Influence of changes in the ECB’s key interest rate on stock
prices of European banks
Name: Floris Pigeaud
University: University of Amsterdam
Faculty: Faculty of Economics and Business
Date: 28/06/2016
Supervisor: D. van Dijk
Bachelor: Economics and Finance
Number of EC: 12
Statement of Originality
This document is written by Student Floris Pigeaud who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text
and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract:
In this thesis the relationship between the ECB’s key interest rate and stock returns of European banks is examined on days of policy announcements. Using hand collected data about the
announcements, the interest rate change is separated in an expected and an unexpected part using the three-month EURIBOR future data. First, regressions are performed to see whether the unexpected and the expected part of the interest rate change influences stock prices and if this influence alters during the crisis. The results of the regressions over the whole sample, show an insignificant relationship with the unexpected part of the interest rate change. However, a negative significant relationship has been found during the crisis. On the contrary, a positive relationship has been found for the expected part, which becomes larger during the crisis. Second, regressions were performed to see if banks react differently to interest rate changes. The results that banks react differently, is suggestive. Last, regressions were performed using ratios to see whether these ratios influence the effect of interest rate changes on stock returns. The results show that banks with a higher solvency ratio and a higher return on average assets react less to the unexpected change of the interest rate.
Table of Contents
1.
Introduction ... 5
2. LITERATURE REVIEW ... 8
2.1MODELS ... 8
2.2THEORIES ... 9
2.3KEY INTEREST RATE AND THE ANNOUNCEMENTS OF THE KEY INTEREST RATE ... 11
2.4PAST RESEARCH ... 11
3. METHODOLOGY ... 15
4. DATA ... 19
5. RESULTS ... 24
6. CONCLUSION AND DISCUSSION ... 30
7. REFERENCE LIST ... 32
1. Introduction
Price stability is the main objective of the European Central Bank. Besides this, the ECB also cares about full employment and stable economic growth. Although the ECB does not have the ability to influence those objectives directly, it is able to influence them indirectly by using conventional and unconventional tools. The most direct effect of those tools (e.g. changing the key interest rate) is on the financial markets, of which the market for equities is the most important. By influencing the financial markets, policymakers try to change the economic behavior of the market participants in such a way that it will help them to achieve their ultimate goals.
As mentioned above, one goal of the ECB is economic growth. This can be influenced by different factors in the economy. One example of those factors is asset prices and returns (e.g. equity prices/stock prices). Those prices and returns affect the economy in different ways. It can change for example the sentiment of economic agents. Chen (2012) argues that that there are two channels. The wealth channel, in which people have greater sentiment because they are wealthier due to stock price increases. The other channel is the leader indicator effect. People see
movements in stock prices as an indicator of future economic activity and potential labour income growth. These channels have positive effects on consumer spending and thus on economic growth. Furthermore, Poterba (2000) concluded that rising stock market surely contributes to rising consumer spending. Moreover, Fariborz, Rebel and Qionging (2008) concluded that a country’s financial sector and the stock market returns affect future economic growth. They also find that bank stock returns are positively correlated with future economic growth.
The ECB(ECB 2010a) also says that it uses the stock market development in the monetary policy-making process, because it shows the development of real economy and the uncertainty of future expectations. Among others things, this is one of the reasons why researchers have investigated the transmission mechanism of monetary policy. As described above, it is therefore crucial to have a good insight in the relationship between stock returns and monetary policy, because it can influence the behavior of market participants and can help the ECB to achieve its goals. It is thus important to understand the transmission mechanism between monetary policy and asset prices (i.e. equity prices).
The results of the scientific research done so far is especially about the overall effect of the FED on the stock market in the United States. Aharony, Saunders and Swary (1986) find that bank stocks exhibit significantly abnormal returns during the announcement week of 1979 FED policy change. Moreover, they find that there is a negative relationship between bank stock returns and unexpected changes in the interest rate. Flannery and James (1984) find that bank stock returns are highly correlated with interest rate changes. Thorbecke (1997) finds that expansionary monetary policy by the FED increases ex post stock returns. Furthermore, he says that there is a significant (at 10%) negative relationship between conventional policy changes and stock prices of banks. Bernanke and Kuttner (2005) look at the unexpected interest rate change of the FFR and they argue that is has a negative relationship with stock returns. Similar results have been found for the U.S. by Gurkaynak, Sack and Swanson (2002), Rigobon and Sack (2004) and Ehrmann and Fratzscher (2004).
Some researchers have looked at the effect on asset prices of changes in monetary policy that take place on particular dates. Rigobon and Sack (2004) look at the effect on days of the FOCM meetings and show that increases in short-term interest rate have negative influence on stock prices. Furthermore, Pearce and Roley (1985) investigate the daily response of stock prices to announcements of the FED. They find a significant influence on stock prices.
There has also been done research about the effect of the policy changes on the stock markets in Europe. Bredin, Hyde and O Reilly (2005) investigate the stock market response to international monetary policy changes in the UK and Germany. Bohl et al (2008) analyse the reaction of European stock market returns to monetary policy. In both articles, there was a negative relationship between interest rate changes and stock returns. Haitsma, Unalmis and de Haan (2015) look at the impact of ECB’s conventional and unconventional monetary policy on stock markets. They make a difference between expected and unexpected changes in the key interest rate to determine this relationship.
Some researchers looked at the difference responses between sectors of the economy to monetary policy changes. Bernanke and Kuttner (2005), Peersman and Smets (2005), Ehrmann (2003), Ehrmann and Fratzer (2004), Kholodilin, Montagnoli, Napolitano and Siliverstovs(2009) and Bredin et al (2005) all conclude that sectors in the economy react differently to monetary policy changes depending on their characteristics.
Following the ECB, changes in the ECB’s key interest rate have direct impact on the sector banks. It influences the lending and deposit rates of the banks, by changing its key interest rate. Knowing that this relationship exists, it is the question how banks are influenced by certain policy changes. A measure of the level of influence could be the stock price/returns.
Ricci (2014) examines the effect of announcements of the ECB on stock returns of large European banks during the crisis. He finds that contractionary policy has a negative effect on bank stock returns.However, he does not look at the differences between banks within the sector. So the question remains how banks within their sector react differently. This is the main topic of this thesis. Are banks reacting differently to monetary policy and why? So the main question is: What is the effect of changes in the ECB’s key interest rate on the stock prices of the large European banks?
This thesis is an empirical study to investigate the relationship between the conventional monetary policy, i.e. decisions in the key interest rate, and the stock returns of European banks on the announcement day of future monetary policy. I look at the response on the announcement day, because on that day new information was transmitted to the market. However, the response of equity prices is complicated, because it is unlikely that equity prices respond to news that was already anticipated by the market. If the news of the ECB on the announcement day was not anticipated, markets would immediately react. It is thus important to distinguish between expected and unexpected policy actions to see the effects. The method suggested by Kuttner (2001) is used to do this. The idea behind this is that future prices reflect market expectations for future policy rates. A change in the future price as a response to monetary policy changes, means that the change was not expected by the market participants. Bernoth and von Haagen (2004) show that the three-month EURIBOR future rates are a reliable predictor for the policy rates of the ECB. These future rates will therefore be used to determine the expected and unexpected parts of policy changes.
After the determination of the expected and unexpected changes, regressions will be performed. The data is hand collected from the website of the ECB. April the 4th 1999 will be taken as the first announcement date. The last announcement date is 10 March 2016.
First there is a literature review in which the theory about the relationship between stock returns and the interest rate and previous studies are discussed. Second the methodology is
discussed. Third, the data will be discussed. Last the results are presented and using these results there will be a conclusion and discussion.
