Market Reaction to Unexpected Monetary Policy Announcement Jean-Yves Filbiena

Fabien Labondanceb

First Draft

June 16, 2009 Abstract This paper examines the daily response of euro area stock markets to unexpected ECB monetary policy announcements. We define ECB’s unexpected decisions in analyzing the consensus in the specialized press the days before the announcement. Our preliminary results show that very few ECB’s announcement’s are unexpected by ECB watchers, this is a sign that the ECB’s monetary policy is very predictable. We also find little responses of stock markets to unexpected monetary policy decisions. If Eurozone equity markets react homogeneously to unexpected monetary policy decisions, we can see that reactions are higher in France than in Germany. We find significant and heterogeneous responses from sectorial indexes which indicate that the financial channel of monetary policy is different following the sector. Finally, we find that markets are more sensitive to good news in bad times, but there is no evidence of the opposite.

JEL Classification: E52, E58, E61 Keys Words: Monetary Policy, Stock Prices, Event Study, ECB

a b

Jean-Yves Filbien, Louvain School of Management & FUCaM and Universit Nord de France, e-mail: jean [email protected].

Fabien Labondance (corresponding author), Louvain School of Management & FUCaM, Grenoble Universit, e-mail:

[email protected]. We thank Isabelle Platten, Natacha Gilson, Eric de Bodt, and Etienne Farvaque for their comments. Remaining errors are ours.

1

Introduction “The most direct and immediate effects of monetary policy actions, such as changes in the Federal funds rate, are on the financial markets; [...]”–Bernanke and Kuttner (2005), p. 1221. The financial crisis started in 2007 renews the question of the efficiency of monetary policy. Some

critics focus on the role of the Fed in the real state bubble. The actual situation leads us to analyze the impact of monetary policy on asset prices. The literature around the question of the impact of monetary policy on financial assets is rich (Bernanke and Kuttner, 2005; Bomfim, 1991; Rigobon and Sack, 2004; Wongswan, 2009), but mainly focused on the United States, and very few works have been done on the euro area. In the literature, the question of whether central banks should react to stock price evolutions is also very much debated. There is thus a problem of endogeneity and it appears very hard to determine who is the most powerful. In other terms, is it the monetary policy which influences stock prices or the opposite? According to their status, central banks should not target asset prices. And Jean-Claude Trichet, European Central Bank (ECB) Chairman, justifies that “Today there is something close to a consensus among economists and central bankers that targeting financial asset prices is a bad idea”1 . However, he adds “with respect to the ECB monetary policy strategy, the important point to note is the following: The ECB’s monetary policy strategy does allow for taking account boom developments without any amendments to the strategy and without providing any additional role to asset prices”2 . ECB uses the stock prices in order to build its strategy. Under the efficient market hypothesis, only a truly new information has an impact on stock prices. It is also important to establish that the news is unbiased. Using the event study methodology, over the 1999-2008 period, we analyse the impact of unexpected monetary policy decisions on the euro area stock markets. The ECB organizes a conference after each Council of Governors. It is hard to determine the next monetary policy decision only from the speech. Between Councils of Governors, macroeconomics news is released, news speeches are done by members of the ECB. That is the reason why we focus on the investors’ expectations about these announcements in order to identify their unexpected components. Only surprise monetary policy decisions move stock prices. This paper extends the existing literature in several dimensions. First, we apply an event study which is a classical and robust financial method to economic issues. Second, to our knowledge, there are few papers which examine the relationship between ECB monetary policy and euro area stock markets. Third, we decompose the impact of unexpected ECB announcements according to equity markets indexes, business conditions and industry. The main findings can be summarized as follows. First, from our descriptive statistics, we observe 1 Trichet, Jean-Claude (2005), ”Asset price bubbles and monetary policy”, Speech at the Monetary Authority of Singapore, Singapore, 8 june. 2 ibid.

1

that the ECB’s monetary policy decisions are more and more predictable by the economists and the investors. Second, we do not find a significant response from stock markets to unexpected monetary policy decisions. However, our results suggest that stock markets do not react homogeneously to unexpected monetary policy decisions according to countries, business conditions and industry. The paper is organized as follows. Section 2 introduces the literature and presents four testable hypotheses that distinguish the impact of monetary policy in accordance to financial structure, economic conditions and the industry. Section 3 describes our data and our method. Section 4 reports the empirical results. Section 5 concludes.

