Unconventional Monetary Policies in the Eurozone

(1)

Unconventional Monetary Policies in the Eurozone

Author: Ioannis Konstantinidis

Student number: S3010554

Supervisor: Dirk Bezemer

(2)

Following the latest financial crisis, all the major central banks around the world decided to decrease the key interest rates to historically low levels. After a period of extreme financial instability and sluggish recovery the Fed and in the following years the BOE, the BOJ and the ECB started to experiment with a new form of monetary policy widely referred to as Quantitative easing policies (QE). Even though the objective of expansionary monetary policy in previous decades is well documented and understood, the 2008 crisis was a paradigm shift the recent data challenge all the preconceived notions of monetary rules. In this study we employ a SVAR model and study the effects of the unconventional monetary policy on the mandate of the ECB.

1. Introduction

In the aftermath of the worst global financial crisis since the great depression, central banks around the world responded by reducing policy interest rates sharply. First the Federal Reserve reduced the federal fund rate to 0.25% by December 2008, the Bank of Japan reduced the already low 0.3% in 2009 to -0.1% in January of 2016. Similarly the ECB gradually decreased its key interest rate from 3.75% at the height of the crisis to 0% in March 2016.

The historical testimony of the past decades indicates that conventional monetary policy managed to achieve low and stable inflation but it did not prevent asset bubbles from occurring. Nonetheless when a financial crisis occurred central bankers were able to reduce the interest rates and stimulate the economy into recovery. The inability of central banks to address the crisis has been partially explained by the apparent secular trend decline in real interest rates and the uncharted waters of the zero lower bound (Geneva Report 2016). As Woodford (2003) clarifies, the conduct of monetary policy is essentially the setting of short-term nominal interest rates by the central bank in some form of market operations. In the period just before the Great Recession widely known as the Great Moderation (1987-2007) the guiding principles of central banks has been models using the Taylor principle (Davig et. al. (2007), Bernanke (2004)). That in turn would recommend negative nominal interest rates in the post crisis era. Indeed, at the March meeting of 2009 Janet Yellen stated that “the optimal policy rate would require the federal funds to go negative 6 percent if it could, and because it can’t, I think we have to do everything we possibly can to use our other tools to compensate.”1

Primarily those tools in the case of the ECB have been negative interest rates on the deposit facility of the ECB, monthly large asset purchasing programs otherwise known as Quantitative easing. There are three caveats that apply specifically to the Eurozone case. First, as Garber (2010) and Joyce et al. (2012) note the stresses in the euro area particularly in 2011 and into 2012 has led to substantial outflow of euro deposits from periphery countries into the core countries. This becomes more pertinent when one looks at the Target system imbalances (Sinn and Wollmershauser, 2011). Second, the expansion of the European central Bank balance

(4)

sheet has come about largely through repo operations that were implemented to accommodate these imbalances and difficulties of the banking sector during the European sovereign debt crisis. Third, in 2009 and 2011 the ECB announced two Covered Bond Purchase Programmes (CBPP and CBPP2) of 60 and 40 billion euros each as well as the Asset-backed securities purchasing programme (ABSPP) that amounted to less than 3 billion euros at the end of January 2015 before the announcement of APP. The size of these programmes are small considering the size of the Eurozone economy as well as similar bond purchasing programmes carried out by the Federal Reserve in the period 2008-2011 and the Bank of Japan in period 2010 and 2011.

Since the beginning of the crisis the Governing council, the main decision making body of the ECB, has repeatedly adjusted its three key interest rates (the rate for the main refinancing operations, the rate for the deposit facility and the rate for the marginal lending facility) downwards. In 2014 as the ECB was reaching the Zero Lower Bound (Figure 2), the inflation expectation for 2014 fell from 1.5%2_{in Q4 of 2013 to 0.9% in Q2 of 2014. At that time in the}

Monthly Bulletin of June 2014, published by the ECB, there was an article explaining the risk of deflation. It was stated there that the euro area economy was not in an immediate risk of deflation and if that was the case that would be a source of concern and it would require appropriate policy response, providing at the same time historical evidence that prolonged deflation is accompanied by stagnant economic activity, the rising burden for debt servicing and the creation of additional negative feedback loops between the real economy and the price setting mechanisms.

As the Vice-President of the ECB Vitor Constâncio mentioned in his October 2015 speech at the Financial Stability Conference:

Mario Draghi (2015a) argued that the trend of low inflation on the micro level reflects the underlying weak aggregate demand. The wage setting level as well as the pricing power among firms is weak. Additionally to this the 2016 UN report on the World economy and perspectives estimates that persistent deflation may inflate the fiscal deficits and the public debt to GDP ratios. At the end of 2014 when the official data came out, inflation had been downgraded 3 times more to the actual 0.4%. As is it apparent that raised concerns. With conventional monetary policy exhausted, the Eurozone economy in sluggish growth and on the verge of deflation, the ECB decided to act to achieve its mandate. This is defined as achieving a year-on year increase on the Harmonized Index of Consumer prices (HICP) of below but close to 2% over the medium term.

At January 22 of 2015 the president of the ECB Mario Draghi announced the Extension of the Asset Purchase Programme widely referred to as APP. The programme included the continuation of the Asset-backed securities programme (ABSPP), the third covered bond purchase programme (CBPP3) to include the purchases of sovereign bonds (PSPP) with a total

2_{For inflation I mean Harmonized Index for Consumer Prices (HICP), the HICP and inflation expectations}

is published by the ECB survey of professional forecasters (SPF) and Eurostat

(5)

monthly figure of 60 billion. The programme was initially designed to run until September 2016 or until the 2% objective was reached. The implementation of the programme began on the 9th of March 2015. A year later on March 2016 Mario Draghi announced the continuation of the APP programme increasing the monthly purchases from 60 to 80 billion euros and at the same time decreasing the interest rate on the main refinancing operations, the interest rates on the marginal lending facility and the deposit facility at 0%, 0.25% and -0.40% respectively. The programme was extended until March 2017. In December 2016 the monthly purchases were extended and deduced from 80 billion per month to 60 billion until December 2017. On the 26th_{of October 2017}

the APP was extended from January 2018 until September 2018 with the pace of 30 billion per month.

As of December 2017 the ECB has bought 2.3 trillion euros worth of government (PSPP), corporate (CSPP) and covered bonds (CBPP, CBPP2, CBPP3) and some mortgage securities (ABSPP). The question I will try to answer in this paper is whether or not the purchases carried out by the ECB contributed in raising inflation, achieving therefore the objective that the ECB and Mario Draghi have put forward.