2. Literature review
First of all, the theory about the relationship between interest rates and stock prices is discussed. Secondly, the past literature is discussed.
2.1 Models
There are several theories about stock pricing. Two of them are the Discounted Free Cash Flow model and the Dividend Discount model (Berk & DeMarzo, 2014). The value of a stock, according to the Law of one price, can be calculated by determining the expected cash flow an investor will receive from owning that specific stock. The Dividend-Discount model sees the dividends of the security as cash flows and uses them to come up with a price. It values stock by calculating the present value of all expected future dividends. This can be done with the
following formula:
𝐷𝑖𝑣$
(1 + 𝑟))$ +
$,-Where 𝐷𝑖𝑣$ is the dividend of period n. And 𝑟) is used as a discount factor, because one cannot be sure about the predictions of the future dividends, what implicates that the cash flows are risky. Due to this fact, instead of the risk free rate, the equity cost of capital is used as discount factor. The equity cost of capital can be calculated using CAPM.
Another model to value stocks is the Discounted Free Cash flow model. This model uses the free cash flow of a firm to determine its share price. The free cash flow is the amount of cash a firm generates before any payments to debt or equity holders are made:
Free Cash Flow = 𝐸𝐵𝐼𝑇 ∗ 1 − 𝜏 − Net Investment − Increases in Net Working Capital
To determine the share price, the firm value must be calculated by discounting the FCF.
𝑉J = 𝐹𝐶𝐹 -1 + 𝑟NOPP + 𝐹𝐶𝐹Q (1 + 𝑟NOPP)+ ⋯ + 𝐹𝐶𝐹$+ 𝑉S (1 + 𝑟NOPP)S 𝑃J = ( 𝑉J + 𝐶𝑎𝑠ℎJ− 𝐷𝑒𝑏𝑡J ) 𝑆ℎ𝑎𝑟𝑒𝑠 𝑂𝑢𝑡𝑠𝑡𝑎𝑛𝑑𝑖𝑛𝑔J
A key difference between the Dividend-discount model and the Discounted Free Cash Flow model is the discount rate. The first model uses the equity cost of capital, because cash flows to equity holders were considered. The second model considers cash flows to both debt and equity holders, and thus requires us to use the weighted average cost of capital. This is the average cost of capital a firm must pay to all investors.
2.2 Theories
In this section theory about the relationship between interest rates and stocks will be discussed.
Following to Palmon and Yaari (1981),there are two theories which try to explain the effect of interest rate changes on stock returns. Both theories use the Fisher Equation to make statements about the effect (Fisher (1930)). The Fisher Equation is as follows:
Nominal interest rate = real interest rate + inflation
The first theory, is the Expected Real Interest Rate theory. This states that the stock prices are affected by the real component of nominal interest rates. The real interest rate is the interest rate that an investor gets after correcting for inflation. It influences stock prices in two ways: directly and indirectly. It has a direct effect on the discount rate, because when the nominal interest rate increases, real interest rates could go up too. When the real interest rate increases, the real
discount rate at which future cash flows are discounted, is also expected to increase. Considering the valuation models, this affects the stock price negatively.
The indirect effect of the real interest rates is due to the negative relationship between real output and higher real interest rates. This relationship states that future free cash flows drop, when the real interest rate rises. According to the Discounted Cash Flow model, stock prices should decrease. However, there are scientists that do not agree with the Expected Real Interest
Rate theory. Bernake and Kuttner (2003) conclude that the reaction of equity prices to monetary policy is, for most part, not directly attributable to the reaction of real interest rates due to policy changes.
The second hypothesis is the Expected inflation theory. Looking at the Fisher equation, the inflation premium in nominal interest rates can increase, when nominal interest rates increase. This theory states that due to the inflation premium, the after tax real dividends decrease, which decreases stock prices according to the Dividend Discount model. This inflation premium is caused by a progressive tax system. Following Feldstein (1982), the tax system is based on nominal income, not on real income. The interaction between this system and inflation, cause tax rates and thus tax liabilities to increase when inflation increases. This can decrease the real dividends and thus the stock prices.
He also argues that an increase in the rate of inflation raises the effective tax rate on equity earnings relative to tax rate on other type of investment income. Individuals and financial institutions will only hold stock for a lower price as a form of compensation. So this could be an explanation too. Other researchers who investigated the relationship between inflation and stock returns, were Fama and Schwert (1977). They found a negative relationship between inflation and stock returns, when monetary authority conducts a counter-cyclical monetary policy. But when it conducts a pro-cyclical monetary policy, there is a positive relationship between interest rates.
Following Bernanke and Kuttner (2005) changes in policy rates may influence stock prices through: the expected future cash flows, the expected future risk-free rate used to discount the cash flows and the risk premium. The impact on future free cash flows is strongly influenced by firm specifics, because Ehrmnn and Fratzscher (2004) find, studying the impact of the Federal Reserve monetary policy on the portfolio comprising the S&P500 over 1996-2003, that firms smaller in size, with low cash flows, low debt to capital ratios, poor credit ratings, high Tobin’s q and high price-earnings ratio are more affected than others by monetary policy.
The ECB (2008) suggests that, focusing on the risk premium component, that monetary policy exerts an immediate reaction on the perceived riskiness of certain assets and the risk compensation required by investors. The perceived riskiness could be caused by an increase of funding cost and the weakening of firms’ balance sheet. Another cause of the risk premium following the ECB, could be the change in the risk appetite of the markets’ participants.
2.3 Key interest rate and the announcements of the key interest rate
It is also of crucial importance that one knows what the key interest rate is and how it is determined. That is what will be discussed in this section.
The Governing Council of the European Central Bank sets the key interest rates for euro area. Three interest rates are dependent on this interest rate: The rate of the Deposit Facility (DF), which is the rate banks get when they make overnight deposits with the Euro system. The
interest rate on the Main Refinancing Operations (MRO), which provides liquidity to the banking system. And the rate on the Marginal Lending Facility (MLF), which offers overnight credit to banks of the Euro system.
These rates are all determined on the first Thursday of each month. The Governing Council makes decisions about the level of key interest rate on that day. On the same day the ECB publishes this decision on 13:45 CET. After about 45 minutes, at 14:30 CET, the president and the vice-president hold a binary press conference. First, they go through the arguments of their decision. Hereafter, the audience has the right to ask some questions. (De Haan (2008)).
2.4 Past research
In this section past research will be discussed. First the research that has been done for the U.S. will be discussed. Second, the research that has been done for Europe will be discussed.
In the past, researchers have concluded that there is a correlation between interest rates and stock returns. Choi, Elyasiani and Kopecky (1982) say that the interest rate is an important variable for the valuation of common stock, especially for financial institutions. This is because the
dependence of the returns and cost of these financial institutions. Furthermore, Hardouvelis (1987) says that stock prices primarily respond to announcements of monetary policy. Stocks that are the most sensitive, are the ones of financial companies, because of the relation with the interest rate and the direct influence of changes in the key interest rate on commercial banks (ECB).
U.S.
The direction has been further investigated by researchers, which is especially done for the US. Thorbecke (1997) investigates the relationship between stock returns and the monetary policy of the Fed. He concludes that expansionary policy increases ex-post stock returns.
Furthermore, Rigobon and Sack (2002) demonstrated that asset prices and market interest rates response to changes in monetary policy shocks of the Fed that take place on particular dates. They argue that increases in the short-term interest rate have negative influence on stock prices and a significant positive effect on market interest rates, with the largest effect on the shortest maturities. Also Ehrmann and Fratzscher (2004) look at the effect U.S. monetary policy on stock markets. As a result of their paper, they found that monetary policy affects stocks in a
heterogeneous way. Studying the impact of the Federal Reserve monetary policy on the portfolio comprising the S&P500 over 1996-2003, they find that firms smaller in size, with low cash flows, low debt to capital ratios, poor credit ratings, high Tobin’s q and high price-earnings ratio are more affected than others by monetary policy.