2

Literature et Hypotheses

The main instrument in the hands of monetary policy makers is the main refinancing operation interest rate. Through the change of this interest rate, central banks can affect the whole economy. The literature on the monetary policy transmission is rich. This channel is the standard studied in all Keynesian textbook: an expansionary monetary policy decreases the real interest rate and then leads to an increase in investment spending and in output. But this framework seems to be deficient to explain all the impacts of monetary transmission. In particular, it is well known that the banking system and asymmetric information play an important role in the way that economic agents get financing. In order to find answers to these dissatisfactions in this initial framework, during the 80’s, researches developed the credit view. Following the work of Bernanke and Blinder (1988), this way of research finds evidence that banks and their balance-sheet played a major role in the monetary policy transmission channel. And it exists a third channel which give us explanations to understand what is going on in the black box called economy, through the reactions of other assets. Works on monetary transmission mechanisms are interested on three categories of assets: stock market prices, real estate prices and exchange rates. Precisely we investigate the link between ECB monetary decisions and stock market prices. Stock markets have an impact on the aggregate economy through several mechanisms involving firms and households. The central idea is always that an expansionary monetary policy lowers interests rates, increases the demand of assets and thus raises their prices. From that logical hypothesis, we highlight at least three pass-throughs. The first one is the effect on the investment defined by the Tobin q-theory (Tobin, 1969). Tobin’s q is calculated as the market price of firms divided by the replacement cost of capital. With a high q, this means that a future investment is cheap relative to the market value. Companies can thus easily finance their new infrastructures in issuing stocks. Monetary policy can also affect aggregate spending through a firm balance-sheet effect. This effect is the same as the one mentioned previously as the credit view. When firms own a low net worth, as adverse selection or moral hazard can appear and banks could be reluctant to lend to these firms. Lower net worth means less collateral, when it declines, banks could be more severe and a phenomenon

2

of credit rationing emerges (Stiglitz and Weiss, 1981). Conversely, an expansionary monetary policy which produces a rise in stock prices, reduces the asymmetric information problem, generates a higher net worth and leads to more lending. Balance-sheet channel operates through household wealth effect. According o the life cycle model, the consumption is determined by the lifetime resources of consumers and for some households stocks are an important component of it (Ando and Modigliani, 1963). An expansionary monetary policy that increases stock prices raises the household wealth and finally causes a rise of consumption. Asset prices and more precisely stock prices are an important channel for monetary policy. The figure 1 sums up the link mentioned above between stock prices and the aggregate economy. And because of these links, we have to carry empirical researches to quantify these relationships and the ability of the ECB to influence European stock markets. Literature on this subject is mainly focused on the USA (Pearce and Roley, 1983, 1985; McQueen and Roley, 1993; Thorbecke, 1997; Roley and Sellon, 1998; Bomfim, 1991; Bernanke and Kuttner, 2005; Rigobon and Sack, 2004; Gurkaynak et al., 2005). Over a 1977 - 1982 period, Pearce and Roley (1983) find that the change of discount rate has a negative impact on stock prices. From September 1977 to October 1982, Pearce and Roley (1985) use a survey date to measure expectations and find that daily stock prices respond to monetary information. From September 1977 to May 1988, McQueen and Roley (1993) find that the S&P 500 Index falls by 0.31% in response to an unanticipated increase in unanticipated change in the Federal Reserves discount rate of 1 percentage point. From January 1967 to December 1990, Thorbecke (1997) observes that expansionary policy increases ex-post stock returns. Roley and Sellon (1998) examine how U.S stock prices respond on Federal Open Market Committee (FOMC) meeting dates when expected policy actions do not occur. Contrary to the previous studies, Roley and Sellon observe a lack of significant responses from a long-term interest rates and stock prices to non announcements. From June 1989 to December 2002, Bernanke and Kuttner (2005) find, that, on average, a hypothetical unanticipated 25 basis point cut in Federal funds rate target is associated with about a 1% increase in broad stock indexes. Bomfim (1991) indicates that for each basis point increase in the expected average daily values of the funds rate in following month, daily stock market returns are reduced by 0.04 percentage point. Over a January 1994 - November 2001 period, Rigobon and Sack (2004) find that the U.S stock indexes have a significant negative reaction to monetary policy. Their results suggest that an unanticipated 25 basis point increase in the short-term interest rate results in a 1.7% decline in the S&P index. Using intraday data, over 1990 - 2004 period, Gurkaynak et al. (2005) find that, on average, a surprise 25 basis point tightening in the federal funds rate leads to a little more than a one percent significant fall in the S&P500. Even if we find very few works done on the euro area, the results are similar. From January 1997 to June 2002, Funke and Matsuda (2002) find that the interest rates have a negative impact on German stock indexes. This is the Gordon-Shapiro rule. This negative relationship should be observed, at least in the short 3

run. A negative policy surprise (i.e. an unexpected rise of the marginal refinancing operation rate) should decrease the European stock prices. This is the first hypothesis that we investigate. But, in the case that this hypothesis is not fulfilled, we decide to investigate further questions for more details.