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 .10 .15 .20 .25 .30 .35 .40 .45 .50 2009 2010 2011 2012 2013 2014 2015 2016 2017 ECB Assets as a % of GDP (rhs) Main refinancing operation (lhs)

(6)

-30 -20 -10 0 10 20 30 40 50 60 .00 .02 .04 .06 .08 .10 .12 .14 .16 .18 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

ECB Assets (% change yoy; lhs) CISS (rhs)

Figure [2], Change in ECB Assets and Composite Indicator of systemic stress (CISS)

-1 0 1 2 3 4 5 6 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17

Marginal facility rate Main refinancing operations Deposit facility rate

%

(7)

The structure of the paper is as follows: In the next section, I will review the literature regarding the transmission channels of unconventional monetary policy. In Section 3 I will present the literature critical to the Unconventional monetary policies both in terms of side effects of the policies and in terms of efficacy considering the objectives that central banks have put forward. I Section 4 I will try to give a qualitative account of the euro-area banking and monetary system. I Section 5 I will present my SVAR and discuss the results. Section 6 concludes.

2. Unconventional monetary policy: The transmission channels

Quantitative easing programmes work through various transmission channels from the financial sector into the real economy. We group them into three main categories: the signalling channel (forward guidance) (Andrade et al 2016, Krishnamurthy and Vissing-Jorgensen 2011, Bernanke et. al. 2004), the portfolio rebalancing channel (Andrade et al 2016, Krishnamurthy et al. 2011, Joyce et al 2012), which is often referred to as asset valuation channel; and the re-anchoring channel. Additionally there is the wealth and confidence channel (Bluwstein et al. 2016, Haldane et al 2016) which in conjunction with the other channels aim to exert positive effects on risky assets as well as increase the overall investment and consumption.

We expect that under QE the cost of borrowing will decrease for both household and non-financial corporations leading to an increase in lending (Butt et al. 2014). It is of paramount importance that this channel works properly. As Mario Draghi (2015b) stated, the reduction of the lending rates through monetary policies breaks the vicious cycle of high risk premia to healthy enterprises. This phenomenon was a result of the general macroeconomic uncertainty and the fear of bank insolvency. By breaking this feedback loop, lending to non-financial corporations and households resumes as well as the macroeconomic picture improves as a result of that.

Similar views are expressed by the vice President of the ECB, Vítor Constâncio. He argues that the lending channel is particularly important channel for the euro area firms and households3_.

He explains that the increase of investments is prerequisite in the reduction of economic slack and unemployment in the euro area economy. Therefore the real cost of capital should lower in order for various investment business plans to be encouraged and be profitable.

Finally as von Beschwitz et al. (2016) show, the bond market access for firms is not a significant driver for investment in the euro area, as it is in the United States, since most firms have access to bank capital. However after the 2008-09 crisis there is an increase in the importance of corporate bond market since banks are reducing the lending.

(8)

Figure [4], The Transmission mechanism of Unconventional Monetary Policy

2.1 Signaling channel

It’s undoubted that Central Bankers should act in calm and decisive manner especially in turbulent times. This became more apparent when in July 2012, in the middle of the European sovereign debt crisis, the president of the ECB Mario Draghi declared the following “the ECB will do whatever it takes to preserve the euro, and believe me it will be enough”. It is naturally to assume that the first tool to use in order to accommodate the markets and lead them to a more desirable equilibrium is communication. According to Andrade et al 2016 and Pattipeilohy et al 20134_{the signaling channel, which is a particular form of communication, plays an integral role in}

the transmission of unconventional monetary policy. The signaling channel or forward guidance, as it usually referred to, aims to explain the plans and the timetable in monetary policy. The evidence suggest that Central Banks can gain credibility and therefore influence market expectation in the medium term if they commit to calendar based guidance. Bluwsteina et al. 2016 and Krishnamurthy et al. 2011 argue that the credibility and the commitment of such policies may reduce uncertainty and financial risks, ensuring long term stability. Additionally the commitment of a central bank into buying long maturing assets adds to the credibility of the central bank since an unexpected increase of the key interest rates would expose the balance sheet of the central

4_{Paper Unconventional monetary policy of the ECB during the financial crisis: An assessment and new}

(9)

bank into losses. As Krishnamurthy and Vissing-Jorgensen (2011) argue the commitment should last until the economy recovers and the central bank can sell the purchased asset in the market.

2.2 Portfolio rebalancing channel

One of the main and most important transmission channels of Quantitative easing has been the portfolio rebalancing channel (see Andrade et al (2016), Krishnamurthy and Vissing-Jorgensen (2011), Joyce et. al. (2012), Pattipeilohy et. al. (2013), Demertzis et. al. (2016), Joyce et. al. 2014). This channel operates continuously, investors rebalance their portfolios as monetary policy influence the yields on various assets. More specifically when an investor sell their assets to the ECB they initially exchange these long maturity assets for a short dated asset: bank deposits. At the same time there will be a rise of assets prices and decline of yields (in the case of PSPP, government bonds) leading investors to rebalance their portfolios, searching for yield in acquiring higher yielding assets. Since institutional investors like Pension funds and Insurance companies would like to replace their newly sold assets with similar long-dated assets we would expect to see a purchase of corporate bonds as well as stocks. All these tend to make it easier for many companies to raise capital and issue bonds, easing therefore credit conditions. Additionally as Andrade et al. (2016) argue the policy of asset purchases mitigates the risk in the balance sheet of banks leading them to increase risky loans and reduce the lending rates. In that sense portfolio rebalancing channel as well as the lending channel play a key role in stimulating the macroeconomy.

2.3 Re-anchoring channel

(10)

3 Critical Assessment of Quantitative easing programmes.

3.1 Overview of the conceptual criticism and unintended side effects of QE.

Unconventional monetary policies similar to the APP has been tried before in the United States, Japan and United Kingdom. Even though these policies are relatively new there have been a significant number of papers that critically assess these programs. The critique focuses on the following areas: the efficacy of such programs, the effect on bank profitability as well as the probable distribution effects on income and wealth inequality.

First to critically assess negative interest rates and quantitative easing programs have been Keynes (1936) and Hicks (1937). As Krugman (1998) notes these posthumous economists argued that an economy with the nominal interest rates close to zero fall into the liquidity trap. This is a situation where monetary policy reaches its limits and becomes ineffectual. An increase in the monetary base fails to increase either prices or output. This seems quite peculiar but if we consider liquidity preference as our analytical tool, it makes sense to assume that when the price of money -that is interest rate- fall near or below zero investors prefer to hold money rather than buy these near zero rate assets that carry some risk. To put it differently if the cost of holding money is zero investors would not have the incentive to lend it out or invest it. In section 4 and Figure [6] it appears that there are evidence of that.

Montecino and Epstein (2014) assessed whether LSAP (large scale asset purchase program) affected bank profitability. They use a panel data set between 2008Q1 and 2009Q4 for the banks that participated in the purchase of mortgage backed securities (MBS), the basic pillar of LSAP. They argue that there is statistically significant increase in bank profitability for the banks that participated, controlling for variables that affect bank performance. On the other hand in the Euro area economy the concern for bank profitability has become an increasingly poignant issue. From September 2016 until November 2016 there have been at least 24 speeches by the members of the executive committee of the ECB addressing that. According to Yves Mersch5_the

ECB estimates that the monetary measures have a positive impact on bank profitability, nevertheless the continuation of the very low interest rate environment over the longer term can have adverse effects. Similarly Mario Draghi(2016) in a speech that he gave addressing the European Banking Congress stated that “the challenge of bank profitability is due to the low growth and inflation environment”. He concluded that ECB will continue to act in order to achieve its mandate. In the same spirit, Demertzis and Wolff (2016) suggest that the fall in term premium affect bank profitability and the ability to transform short term deposits into long term loans. They found that between January 2014 and April 2016 there is a positive correlation between term spreads and bank lending spreads in most Eurozone countries suggesting that QE might influence bank profitability.