Aharony, Saunders and Swary (1986) find that bank stocks exhibit significantly abnormal returns during the announcement week of 1979 FED policy change. Moreover, they find that there is a negative relationship between bank stock returns and unexpected changes in the interest rate.
There are also papers which make a distinction between expected and unexpected policy changes of the Fed. Pearce and Roley (1984)investigate the daily response of stock prices to
announcements of the Fed about money supply, inflation, real economic activity and the discount rate. Surveys about market participants’ expectations, except for discount rate, are used to
determine the unexpected part of the announcement and to test the hypothesis that only the unexpected part of the announcement influences stock prices. Empirical results support this hypothesis. They find a significant influence on stock prices.
Bomfin (2003) makes also use of expectations. He finds that near-term revisions in policy expectations have a negative and significant effect on stock prices in the U.S. The daily stock returns are reduced by 0.04% point, for each basis-point increase in the expected average daily value of the FFR in the following month. Moreover, Bernanke and Kuttner (2005) try to explain the stock market’s reaction to Federal Reserve policy. They look at the reaction of the unexpected
part of the FFR, and conclude a 25 basis point cut in this rate is associated with a 1 percent increase in the level of stock prices. On the contrary, they find that expected changes have a positive significant effect. Also Kontonikos, Macdonald and Saggu (2013) examine the impact of FFR surprises on stock returns in the United states over the period 1989-2009. They conclude that before the crisis, stock prices increased as response to unexpected FFR cuts. Although during the crisis market participants did not react positively to FFR cuts. They used the method
suggested by Kuttner (2001), which will be used in this thesis too.
Europe
The research about the effect of interest rate change has also been done for Europe. Bredin, Hyde and O Reilly (2009) investigate the stock market response to international
monetary policy changes in the UK and Germany. They specifically investigate the influence of the (un)expected changes in the policy rates of both countries on the aggregate and sectoral stock returns. They find that the UK monetary policy surprises have a negative effect on the stock returns. The findings for the German surprises are insignificant, so no statements are made for this country. Also Bohl, Siklos and Sondermann (2008) analyse the reaction of European stock market returns to monetary policy actions. They find a negative and significant relationship between unexpected ECB policy changes and European stock market performance. Another conclusion in this paper, is that policy changes are well expected by the market participants, so the ECB communicates successfully.
Fiordelisi, Gallopo and Ricci (2014) investigate the effect of the central banks in the Euro Area, Japan, the US, the UK and Switzerland between 2007 and 2012. They show that
expansionary monetary measures have a positive effect on the stock market, while contractionary monetary measures have a negative effect on the stock prices. However, they find that interest rate cuts, as expansionary measures, do not have a significant positive reaction to the stock market. They only find that the central bank’s decisions to change the interest rates have a
statistically market reaction only in cases of unchanging rates or increasing interest rates. Another paper that examines the effect in Europe is the paper of Haitsma, Unalmis and de Haan (2015). They look at the impact of ECB’s conventional and unconventional monetary policies on stock markets between 1999-2015 using an event study. They make a difference in the effect of expected changes and unexpected changes, using the method suggested by Kuttner (2001). This
method, which will also be used in this thesis, is discussed in the next section. Kuttner examines the impact of changes in the ECB policies on returns of several portfolios sorted on firm
characteristics and past performance. Their results do not provide evidence for the interest rate channel. Although they find that stocks of different sectors respond differently to monetary policy surprises, these differences are hardly linked to differences with respect to their interest rate sensitivity. Besides the conclusion of this paper, there has also been done some more research to look at the effect on different sectors in the economy.
Bernanke and Kuttner (2004) find that the effect of monetary policy surprises tends to differ across sectors. High-tech, telecom and durable goods react quite strongly to unanticipated Fed policies. Other sectors, like the energy sector and utilities, seem not to respond significantly to monetary policy. Ehrmann (2003) finds the same results. He argues that stock prices of consumer goods, telecommunications, technology and finance firms seem most sensitive to monetary policy surprises. Moreover, Peersman and Smets (2005) argue that the reaction to monetary policy surprises should differ across sectors depending on the interest-elasticity of the demand for their products. They also argue that the credit channel implies that when sectors are more dependent on bank funding, that they are more affected by monetary surprises. Kholodilin, Montagnoli, Napolitano and Siliverstovs (2009) find that, depending on the sectors an increase of the interest rate with 25 basis points results in a decrease in the stock market between 0.3% and 2% on the day he monetary policy is publically announced. On the contrary Bredin et al (2009) argue that sectoral indices do not react significantly to expected and unexpected policy rates of the German Bundesbank and the ECB.
Taking the previous literature into account, the stocks of financial institutions are part of the sector that is most influenced by changes in interest rate. There has been done some research about the effect on bank stock returns in particular. Ricci (2014) examines the effect of
announcements of the ECB during the crisis on the stock of large European banks. He finds that contractionary policy has a negative effect on bank stock returns. Furthermore, Grammatikos et al (2013) investigate the effect of both conventional and unconventional monetary policy decisions on stock prices of globally systematically important financial institutions. They find that US policy announcements had a stronger impact on the European and US banking industry than the impact of announcements of the ECB. Aharony, Saunders and Swary (1986) find that bank stocks
exhibit significantly abnormal returns during the announcement week of 1979 FED policy
change. Flannery and James (1984) find that bank stock returns are highly correlated with interest rate changes.
As far as I know, not a lot of research is done to look at the differences of the effects within one sector. This thesis tries to explain to what extent bank stock returns react to expected and unexpected policy changes and if banks react asymmetrically to those changes.
3. Methodology
In this section the method will be discussed. First the method of distinguishing the expected and unexpected part of the interest rate change will be discussed. Secondly, the regressions will be discussed. The abbreviations with their corresponding explanations can be found in table 1 of the appendix.
As already stated, it is important to make a difference in expected and unexpected monetary policy changes, because Bredin et al (2009) conclude that empirical work that fails to distinguish between expected and unexpected interest rate changes is likely to lead to biased results due to an errors in variables problem. In particular, if we assume that the Efficient Market Hypothesis holds. The Efficient Market Hypothesis states that all relevant information is
incorporated in stock prices (Body, Kane and Markus, 2011). If some new information is available, the market will incorporate this and the stock price will change accordingly. Anticipated news is already reflected in the stock price, otherwise there would have been an arbitrage opportunity and the markets would not have been efficient. This means that on the announcement day of a monetary policy change, the asset price only reacts to the unexpected element.
On the contrary, the anticipated changes should not affect the asset prices on the
announcement day, because this information should already have been priced in when the market participants came aware of this information prior to the announcement date. If this was not the case, arbitrage opportunities would exist and the market would be inefficient. However, this does not mean that stock prices do not react to anticipated changes in monetary policy. Equity prices will also react to revisions of expectations of future policy. The unexpected changes can be merely seen as exogenous shocks that affect stock prices (Benanke and Kuttner 2005). So if a
distinction between anticipated and unanticipated changes is made, one can be more precise about the effect the changes have on stock prices.