H1: ECB unexpected policy announcements have a negative impact on the Eurozone equity markets returns.

Traditionally, literature distinguishes the market-based economies and the bank-based economies. Recent studies incorpore a more precise classification and extend the typical one (Amable, 2005). However, we take into account the recent evolution of financial structures. The frontier across the market-based and the bank-based economies are more and more narrow. From a 1957-2000 sample period and using a sample of 16 countries, Durham (2003) finds that the relationship between monetary policy and stock price returns is weak or nonexistent. However, Durham observes that data for United States only support that monetary policy easing correlates with higher excess returns. It is interesting to isolate the impact of unexpected monetary policy decisions on national stock market indexes. We expect that the stock market reaction depends on the characteristics of local financial system and will therefore not be homogenous.

H2: The effect of the ECB key rates announcements depends on financial structure.

Previous researches find few evidence from investors’ reactions around monetary policy announcements. However, we notice that the previous literature assumes that the investors’ reaction is the same over different stages of the business cycles. Analyzing the daily percentages changes in closing values of the S&P500 index on the September 1977-May 1988 period, McQueen and Roley (1993) find a stronger relationship between stock prices and news. They observe that when the economy is strong, the stock market responds negatively to news about higher real economic activity. The authors explain this inverse relation by the larger increase in discount rates. Veronesi (1999) shows that the stock markets overreact to bad news in good times and underreact to good news in bad times. When business conditions are well, a bad piece of news increases the discount over expected future dividends in order to bear the risk of higher uncertainty. Besides, a good piece of news in bad times tends to increase the expected future dividends, but it also increases the discount investors demand to hold the asset. To sum it up, the reaction of prices to news tends to be high in good times and low in bad times. Funke and Matsuda (2002) analyze the impact of a broad set of macroeconomic news on stock prices according to the business conditions of the economy. Theirs results suggest some evidence for asymmetric effects of macroeconomic news. They find that, in a boom period, bad economic news may be good news for stock prices and inversely. They also find that the impact of FED target rates varies 4

according to business cycles. When economy is growing, a higher-than-expected federal funds rate tends to lower stock prices. Higher interest rates may have a direct impact on stock prices by reducing the value of discounted future earnings. When the economy is in recession, a higher-than-expected federal funds rate tends to lead a rise in stock prices. Higher-than-expected interest rates may imply a better assessment of the monetary authorities of the future outlook of the economy, and thus, may lead to higher expectations of growth.

H3: The effect of the ECB key rates announcements depends on business conditions.

The question of monetary policy transmission is a key interest for monetary policy, mainly in a new and heterogeneous monetary union as the European one. Most of the time, researches focus on the aggregate level of the economy. However, such works ignore asymmetries on disaggregate levels which could lead to asymmetric effects of monetary policy across regions or sectors. This intuition known as the ”Krugman and Venables hypothesis” (Krugman and Venables, 1993) indicates that a monetary integration can promote specialization patterns. Such an argument implies that following some regions’ industrial portfolios, an industry shocks which affects only some industries could lead to asymmetrical effects among regions. In this case, monetary policy will be impotent to stabilize the whole economy. Carlino and DeFina (1998) look the effects of monetary policy shocks at state variations for the USA. And they find that heterogeneities among states are significantly related to sectoral structure variables. Hayo and Uhlenbock (2000) find that German Lander are asymmetrically affected by monetary shocks if it exists large differences in their regional industry portfolios.

H4: The effect of the ECB key rates announcements depends on industry.

3

Data and Method

3.1

Data

To maintain price stability is the primary objective of the Eurosystem. ECB mainly uses the main refinancing operations rate into its strategy. ECB Council of Governors have an official meeting once a month on the first Thursday of each month.3 It determines the key rates, comments the macroeconomic environment, and sends a message to investors about its futur monetary policy decisions. We collected all ECB Council of Governors announcements from January 1, 1999, to December 31, 2008. We described the announcement of the ECB main refinancing operations (MRO) rates (∆M RO) in equation 1: 3 Before

2002, ECB had 2 official meetings.