5_{“The causes of monetary policy measures and their impact” – a Speech by Yves Mersch, Member of the}

(11)

In Figure [6] we can see that the minimum reserves of all the credit institutions in the euro area since the creation of the euro until January of 2012 were at 2%. That is, that the commercial banks have to hold at least 2% of their liabilities, mainly customer’s deposit, at their national banks as reserves. From February 2012 the ECB lowered the minimum reserves ratio to 1%. The ECB pays the commercial banks the main refinancing operation rate (MRO) for their minimum reserves while the excess reserves is subjected to deposit facility rate. As we can see in the Figure [3] the ECB has lowered both mro and the deposit facility rate at 0% and -0.4% respectively, in order to stimulate the banks to lend money to the real economy. As we can see from the creation of the euro until the crisis, banks have always managed to create loans increasing the assets-loans and liabilities-deposits so that the excess reserves fluctuates a bit above zero. From the graph we can conclude that more than 1 trillion euros of reserves are returning -0.4% to the European banks, a direct effect of the extension of the QE programme that started in March 2015. At the same time it is hard to assess whether that alone hurt the profitability of the banks since the initial price at purchasing of the bonds from the commercial banks and later selling to the ECB has a decisive role.

There is a vast range of literature that focuses on income and wealth inequality but limited literature on the distributional effects of QE programs. As I explained in the introduction the main issue that central banks that the Euro area and Japan have, is abnormally low inflation and in the other two countries that initiated QE programs, UK and United States, inflation is well below the 2% inflation target. Therefore it is reasonable to focus on distributional consequences of monetary policy in the Post Great Recession era. The idea behind this critique is that as a central bank try to stimulate an economy, the recipients of the stimulus are disproportionately the wealthiest parts of a society.

Bell et. al. (2012), in the quarterly report from the Bank of England argue that the benefits from the expansive monetary policy were not evenly distributed. It is argued that since the median household has £1000 in current accounts, the lowering of the interest rates would hurt these households. Additionally the benefits from the appreciation of assets prices following the QE programs are being accrued by the richest households. This is indicative of the fact that the top 5% households hold 40% of the financial assets. In contrast to the argument of the distributional effects of Quantitative easing studies like Joyce, Tong and Woods (2011) suggest that in the counterfactual scenario that the QE had not been implemented the GDP would have been 1 and a half to 2% lower and inflation would be around 1% lower. This means that in the absence of QE the UK economy would be in a much deeper recession.

(12)

effects on income inequality. It is argued however that given the structure and the state of the economy, tight monetary policy would also be dis-equalizing since it would hurt employment growth and the refinancing of mortgages suggesting therefore more progressive fiscal policy. Hence as the authors suggest the paradox in the absence of fiscal policy is that either expansionary or contractionary monetary policy has dis-equalizing effects.

Mario Draghi (2015b) expressed similar views on the concern of distributional consequences of quantitative easing. On the one hand, he points that the rising asset prices might benefit wealthy households disproportionately, increasing therefore wealth inequality. On the other hand he emphasized that monetary policy inaction would lead potentially disastrous economic consequences on inflation. According to ECB’s data the most affected by low of negative inflation would be younger households, net debtors, since debt denominated in nominal terms would rise in terms of real debt burdens. Additionally to this, he defended the dis-equalizing effects of his policy by contrasting the benefits of lowering the cost of equity and borrowing rates to spur investment consumption and growth.

3.2 Literature review on the effectiveness of Quantitative easing

These recent monetary developments pose certain question on the effectiveness of these programmes. As I mentioned in Section I, when the negative interest rates reached its limits (Figure [2]) central banks around the world initiated various quantitative easing programs. Since such unconventional policies were implemented for the first time, the framework for evaluating such programs were -and still is- an uncharted territory. Even though there is a widespread consensus on the transmission mechanisms, quantifying and comparing with the counterfactual scenarios seems to be a very thorny task. There are various methodologies being developed so far from event studies to SVAR and DSGE models. It is important to note here that the vast majority of the literature consistently points out the transmission of such programmes to long term yield rates. In this section I will introduce the various research papers and the methodologies that were implemented, categorized by each country of focus.

3.2.1 Japan

(13)

and Koeda (2014) found that that in the counterfactual scenario that Japan’s loose monetary policy would continue the result appears to lower inflation and GDP. They use a SVAR accounting for inflation, output gap, policy rates and excess reserves while utilizing Impulse responses (IR) to regime changes. Indeed Japan’s exit of the zero interest rate policy proved expansionary in reality until the 2008 global crisis. Ever since April 2013 Japan has embarked an ongoing Quantitative and Qualitative monetary easing program purchasing 50 trillion yen annually. According to Kuroda (2013) the program will continue until the 2% inflation target is reached. He added that the inflation target will be achieved in the time horizon of two years. Despite the increase in the annual purchases in October 2014 to 80 trillion annually, the national CPI in Japan is hovering around 0% year-on-year in November 2016.

3.2.2 United States of America

The economy of the United States incurred much deeper recession with sharp loss of hundreds of thousands of jobs every month and a significant drop in GDP growth. Consequently the response from the Obama administration has been the American Recovery and Reinvestment Act of 2009 as the main of the fiscal stimulus package and the Large Scale Asset Purchases as the main monetary stimulus. Gagnon et. al. (2010) explain that these asset purchases by the Federal Reserve reduce the supply of long term risky assets in the market, reducing therefore the long term interest rates and increasing the bank reserves. In order to estimate the effectiveness of the program they employ first an event-study methodology, in particular examining the long term effects of the 2-year and 10-year Treasury bills, as well as, the current-coupon 30-year agency MBS yield, the 10-year swap rate, and the Baa corporate bond index yield. They estimate that within the timeframe of 8 selected events that the yield did decrease significantly. Additionally they use another methodology called Dynamic ordinary least squares (DOLS), regressing the long term Treasury yields with Cpi, unemployment gap, differences in inflation expectations and the 6month realized volatility on the 10-year bil. The verdict of both of these methodologies are that the term premium for long term T bonds reduced by 30-100 basis points while the yield of corporate bonds decreased by 400 basis points. All in all, the LSAP seems to be quite successful in reducing the long term yields in both Treasuries and corporate bonds an intended outcome for the portfolio balance channel to take hold and increase investment and growth. Similarly, Krishnamurthy and Vissing-Jorgensen (2011) conducted an event study on periods between 2008-2009 QE1 and 2010-2011 QE2. They find that the nominal interest rates of the Treasury bonds decreased considerably. It is interesting to note that the QE2 period had lower effects in reducing yields for long term assets and low-grade corporate bonds indicating diminishing effects on the effectiveness of the program. Summers (2015, 2016) agrees with that view and argues that QE works only in turbulent times and after that the effectiveness is diminishing.