Following Bredin et al (2009) the most popular method to derive the expected and unexpected changes is by looking at future markets data. They say that future markets have dramatically increased in liquidity and in the range of instruments on offer. Therefore, one can easily derive a measure of the expectation on a continual basis. The method suggested by Kuttner (2001) is used to distinguish between expected and unexpected policy changes. The idea behind this is that future prices reflect market expectations of future policy rates. A change in future prices as a response to policy changes, means that the change was not expected by the market. Bernoth and von Haagen (2004) show that continuous three-month EURIBOR future rates are a reliable predictor for the policy rates of the ECB. They analyze the efficiency of the three-month EURIBOR interest rate futures market. Their results show that the EURIBOR future rates with a forecast horizon of up to four months are unbiased and informationally efficient predictors of future spot rates. This is the reason why the three month continuous EURIBOR future rate is used to determine the expected and unexpected part.
The expected and unexpected parts of the interest rate change is determined as follows. Following Bredin et al (2009), the discrepancy between the daily future rates, to determine the unexpected part of the interest rate change, must be used:
Δ𝑟bc = 𝑓
e,b− 𝑓
e,bg-Δ 𝑟bh = Δ𝑟
b− Δ𝑟bc
Where Δ𝑟bc represents the unexpected policy change at day t, 𝑓
e,b− 𝑓e,bg- represent the
discrepancy between the future spot rate at day t, and the future rate the day before the
announcement day, t-1. This indicates the one-day change in the three-month EURIBOR future price. Using the daily settlement prices of the futures, the future rates can be determined. This is done by subtracting the price from a hundred, which gives us the implied expectation for the policy rate. The expected part, indicated as Δ 𝑟bh, is represented by the difference between the
actual interest rate change Δ𝑟b and the unexpected interest rate change Δ𝑟bc, which is calculated
Sometimes the interest rates of the ECB are changed differently. This means that the expected change in some cases is different. This problem is solved by making three different expected interest rate changes: the expected change in the Deposit Facility rate, the expected change in the Main Refinancing rate and the expected change in the Marginal Lending rate. The data about the announcements with the interest changes and the data about the expected and unexpected parts can be found in table 2 and 3 respectively in the appendix.
After the determination of the expected and unexpected interest rate changes for each
announcement day, the following regression will be used to look at the effect of these changes in the interest rate on stock returns:
𝑅b= 𝛼 + 𝛽-Δ𝑟bl+ 𝛽QΔ𝑟bl𝐶𝑟𝑖𝑠𝑖𝑠 + 𝛽mΔ𝑟bh+ 𝛽nΔ𝑟bh𝐶𝑟𝑖𝑠𝑖𝑠 + 𝛽o𝐶𝑟𝑖𝑠𝑖𝑠 + 𝜀b
Where 𝑅b is the return of a specific stock at day t and 𝛼 is a constant. Δ𝑟bh represents the expected
interest rate change and Δ𝑟bl represents the unexpected interest rate change. Crisis represents a
dummy variable for the crisis (1 during crisis, 0 otherwise). This variable is added to see if reaction to monetary policy changes differ during crises. 𝛽- indicates the effect of the
unexpected monetary policy on stock returns. 𝛽Q represents the effect of the interaction variable of the unexpected interest rate change and the crisis. This coefficient shows if the reaction to interest rate changes differs during the crisis. 𝛽m represents the effect of the expected monetary policy on stock returns. 𝛽n represents the effect of the expected interest rate change during the crisis. 𝛽o represents the effect of the crisis on stock returns. 𝜀b is the error term and contains al the other effects on stock returns.
Haitsma, Unalmis and de Haan (2015) define the ECB’s announcement of the first unconventional tool as the start of the crisis. This was on 22/08/2007. This will also be used in this thesis. The period that is used to look at the effect is from 08/04/1999 until 10/03/1016.
First, three regressions will be performed for whole the sample of banks. They differ in use of the expected interest rate change. The results are used to determine the effect of the sector banks. It also determines if the effect is different during the crisis. Hereafter, the regressions of the individual banks will be performed using only the expected interest rate change of the Deposit Facility. These regressions will tell us if one can determine the effect of the interest rate changes
on individual bank stock returns. If the coefficients of these regressions are almost all significant, then one can try to make inferences about the reason of the difference in reaction. The regression that will be performed for the individual banks, is the same as the regression that is used to determine the effect using the whole sample.
Another goal of this thesis is to determine if banks react differently to interest rate changes and if so, why do they react differently? Using four ratios that were available, regressions are performed. The following ratios are used: The solvency ratio, the Net Interest Margin, the Return On Average Assets and the Return On Average Equity. The solvency ratio is a measure of the ability of a company to meet its long term debt. The Net Interest Margin is a measure of the difference between the interest income generated by the bank and the amount of interest paid out to its lenders, relative to the amount of their interest earning assets. The Return On Average Assets, is the net income divided by the average total assets. The Return On Average Equity, is the net income divided by the average stockholders’ equity outstanding.
The formulas of the ratios are as follows:
𝑆𝑜𝑙𝑣𝑒𝑛𝑐𝑦 𝑟𝑎𝑡𝑖𝑜 = 𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒 + 𝐷𝑒𝑝𝑟𝑒𝑐𝑖𝑎𝑡𝑖𝑜𝑛 𝑆ℎ𝑜𝑟𝑡 𝑡𝑒𝑟𝑚 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 + 𝐿𝑜𝑛𝑔 𝑡𝑒𝑟𝑚 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑁𝑒𝑡 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝑀𝑎𝑟𝑔𝑖𝑛 = 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝑅𝑎𝑡𝑒 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑑 − 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝐸𝑥𝑝𝑒𝑛𝑠𝑒𝑠 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝐸𝑎𝑟𝑛𝑖𝑛𝑔 𝐴𝑠𝑠𝑒𝑡𝑠 𝑅𝑒𝑡𝑢𝑟𝑛 𝑂𝑛 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑠𝑠𝑒𝑡𝑠 = 𝑁𝑒𝑡 𝐼𝑛𝑐𝑜𝑚𝑒 𝑇𝑜𝑡𝑎𝑙 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑠𝑠𝑒𝑡𝑠 𝑅𝑒𝑡𝑢𝑟𝑛 𝑂𝑛 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐸𝑞𝑢𝑖𝑡𝑦 = 𝑁𝑒𝑡 𝐼𝑛𝑐𝑜𝑚𝑒 𝑇𝑜𝑡𝑎𝑙 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐸𝑞𝑢𝑖𝑡𝑦
The following regressions are performed to test whether the ratios influence the reaction to the expected and unexpected part of the interest rate changes:
𝑅b = 𝛼 + 𝛽-Δ𝑟bl+ 𝛽 QΔ𝑟bh+ 𝛽m𝑆𝑅𝑡+ 𝛽nΔ𝑟bl𝑆𝑅𝑡+ 𝛽oΔ𝑟bh𝑆𝑅𝑡+ 𝜀b 𝑅b = 𝛼 + 𝛽-Δ𝑟bl+ 𝛽 QΔ𝑟bh+ 𝛽m𝑁𝐼𝑀𝑡+ 𝛽nΔ𝑟bl𝑁𝐼𝑀𝑡+ 𝛽oΔ𝑟bh𝑁𝐼𝑀𝑡+ 𝜀b 𝑅b = 𝛼 + 𝛽-Δ𝑟bl+ 𝛽 QΔ𝑟bh+ 𝛽m𝑅𝑂𝐴𝐴𝑡+ 𝛽nΔ𝑟bl𝑅𝑂𝐴𝐴𝑡+ 𝛽oΔ𝑟bh𝑅𝑂𝐴𝐴𝑡+ 𝜀b 𝑅b = 𝛼 + 𝛽-Δ𝑟bl+ 𝛽 QΔ𝑟bh+ 𝛽m𝑅𝑂𝐴𝐸𝑡+ 𝛽nΔ𝑟bl𝑅𝑂𝐴𝐸𝑡+ 𝛽oΔ𝑟bh𝑅𝑂𝐴𝐸𝑡+ 𝜀b
𝛽- indicates the influence of unexpected monetary policy on stock returns.𝛽Q represents the effect of expected monetary policy on stock returns. 𝛽m represents the relationship between the ratio and the stock returns. 𝛽n represents the interaction coefficient, which gives the effect between the ratio and the unexpected part of the policy change.𝛽o is the interaction coefficient, which indicates the effect between the ratio and the expected part of the interest rate change. If the last two coefficients are significant, then one can say that there is a relationship between the ratios and the reaction to interest rate changes.
