Asset price bubbles : is Quantitative Easing in the Eurozone likely to lead to counterproductive outcomes?

1 20-02-2016

E. Jakucionyte 2015-2016

Bachelor Thesis Economics and Business Semester 1

Asset price bubbles: Is Quantitative Easing in the Eurozone likely to lead to counterproductive outcomes?

2

1. Introduction

As of March 2015 the European Central Bank (ECB) launched a large scale asset buying program. The ECB will buy up to 1.1 trillion Euros worth of government bonds in order to resolve pressing economic problems such as impeding deflation and financial instability in the Eurozone. This program is widely known as Quantitative Easing. Even though the program is aimed at reaching financial stability Palley (2011, p. 82) expressed concern over whether Quantitative Easing could actually induce an asset price bubble, thereby producing a counterproductive outcome for the Eurozone.

This is an interesting topic to research, whether the monetary policy in the Eurozone may actually lead to a worsening of economic circumstances. The research question of this thesis is therefore: Can the ECB’s quantitative easing policy lead to asset price bubbles? To answer this question a literature review will be conducted along with a brief data analysis exercise. This due to the fact that there exists very little academic literature on Quantitative Easing in relation to asset price bubbles inflation. Therefore, the effects of conventional monetary policy on asset price bubble inflation will be examined. These results will be extrapolated to Quantitative Easing in order to answer the research question.

This thesis will be constructed as follows: First, Quantitative Easing and its effects. Second, the emergence and consequences of asset price bubbles and conventional monetary policy in relation to asset price bubbles. Third, monetary policy in relation to the stock market. Fourth, monetary policy in relation to the housing market. Fifth, can Quantitative Easing create an asset price bubble? Sixth, can Quantitative Easing lead to an asset price bubble in the housing market? Seventh, can Quantitative Easing create an asset price bubble in the stock market? Finally, a conclusion will be established on whether Quantitative Easing can lead to asset price bubbles.

3

2. Quantitative Easing

Different central banks have used conventional monetary policy with one common goal: Low and stable inflation. Using the inflation-targeting approach with short-term nominal interest rates as their tool, central banks tried to limit the volatility of prices (Bernanke, 2000, pp. 2-3). The setting of the interest rate was approximated by central banks using the Taylor Rule, where interest rates responded to changes in both inflation and output gap (Joyce, 2012, p. F271). However, during the aftermath of the 2007-2008 financial crisis conventional monetary tools proved to be ineffective to curb the economic problems such as impending deflation. Deflation will, according to Fisher’s debt-deflation theory, increase debt burden, amplify economic downturns and cause financial distress (Kuttner, 2011, p. 3). Moreover, by 2009 the supply of low rate loans, the ECB’s covered bond purchase program, could not overcome the dried up interbank lending in Europe due to fears over counterparty risk (Fawley, 2013, p. 68).

As the official rate reached a zero lower bound there occurred a disconnection between the official rates and market rates, meaning conventional monetary policy failed to be effective (Joyce, 2012, p. F276). This caused the economy to be stuck in a liquidity trap, a situation where the provision of short term funds will not stimulate economic activity (Fawley, 2013, p. 53). As of January 2009 the Bank of England (BoE) resorted to so-called unconventional monetary policy and started buying mainly British government bonds, therewith providing the economy with ample liquidity. This permanent increase in money stock, Krugman (1998, p. 161) argues, will increase spending, raise inflation expectations, restore market function and will additionally affect the medium-term inflation rate (Joyce, 2011, p. 113) enabling the economy to escape the liquidity trap. Using this policy the central bank signals a credible commitment to economic recovery in the future (Ugai, 2007, p. 36). This unconventional policy is known as Quantitative Easing (QE) and has since the crisis been used by several central banks such as the Federal Reserve (FED), the Bank of Japan (BoJ) and the European Central Bank (ECB). By February 2010 the BoE had already bought 200 billion pounds worth of assets, primarily British government bonds (Joyce, 2011, p. 113). The central banks have since then served a dual mandate of both price stability and securing financial stability in the economy.

The implementation of QE comes about as central banks expand their balance sheet by buying both government and corporate bonds and in turn providing the economy with central bank money (Joyce, 2011, p. 114). This provision of liquidity affects the real economy through different transmission channels. Joyce (2011, pp. 116-119) estimated the following tranmission channels through several

4 econometric models and these established channels will be leading in the following study of QE’s effects.

First, the portfolio balance channel reflects the impact QE has on asset prices. The buying of long-term dated bonds causes a reduction in private porfolio risk and leaves investors with a short-dated asset, bank deposits (Joyce, 2012, p. F278). Investors consider bonds and bank deposits to be imperfect substitutes, forcing them to rebalance their portfolios by acquiring new long-term assets in search of new risk (Joyce, 2011, p. 117). The change in relative supply and demand raises prices of these financial assets, increasing households’ value of collateral, thereby easing their credit constrictions and boosting consumption (Ugai, 2007, p. 15). However, the shift to risky assets may also be towards foreign assets, impeding the effect on domestic markets (Fratzscher, 2013, p. 5). The Fed’s implementation of QE during 2010 induced portfolio rebalancing that affected assets issued by emerging markets economies, in turn causing the dollar to depreciate and not lowering domestic yields (Fratzscher, 2013, p. 26). Moreover, contrary to the earlier mentioned views Eggertsson and Woodford (2003, p. 160) state that this channel does not impact asset prices due to the fact that the near zero interest rate drives the investors’ marginal utility from additional liquidity down to zero.