(14)

The policy of asset purchases as it was implemented in the United Kingdom carries similar features with such programs. One key difference comparing with the ECB is that the purchases were mainly focused on the non-bank sector. The main administrative body, of the Bank of England, purchased both private and public assets, expanding the balance sheet of the BOE, aiming to increase the amount of money into the economy and meet the inflation target of 2% over the medium term. Joyce, Lasaosa, Stevens et. al. (2011) conducted an event study in order to determine whether or not the announcements of monetary actions by the BOE had the intended effects. The authors estimate that the long-term yields of gilts decrease by about 100 basis points. Even more impressive is the decrease in the corporate bonds. The A-graded bonds saw a decrease of 70 points and the high yield bonds 150 points. It is important to note that these changes had a sustained effect. Additionally, their estimation suggests that the main transmission mechanism is through the balance portfolio channel. This something which becomes pertinent if we look at Joyce et. al. (2014) paper on the rebalancing effect of institutional investors. The authors rather than estimating the impact of QE on the government yields or other asset prices they looked into the balance sheets of insurance companies and pension funds. They found that insurance companies and pensions funds reduced their holdings of gilts and increase their holdings of corporate bonds as expected. By contrast, their equity holdings reduced during that period. Joyce et. al 2011 estimated that the initial 200 billion of QE increased GDP by 1.5-2% and inflation by 0.75-1.5%. The authors look into the changes of the asset prices in an event study analysis and then put those results in a macroeconomic model. The efficacy of these Unconventional policies is still debated in the academic community. Another factor that contributes to complications of the British economy is the aftermath of the Brexit vote. On August 4, 2016, after the Brexit vote, the BOE decided to decrease the key from 0.5% to 0.25% while increasing the government purchases by 60 billion pounds over the next 6 months and corporate bond buying program by 10 billion over the next 18 months. This postponed the tapering process by the BOE leaving a lot of issues about the future of the British economy to be answered.

3.2.4 Euro Area

(15)

finding of the paper suggest that VIX is the main explanation for the shocks in output, inflation and central bank assets and that inflation affected 3 times less than output due to monetary policy, something which is contrary to the pre-crisis monetary literature. The effects of unconventional monetary policy for the euro area suggest that inflation and output were positively affected by the purchases of the ECB. Nevertheless it is interesting here to note two things, Even though there is dearth of historical examples to draw conclusions from, Joyce et al.(2012), it appears that the effects of quantitative easing announcements are more effective during periods of financial uncertainty while at the same time it appears that there are diminishing returns to the effects of these programs (Krishnamurthy and Vissing-Jorgensen (2011), Summers L. (2015), (2016)).

4. Qualitative analysis of the monetary and banking system of the euro area

In this section I will present two alternative methods for understanding the monetary policy in the euro. The first one is called the shadow rate, a rate which calculates the appropriate key of the Central Bank should be without the restraint of the zero lower bound. The second methodology explores the relationship between QE in the context of the euro area, the balance sheet of the commercial banks and the portfolio rebalancing channel.

4.1 Shadow rate

Wu and Xia (2016) estimated a rate which they call the “shadow rate” which tracks on par the changes of the key interest rate from 2004 until now. In the period after the crisis the key rate and the shadow rate and diverging as we see in Figure [5]. The shadow rate includes 97 macroeconomic variables the most important of which are: the policy rate, industrial production index, consumer price index, capacity utilization, unemployment, and housing market. As we can observe from both Figure [2] and [5], after the Mario Draghi’s speech the shadow rate converge with the mro rate in the following year while the system stress indicator (Ciss) subsided for the following two years until inflation in the euro area reached negative territory.

Similarly, the authors calculate the shadow rate for the federal funds. Their results suggest that on that the shadow fed funds rate reached an all-time low of -3% in May 2014 while rebounded back at 0% on November 2015. Indeed two months later the Federal Open markets committee headed by Janet Yellen decided to raise rates starting the process of normalizing monetary policy.

As of December 2017 it appears that shadow rate for the euro zone is the lowest its ever been since the beginning of the crisis.

4.2 Balance sheet analysis

(16)

government debt from a seller, that means that the seller swaps the security for a deposit on their balance sheet, while the commercial bank increase their deposits (liability) and their reserves (the IOUs that the Central Bank paid). Finally the Central Bank increases both their assets with the security that they bought and their liability with the money they issued. In Figure [7] we can see how that process takes place in the balance sheet of the participating actors. Two things stand up from Figure [7], firstly the newly created reserves on the commercial banks do not necessarily imply that these banks will have an incentive to lend since banks do not lend their reserves but they rather lend depending on the prospects of financing successful ventures. Secondly, the deposits of the sellers will have to either purchase riskier assets (rebalance portfolio) or buy shares of companies reducing the cost of financing.

In Figure [6] we can see the minimum reserve requirements and the excess liquidity of all the euro zone commercial banks. Up until January 2012 commercial banks had to hold 2% of their liabilities (mainly customer deposits) with the National Central Bank. This can also be seen in Figure [7], row 3. The minimum reserve requirements yield the mro rate for the bank and the excess liquidity: the deposit facility rate (currently at 0% and -0.4% respectively). As Figure [7] and Baldo et al. (2017) indicate the APP purchases are the main driver of the excess liquidity of the banking system. The reasons that the authors trace, is not system risk as it was the case for 2012 but rather the financial structure, bank business models and regulation. As the authors argue that since there is a “home bias” in government bond investment, there is a propensity to investment in the government bonds where each bank is domicile in. Meaning since the interest rates throughout the AAA bond market have been reduced to such an extent there is an incentive for banks to hold such the liquidity (from the Central Banks QE program) and incurring -0.4% rather than investing in a more illiquid and riskier assets classes.

Additionally the authors claim for the banks of lower-rated countries like Italy, Spain, Portugal, Greece the there is an incentive to take advantage of the yield opportunities in domestic assets (home bias) that yield higher that the deposit facility rate. Hence, there is only a small fraction of the excess liquidity belong to the banks of these countries.

The differences in banking behavior becomes more pertinent by the fact that more than 50% of the Sellers of PSPP bonds were investors with headquarters domiciled outside the euro area (Baldo et al. (2017). This also comes with the natural tendency of those institutional investors to have subsidiaries either Frankfurt, Paris, Amsterdam or Luxembourg. From the 2.2 trillion euros in asset purchases 1.2 trillion euros stay at the Central banks in the form of excess liquidity. Around 80-85% of those 1.2 trillion euros as of December 2017 are held within the banking system of Germany, France, the Netherlands and Finland yielding -0.4%. Which in turn helps also explain the fact that 60% of all the QE transactions were settled by the Bundesbank via the Target 2. Another unintended consequence of these purchases is the divergence of the Target 2 balances (Figure [9]). Peter Praet6_{, member of the governing council of the ECB, explained in a speech he}

gave that when the Banco de España buys Spanish bonds from a German counterpart, or a UK based bank via the Bundesbank there is a flow of money from Spain to Germany.