4. Data
In this section the data will be described. First the sample of banks will be described. Secondly there are summary statistics given for the daily stock returns of the banks and the expected and unexpected interest rate changes. Third, the key interest rate changes with their corresponding announcement dates are discussed. At last, the ratios will be discussed.
In the appendix, table 4 consists of the sample of 31 European banks. They are retrieved from DataStream using the following criteria: The bank is based in a member country of the European Monetary Union and the bank is listed on the stock exchange. For some banks, there was no stock data available for our whole sample period.This is the reason why the first date of the stock returns is presented too. The stock returns of the European banks and the future data is also retrieved from DataStream. The daily stock returns were calculated using the retrieved data. The
summary statistics are presented in table 5. This table also summarizes the data for the
unexpected part, the expected part using the DF rate, the expected part using the MRO rate and the expected part using the MLF rate. These values will be discussed next.
The mean of the daily stock returns is 0.0136% with a standard deviation of 2.866%. The minimum value, is a loss of 58.62% and the maximum value, is an increase of 49.91%.
The announcement dates with the corresponding announcements can be seen in table 2 in the appendix. The change in the key interest rate is also calculated for each announcement date. These changes can also be seen in table 2. This table contains the announcement data and the changes of the rates of the Deposit facility (DF), the Main Refinancing Operations (MRO) and the Marginal Lending Facility (MLF). The changes of the key interest rate are divided into three groups, because sometimes the interest rates are changed differently. As one can see in the table, there is a fixed and a variable MRO. On 8 June 2000 the ECB announced that, starting from 28 June, the main refinancing operations of the Euro system would be conducted as variable rate tenders. The minimum bid rate refers to the minimum interest rate at which counterparties may place their bids. This is why the interest rate of the Main Refinancing Operations is divided into a fixed column and a variable one.
The unexpected and the expected interest rate changes can be found in table 3. These parts are calculated using the method that is described in the methodology. In some cases, there was no change in one of the three interest rates, so no value could be calculated. Table 5 summarizes the separated parts of the interest rate changes.
Table 5
Variable Mean Std. Dev. Min Max
Daily Stock Returns (%)! 0.000136 0.02866 -0.5862 0.4991 %"""Δ$%&! 0.0017 0.05698 -0.0700 0.3150 %"""Δ"$%'"()! -0.0778 0.3746 -1.0600 0.5500 %"""Δ"$%'"MRO! -0.0731 0.34264 -0.6800 0.5500 %"""Δ"$%'"MLF! -0.08085 0.35162 -0.6800 0.5500
Source: DataStream and Osiris
The mean of the unexpected change is 0.001744% with a standard deviation of 0.0570%. The minimum value of the unexpected change is a decrease of 0.07%. The maximum value of the unexpected part is an increase of 0.315%.
The mean of the expected interest rate change using the Deposit Facility as key interest rate, is -0.078% with a standard deviation of 0.375%. The minimum value is -1.06% and the maximum value is 0.55%.
The mean of the expected interest rate change using the Main Refinancing Operations as key interest rate, is -0.0731% with a standard deviation of 0.346%. The minimum value is -0.68% and the maximum value is 0.55%.
The mean of the expected interest rate change using the Marginal Lending Facility as key interest rate, is -0.0809% with a standard deviation of 0.352%. The minimum value is -0.68% and the maximum value is 0.55%.
In order to determine the effect of announcements of the ECB on the stock prices of European banks, one has to collect the announcement dates with its corresponding interest rate changes. The announcement dates with its corresponding announcements of the key interest rate, had to be manually retrieved from the website of the ECB. For every announcement, there is one document in which the ECB states whether it changed the interest rate or not. These documents were
downloaded and opened individually, to create table 2 and 3. There was a total of 236
announcements from 08/04/1999 until 16/03/2016. However, not every announcement contained a change in the key interest rate. The changes in the key interest rate, which lead to changes in the interest rate of the DF, the MRO and the MLF, are summarized in table 6,7 and 8. The number represents the amount of changes for a specific percentage. As you can see in the table and as already stated, the numbers of interest rate changes differ in some cases. This means that the interest rates of the facilities are sometimes changed differently.
Table 7
!"#$%"MRO" Downwards" Upwards" Unchanged" !
0.00%$ %$ %$ 194$ $ 0.05%$ 2$ %$ %$ $ 0.10%$ 2$ %$ %$ $ 0.25%$ 10$ 16$ %$ $ 0.35%$ %$ %$ %$ $ 0.50%$ 9$ 2$ %$ $ 0.75%$ 1$ %$ %$ $ 1.00%$ %$ %$ %$ $ Total$ 24$ 18$ 194$ 236$ $ ! Table 8
!"#$%"MLF" Downwards" Upwards" Unchanged" "
0.00%! %! %! 195! ! 0.05%! 2! %! %! ! 0.10%! 1! %! %! ! 0.25%! 8! 16! %! ! 0.35%! 1! %! %! ! 0.50%! 10! 2! %! ! 0.75%! 1! %! %! ! 1.00%! %! %! %! ! Total! 23! 18! 195! 236! Table 6
!"#$%"&'" Downwards" Upwards" Unchanged" !
0.00%$ %$ %$ 196$ $ 0.05%$ %$ %$ %$ $ 0.10%$ 5$ %$ %$ $ 0.25%$ 7$ 16$ %$ $ 0.35%$ %$ %$ %$ $ 0.50%$ 8$ 2$ %$ $ 0.75%$ 1$ %$ %$ $ 1.00%$ 1$ %$ %$ $ Total$ 22$ 18$ 196$ 236$ $
In order to give the reader a good representation of the changes in the key interest rate, a graph is created using the manually collected key interest rate changes of the Deposit Facility. In graph 1, one can see the changes of the Deposit Facility. In the graph, the unexpected and the expected parts of the interest rate change are also shown. Looking at the graph, one can conclude that the expected interest rate change lies very close to the total interest rate change. The unexpected part is small compared to the expected part.
Graph 1
The last section of regressions uses ratios to determine whether the characteristics of the ratios influence the sensitivity to changes in the key interest rate. The ratios were retrieved from Osiris. There were four ratios available from 2011 up to and including 2014. The information about the ratios can be found in table 9.
-‐1,2 -‐1 -‐0,8 -‐0,6 -‐0,4 -‐0,2 0 0,2 0,4 0,6 0,8
Change in the interest rate of the Deposit Facility Unexpected part of the change
5. Results
This section presents the results that are found using the regressions. First, the results of the three regressions for whole the sample of banks will be presented. They differ in the use of the
expected interest rate change. These regressions are performed to look at the effect of the interest rate change on the whole sample. Hereafter, the results of the regressions of the individual banks using only the expected interest rate change of the Deposit Facility are discussed. These
regressions will tell us if one can determine the effect of the interest rate changes on individual bank stock returns. Last, using the results of the regressions that contain the ratios, the effect of the ratios on the reaction to interest rate changes is presented.