Secondly, the liquidity premia channel aids investors by lowering the cost of selling assets. During distressed economic circumstances liquidity premia may be substantial. Buying government bonds, which are considered to be liquid assets, by the central bank provides the capital market with liquidity and raises yields on these liquid assets due to increased demand for government bonds. Yields for these liquid assets will rise relatively to less liquid assets of a similar maturity, such as corporate bonds. Trade in corporate bonds will therefore increase and prices of these less liquid assets will increase resulting from the reduction in liquidity premia (Krishnamurty, 2011, p. 6). Krishnamurty finds in his event study analysis that liquidity premia in the United States did decreased substantially QE (Krishnamurty, 2011, p. 18).

Thirdly, the bank lending channel creates an accommodative financial environment where liquidity positions and lending access improve. Ugai (2007, p. 15) theorizes that ample supply of funds creates an accommodative environment for corporate lending and mitigates firms’ liquidity constraints, shifting liquidity flows from saving towards provision of credit eventually resulting in an increase of investment. In addition, King (2002, p. 172) argues based on economic theory that provision of liquidity will lower transaction costs of financial assets by reducing the scarcity of liquidy, thereby easing the cash-flow constraints of the private sector. Likewise, the increased demand for government bonds will lower yields on corporate bonds therby easing the price paid for corporate lending. This increased firm access to capital markets will increase investments (Joyce, 2012, p. F274).

5 Lastly the central bank provides forward guidance through the policy signaling channel, which provides transparity towards economic agents about the course of future policy. By committing to a long term policy rate Ugai (2007, p. 9) expects the central bank will reduce uncertainty in the private sector about the development of future short-term interest rates and liquidity premia. In his empirical analysis he verified this expectation during the implementation of QE in Japan (Ugai, 2007, p. 29). If these signals are conveyed successfully, the expectations on future rates will change toward the set long-term rate (Fratschzer, 2013, p. 10). Krishnamurty (2011, pp. 3-4) concludes that during the QE announcement in the United States the signaling channel effectively lowered all yields on bonds, however it is expected to have a larger impact on intermediate maturity rates than longer maturity rates due to commitment to keep rates low up to economic recovery. Joyce (2011, p. 129) expects the impact of QE on yields to occur not when purchases are made but when expectations on those purchases are formed. The event analysis showed Joyce’s expectations to be true when fifteen- to twenty-year yields fell significatly after the BoE announced the large-scale asset buying programme (Joyce, 2011, p.131). Moreover will these expectations increase consumption because when the expected inflation is bound to rise it will boost consumption due to households bringing forward future consumption as a result of higher anticipated future prices (Palley, 2011, p. 71). In the United States the large scale asset purchase announcement mainly exhibited its effects mainly through the signaling channel according to the econometric study of Christensen (2012, p. F397).

How long the central bank will pursue QE depends on the pace of economic growth. Even when there is economic recovery the Central Bank will continue QE and will return only to the point of nominal interest rate under normal circumstances when monetary easing threatens to overheat the economy and price progression (Ugai, 2007, p. 7). ECB chairman Draghi stated that QE within the Eurozone will persist for as long as necessary to uphold the price stability goal of 2 percent inflation (ECB, 2015b).

Even though the empirical research seems to be pointing out QE does affect real economic activity there is uncertainty about its effects. It is unclear what the contribution of each individual transmission channel is in the overall effect and what the duration of these effects are (Joyce, 2012, p. F287). Moreover, the effects of the transmission channels may be subjected to long lags (Joyce, 2011, p. 115). However, QE has in the case of Japan proved to have a more significant impact on lowering yields than zero interest rate policy (Ugai, 2007, p. 35).

In conclusion, centrals banks implemented QE in order to resolve economic problems where conventional monetary policy failed to do so. The large scale assets purchasing by central banks would result in regaining both price and financial stability. The effects of QE can be divided into four separate

6 transmission channels: The portfolio balance channel, the liquidity premia channel, the bank lending channel and the policy signaling channel.

3. Asset price bubbles and monetary policy

Asset price bubbles can pose a threat to the economy. In the coming section will be explained why this is the case and how asset price bubbles come about. Asset prices are variables influenced by underlying factors such as level of productivity and net worth to owner. However, prices may be subjected to exogenous factors causing an asset price bubble (Gilchrist, 2002, p. 83). An asset price bubble is in essence a sustained overvaluation of an asset compared to its fundamental value. Bubbles are able to prolong because they are expected to grow, driven by investors’ optimism about their ability to sell the asset in the future at a higher price, otherwise known as animal spirits (Brunnermeier, 2013, p. 12). The feedback theory stipulates that the increase in asset values will attract more investors seeking profit, resulting in an even higher rise in asset prices and therefore inflation of the bubble theorizes Roubini (2010, p. 43). According to economic theory will optimistic investors cause the price to increase and pessimists will not be able to counterbalance this development due to short-sale constraints (Brunnermeier, 2013, p. 24).