The literature on the transmission of monetary policy ignores the peculiarities of the European banking system, but it becomes worrying that if half of the asset purchases stays in the books of the banks with the Central Bank, that questions the very effectiveness of the portfolio

(17)

rebalancing channel and the transmission of monetary policy from the Central bank into the real economy. The effectiveness of the monetary policy, as a whole, based on the mandate of the ECB will be investigated in the next section.

800,000 1,200,000 1,600,000 2,000,000 2,400,000 2,800,000 3,200,000 3,600,000 4,000,000 4,400,000 -6 -4 -2 0 2 4 6 04 05 06 07 08 09 10 11 12 13 14 15 16 17

ECB's Balance sheet (in millions of euro, lhs) MRO rate % (rhs)

Shadow rate % (rhs)

Black line: "Whatever it takes" speech by Mario Draghi Shaded area: The introduction of APP

Figure [5], The Balance sheet of the ECB, MRO rate and the shadow rate (Wu and Xia, 2016)

0 200 400 600 800 1,000 1,200 04 05 06 07 08 09 10 11 12 13 14 15 16 17

Total excess reserves of credit institutions in the euro area Total required reserves of credit institutions in the euro area Unit: Billions of euros

(18)

Figure [6], Total excess and required reserves of credit institutions in the euro area.

Figure [7], Balance sheet changes regarding QE purchases and issuing of loans by commercial banks

4.3 Conclusion

According to Wu and Xia (2016) the shadow rate for the euro area appear to be lowest levels yet, in contrast with the US economy where the authors calculate that the shadow rate surpassed the zero lower bound on November 2015. Two months later the Fed start raising rates again, after seven years of 0% federal funds rate.

(19)

following section I will try to create a model in order to estimate whether these purchases where effective in helping the ECB to achieve its mandate.

5. Euro area SVAR model

5.1 Introduction

The large-scale macroeconomic models prior to the 1980’s were predicated upon a set of a priori assumptions that were dictating to a very large extent the outcome of the models reinforcing the assumptions of the modeler. In other words these models according to Sims (1980) were over-identified. In that paper Sims propose a new way of macroeconomic modelling without the need of specifying every interaction among the variables. These models are multivariate generalizations of univariate autoregressive equations and can be employed in order to model the macroeconomic effects of policy innovations. The mathematical intuition of these models is to analyze the dynamic impact of random disturbances on a system of variables. Structured VARs (SVAR) models play an important role in interpreting monetary policies by imposing some restrictions on these equations with the purpose of producing meaningful impulse response functions and variance decomposition without distorting the dynamic structure of the model.

Following the recent literature on macroeconomic modelling of conventional and unconventional monetary policy (eg. Jayaraman et. al. (2009), Boeckx et. al. 2017, Elbourne, Ji and Duijndam (2017)), I will attempt to employ a VAR model and impose restrictions in order to produce impulse response functions and variance decompositions.

The benchmark VAR model that we consider has the following representation:

Compact form: Yt = α + D(L)Yt+ εt (1)

Y1,t = d01 + d11Y1,t-1 +d21Y2,t-1 + d31Y3,t-1 + d41Y4,t-1 + d51Y5,t-1 + ε1,t

⋮

(20)

Where t = 1,…, T denotes time, Yt is an 5-dimensional vector of endogenous variables, α is a

vector of constants, D(L) is a matrix lag polynomial, and εt is the mutually uncorrelated error

disturbances.

The VARs in this study are estimated in log levels in order to allow for the cointegration of the variables.

5.2 Data

The benchmark specification includes the main macroeconomic, financial and monetary variables that capture the interactions during the financial crisis. The log of seasonally adjusted GDP was transformed by interpolating the quarterly data into monthly series. The second variable is the log of the hicp an important variable for estimating the relative success of the unconventional monetary policy, fulfilling the mandate of the ECB.

Yt is a vector that contains 5 variables: the log of central bank assets, the log of seasonally

adjusted real GDP, the log of seasonally adjusted harmonized index of consumer prices, the level of financial stress (CISS) and the difference between the marginal refinancing rate (MRO) and the interbank rate (Eonia).

In order to capture the central bank’s monetary policy we employ the balance sheet of the central bank as a proxy (Gambacorta and Hofmann (2012), Boeckx, Dossche et. al. (2014), Gambacorta et. al. (2014)). Then I include gross domestic product and hicp (Gambacorta and Hofmann (2012), Elbourne et. al. (2017)). The fourth endogenous variable is the CISS indicator which is the composite indicator of the bond market uncertainty (Holló et. al. (2012), Boeckx et. al. (2014). Since the euro area is quite diverse across the different countries it is essential to include an indicator for the relative stability of the euro area as a whole. Lastly, the benchmark VAR model includes the difference between the main refinancing rate (mro) and the EONIA rate in order to capture the interbanking liquidity of the Eurozone banking system (Haldane et. al. (2016), Burriel et. al. (2016), Boeckx, Dossche et. al. (2014)).

I gathered all the data from the ECB statistical data warehouse.

5.3 Benchmark estimation results

(21)

After the estimation of the reduced-form VAR we check the root of our characteristic polynomial. We found that all the unit roots are below 1, meaning that our VAR is stable.

That means that D(L) is an invertible table and therefore we can express Yt in equation (1) as the

sum of the past white noises et. (VMA (∞) representation). Now we can rewrite equation (1) in the

SVAR form accordingly: A [ 1 ⋯ 𝛼15 ⋮ ⋱ ⋮ 𝛼51 ⋯ 1 ] [ 𝑌1, 𝑡 ⋮ 𝑌5, 𝑡 ]= [ 𝛼01 ⋮ 𝛼05 ] +[ 𝛽11 ⋯ 𝛽15 ⋮ ⋱ ⋮ 𝛽51 ⋯ 𝛽55 ] [ 𝑌1, 𝑡 − 1 ⋮ 𝑌5, 𝑡 − 1 ] + … +[ 𝑒1𝑡 ⋮ 𝑒5𝑡 ] (2) AYt = α + Β1Yt-1 + … + B5Yt-4 + εt

Compact form: AYt = α +B(L)Yt + εt, εt ~ Ν(0,1)

Now we premultiply with A-1_{and we have:}

AA-1_Y

t = A-1α + Α-1Β(L)Yt + A-1εt

Υt = γ + Φ(L)Yt + ut, ut = A-1εt, ut ~ N(0,Συ) (3)

Without any restrictions, the parameters of SVAR are not identified. We have 15 equations, because the variance covariance matrix (Συ) is symmetric and 25 (52) unknowns (A-1).

Συ = Α-1Α-1ʹ

Now in order to identify it the SVAR model we need to impose some restrictions. That means that we need to impose 10 restrictions on the upper triangular A matrix.