First, table 10 shows us the results of the three regressions, each using another expected key interest rate change. In the three regressions, the coefficients of the unexpected interest rate change are insignificant. This means that the hypothesis that the unexpected part of the interest rate affects stock returns, can be rejected. However, the coefficients of the interaction variable of the unexpected interest rate change and the crisis dummy, are significant. These values are negative and indicate that during the crisis, the unexpected part influences stock returns in a negative way. Furthermore, the effects of the expected parts of the interest rate changes, are significant. These coefficients indicate a positive effect on stock returns. The interaction variables of the expected part and the crisis dummy are also significant. These coefficients are positive, which means that the effect of the expected part during the crisis is larger. Furthermore, the coefficient of the crisis dummy is negative and significant. Stock returns during the crisis are lower.
The findings are in line with prior research. For example, Bernanke and Kuttner (2005) find also that there is a negative relationship between the unexpected part and the stock returns. Moreover, the find that there is a positive relationship between the expected part and the stock returns. Furthermore, Kontikos (2013), Bredin et al (2005) and Bohl, Siklos and
Sondermann(2008) find also a negative relationship between the unexpected part and the stock returns. Haitsma, Unalmis and de Haan (2015) find the same results: a negative relationship between the unexpected part and stock returns and a positive relationship with the expected part. They also find that during the crisis, the effect of expected and unexpected changes become larger.
Looking at the regression for DF, a 1%- point increase in the unexpected part of the interest rates, lowers the stock returns by 42.27% during the crisis. This value is calculated by adding the coefficient of the unexpected change and the coefficient of the unexpected change during the crisis. A 1%-point increase in the expected interest rate, increases the stock returns with 0.583%. During the crisis, this effect is larger and in this case a 1%-point increase in the expected part, leads to a 3.549% increase in stock returns. During the crisis, stock returns are 0.0651% lower.
Looking at the regression of MRO, a 1%-point increase in the unexpected part, leads to a
decrease of 48.60% in stock returns during the crisis. On the contrary, a 1%- point increase in the expected part of the interest rate,leads to a 0.583% increase of stock returns. During the crisis, this effect is larger and becomes 4.743%. The coefficient of the crisis dummy, is almost the same. During crises, stock returns are 0.063% lower.
Looking at the regression of MLF, a 1%-point increase in the unexpected part of the interest rate change, results in a decrease of stock returns of 55.585%. However, this same change in the expected part of the interest rate change, leads to an increase of 0.583%. Again, this effect is larger during the crisis: It becomes 4.208%. Furthermore, the stock returns during the crisis are 0.060% lower. Table 10 Return' DF' MRO' MLF' Δr#$! -0.00490 -0.0049 -0.00490 ! (0.774) (0.774) (0.774) Δr#%Crisis! -0.41781*** -0.4812*** -0.55095*** ! (0.000) (0.000) (0.000) Δ)r#*! 0.00583* 0.0058* 0.00583* ! (0.081) (0.081) (0.080) Δ)r#*)Crisis! 0.02966*** 0.0416*** 0.03625*** ! (0.000) (0.000) (0.000) Crisis! -0.00065*** -0.00063*** -0.00060*** ! (0.000) (0.000) (0.000) Constant! 0.00037** 0.00037** 0.00037** ! (0.002) (0.002) (0.002) !
To translate these numbers to real economic values, the summary statistics of the unexpected changes and expected changes are used to determine the effect on stock returns after the announcement of the ECB. For example, the regression using the interest rate of the Deposit Facility as key interest rate, a 1%-point increase in the unexpected part, results in a decrease of 42.27% in stock returns. This seems a bit much, but if we look at table 5 with the summary statistics, we see that the mean of the unexpected change is 0.001744%. This means that stock returns on average, decrease by 0.001744 times 42.27%, which is 0.0737%. This is the change in stock prices due to changes in the unexpected part of the interest rate change during the crisis.
The effect on the expected part of the interest rate change of the Deposit Facility, is different. A 1%-point change in the expected part of the interest rate of the Deposit Facility, results in an increase of 0.583% in stock returns before the crisis. Using table 5, on average stock returns are decreased by 0.07775 times 0.583%, which is 0.0453%. During the crisis, the stock returns are increased by 3.549% if the expected interest rate goes up by 1%-point. This change leads on average to a decrease of 0.2759% in stock returns. During the crisis, the total effect on stock returns is decreased by 0.065%.
The changes in stock prices could be larger or smaller, depending on the change of the interest rate. The minimum and maximum values of the changes is used to provide the reader a range. Using table 5, the effect of the minimum of the interest rate changes is as follows: During the crisis the unexpected part of the interest rate change, is calculated by multiplying the
minimum value with the percentage point change in stock prices. This is calculated by: -0.07 times -42.27%, which is an increase of 2.96%. This effect is much larger than the effect that is calculated using the average unexpected change. The effect of the expected changes, is calculated by the multiplication of -1.06 with 0.0583%-point, which is a decrease of 0.0618% in stock returns. During the crisis, this effect is larger, then it becomes -1.06 times 3.549%-point, which is a decrease of 3.762%.
The rest of the outcomes using the maximum, mean and the minimum are presented in the tables 11, 12, and 13. The values are calculated in the same way as described above.
The most interesting thing about the tables above, is that the sensitivity to interest rate changes is much larger during the crisis. This can be seen if one compares the percentage change between the period prior the crisis and during the crisis. Furthermore, the table also gives the reader a good representation how the interest rates are influenced by the two parts of the interest rate change. Another thing that the table shows, is that before the crisis stock prices are not influenced by changes in the unexpected part. Furthermore, the reactions differ due to the difference in the choice of the interest rate of the facilities.
Next, table 14 in the appendix represents the results of the individual regressions for the whole sample of banks. A lot of coefficients have a p-value above 10%, which indicates that the hypothesis that a certain part of the interest rate change influences stock returns, can be rejected. However, there are some coefficients significant. The coefficients of the unexpected part of the interest rate change, are in all the regressions insignificant, so no conclusions can be made on this. The significant coefficients of the interaction variable of the unexpected part and the crisis dummy, are all negative. During the crisis, an increase in the unexpected part of the interest rate
Table 11 Returns( )*+,( )* + -.*/0/0( )(* + 1(23( )(* + 1(23(.*/0/0( Maximum 0 -13.3154% 0.3207% 1,9520% Mean 0 -0.0737% -0.0453% -0.2759% Minimum 0 2.9590% -0.0618% -3,7620% ! Table 12 Returns( )*+,( )* + -.*/0/0( )(* + 1(MRO( )(* + 1(MRO(.*/0/0( Maximum 0 -15.3103% 0.3207% 2.6087% Mean 0 -0.0848% -0.0426% -0.3467% Minimum 0 3.4023% -0.3964% -3.2252% ! Table 13 Returns( )*+,( )*+-.*/0/0( )(*+1(MLF( )(*+1(MLF(.*/0/0( Maximum 0 -17.5093% 0,3207% 2,3144% Mean 0 -0.0969% 0,0471% 0.3402% Minimum 0 3,8910% -0.3964% -2.8614% !
change, leads to lower stock returns. There is only one significant coefficient of the expected part and it is positive. The coefficient of the interaction variable of the expected part and the crisis dummy is, in the case of significant values, positive. During the crisis, stock returns are more positively affected by changes in the expected part of the interest rate. The coefficient of the crisis dummy is in three cases significant and negative.
Now a few examples will be given of the effect of interest rate changes on the stock returns of individual banks. The results for the Allied Irish banks is as follows: During the crisis, a 1%-point increase in the unexpected part of the interest rate change leads to a 70.1993% decrease in stock return. On the contrary, a 1%-point change in the expected part of the key interest rate, results in an 8.499% increase of the stock return during the crisis. For BBV Argentaria, only one coefficient was significant. During the crisis, a 1%-point increase in the unexpected part of the key interest rate change, leads to 34.75% lower stock return. The effect on BBV Argentaria, is half of the effect on the stock returns than the effect on the stock returns of Allied Irish Banks.