The sudden collapse of these bubbles, known as ‘The Minsky moment’, often causes damages larger than the gains obtained during the bubble (Okina, 2001, p. 442). The collapse induces direct effects such as large drops in both markets and real estate prices, thereby impairing the balance sheets of financial institutions and leading to banks cutting down lending. This in turn leads to a decline in real economic activity and eventually causes a recession (Brunnermeier, 2013, p. 8). Moreover, it inflicts indirect effects such as bank runs, where short-term creditors demand their funds at the banks, thereby amplifying the impact of the collapse (Brunnermeier, 2013, p. 5). The study of these bubbles is also important because prices determine allocation of capital within the economy, meaning overvaluation could lead to sub-optimal allocation of means. Businesses will accumulate excessive capital stock during investors’ optimism, leading to costs associated with liquidation during price adjustments (Yamaguchi, 2002, p. 2) and distortion of consumption-investment decisions (Kuttner, 2011, p. 13).

Bubble economies distinguish themselves through the following characteristics: A rapid rise in asset prices, overheating of economic activity and sharp expansion of both money and credit supply (Okina, 2001, p. 397). During the 1980’s Norway, Finland and Sweden experienced a boom-bust cycle

7 as a result of liberalization of the financial sector. Norwegian banks used this freedom to expand lending thereby causing a real estate boom which collapsed after monetary tightening in 1986 (Bordo, 2013, p. 14).

The emergence of an asset price boom, otherwise known as the run-up phase of the boom-bust cycle, is often the result of an innovative change within the economy such as a technological change, financial liberalization or financial innovation (Brunnermeier, 2013, p. 3). After technological development there is often uncertainty regarding potential increases in productivity and learning, making it difficult to value the new development (Brunnermeier, 2013, p. 27).

The increases in asset prices are often paired with booms in consumption and lending. The rise in asset prices increases agents’ financial position due to an increase in the value of collateral, lowering the costs of lending for borrowers and improving access to credit, therewith boosting credit supply (Bernanke, 2000, pp. 8-9). The amplification of the consequences of bubbles is stronger when the bubble is fueled by credit, such as the housing bubble of 2007-2008, in contrast to bubbles which are driven solely by animal spirits like the dot com bubble in early 2000 (Brunnermeier, 2013, p. 5). While credit does contribute to asset price bubbles, the expansion of credit does not necessarily imply that there will be inflation of an asset price bubble.

The initiation of an asset price bubble also depends on the created economic environment. Earlier boom-bust cycles sprung in an environment of easy bank and non-bank credit access (Bordo, 2012, p. 16). Credit- driven bubbles cause their own demise when investors realize that the value of their asset dropped below the borrowed amount. Brunnermeier (2013, p. 23) conjectures based on economic modelling this will cause these investors to apply a gamble for resurrection-strategy, meaning additional risk will taken up in an attempt to eradicate initial losses, therby ultimately reinforcing the effects of the bubble collapse.

Asset price bubbles thrive when accompanied by an accommodative monetary policy such as low inflation and credibility of the central bank to commit to stabilizing prices (Bordo, 2012, p. 6). Protracted monetary policy is considered to be one of the driving forces behind Japan’s 1980’s bubble. This monetary easing reduced funding costs for large-scale speculation, supported the rise in stock prices through enhanced capital financing and increased funding abilities through an increase in collateral value (Okina, 2002, p. 416). In addition, an environment of low volatility helps inflating the bubble as well, for investors will seek a higher return in riskier assets (Brunnermeier, 2013, p. 4).

In previous sections was established what elements influence the formation of asset price bubbles. In order to extrapolate the effects of conventional monetary policy to QE and look at the

8 effects it has on asset price bubbles in the two chosen markets, the next section will look at the effect of conventional monetary policy on asset bubble inflation.

Conventional monetary policy can be defined as the central bank adjusting the overnight interest rate on the interbank money market through open market operations. These liquidity provisions take form of central banks purchasing government bonds. These provisions, however, are of a short-term nature; after a certain amount of time the central bank reverses them through a repurchasing agreement. The central bank thereby influences the monetary base and interest rates on a short to medium term (Smaghi, 2009, p. 2). Through these operations the central bank affects asset prices, for instance stock prices, in an upward direction and lowers lending costs, thereby stimulating investment and consumption in the economy (Fawley, 2013, p. 53). QE has similar aims like conventional monetary policy. Through purchasing government bonds the middle to long term interest rate is lowered, the monetary base is expanded and liquidity is provided to increase supply of credit. The way QE differs from conventional monetary policy has to do with its duration and implementation. Central banks implement QE through the expansion and change in composition of their balance sheets in order to improve financing conditions and lower inflation expectations for an extended period beyond short-term interbank interest rates (Smaghi, 2009, p. 3). This because the liquidity provision, in contrast to conventional monetary policy, is not reversed through a repurchasing agreement.

As mentioned earlier, in this research the effects of conventional monetary policy on asset price bubbles will be used as a tool to extrapolate these effects to the possible outcomes of QE. As we saw in the previous section the variables affected by conventional monetary policy, such as the short-term interest rate, differ from the variables QE affects, like long-short-term interest rates and long-short-term inflation expectations. In this manner does QE affect longer-term investment decisions (Smaghi, 2009, p. 6). However, both conventional monetary policy and QE do affect similar economic variables to a certain extent. For instance, both policies affect asset price levels and are aimed at achieving price stability. Additionally, they aid troubled financial markets by providing liquidity and thereby stimulating supply of credit. This way the effects of conventional monetary policy can be used to look at the possible outcomes of QE, although the impact of both policies on the economy may differ.