Matrix A = [ 𝑎11 0 0 0 0 𝑎21 𝑎22 0 0 0 𝑎31 𝑎32 𝑎33 0 0 𝑎41 𝑎42 𝑎43 𝑎44 0 𝑎51 𝑎52 𝑎53 𝑎54 𝑎55]

By imposing 10 restrictions as seen above we perform a regular recursive Choleski identification. The residuals will be as follows:

uECBassets, t = a11εECBassets, t

uGDP, t = a21εΕCBassets, t + a22εGDP, t

uhicp, t = a31εECBassets, t, + a32εGDP, t + a33εhicp, t

uCISS, t = a41εECBassets, t + a42εGDP, t + a43εhicp, t + a44εCISS, t

(22)

The mathematical meaning of the recursive Choleski ordering is that the residuals ut of the first

equation is not affected contemporaneously by the residuals of second or any other equation. The residuals of the GDP are only affected contemporaneously by the residuals of the first equation. Lastly, the residuals of the fifth equation is affected contemporaneously by the residuals of all the other equations.

Since the ECB assets purchases are the most exogenous variable and affect all the other variables contemporaneously, it is ordered first. This is due to the fact that the markets are aware of the monetary policy schedule of the ECB months before implementation of the purchases (signaling and re-anchoring channel). The second ordering is GDP the residual of which affects contemporaneously all the variables except the ECB assets, following with inflation, the financial uncertainty index, and the spread between the MRO and the eonia rate (Ciccarelli et. al (2015), Boeckx et. al (2014)).

5.3 Variance decomposition

We calculate the variance decomposition for our identified VAR model. The variance decomposition show the relative variance of each endogenous variable to each own shock and in response to other endogenous variables. Therefore with variance decomposition we can tell the proportion of movement of each endogenous variable to each other endogenous variable. In Table 1 we can that the variance of the ECB assets is explained by 20% by inflation whereas in Table 2 inflation is explained only by 3% by the ECB assets prices.

In Table 3 we can confirm that the use of the balance sheet of the central bank for the purposes of monetary easing help to explain 10% of the CISS financial uncertainty index. That seems to provide evidence that the signaling channel of the ECB is working. Similar event studies provide evidence towards that direction as well.

Nevertheless, the variance decomposition analysis provides a playing level field understanding our endogenous variables but it creates difficulties deriving at concrete conclusions about the nature of the relationship of them.

5.4 Granger causality

(23)

past and the present may cause the future but the future cannot cause the past. After all, the question of causality is a deeply philosophical question, difficult to disentangle, just because something preceded something that doesn’t mean that the first event is responsible for the second.

In Table [4] we can notice that inflation Granger cause the ECB assets price at the 10% confidence level something that is in line with our expectations as well as with numerous speeches of the members of the governing council of the ECB. At the same time in Table [6] we can see that the opposite is not true. We cannot reject the null hypothesis that ECB Assets do not Granger cause inflation. These results does not point to the narrative of the ECB that QE helps to achieve the 2% inflation mandate.

5.5 Impulse response functions

We produce the impulse response function with Cholesky ordering just like the variance decomposition. Graph [7] shows that the standard responses of the ECB balance sheet to inflation and gdp. The blue line shows the median impulse responses and the doted red lines represent ±2 standard deviations.

Impulse Responses of inflation and GDP to Balance sheet shocks show that the unconventional monetary policy of the ECB has not been very effective boosting inflation and growth.

-.1 .0 .1 .2 .3 .4 1 2 3 4 5 6 7 8 9 10 11 12 Inflation -0.02 -0.01 0.00 0.01 0.02 0.03 1 2 3 4 5 6 7 8 9 10 11 12 GDP

Figure [8], Impulse response of Balance sheet assets to the inflations and GDP

6. Conclusion

(24)

the ECB has bought 2.1 trillion euros worth of bonds. The focus of the paper is whether or not the expansion of the balance sheet of the central bank were on aggregate raised inflation and GDP. In the SVAR model that I employed I found that the expansion of the balance sheet had minimal or no effects on inflation and GDP. Earlier works by Haldane et al. (2016) and Summers L. (2016) confirm that view noting that quantitative easing is more effective in times of financial uncertainty. It appears that even though inflation Granger caused the ECB assets the opposite is it not the case. This finding confirms the institutional literature that I demonstrate thought-out the paper that deflation has been the main reason why the ECB decided to act.

Nonetheless the monetary policy that the ECB embarked upon did not yield significant results. According to the findings in Section 4, almost half of the APP program, around 1.2 trillion euros were not reinvested in the economy as the theory of portfolio rebalancing channel would suggest. Clearly, the method and analysis presented here is just the beginning of understanding the broader macroeconomic consequences of unconventional monetary policies. The limitation of this study include but is not limited to: the divergence of euro-area economies, the absence of debt and spillover effects of the quantitative programs of the Fed, BoJ, BoE and ECB to one another.

7. Appendices

Table [1], Variance decomposition of ECB Assets

Variance Decomposition of ECB ASSETS:

Period S.E. ECB ASSETS GDP HICP CISS MRO EONIA

(25)

Table [2], Variance decomposition of HICP

Variance Decomposition of HICP:

1 0.264580 1.213749 1.592157 97.19409 0.000000 0.000000 2 0.274535 3.315245 4.623704 90.56282 0.763101 0.735130 3 0.281394 3.579523 4.609989 89.81686 0.764643 1.228989 4 0.283610 3.524928 5.168335 88.88754 0.760622 1.658579 5 0.287935 3.547768 5.416093 88.11133 0.739068 2.185745 6 0.288694 3.584933 5.732252 87.67763 0.750948 2.254242 7 0.289921 3.603037 5.885863 87.48177 0.785888 2.243443 8 0.290060 3.616597 5.881304 87.43442 0.814422 2.253254 9 0.290273 3.615766 5.888642 87.37995 0.826956 2.288691 10 0.290502 3.686563 5.898488 87.27164 0.852639 2.290666

Table [3], Variance decomposition of CISS

Variance Decomposition of CISS:

1 0.010813 2.352574 2.673827 0.138321 94.83528 0.000000 2 0.014150 2.645883 1.561442 0.313769 95.47889 1.23E-05 3 0.016191 2.206095 1.293697 0.633626 95.58508 0.281503 4 0.017777 4.246963 1.080617 0.946126 92.90130 0.824997 5 0.018766 6.129201 0.972531 1.491359 90.08494 1.321968 6 0.019308 7.534247 1.041750 1.801197 88.14807 1.474732 7 0.019771 8.778632 1.397627 2.303643 86.02997 1.490124 8 0.020093 9.504961 1.660724 2.669038 84.65398 1.511295 9 0.020287 9.716346 1.692783 3.111009 83.96246 1.517398 10 0.020416 9.756355 1.692356 3.457106 83.57203 1.522151

(26)

Dependent variable: ECB ASSETS

Excluded Chi-sq df Prob.