For the Erste Group Bank, a 1%-point increase in unexpected part during crisis, leads to a decrease of 91.50% in stock returns. This same change for Nordea Bank, results in 45.49% lower stock return. For ING Group, the coefficient of the interaction variable of the expected part and the crisis, is the only coefficient that is significant. An increase of 1%-point, results in 5.44% higher stock return during the crisis.
The only significant coefficient of the expected part, was that of Caixabank. A 1%-point increase of the expected part, results in 2.63% higher stock returns. For Unicredit, two
coefficients were significant. An increase of 1%-point in the unexpected part, results in 46.65% lower stock returns. An increase of 1%-point in the expected part leads to 7.265% higher stock returns. The significant coefficients differ, however only a part of the coefficients are significant. Due to the difference in significant coefficients, one can suggest that banks react differently to interest rate changes.
The next section of regressions was performed to see why banks react differently to changes in the key interest rate. Using the solvency ratio, Net Interest Margin, the Return On Average Assets and the Return On Average Equity, regressions were performed as described in the methodology. Table 12 shows the results of the regressions. First, the regression using the 𝑆𝑅b is discussed. Second, the regression using the 𝑁𝐼𝑀b is discussed. Third, the regression using the 𝑅𝑂𝐴𝐴b is discussed. Last, the regression using 𝑅𝑂𝐴𝐸b is discussed.
The coefficient of the interaction variable of the solvency ratio and the unexpected part, is positive and significant. When the solvency ratio increases, the reaction to changes in the unexpected part becomes less negative or even positive. The other coefficients are insignificant, so no statements can be made on this.
The interaction variables using 𝑁𝐼𝑀b are insignificant, which means that the hypothesis that the
NIM has an effect on the reaction to changes in the expected and unexpected part, can be rejected. The effect of the NIM on stock returns is negative and significant: an increase of
1%-Table 12 Return' !"#' $%&#' "'((#' "'()#' Δ+#,' (2.3317***' (0.107643' (0.043134' 0.057635' ' (0.001)' (0.848)' (0.844)' (0.788)' Δ-+#.-/0' 0.015454' 0.003591' (0.007097' (0.006729' ' (0.664)' (0.194)' (0.590)' (0.602)' !"#' 0.000016' (' (' (' ' (0.871)' (' (' (' $%&#' (' (0.000569**' (' (' ' (' (0.036)' (' (' "'((#' (' (' 0.000774***' (' ' (' (' (0.000)' (' "'()#' (' (' (' 0.000018***' ' (' (' (' (0.000)' Δ+#1-"2345' 0.404425***' 0.06798' 0.424763**' 0.007155' ' (0.001)' (0.837)' (0.039)' (0.395)' Δ+#./0-"2345' (0.004467' (0.01061' 0.005776' 0.000341' ' (0.500)' (0.557)' (0.512)' (0.180)' Constant! (0.000039' 0.000971**' 0.000087' 0.000109' ' (0.944)' (0.042)' (0.636)' (0.549)' '
The effect of ROAA on stock returns is positive: an increase of 1 in the ROAA, leads to 0.0774% higher stock returns. The interaction variable of ROAA and the unexpected part, shows that when ROAA increases the effect of the unexpected part on stock returns becomes positive.
In the last regression, the only coefficient that is significant, is that of ROAE. An increase of ROAE has a positive effect on stock returns. The other coefficients are insignificant, so no other conclusions can be made.
6. Conclusion and Discussion
In this study, I tried to determine what the relationship is between announcements of ECB’s conventional tools (e.g. key interest rate decisions) and European bank stock returns. This was done by looking at the effect of expected and unexpected interest rate changes on stock returns. There has been done research about this effect, but this thesis tries also to investigate whether banks react differently to changes in monetary policy and what kind of factors could drive this difference in reaction.
First, looking at the results of the regression over the whole sample, there is no significant relationship between the unexpected part of the interest rate change and the stock returns before the crisis. However, there is a significant negative relationship between the unexpected part and the stock returns during the crisis. If the unexpected part of the interest rate goes up, stock returns decrease during the crisis. The effect of the expected part of the interest rate change on the stock returns, is significant and positive before the crisis. During the crisis, this effect becomes larger, which means that the response to the expected part of interest rate changes affect stock prices more. An increase of the expected part, results in an increase of the stock prices. Another conclusion that can be made, is that during crisis stock returns are lower.
The tables that contain the percentage changes using the mean, maximum and minimum values of the changes, show that the effect on stock prices is small before the crisis. However, the effect on stock prices is much larger during the crisis.
Second, the regressions of the individual banks are performed. Although a lot of coefficients were not significant, the results show the same outcome as the regression over the whole sample. The unexpected part of the interest rate change has no effect on stock returns before the crisis. Furthermore, the expected part has a positive effect on stock returns and this
effect becomes larger during the crisis. During the crisis, stock returns are lower. If we look at the difference in the significant coefficients, the answer if banks react differently is suggestive.
Third, the results of the regressions containing the four ratios show that the solvency ratio and the Return On Average Assets influence the reaction of stock prices to changes in the
unexpected part of the interest rate changes positively. Banks that have a higher solvency ratio, react less to changes in the unexpected part. However, banks that have a higher ROAA ratio, react more severe to interest rate changes. The Net Interest Margin and the Return On Average Equity do not influence the effect of the unexpected part and the stock price. Furthermore, there is no effect of the ratios on the influence of the expected part of interest rate changes. This means that these regressions tell us that there is not a large influence of the ratios on the sensitivity to key interest rate changes.
The reason why a lot of results were not significant may lie on the fact that the sample or time period was too small. If one could test with a larger sample and over a longer period of time, then the results may would have been more significant and more inferences could have been made. This especially yields for the test whether banks react differently to policy changes and if so, why banks would react differently.
Researchers could look better at the effect of firm characteristics, using ratios or other methods, to determine whether these characteristics influence the reaction to interest rate
changes. If they do so, one can be more precise about the expectation of which companies will be affected the most.
It is still a very interesting topic to investigate and it is still an important field. The effect on the whole sample is, not only in this study, but also in a lot of other studies investigated. However, the answer to the question if banks react differently is suggestive. Further research has to be done, to answer this question and to give policy makers of the ECB a better insight in the transmission mechanism in order to make their policies more effective and efficient.
7. Reference list
Aharony, J., Saunders, A., Swary, I., (1986). The effect of the shift in monetary policy regime on the profitability and risk of commercial banks. Journal of Monetary Policy, 17 (3).
Berk, J. B., DeMarzo, P. M. (2014). Corporate finance (Third edition, Global ed.). Boston, Mass.; London: Pearson.
Bernanke, B. S., and K. N. Kuttner (2005), `What Explains the Stock Market’s Reaction to Federal Reserve Policy? ́, The Journal of Finance, Vol. 60 (3), pp. 1221-1257.
Bernoth, K., and J. Von Hagen (2004), `Euribor Futures Market: Efficiency and the Impact of ECB Policy Announcements ́, International Finance, Vol. 7(1), pp. 1-24.
Blanchard, O. (1993). Consumption and the recession of 1990-1991. The American economic
review, 83 (2).
Bodie, Z., Kane, A., & Marcus, A. J. (2011). Investments. New York: McGraw-Hill/Irwin
Bohl, M. T., P. L. Siklos, and D. Sondermann (2008), `European Stock Markets and The ECB’s Monetary Policy Surprises ́, International Finance, Vol. 11(2), pp. 117-130.
