboard. viewed 23 January

2014. http://www.oecd.org/cfe/ smes/Scoreboard_2013_extract_chapter2.pdfPuri, M., Rocholl, J., & Steffen, S. (2011). Global retail lending in the aftermath of the US financial

crisis: Distinguishing between supply and demand effects. Journal of Financial Economics, 100(3), 556–578.

Romer, D. (2013). Short-run fluctuations. viewed 23 January 2014. http://emlab.berkeley.edu/users/webfac/cromer/e134_sp13/ Romer%20Short-Run%20Fluctuations%20January%202013.pdf

The Economist (2013). The Economist, Broken transmission—The European Central Bank has lost control of interest rates in Spain and Italy’ 2013, The Economist Newspaper Limited, 4 May 2013, viewed 23 January 2014. http://www.economist.com/news/finance-and-economics/21577060-european-central-bank-has-lost-control-interest-rates-spain-and.

Williams, B. R., & Donnelly, J. M. (2012). U.S. International Trade: Trends and Forecasts. viewed 23 January 2014. http://www.fas.org/ sgp/crs/misc/RL33577.pdf