GDP 2.443312 4 0.6548 HICP 8.156465 4 0.0860 CISS 2.580396 4 0.6303 MRO EONIA 3.348286 4 0.5013

All 28.64116 16 0.0265

Table [5], X Granger cause GDP

Dependent variable: GDP

ECB ASSETS 5.184214 4 0.2689 HICP 2.043259 4 0.7278 CISS 0.271171 4 0.9916 MRO EONIA 2.426018 4 0.6579

All 12.48034 16 0.7103

Table [6], X Granger cause HICP

Dependent variable: HICP

ECB ASSETS 2.848733 4 0.5835 GDP 4.988906 4 0.2884 CISS 1.409695 4 0.8425 MRO EONIA 3.006636 4 0.5567

All 12.69173 16 0.6951

Table [7], X Granger cause CISS

Dependent variable: CISS

(27)

ECB ASSETS 7.251153 4 0.1232 GDP 6.537902 4 0.1624 HICP 3.058778 4 0.5480 MRO EONIA 1.985652 4 0.7384

All 17.84669 16 0.3330

Table [8], X Granger cause MRO-EONIA spread

Dependent variable: MRO-EONIA spread

ECB ASSETS 7.587500 4 0.1079 GDP 8.611569 4 0.0716 HICP 3.519998 4 0.4748 CISS 18.11456 4 0.0012

All 44.82148 16 0.0001

Vector Autoregression Estimates Date: 11/29/17 Time: 18:02 Sample: 2009M01 2017M06 Included observations: 102

Standard errors in ( ) & t-statistics in [ ]

ECB ASSETS GDP HICP CISS MRO_EONIA

ECB ASSETS(-1) 1.205389 -31152.16 1.384473 -0.000549 0.743630 (0.10974) (82178.6) (1.09723) (0.04484) (0.44928) [ 10.9839] [-0.37908] [ 1.26179] [-0.01223] [ 1.65515] ECB ASSETS(-2) -0.135884 162912.4 -0.707528 -0.021514 -0.898543 (0.16837) (126085.) (1.68345) (0.06880) (0.68932) [-0.80704] [ 1.29209] [-0.42029] [-0.31270] [-1.30351] ECB ASSETS(-3) 0.173115 -235821.3 -1.236692 0.106135 0.680961 (0.14435) (108096.) (1.44326) (0.05899) (0.59098) [ 1.19927] [-2.18160] [-0.85687] [ 1.79935] [ 1.15226] ECB ASSETS(-4) -0.248672 108138.5 0.462487 -0.091251 -0.611478 (0.08861) (66352.1) (0.88592) (0.03621) (0.36276) [-2.80648] [ 1.62977] [ 0.52204] [-2.52027] [-1.68564]

(28)

GDP(-2) 1.76E-08 0.037879 -3.71E-07 4.61E-08 6.29E-07 (8.2E-08) (0.06122) (8.2E-07) (3.3E-08) (3.3E-07) [ 0.21470] [ 0.61877] [-0.45425] [ 1.37936] [ 1.88006]

GDP (-3) -2.28E-08 -0.904577 -7.30E-07 3.71E-08 4.38E-07 (8.3E-08) (0.06180) (8.3E-07) (3.4E-08) (3.4E-07) [-0.27580] [-14.6371] [-0.88480] [ 1.10161] [ 1.29754]

(29)

(0.19908) (149077.) (1.99044) (0.08135) (0.81503) [ 0.42872] [-0.36285] [ 0.71326] [ 1.42929] [ 1.42865]

R-squared 0.989378 0.810455 0.209076 0.848792 0.413830 Adj. R-squared 0.986755 0.763654 0.013787 0.811457 0.269096 Sum sq. resids 0.056721 3.18E+10 5.670225 0.009471 0.950710 S.E. equation 0.026462 19816.19 0.264580 0.010813 0.108338 F-statistic 377.2200 17.31694 1.070595 22.73427 2.859257 Log likelihood 237.4918 -1142.189 2.645215 328.7777 93.71971 Akaike AIC -4.244938 22.80763 0.359898 -6.034856 -1.425877 Schwarz SC -3.704502 23.34806 0.900333 -5.494421 -0.885441 Mean dependent 14.69705 4292.822 -0.002941 0.055812 0.000980 S.D. dependent 0.229933 40761.07 0.266423 0.024903 0.126722

(30)

(31)

Name of the Monetary Policy Start End Volume Covered bond purchase program

(CBPP1)

2-Jul-09 30-Jun-10

Total: 60bn. March 2017 10bn in holdings

Covered bond purchase program 2 (CBPP2)

Nov-11 Oct-12 Plan for 40bn. Actual 16bn. March 2017 6.5 bn in holdings

Covered bond purchase program 3 (CBPP3)

20-Oct-14 ongoing Total: 214 bn so far.

Securities Markets Programme 10-May-10 6-Sep-12 Total: 212.1 bn. March 2017: 100 bn in holdings

Asset-backed Securities Purchase Programme (ABSPP)

21-Nov-14 ongoing Total: 24 bn so far

Public Sector Purchase Programme (PSPP)

9-Mar-15 ongoing Total: 1435 bn so far

Corporate Sector Purchase Programme (CSPP)

8-Jun-16 ongoing Total: 72 bn so far

Targeted long-term operations I (TLTRO-I)

Sep-14 Mar-16 Total: 401 bn.

Targeted long-term operations II (TLTRO-II)

(32)

(33)

8. References

● Adam, K., Tzamourani, P., & Bundesbank, D. (2015). Distributional Consequences of Asset Price Inflation in the Euro Area. Deutsche Bundesbank, Discussion Paper, No 27/2015.

● Andrade, P., Breckenfelder, J. H., De Fiore, F., Karadi, P., & Tristani, O. (2016). The ECB's asset purchase programme: an early assessment, Working Paper Series No. 1956, European Central Bank.

● Baldo, L., Hallinger, B., Helmus, C., Herrala, N., Martins, D., Mohing, F., & Wang, Y. (2017). The distribution of excess liquidity in the euro area, No 200 / November 2017.

● Bank of Japan (2016). Comprehensive Assessment: Developments in Economic Activity and Prices as well as Policy Effects since the Introduction of Quantitative and Qualitative Monetary Easing (QQE). Bank of Japan. September 21, 2016.

● Bean, C., C. Broda, T. Ito and R. Kroszner (2015), Low for Long? Causes and Consequences of Persistently Low Interest Rates, Geneva Reports on the World Economy 17, ICMB and CEPR. ● Bell, V., Joyce, M., Liu, Z., & Young, C. (2012). The distributional effects of asset purchases. Bank

of England Quarterly Bulletin, Q3.

● Ben S. Bernanke remarks At the meetings of the Eastern Economic Association, Washington, DC February 20, 2004

● Bernanke, B. S., & Reinhart, V. R. (2004). Conducting Monetary Policy at Very Low Short-Term Interest Rates, Speech at the meetings of the American Economic Association, San Diego, California.

● Bernanke, B., Reinhart, V., & Sack, B. (2004). Monetary policy alternatives at the zero bound: An empirical assessment. Brookings papers on economic activity, 2004(2), 1-100.

● Bluwstein, K., & Canova, F. (2015). Beggar-thy-neighbor? The international effects of ECB unconventional monetary policy measures (No. 10856). CEPR Discussion Papers.