5

   = −1 if    ∆M RO =0 if      = +1 if

M ROt < M ROt−1 M ROt = M ROt−1

(1)

M ROt > M ROt−1

We obtain a final sample of 157 announcements. Table 1 reveals the frequencies of ECB monetary announcements. We observe more increases (#:16) than decreases (#:11).

3.2

Method

Measuring the Unexpected Monetary Policy Announcement Under the efficient market hypothesis, only truly new information has an impact on stock prices. We define the surprise element of policy actions as the difference between the investors’ expectations, released in the press just before the ECB Board of Governors meetings, and the decisions of ECB. From the Factiva database, we use three different press sources: la Tribune, les Echos, and the Financial Times to estimate the market expectations. We estimate the unexpected ECB monetary announcement from 2 variables: ∆M RO∗Sign and ∆M RO∗Degree , described in equations 2 and 3. ∆M RO∗Sign corresponds to the sign of the monetary policy surprise and ∆M RO∗Degree indicates the degree of the unexpected news.   \  = −1 if ∆M RO < ∆M RO    ∗ \ ∆M ROSign =0 if ∆M RO = ∆M RO      = +1 if ∆M RO > ∆M \ RO   \  = −1 if M RO < M RO    ∗ \ ∆M RODegree =0 if M RO = M RO      = +1 if M RO > M \ RO

(2)

(3)

In tables 2, 3, 4, and 5, we present the occurrence matrix of these variables. We have 17 and 11 unexpected ECB monetary policy announcements in accordance with the degree and in accordance with the sign, respectively. ECB monetary policy decisions are well anticipated by the market. It implies that ECB has successfully communicated its monetary policy, mainly in the recent years. Stock Prices

Daily stock prices of equity markets come from Datastream. We study seven euro

area equity market indexes: the AEX Index (Netherlands), the BEL20 Index (Belgium), the CAC40 Index (France), the Performance DAX30 Index (Germany), the IBEX35 Index (Spain), the Milan MIB30 Index (Italy), and the OMX Helsinki Index (Finland). In table 6, we present the performance of the euro area stock markets (CAC40, DAX30, AEX, IBEX35 and MIB30) from 1999 to 2008.4 Each year, in average, the performance are negative going from -0.39% for the DAX30 index to -7.54% for 4 We

Pi,t

calculate the returns as follows: Ri,t = ln( P

i,t−1

), where Pi,t is the price of stock i on day t.

6

the AEX index. We also present the performance of sectorial Stoxx indexes in table 7. We observe a large heterogeneity of performance among the industries. Event Study Methodology

We calculate the returns using from the event study methodology

introduced by Fama et al. (1969). Abnormal returns can be estimated by three different models: constant mean returns model, market model, and adjusted return risk market. The measure of abnormal returns is robust of the choice of the model (Brown and Warner, 1985). We select the mean constant returns model to estimate the abnormal component of returns of stock market index i at date t: ARi,t = Ri,t − Ri,t , with Ri,t

−10 1 X = Ri,t , 238 t=−244

(4)

(5)

where ARi,t is the abnormal return of the stock market index i at day t, Ri,t reveals the observed return of the stock market index i at day t and Ri,t indicates the average of returns of index i over the estimation period. To avoid the contamination of estimation period, we use an estimation period going from 244 days to 10 days before the announcement date5 . We calculate the cross-sectional average abnormal return:

n

ARt =

1X ARi,t , n i=1

(6)

where n indicates the number of announcements containing our sample. Then, we estimate the cumulative abnormal returns firms over the 5 trading days that surround the announcement dates [-2 days; +2 days]: CAR =

2 X

ARt .