Data-analysis of 1980’s Japan brings Okina (2000, p. 428) to the conclusion that during economic expansion low interest rates and long-term expectations about that low interest rate amplified inflation of the Japanese bubble. Furthermore, econometric analysis showed monetary expansion played a prominent role in the emergence of housing booms and busts (Bordo, 2012, p. 5). Sticky prices and interest rates as a result of monetary policy tend to cause an increase in expected inflation, following a reduction in real interest rate in turn leading to a sharp rise in asset prices

9 according to economic model simulations done by Gilchrist (2002, p. 76). In addition, the initial boom becomes even larger when access to credit depends on the value of the collateral, because the negative correlation between borrowing rate and value of collateral causes the borrowed amount to be larger after a boom increases asset prices and therefore value of collateral (Gilchrist, 2002, p. 85). Economists from the Austrian school view an environment where asset bubbles thrive to be one of stable and low inflation (Bordo, 2012, p. 6). Kuttner opposes Bernanke’s (2000, p.4) view that monetary policy shouldn’t take asset bubbles into account. He finds not taking asset price developments in to account dangerous because expansionary monetary policy can create asset price inflation even when observable prices of goods and services stay put at their current level (Kuttner, 2011, p. 9).

Inflation of asset price bubbles as a result of monetary policy is put forward in the following examples. The housing and stock market boom from 1971 to 1974 in the UK sprung from an increase of broad money. Moreover, the British commodity price boom of the 1970’s as a result of expansionary monetary policy. Finally, global inflation as a result of the 1974 and 1978 OPEC oil price shocks was aided by expansionary US monetary policy (Bordo, 2013, pp. 13-17). Asset prices bubbles however, do not necessarily have to be preceded by expansionary monetary policy. An example is the dot com bubble of 2000 which was embedded in a rapid growth of productivity and newly developed technologies (Bordo, 2012, p. 15).

As mentioned earlier does easy credit access influence asset prices. Easy credit, propagated by the central bank, induces asset price booms according to cross country econometric analysis done by Borio (2002, p. 11). In the New Keynesian model used by Bean (2004, p. 17) monetary policy did not increase the probability of the occurrence of a bubble, however it did influence the accumulation of debt as a result of its effects on the future output gap and with that raising future costs of a potential credit crunch.

To sum up, asset price bubbles reflect investors’ self amplifying optimism about the price trajectory of an asset. Different factors such as technological and financial innovation, eased credit supply and accomodative monetary policy contribute to inflation of these bubbles. The moment of bubble collapse will cause a sharp reduction in the asset price, thereby inducing harm to the rest of the economy. Moreover, monetary policy has been associated with asset price bubble inflation. Empirical analysis has shown expansionary monetary policy and low interest rates to have influenced bubbles in several countries. However, implementation of expansionary monetary policy does not necessarily mean it will be followed by asset price bubble inflation. Finally, the increased access to credit can influence bubble inflation, but does not necessarily lead to a bubble.

10

3a. Monetary policy effects on the stock market

During the 1920’s Wall Street boom increased investments in developed technologies, such as electricity and automobiles, and innovations in the financial sector caused a bubble in the stock market. The Fed’s monetary policy is considered to have fostered this boom by creating an accommodative monetary policy as well as an easy credit access environment (Bordo, 2012, p. 12). According to economic modelling done by Gilchrist (2002, p. 94) monetary policy could have avoided this boom-bust cycle if the Fed tightened interest rates. Moreover, regression analysis has shown monetary policy in the United States to have had a quantitatively meaningful effect on the stock market, leading Kuttner (2013, p. 19) to the conclusion that a consecutive increase of interest rates could have dampened the large increases in stock prices four years prior to the 2007-2008 crisis. This conventional view however is contradicted by Gali’s (2013, p. 16) vector-autoregression analysis from the United States their GDP dating from 1960 to 2011. From this analysis he concluded that stock prices consistently increased during a period of protracted tightening in monetary policy. Contrarily,

QE is, as stated earlier, monetary policy with certain characteristics such as liquidity provision. In the coming section will be looked at how these characteristics affects stock prices. Kurihara (2006, p. 377) states in his empirical analysis that due to Japanese interest rates under QE being near zero that these on their own don’t affect the prices of stocks. Joyce (2011, p. 116) nonetheless argues based on economic theory that QE’s portofolio rebalancing will induce increased demand for corporate bonds. His event study-analysis shows a reduction of corporate bond yields and an increase in stock prices as a result of QE (Joyce, 2011, p. 141). Moreover, Japanese stock prices seem to fluctuate together with the exchange rate due to the fact that Japan is an export-based country (Kurihara, 2006, p. 377). A depreciation of the currency will therefore stimulate exports in turn leading to an increase in stock prices (Kurihara, 2006, p. 379). Moreover, Kurihara’s (2006, p. 385) empirical analysis found QE to have influenced Japanese stock prices through liquidity provision and the resulting depreciation of the currency.

In brief, empirical studies on monetary policy in relation to stock market bubbles in the United States vary in results. Empirical studies conducted in Japan however have found results in favor of stock market bubble inflation. Economic theory on QE expects demand for stocks and prices of stocks to increase in turn possibly fostering a bubble.