Bomfim, A. (2003). Pre-announcement effects, news effects, and volatility: Monetary policy and the stock market. Journal of Banking and Finance, 27(1), 133-151.
Bredin, D., Hyde, S., Nitzsche, D., & O'Reilly, G. (2009). European monetary policy surprises: The aggregate and sectoral stock market response. International Journal of Finance &
Economics, 14(2), 156-171.
Chen, S.–S. (2012). Consumer confidence and stock returns over market fluctuations.
Choi, Elyasiani, & Kopecky. (1992). The sensitivity of bank stock returns to market, interest and exchange rate risks. Journal of Banking and Finance, 16(5), 983-1004.
Chuliá, H., M. Martens, and D. van Dijk (2010), `Asymmetric Effects of Federal Funds Target Rate Changes on S&P100 Stock Returns, Volatilities and Correlations ́, Journal of Banking and
Finance, Vol. 34(4), pp. 834-839.
Ehrmann, M., and M. Fratzscher (2004), `Taking Stock: Monetary Policy Transmission to Equity Markets ́, Journal of Money, Credit, and Banking, Vol. 36(4), pp. 719-737.
Ehrmann, Michael. (2003). Monetary policy transmission in the euro area: Any changes after EMU? 0240.
European Central Bank (2010a). The ECB’s monetary policy strategy [online]. [cited 2.2.2010]. Available from World Wide
Web:<URL:http://www.ecb.europa.eu//mopo/strategy/html/index.en.html>
Fama, & Schwert. (1977). Asset returns and inflation. Journal of Financial Economics, 5(2), 115-146.
Fariborz, M., Rebel, A. C., Qionging, W. (2008). Bank stock returns and economic growth.Journal of Banking and Finance, 32 (6). Doi:10.1016/j.jbankfin.2007.07.006
Feldstein, M. (1982). Inflation and the Stock Market: Reply. The American Economic
Review, 72(1), 243-246.
Fiordelisi, F., G. Galloppo, and O. Ricci (2014), `The Effect of Monetary Policy Interventions on Interbank Markets, Equity Indices and G-SIFIs during Financial Crisis ́, Journal of Financial
Stability, Vol. 11, pp. 49-61.
Flannery, M. J., James, C. M. (1984). The effect of interest rate changes on the common stock returns of financial institutions. Journal of Finance, 39 (4).
Grammatikos, Lehnert, & Otsubo. (2015). Market perceptions of US and European policy actions around the subprime crisis. Journal of International Financial Markets, Institutions & Money, 37, 99-113.
Gurkaynak, R.S., Sack, B. and Swanson, E. (2002) “Market Based Measures of Monetary Policy Expectations”, Finance and Economics Discussion Series Working Paper 2002-40, Federal Reserve.
Haan De J. (2008). The effect of ECB communication on interest rates: An assessment. The
Review of International Organizations 3:4, 375–398.
Haitsma, R., Unalmis, D., de Haan, J., (2015). The impact of the ECB’s conventional and unconventional monetary policies on stock market. DNB Working Paper
Hardouvelis, G. (1987). Reserves Announcements and Interest Rates: Does Monetary Policy Matter? Journal of Finance,
Hussain, S. M. (2011). Simultaneous monetary policy announcements and international stock market response: An Intraday analysis. Journal of Banking & Finance 35
Kholodilin, K., A. Montagnoli, O. Napolitano, and B. Siliverstovs (2009), `Assessing the Impact of the ECB’s Monetary Policy on the Stock Markets: A Sectoral View ́, Economic Letters, Vol. 105(3), pp. 211-213.
Kontonikas, Macdonald, & Saggu. (2013). Stock market reaction to fed funds rate surprises: State dependence and the financial crisis. Journal of Banking and Finance, 37(11), 4025-4037.
Kuttner, K. N. (2001), `Monetary Policy Surprises and Interest Rates: Evidence from the Fed Funds Futures Market ́, Journal of Monetary Economics, Vol. 47(3), pp. 523-544.
Palmon, D., & Yaari, U. (1981). Share Values, Inflation, and Escalating Tax Rates. Journal of
Banking & Finance, 5, 395-403.
Pearce, D., & Roley, V. (1985). Stock Prices and Economic News. Journal of Business, 58(1), 49-67.
Peersman, G., and F. Smets (2005), `The Industry Effects of Monetary Policy in the Euro Area ́,The Economic Journal, Vol. 115(503), pp. 319-342.
Poterba, J. M. (2000). Stock market wealth and consumption. Journal of Economic Perspectives,
14 (2).
Ricci, O. (2014). `The impact of Monetary Policy Announcements on the Stock Price of Large European Banks during the Financial Crisis ́. Journal of Banking & Finance. Advance online publication.
Rigobon, R., and B. Sack (2004), `The Impact of Monetary Policy on Asset Prices ́, Journal of
Monetary Economics, Vol. 51, pp. 1553–1575.
The European Central Bank. Key ECB interest rates. Retrieved 02/06/2016, from https://www.ecb.europa.eu/stats/monetary/rates/html/index.en.html
The European Central Bank. Objective of monetary policy. Retrieved on 04/17/2016, from https://www.ecb.europa.eu/mopo/intro/objective/html/index.en.html
The European Central Bank. Transmission.Retrieved on 04/17/2016, from https://www.ecb.europa.eu/mopo/intro/transmission/html/index.en.html
Thorbecke, W. (1997). On stock market returns and monetary policy. American Finance
8. Appendix
Table 1Abbreviation Explanation
ECB European Central Bank FED Federal Reserve
DF Deposit Facility
MRO Main Refinancing Facility MLF Marginal Lending Facility
!"#$#$ Crisis dummy (1 during crisis, 0 otherwise) Δ"& Key interest rate change
Δ"&' Unexpected part of the key interest rate change
Δ("&) Expected part of the key interest rate change
Δ("&)*+ Expected part of the key interest rate change, using the rate of
the DF as Key interest rate
Δ("&)MRO Expected part of the key interest rate change, using the rate of
the MRO as Key interest rate
Δ("&)MLF Expected part of the key interest rate change, using the rate of
the MLF as Key interest rate
Δ"&,!"#$#$ Unexpected part of the key interest rate change, during the
crisis (interaction variable)
Δ"&)!"#$#$ Expected part of the key interest rate change, during the crisis
(interaction variable)
Δ("&)(*+(!"#$#$ Interaction variable of the expected part of the key interest rate
change, using the rate of the DF as key interest rate during the crisis
Δ("&) MRO(!"#$#$ Interaction variable of the expected part of the key interest rate
change, using the rate of the MRO as key interest rate, during the crisis
Δ("&) MLF(!"#$#$ Interaction variable of the expected part of the key interest rate
change, using the rate of the MLF as key interest rate, during the crisis
-.,& Future rate at day t
-.,&01 Future rate at day t - 1
SR Solvency ratio
NIM Net Interest Margin (%) ROAA Return On Average Assets ROAE Return On Average Equity
Δ"&,23 Interaction variable of the unexpected part of the key interest
rate change and the solvency ratio
Δ"&,456 Interaction variable of the unexpected part of the key interest
rate change and the Net Interest Margin
Δ"&,3788 Interaction variable of the unexpected part of the key interest
rate change and the Return On Average Assets
Δ"&,3789 Interaction variable of the unexpected part of the key interest
rate change and the Return On Average Equity
Δ"&)*+(23 Interaction variable of the expected part of the key interest rate
change, with DF as key interest rate and the solvency ratio Δ"&)*+(456 Interaction variable of the expected part of the key interest rate