(7)

t=−2

This window of 5 trading days control possible news leaks and allows investors time to gather additional information. We use a matching sample to control for the industry and market impacts. We select the New-York Stock Exchange and the DJEurostoxx50 as matching universe of the euro area stock markets and the euro area sectorial stock markets, respectively6 . 5 Our

results are robust to the length of estimation period. apply the significance test of Corrado (1989). The first step is to transform each security’s times series of abnormal returns into their respective ranks. Let Ki,t denote the rank of the excess returns ARit in stock i’s times series of 240 excess returns: Ki,t = rank(ARi,t ), (8) 6 We

t=-233,...,+2 where ARi,t > ARi,j implies Ki,t > Ki,j and 250 > Ki,t > 1 By construction, the average rank is one-half plus half the number of observed returns, or 120.5. The rank statistics for the abnormal returns at the event day is: 1 PN i=1 (Ki0 − 120.5) (9) Z= N S(K)

7

Classification of Business Conditions

Following McQueen and Roley (1993), we use the season-

ally adjusted monthly industrial production index of each country to define business conditions. First, we estimate a trend in the log of industrial production on a constant and a time trend from January 1, 1999, to December 31, 2008. Then, we add and subtract a constant from a trend, creating the upper and lower bounds. We choose the constant for each index so that the log of industrial production is above the upper bound, denoted as ”high” business activity, 25 percent of the time. The log of industrial production is below the lower bound, indicating ”low” economic activity, about 25 percent of the time as well. ”Medium” economic activity is represented by the remaining observations between the bounds. Classification of Financial Structure

Demirguc-Kunt and Levine (2000) examine the financial

structure for a cross-section of up to 150 countries. They construct a conglomerate index of financial structures based on measures of size, activity and efficiency. Classification of Industry The effect of monetary policy news changes with the type of stock (Krugman and Venables, 1993; Funke and Matsuda, 2002). We carry out our event study according to the European supersectors stock indexes, based on the industry classification benchmark (ICB) published by Stoxx Ltd. The ICB define 10 industries divided in 19 supersectors7 . Farrell (1975) suggests that stocks tend to group naturally according to their price behavior. We apply a cluster analysis on the Stoxx classification. We present the results in figure 2. We distinguish five panels of sectorial equity indexes. The ”non cyclical” stocks group aggregates the following supersectors: oil & gas, chemicals, basic ressources, foods and beverage, health care and utilities. Construction & materials, personal & house goods, industrial goods & services, retail, and travel & leisure compose the ”cyclical” stock group. The ”growth” stocks group aggregates media, technology, and telecommunication. We also cluster the financial sectorial indexes. Finally, the automobile supersector forms a panel by itself.

4 4.1

Results Market Reaction to Unexpected ECB Monetary Policy

Table 8 presents the correlation coefficients between the returns of the euro area stock markets, the ECB refinancing operations rates and the European short interest rates. We observe that the stock The standard deviation S(K) is calculating using the entire 234-day sample period: v u +2 N X u 1 1 X S(K) = t ( (Ki,t − 120.5))2 240 t=−233 N i=1

(10)

The ranking procedure transforms the distributions of security excess returns into a uniform distribution across the possible rank values regardless of any asymmetry in the original distribution. 7 Data unavailability for real estate supersector required us to delete it.

8

market returns are significatively and strongly correlated. Moreover, we find that the European short interest rates are also significant and strongly correlated. However, abnormal returns do not indicate an evidence from the inverse relationship between the abnormal returns and the unexpected ECB monetary policy decisions. Table 9 shows the cumulative abnormal returns according to the level of the changes of unexpected ECB monetary policy announcements (from -0.75 b.p. to +0.5 b.p. for the euro area stock markets. No result do not appears significatively and the relationship is not inverse. H1 is rejected. This stylized fact is confirmed by table 10 which analyses each day composing the event period (-2 days; +2 days).

4.2

Market Reaction to Unexpected ECB Monetary Policy according the Financial Structure

According to H2, we expect to find significant differences between national stock markets. Because in their monetary channel structure, we supposed that in countries where economy is more financed by the markets, the abnormal returns should be higher than in the others which are more turned to the banking sector. However, from table 10, we do not find significant heterogeneous responses. This implies a high degree of Eurozone financial markets integration. Further investigations should be conducted in order to be certain that there are no national effects. But for now monetary policy does not seems to be a source of asymmetries among national stock markets.

4.3

Market Reaction to Unexpected ECB Monetary Policy according to business conditions

The moving correlation allows us to evaluate the level of the correlation between two variables in the course of time. From figure 3, we observe that the correlation between the Euribor and the DjEurostoxx50 index is not stable along the period. Furthermore, it appears some states of positive and negative correlation. This result suggests some periods where the decisions of monetary policy had a pro-cyclical effect among the European financial market. We split our sample period into three economy states: low, medium, and high. From table 11, H3 is not verified. Contrary to the literature, investors’ reactions are more important with good market conditions than when market conditions are bad.