11

3b. Monetary policy effects on the housing market

The second market analyzed, the housing market, exhibits an interesting conjunction with monetary policy. A bubble within the market makes financing cheaper, thereby increasing the housing demand (Gilchrist, 2002, p. 94). The buoyancy in the housing market leading to the housing market bubble is according to data-analysis across different countries in Europe, Northern-America and in Japan done by Ahrend (2008, p. 5) the result of accommodative monetary policy and interest rate set below the Taylor Rule recommended policy. Many of the economies fostering a housing bubble in the beginning of the twenty-first century, such as the US, Denmark and Australia exhibited interest policy rates below the Taylor rule (Bordo, 2012, p. 5). This loose monetary policy in combination with financial deregulation and innovation lead to the overheating of economic activity, a run-up in asset prices and fiscal imbalances which contributed to the 2007-2008 financial turmoil (Ahrend, 2008, p. 21). The view that monetary policy played a key role in the commencement of housing booms and busts is according to Bordo (2012, p. 5) widely agreed upon by several economists. Moreover, the housing bubble does not instantaneously show up in the price of real estate because housing prices display a lag. Economic theory postulates a shift in demand only reflects in price after a period of time, giving the bubble an opportunity to be protracted, which is confirmed by Gilchrist (2002, p. 80) his data analysis.

The influence of interest rate on housing prices is not conclusive. Kuttner (2011, p. 19) states that only a small portion of the rise in American housing prices before 2007 can be explained by the low interest rate. However, he does conclude from econometric analysis on developed and emerging economies that countries with a higher short-term interest rate seem to have a smaller increase in asset prices (Kuttner, 2011, p. 26).

QE in the United States influenced the housing market in the following way: The operations of the Fed buying mortgage backed securities increased the supply of credit for potential house owners (Fratzscher, 2013, p. 17). Krishnamurty (2011, p. 3) also finds a drop in mortgage rates as a result of the Fed’s buying of mortgage backed securities. In addition, Krishnamurty (2011, p. 37) found in his regression analysis the earlier mentioned duration effect of the signaling channel to have had a particularly strong effect on long duration assets, such as real estate. The increased inflation expectations due to policy signaling from the central bank will lower real lending rates for borrowers suggesting stimulation of the housing market.

Concluding, countries where interest rates were set below Taylor Rule recommended policy often coped with housing market bubbles. Moreover, the results from empirical analysis suggest QE could increase housing prices thereby contributing to a housing bubble.

12

4. Can the ECB QE create an asset price bubble?

Looking at the previously explained effects of QE and the circumstances in which asset price bubbles emerge in the markets of interest we can now examine whether the ECB’s monetary policy actually can create such a bubble.

Firstly, the influence of interest rates and credit supply on asset price bubble formation will be scrutinized. The goal of maintaining price stability and preventing deflation from occurring will keep the interest rate at the current zero lower bound during the implementation of the QE program. In this manner could QE create an environment where bubbles can be brought into being. This because of lowered lending rates and provision of liquidity through QE is aimed at increasing lending to firms and households, thereby facilitating eased access to credit. The increase in credit access will depend on the creditworthiness of borrowers and the amount of risk lenders are willing to take. According to a bank survey conducted by the ECB (2015a, pp. 9-11) borrowing conditions for supplying loans to enterprises have eased, the rejection rate of loans to enterprises has decreased and the enterprises’ demand for loans is increasing since the implementation of QE in March 2015. One of the drivers behind these improvements is an increase in inter-bank competition, but moreover did the banks’ cost of lending and balance sheet constraints improve the credit supply which could be a result from ECB policy (2015a, p. 9). Household credit supply also improved since the implementation of QE due to a decrease in rejection rate, an increase in demand for housing loans even in spite of European banks tightening their credit standards for housing loans (ECB, 2015a, pp. 13-16). Banks have stated that their liquidity positions have improved as a result from the ECB’s purchasing program and will continue to improve in the coming months. These improvements of the banks’ capacity will, as stated by banks in the survey, be used to grant loans to both enterprises and future house owners (ECB, 2015a, p. 24-25). It can be concluded that the short term access to credit has indeed increased, but this will not inevitably lead to over-optimistic investors and asset price bubbles. Moreover, the fact that short-term credit supply has increased does not imply it will increase over a longer amount of time as well.

Secondly, the influence of central bank inflation policy on asset price bubbles will be examined. The probability of the occurrence of asset price bubbles during conventional monetary policy increases when the central bank credibly committed to stabilizing prices according to economic theory of the Austrian school (Bordo, 2012, p. 6). The ECB is using QE to signal long-term commitment of doing everything within their means to maintain price stability. Hereby stickiness of both inflation and interest rate can be assumed during the course of QE. These conditions could also contribute to bubble inflation.

13 Thirdly, as the literature in section four stated does expansionary monetary policy not inevitably lead to an asset price bubble, but the environment created by expansionary monetary policy can however foster a bubble. The ECB’s QE policy of buying of treasury bonds will provide the European economy with 1.1 trillion Euros worth of liquidity. As earlier mentioned the emergence of bubbles was often during a period of expansionary monetary policy. QE will create an accommodative monetary environment, therewith stimulating speculation which could result in increasing asset prices. Moreover, the portfolio rebalancing resulting from QE forces investors to undertake new risky investments, thereby boosting asset prices of these alternative investments. The two mentioned effects will affect the allocation of capital in such a way that the price increases will attract more investors to undertake these risks. Thereby they are creating a self-amplifying asset price increase, possibly with as a result an asset price bubble within that particular sector.