● Boeckx, J., Dossche, M., & Peersman, G. (2014). Effectiveness and transmission of the ECB's balance sheet policies.

● Burriel, P., & Galesi, A. (2016). Uncovering the heterogeneous effects of ecb unconventional monetary policies across euro area countries.

● Butt, N., Churm, R., McMahon, M. F., Morotz, A., & Schanz, J. F. (2014). QE and the bank lending channel in the United Kingdom. CAGE Working Paper No: 244

● Campbell, J. R., Evans, C. L., Fisher, J. D., & Justiniano, A. (2012). Macroeconomic effects of Federal Reserve forward guidance. Brookings Papers on Economic Activity, 2012(1), 1-80. ● Ciccarelli, M., Maddaloni, A., & Peydró, J. L. (2015). Trusting the bankers: A new look at the credit

channel of monetary policy. Review of Economic Dynamics, 18(4), 979-1002.

(34)

● Davig, T., & Leeper, E. M. (2007). Generalizing the Taylor principle. The American economic review, 97(3), 607-635.

● De Santis, R. A. (2016). Impact of the asset purchase programme on euro area government bond yields using market news, working paper series, No 1939 / July 2016.

● Demertzis, M., & Wolff, G. B. (2016). The effectiveness of the European Central Bank’s Asset Purchase Programme. Bruegel Policy Contribution issue 2016 10 June 2016.

● Draghi, M. (2015a). The ECB’s recent monetary policy measures: Effectiveness and challenges. Camdessus lecture, IMF, Washington DC, 14 May 2015.

● Draghi, M. (2015b). Monetary Policy and Structural Reforms in the Euro Area. Prometeia40, Bologna, Italy, December, 14.

● Draghi, M. (2016). The state and prospects of the euro area recovery” Speech by the President of the ECB, at the European Banking Congress, Frankfurt, 18 November 2016.

● Draghi, M. (2017), Frankfurt, 11 May 2017. Re: Your Letter (QZ-022). Letter.

● ECB Monthly Bulletin June 2009, Chapter 2 “Monetary Financial developments”, p.18-52 ● ECB Monthly Bulletin May 2013, Target balances and monetary operations, p. 103-114 ● ECB Press Release, Covered bond purchase programme completed, 30 June 2010

● Effects of Federal Reserve Forward Guidance, Brookings Papers on Economic Activity 44(1): 1-80.

● Elbourne, A., & Smid, B. (2017). Onderweg naar normaal monetair beleid. CPB

● Gagnon, J., Raskin, M., Remache, J., & Sack, B. P. (2010). Large-scale asset purchases by the Federal Reserve: did they work?. FRB of New York Staff Report, (441).

● Gambacorta, L., Hofmann, B., & Peersman, G. (2014). The effectiveness of unconventional monetary policy at the zero lower bound: A cross‐country analysis. Journal of Money, Credit and Banking, 46(4), 615-642.

● Gambetti, L., & Musso, A. (2017). The macroeconomic impact of the ECB's expanded asset purchase programme (APP).

● Haldane, A., Roberts-Sklar, M., Young, C., & Wieladek, T. (2016). QE: the story so far. Bank of England Working Paper No. 624

● Hausken, K., & Ncube, M. (2013). Quantitative easing and its impact in the US, Japan, the UK and Europe. Springer.

● Hayashi, F., & Koeda, J. (2014). Exiting from QE (No. w19938). National Bureau of Economic Research.

● Hollo, D., Kremer, M., & Lo Duca, M. (2012). CISS-a composite indicator of systemic stress in the financial system, working paper series No 1426, March 2012.

(35)

● Joyce, M., Lasaosa, A., Stevens, I., & Tong, M. (2011). The financial market impact of quantitative easing in the United Kingdom. International Journal of Central Banking, 7(3), 113-161.

● Joyce, M., Liu, Z., & Tonks, I. (2014). Institutional investor portfolio allocation, quantitative easing and the global financial crisis. Bank of England Working paper No. 510

● Joyce, M., Miles, D., Scott, A., & Vayanos, D. (2012). Quantitative easing and unconventional monetary policy–an introduction. The Economic Journal, 122(564), F271-F288.

● Joyce, M., Tong, M., & Woods, R. (2011). The United Kingdom’s quantitative easing policy: design, operation and impact. Bank of England Quarterly Bulletin, 51, issue 3, p. 200-212

● Keynes, J. M. (2007). General theory of employment, interest and money. Atlantic Publishers & Dist.

● Krishnamurthy, A., & Vissing-Jorgensen, A. (2011). The effects of quantitative easing on interest rates: channels and implications for policy (No. w17555). National Bureau of Economic Research. ● Kuersteiner G.M. (2010) Granger-Sims causality. In: Durlauf S.N., Blume L.E. (eds) Macroeconometrics and Time Series Analysis. The New Palgrave Economics Collection. Palgrave Macmillan, London

● Kuroda, H. (2013, April). Quantitative and Qualitative Monetary Easing. In Speech at a Meeting Held by the Yomiuri International Economic Society in Tokyo (Vol. 12).

● Matsuki, T., Sugimoto, K., & Satoma, K. (2015). Effects of the Bank of Japan’s current quantitative and qualitative easing. Economics Letters, 133, 112-116.

● McLeay, M., Radia, A., & Thomas, R. (2014). Money creation in the modern economy. Quartely Bulletin 2014 Q1

● McLeay, Michael, Amar Radia and Ryland Thomas. “Money in the Modern Economy: An Introduction” Bank of England Quarterly Bulletin 2014Q1, 2014

● Monetary Policy. European Economic Review 36 (5), 975-1000.

● Montecino, J. A., & Epstein, G. (2014). Have Large Scale Asset Purchases Increased Bank Profits?. Institute for New Economic Thinking Working Paper Series, (5).

● Montecino, J. A., & Epstein, G. (2015). Did Quantitative Easing Increase Income Inequality?. Institute for New Economic Thinking Working Paper Series, (28).

● Pattipeilohy, C., Van Den End, J. W., Tabbae, M., Frost, J., & De Haan, J. (2013). Unconventional monetary policy of the ECB during the financial crisis: An assessment and new evidence, DNB Working paper, No. 381.

● Sims, C. (1980). Macroeconomics and reality. Econometrica: Journal of the Econometric Society, 1-48.

● Sims, C. (1992). Interpreting the Macroeconomic Time Series Facts: the Effects of

(36)

● Summers, L. H. (2015), "Central Bankers Do Not Have as Many Tools as They Think", commentary in Financial Times, 6 December

● Summers, L. H. (2016). Age of Secular Stagnation: What It Is and What to Do about It, The. Foreign Affairs., 95, 2.

● United Nations (2016). World Economic Situation and Prospects 2016

● von Beschwitz, B., & Howells, C. (2016). Are euro-area corporate bond markets irrelevant? The effect of bond market access on investment. The Effect of Bond Market Access on Investment (2016-06-30). FRB International Finance Discussion Paper, (1176).

Unconventional Monetary Policies in the Eurozone