4.4

Market Reaction to Unexpected ECB Monetary Policy according the Industry

From table 12, we find that the panel A is clearly non-cyclical, but more surprisingly, the panel B seems to be non-cyclical too. The panel D where the financial indexes are aggregated shows us that over the period, these supersectors have always react badly when unexpected ECB announcements 9

occur. Finally, panel C and E react as we hope to. An unexpected decrease of the marginal refinancing operation rate conducts to a rise of the abnormal returns and vice versa. These asymmetric responses of supersectors to monetary policy could be explained by differences across firms in the level to which they are financially constrained.

5

Conclusions

The preliminary results indicate that the surprises of ECB monetary policy do not involve significant abnormal outputs on behalf of the Eurozone equity markets. The event study was also led on sectorial indexes and business conditions. This last point gives more satisfactory results. The transmission channel of the monetary policy by the stock market is thus effective during the high economy periods. These first results can consequently show several stylized facts. It appears that communication policy by the ECB seems efficient because its announcements do not surprise the markets. We also find that these unexpected announcements affect asymmetrically the sectoral indexes. Solutions are certainly in another prudential regulation in the prevention of the crises, and in a policy mix getting more active when crises occur. Solutions which are very difficult to be settled in the current institutional framework.

References Amable, B. (2005). Les Cinq Capitalismes. Le seuil. Ando, A. and F. Modigliani (1963). The Life Cycle Hypothesis of Saving: Aggregate Implications and Tests. American Economic Review 53 (1), 55–84. Bernanke, B. and A. Blinder (1988). Credit, Money, and Aggregate Demand. American Economic Review 78 (2), 435–362. Bernanke, B. and K. Kuttner (2005). What Explains the Stock Market’s Reaction to Federal Reserve Policy? Journal of Finance 60 (3), 1221–1257. Bomfim, A. (1991). Pre–Annoucements Effects, News Effects, and Volatility: Monetary Policy and the Stock Market. Journal of Banking and Finance 27, 133–151. Brown, S. and J. Warner (1985). Usual Daily Stock of Event Studies. Journal of Financial Economics 14, 3–31. Carlino, G. and R. DeFina (1998). Monetary Policy and the U.S. and Regions: Some Implications for European Monetary Union. Federal Reserve Bank of Philadelphia. Working Paper.

10

Corrado, C. (1989). A Non–Parametric Test for Abnormal Security–Price Performance in Event Studies. Journal of Financial Economics 23 (2), 385–395. Demirguc-Kunt, A. and R. Levine (2000). Bank–Based and Market–Based Financial Systems Cross– Country Comparisons. Working Paper. Durham, J. (2003). Monetary Policy and Stock Price Returns. Financial Analysts Journal 59 (4), 26–35. Fama, E., L. Fisher, M. Jensen, and R. Roll (1969). The Adjustment of Stock Prices to New Information. International Economic Review 10, 1–21. Farrell, J. (1975). Homogeneous Stock Groupings: Implications for Portfolio Management. Financial Analysts Journal 31 (3), 50–56+58–62. Funke, N. and A. Matsuda (2002). Macreconomic News and Stock Returns in the United States and Germany. IMF Working Paper . Gurkaynak, R., B. Sack, and E. Swanson (2005). Do Actions Speak Loud than Words? The response of Asset Prices to Monetary Policy, Actions and Statements. International Journal of Central Banking 1, 719–742. Hayo, B. and B. Uhlenbock (2000). Industry Effects of Monetary Policy in Germany. In Regional Aspects of Monetary Policy in Europe, pp. 127–158. Boston: Kluwer. Krugman, P. and A. Venables (1993). Integration, Specialization and Adjustment. C.E.P.R. Discussion Papers. Working Paper. McQueen, G. and V. Roley (1993). Stock Prices, News, Business Condition. Review of Financial Studies 6, 683–707. Pearce, D. and V. Roley (1983). The Reaction of Stock prices to Unanticipated Changes in Money. Journal of Finance 38, 1323–1333. Pearce, D. and V. Roley (1985). Stock Prices and Economic News. Journal of Business 58, 19–67. Rigobon, R. and B. Sack (2004). The Impact of Monetary Policy on Asset Prices. Journal of Monetary Economics 51, 1553–1575. Roley, V. and H. Sellon (1998). Market Reaction to Monetary Policy Nonannouncements. Working Paper. Stiglitz, J. and A. Weiss (1981). Credit Rationing in Markets with Imperfect Information. American Economic Review 71 (3), 393–410.