Recommendation for future research could be not to look at asset bubble formation within the region of QE implementation but outside of these borders. Fratzscher (2013, p. 4) argues that QE will lead to capital outflows from the QE-region which will direct themselves to emerging market economies. This in turn will lead to appreciative pressures on these emerging economies’ currencies, fiscal imbalances and credit growth, potentially resulting in asset price bubbles.

The circumstances created by QE would suggest that the economic environment favors the emergence of asset price bubbles. It cannot however be disclosed that there will in fact be a bubble due to a lack of being able to predict and observe in what market and when it will occur. In the next section we can look at the two earlier mentioned markets of interest and whether the circumstances are particularly favorable for the emergence of a bubble in these markets.

5. Can the ECB QE lead to an asset price bubble in the housing

market?

After examining the general impact of QE on asset bubble inflation we can now look at the impact on specific markets, starting with the housing market. The ECB QE policy focuses on purchasing government bonds. This provision of liquidity increases the supply of credit available and could be considered to have a positive influence on real estate prices, given that propensity to take up new loans by households will be positive. The debt-income ratio of Eurozone households is namely 95.66 percent, with extremes as for instance 214.15 percent in The Netherlands (Eurostat, 2015). The policy duration effect of QE is supposed to influence long dated assets strongly (Krishamurty, 2011, p. 37). Hence, it can be expected to also affect housing prices in an upward direction because houses can be

14 considered to be long duration assets. Moreover, risen inflation expectation will lead to a decrease in lending costs, therewith stimulating house ownership and rising property prices.

The development of housing prices in the Eurozone can be seen in Figure 1, which is generated with data from Eurostat (2016). In the graph the second quarter of 2013 serves as the base period. One can infer by looking at the development of housing prices that between the announcement of QE by the ECB in the first quarter of 2015 and the implementation during the second quarter of 2015 housing prices started to increase beyond levels in the years prior to QE. The response of housing prices on implementation of QE could mean that QE can contribute to a housing bubble by raising housing prices. Housing price developments in the five largest economies of the Eurozone show the same increase in housing prices in the beginning of 2015, however, housing price increase in Italy was more tempered compared to the other countries. These graphs can be found in Appendix A.

Figure 1. Development of housing prices in the Euro Area

Concluding, the extent to which credit will contribute to a bubble is uncertain. The policy duration effect however can stimulate bubble inflation. Housing prices developments in the Eurozone show an increase in prices between the announcement and implementation of QE, meaning QE could in fact give rise to an asset price bubble in the housing market.

0 .9 9 1 .0 0 1 .0 1 1 .0 2 1 .0 3 Eu ro a re a Q1 2013 Q3 2013 Q1 2014 Q3 2014 Q1 2015 Q3 2015 Time

15

6. Can ECB QE lead to an asset price bubble in the stock market?

We now examine the effects of QE on the stock market. Expansionary monetary policy could possibly give rise to a bubble within the stock market. The provided financial stability by the ECB will stimulate investor speculation, given their assumption that central banks or government will take measures to aid them when the economy will contract such as bailouts and easy credit access theorizes Kindleberger (2005, p. 204). This moral hazard problem exercised by these investors will therefore push up stock prices even higher. Moreover, monetary easing will induce portfolio rebalancing which will change the composition of relative long-term asset demand. Supply of government bonds will be lower and will therefore lead to increased demand for alternative long-term investments. This demand will push up the prices of these alternative assets. Additionally, a reduction in liquidity premia as a result of QE will reduce transaction costs thereby reducing costs associated with corporate bond trading. This will result in higher demand within the stock market in turn inducing an increase in stock prices. Lastly, the depreciation of the euro as a result of QE could induce higher exports, thereby pushing up stock prices of companies involved in trade outside of the Eurozone (Kurihara, 2006, p. 379). There is however not a particular technological development which could be designated to be the driving factor behind the next bubble, such as the internet and technological stocks during the dotcom bubble.

To look at the response of European stock prices to QE the trajectory of the stock index Eurostoxx 50, depicted in Figure 2, can be used. The Eurostoxx 50 is a stock index consisting the fifty largest and most liquid stocks in the European stock market (Yahoo Finance, 2016). In contrast to housing prices do European stocks seem not to increase after implementation of QE. In fact, after the implementation of QE in the second quarter of 2015 European stock prices have fallen. The effects of QE do not seem to have the expected effect on European stock prices in the short run. The individual Dutch, French and German stock indices depict a similar trajectory, as can be seen in the graphs of Appendix B.

16

Figure 2. Euro Stoxx 50 Index

Summing up, QE can be expected to stimulate investor speculation and portfolio rebalancing, thereby driving up stock prices. Moreover, reduced trading costs will stimulate demand in turn increasing prices. QE can induce a devaluation of the Euro, in turn increasing competitiveness with possibly increasing stock prices of firms operating in the export business. QE can therefore stimulate the inflation of an asset price bubble in the stock market. The development of European stock prices, however, has exhibited an opposite outcome since the implementation of QE in March 2015.

7. Conclusion

This thesis attempted to research whether the implementation of QE could lead to asset price bubbles. After conventional monetary policy failed to resolve problems such as impeding deflation and liquidity trap in the Eurozone the ECB implemented a large scale asset buying program called Quantitative Easing. Through purchasing government bonds the ECB supplied ample liquidity in order to curb these problems. QE affects the economy through four main transmission channels: The

2 8 0 0 3 0 0 0 3 2 0 0 3 4 0 0 3 6 0 0 3 8 0 0 Eu ro St o xx 5 0

Jan 2014 May 2014 October 2014 March 2015 August 2015 January 2016

17 portfolio balance channel, the liquidity premia channel, the bank lending channel and the policy signaling channel.

The emergence of asset price bubbles comes about as a result of investors’ optimism on future asset price developments. Monetary expansion can stimulate bubble inflation, especially through increased access to credit. Empirical research found expansionary monetary policy to increase stock prices in Japan, possibly contributing to stock market bubbles. Additionally, regression analysis has shown expansionary monetary policy in the United States to have had a significant effect on the stock market. Countries with bubbles in the housing market have often exhibited an interest rate below the Taylor Rule recommended rate. Furthermore, empirical analysis has shown QE could increase housing prices, thereby contributing a housing bubble.

The creation of an asset price bubble in the Eurozone as result of QE is no certainty. Credit supplied to enterprises and households has increase since the introduction of QE, but these are tentative results with no conclusive developments for the longer term. Moreover, expansionary monetary policy will stimulate speculation and portfolio rebalancing thereby driving up asset prices and creating a financial environment favorable for emergence of asset price bubbles. An initiation of a bubble in the Eurozone housing market depends on propensity of households to take on new debt. Housing prices in the Eurozone have risen after implementation of QE, meaning a housing bubble could be conceivable. Through portfolio rebalancing and liquidity premia reduction can QE give rise to an increase speculation and stock trade, thereby increasing stock prices. Moreover, depreciation of the euro can induce an additional increase in stock prices. However, European stock prices have dropped after the QE was initiated.

Concluding, QE has created an environment which can foster bubbles in both stock and housing market. However, there is a lack of certainty on whether a bubble will be created. QE can increase the probability of asset price bubble inflation, but no decisive answer can be given on whether one will actually come about.

18

Bibliography

Ahrend, R. B. (2008). Monetary Policy, Market Excesses and Financial Turmoil. OECD Economics

Department Working Papers, 597, 1-86.

Bean, C. (2004). Asset Prices, Financial Instability and Monetary Policy. The American Economic

Review, 94(2), 14-18.

Bernanke, B. a. (2000). Monetary Policy and Asset Price Volatility. Cambridge, MA: NBER.

Blanchard, O. (2000). Bubbles, Liquidity Traps and Monetary Policy: Comments of Jinushi et al, and on Bernanke. In R. M. Posen, Japan's Financial Crisis and its Parallels to US experience (pp. 1-14). Washington: Institute for International Economics.

Bordo, M. (2012). Does Expansionary Monetary Policy Cause Asset Price Booms; Some Historical and Empircial Evidence. NBER working Paper(19585), 1-22.

Borio, C. a. (2002). Asset Prices, Financial and Monetary Stability: Exploring the Nexus. BIS Working

Papers, 114, 1-39.

Brunnermeier, M. (2013). Bubbles, Financial Crises, and Systemic Risk. In M. H. George M.

Constantinides, Handbook of the Economics of Finance, Volume 2 (pp. 1-89). Amsterdam: Elsevier.

Christensen, J. a. (2012). The Response of Interest Rates to US and UK Quantitative Easing. the

Economic Journal, 122(11), F385-F414.

ECB. (2015a). The Euro Area Bank Lending Survey - Third Quarter of 2015. Frankfurt: ECB.

ECB. (2015b). Introductory statement to the press conference (with Q&A). ECB press conference 3

December 2015. Frankfurt am Main: ECB. Retrieved January 28, 2016, from

https://www.ecb.europa.eu/press/pressconf/2015/html/is151203.en.html ECB. (2016). Statistical Data Warehouse. Retrieved january 23, 2016, from ecb.europa.eu:

http://sdw.ecb.europa.eu/home.do?chart=t1.6

Eggertsson, G. a. (2003). The Zero Bound on Interest Rates and Optimal Monetary Policy. Brookings

Papers on Economic Activity, 1, 139-211.

Eurostat. (2015, September 28). Gross debt-to-income ratio of households. Retrieved january 20, 2016, from ec.europa.eu:

http://ec.europa.eu/eurostat/tgm/table.do?tab=table&init=1&language=en&pcode=tec0010 4&plugin=1

Eurostat. (2016, January 20). House Price Index - Quarterly data. Retrieved February 10, 2016, from Eurostat: http://appsso.eurostat.ec.europa.eu/nui/show.do

Fawley, W. a. (2013). Four Stories of Quantitative Easing. Federal Reserve Bank of St. Louis, 95(1), 51-88.

Fratzscher, M. M. (2013). On the International Spillovers of US Quantitative Easing. German Institute

for Economic Research Working Paper, 1-48.

Fujiwara, I. (2006). Evaluating Monetary Policy When Nominal Interest Rates Are Almost Zero.

19 Gali, J. a. (n.d.). The Effects of Monetary Policy on Stock Market Bubbles: Some Evidence. American

Economic Jounral: Macroeconoimcs, American Economic Association, 7(1), 233-257.

Gilchrist, S. a. (2002). Monetary Policy and Asset Prices. Journal of Monetary Economics, 49, 75-97. Joyce, M. (2011). The Financial Impact of Quantitative Easing in the United Kingdom. International

Journal of Central Banking, 7(3), 113-161.