11

Thorbecke, W. (1997). On Stock Market Returns and Monetary Policy. Journal of Finance 52 (2), 635–654. Tobin, J. (1969). A General Equilibrium Approach to Monetary Theory? Journal of Money Credit and Banking, 15–29. Veronesi, P. (1999). Stock Market Overreaction to Bad News in Good Times: A Rational Expectations Equilibrium Model. Review of Financial Studies 12 (5), 975–1007. Wongswan, J. (2009). The Response of Global Equity Indexes to U.S. Monetary Policy Announcements. Journal of International Money and Finance 28, 344–365.

12

Figure 1: Transmission Channels of Monetary Policy

Table 1: Descriptive Statistics: ECB Main Refinancing Operations Rates Announcements This table presents the ECB main refinancing operations rates announcements between 1999 and 2008.

Year 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Total

-0.75 0 0 0 0 0 0 0 0 0 1 1

-0.5 1 0 2 1 1 0 0 0 0 2 7

-0.25 0 0 2 0 1 0 0 0 0 0 3

0 22 18 20 11 10 12 11 7 10 9 130

0.25 0 5 0 0 0 0 1 5 2 1 14

0.5 1 1 0 0 0 0 0 0 0 0 2

Total 24 24 24 12 12 12 12 12 12 13 157

Table 2: Classification of Unexpected Changes of ECB Main Refinancing Operations Rates by the Sign

This table presents the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign takes -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. We present in column the \ analyst’s expectations of the change of ECB main refinancing operation rates (∆M RO) and in row the ECB main refinancing operation rates (∆M RO).

∆M RO

-1 0 +1

-1 0 8 +1 4 +1 0

13

\ ∆M RO 0 -1 3 0 125 +1 3

+1 -1 0 -1 1 0 13

Table 3: Descriptive Statistics: Unexpected ECB Main Refinancing Operations Rates Announcements by the Degree This table presents the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗degree takes -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. We present \ in column the analyst’s expectations of the change of ECB main refinancing operation rates (∆M RO) and in row the ECB main refinancing operation rates (∆M RO).

-0.75

M RO

-0.50 -0.25 0 +0.25 +0.50

-0.75 0 0 +1 0 +1 0 -1 0 +1 0 +1 0

-0.50 -1 1 0 2 +1 1 +1 0 +1 0 +1 0

\ M RO -0.25 0 -1 -1 0 0 -1 -1 3 2 0 -1 1 1 +1 0 4 125 +1 +1 0 3 +1 +1 0 0

0.25 -1 0 -1 0 -1 0 -1 1 0 11 +1 1

0.50 -1 0 -1 0 -1 0 -1 0 -1 0 0 1

Table 4: Descriptive Statistics: Unexpected ECB Main Refinancing Operations Rates Announcements

This table presents the number of unexpected ECB main refinancing operations rates announcements. ∆M RO∗Sign corresponds to the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO ∗degree corresponds to the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign and ∆M RO∗degree take -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase.

Year 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Total

-1 1 1 3 1 0 0 0 0 0 2 8

∆M RO∗degree 0 +1 Total 23 0 24 19 4 24 17 4 24 11 0 12 11 1 12 12 0 12 12 0 12 12 0 12 12 0 12 11 0 13 140 9 157

14

-1 0 1 2 0 0 0 0 0 0 1 4

∆M RO∗Sign 0 +1 Total 24 0 24 20 3 24 18 4 24 12 0 12 12 0 12 12 0 12 12 0 12 12 0 12 12 0 12 12 0 13 146 7 157

Table 5: Descriptive Statistics: Unexpected ECB Main Refinancing Operations Rates Announcements

This table presents the number of unexpected ECB main refinancing operations rates announcements. ∆M RO∗Sign corresponds to the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO ∗degree corresponds to the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign and ∆M RO∗degree take -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. Stock Market ∆M RO∗degree Indexes -1 0 +1 Panel A:Low Economy State AEX 4 26 1 BEL20 5 31 0 CAC40 3 26 0 DAX30 1 21 0 DJESTX 3 28 0 IBEX35 4 24 1 MIB30 3 27 0 OMXH 4 24 2 Panel C: Medium Economy State AEX 2 74 2 BEL20 2 76 6 CAC40 3 81 5 DAX30 5 81 7 DJESTX 3 70 5 IBEX35 4 70 4 MIB30 3 78 6 OMXH 3 76 6 Panel C: High Economy State AEX 2 40 6 BEL20 1 33 3 CAC40 2 33 4 DAX30 2 38 2 DJESTX 2 42 4 IBEX35 0 46 4 MIB30 2 35 3 OMXH 1 40 1