Joyce, M. D. (2012). Quantitative Easing and Unconventional Monetary Policy - An Introduction. The

Economic Journal, 122(11), F271-F288.

Kimura, T. H. (2003). The Effects of the Increase in the Monetary Base on Japan's Economy at Zero Interest Rates: An Empirical Analysis. Bank for International Settlements Conference Series.

19, pp. 330-361. Bank of International Settlements.

Kindleberger, C. a. (2005). Manias, Panics, and Crashes 5th edition. Hoboken: John Wiley & Sons inc. . King, M. (2002). No Money, No Inflation: The Role of Money in the Economy. Bank of England

Quarterly Bulletin(3), 162-177.

Krishnamurty, A. a.-J. (2011). The Effects of Quantitative Easing on Interest Rates: Channels and Implications for Policy. Brookings Papers on Economic Activity Fall 2011 (pp. 1- 43). Cambridge: NBER.

Krugman, P. (1998). It's Baaack: Japan's Slump and the Return of the Liquidity Trap. Brookings Papers

on Economic Activity, 2, 221-237.

Kurihara, Y. (2006). The Relationship between Exchange Rate and Stock Prices during the Quantitative Easing Policy in Japan. International Journal of Business, 11(4), 376-386. Kuttner, K. (2011). Monetary Policy an Asset Price Volantility: Should We Refill the Bernanke-Gertler

Presciption? New Perspectives on Asset Price Bubbles: Theory, Evidence and Policy (pp. 1-35). Chicago: Loyola University.

Kuttner, K. (2011). Monetary Policy an Asset Price Volatility: Should We Refill the Bernanke-Gertler Presciption? New Perspectives on Asset Price Bubbles: Theory, Evidence and Policy (pp. 1-35). Chicago: Loyola University.

Nechio, F. (2011, June 13). Monetary Policy When One Size Does Not Fit All. Federal Reserve Bank of

San Francisco Economic Letter, p. 1. Retrieved January 23, 2016, from

http://www.frbsf.org/economic-research/publications/economic-letter/2011/june/monetary-policy-europe/

Okina, K. M. (2001). The Asset Price Bubble and Monetary Policy: Japan's Experience in the Late 1980's and the Lessons. Monetary and Economic Studies, 19(S1), 395-450.

Palley, T. (2011). Quantitative Easing: a Keynesian Critique. Investigación Económica, 70(277), 69-86. Roubini, N. a. (2010). Crisis Economics: A Crash Course in the Future of Finance. New York: The

Penguin Press.

Smaghi, L. (2009). Conventional and Unconventional Monetary Policy. International Center for

Monetary and Banking Studies Keynote Lecture (pp. 1-14). Geneva: ICMB.

Ugai, H. (2007). Effect of the Quantitative Easing Policy: A Survey of Empirical Analyses. Tokyo: Bank of Japan.

20 YahooFinance. (2016, February 11). STOXX50E. Retrieved February 11, 2016, from Yahoo! Finance:

http://finance.yahoo.com/q/hp?s=%5ESTOXX50E&a=00&b=1&c=2014&d=01&e=10&f=2016 &g=m

Yamaguchi, Y. (2002). Monetary Policy in a Changing Economic Environment. New Challenges for

21

Appendix A

Figure 3: Development of housing prices in Germany

Figure 4. Development of housing prices in The Netherlands

1 .0 0 1 .0 2 1 .0 4 1 .0 6 1 .0 8 G e rma n y Q1 2013 Q3 2013 Q1 2014 Q2 2014 Q1 2015 Q3 2015 Time 1 .0 0 1 .0 2 1 .0 4 1 .0 6 N e th e rl a n d s Q1 2013 Q3 2013 Q1 2014 Q3 2014 Q1 2015 Q3 2015 Time

22

Figure 5. Development of housing prices in Italy

Figure 6. Development of housing prices in France

0 .9 2 0 .9 4 0 .9 6 0 .9 8 1 .0 0 It a ly Q3 2015 Q1 2015 Q3 2014 Q1 2014 Q3 2013 Q1 2013 Time 0 .9 6 0 .9 7 0 .9 8 0 .9 9 1 .0 0 1 .0 1 F ra n ce Q1 2013 Q3 2013 Q1 2014 Q3 2014 Q1 2015 Q3 2015 Time

23

Figure 7. Development of housing prices Spain

0 .9 8 1 .0 0 1 .0 2 1 .0 4 1 .0 6 Sp a in Q1 2013 Q3 2013 Q1 2014 Q3 2014 Q1 2015 Q3 2015 Time

24

Appendix B

Figure 8. Dutch stock index development AEX

Figure 9. French stock index CAC40 development

4 0 0 4 5 0 5 0 0 AEX N L

Jan 2014 May 2014 Sepember 2014 March 2015 August 2015 January 2016

Time 4 2 0 0 4 4 0 0 4 6 0 0 4 8 0 0 5 0 0 0 C AC 4 0 F R

January 2014 May 2014 October 2014 March 2015 August 2015 January 2016

25

Figure 10. German stock index DAX development

9 0 0 0 1 0 0 0 0 1 1 0 0 0 1 2 0 0 0 D AX D E

January 2014 May 2014 October 2014 March 2015 August 2015 January 2016