∆M RO∗Sign -1 0 +1 1 1 1 0 1 1 1 1

30 35 28 22 30 28 29 27

0 0 0 0 0 1 0 2

1 2 1 2 1 3 1 2

75 78 85 86 73 71 81 79

2 4 3 5 4 4 5 4

2 1 2 2 2 0 2 1

41 33 33 38 43 47 36 40

4 3 4 2 3 3 2 1

Table 6: Descriptive Statistics: Returns of the Euro Area Stock Markets This table presents the annual returns (in percent) of the euro area stock markets between January 1, 1999, and December 31, 2008. The average return over the 1999-2008 period is calculated from the geometrical average.

year 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Total

AEX 22.09 -5.17 -22.96 -45.13 4.52 3.04 22.70 12.58 4.04 -74.06 -7.54

BEL20 -5.08 -9.94 -8.36 -31.76 10.28 26.76 19.08 21.23 -6.13 -77.13 -5.92

CAC40 41.29 -0.54 -24.80 -41.17 14.95 7.14 21.02 16.15 1.30 -55.65 -2.01

DAX30 33.00 -7.84 -22.06 -57.88 31.54 7.08 23.96 19.87 20.12 -51.71 -0.39

15

DJESTX50 38.35 -2.73 -22.62 -46.68 14.57 6.67 19.29 14.08 6.57 -58.64 -3.07

IBEX35 16.85 -24.52 -8.14 -33.01 24.81 16.01 16.72 27.61 7.07 -50.14 -0.67

MIB30 20.13 1.68 -30.39 -30.06 11.19 15.58 12.48 16.15 -6.68 -66.17 -5.45

OMXH 96.31 -11.21 -39.22 -42.17 4.35 3.20 27.10 16.43 18.65 -76.38 -0.29

16

Table 7: Descriptive Statistics: Returns of Euro Area Sectorial Stock Markets

year 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Total

auto 0.66 -19.35 -0.72 -35.12 20.61 1.33 18.14 23.33 22.27 -58.11 -2.66

bank 18.51 3.71 -20.46 -31.26 26.99 10.30 23.54 20.31 -9.28 -101.44 -5.74

chem 30.92 3.90 -21.85 -28.47 12.99 11.21 26.93 17.39 23.91 -50.49 2.68

con 31.50 -9.17 1.12 -32.90 21.03 19.79 25.43 31.95 -3.09 -65.58 2.03

fins 6.59 15.61 -25.31 -38.88 9.00 25.83 34.55 39.65 -6.37 -78.62 -1.78

food -18.95 21.81 -12.27 -15.33 -3.15 5.17 20.18 20.52 10.48 -46.14 -1.75

health 2.65 41.48 -6.74 -52.90 6.65 22.67 16.26 6.09 -5.20 -36.10 -0.51

housg 20.13 -11.97 -19.16 -32.33 14.89 -1.40 23.31 10.40 11.07 -52.37 -3.67

ind 65.00 -8.46 -36.75 -39.40 26.61 5.26 22.90 18.92 13.85 -67.27 0.07

ins 5.27 19.24 -33.35 -72.69 15.27 7.13 27.52 15.14 -10.57 -64.14 -8.72

media 53.91 -18.00 -26.03 -74.15 14.91 9.51 12.49 7.24 -3.73 -43.52 -6.51

o&g 29.97 9.22 -4.68 -18.75 4.41 10.28 26.81 6.89 4.05 -50.06 1.83

res 53.67 -31.18 5.96 -28.56 14.24 15.32 10.48 56.49 12.67 -88.17 2.11

retai 20.43 -41.20 -15.37 -49.52 0.44 -1.67 11.93 21.38 8.60 -51.35 -9.18

techn 85.47 -2.61 -46.56 -72.93 25.08 -12.51 19.77 3.33 11.87 -72.66 -5.99

telec 71.92 -55.83 -33.27 -44.68 21.33 17.66 -7.20 10.92 15.47 -32.74 -3.58

trav 8.72 4.15 -16.27 -38.05 19.52 -0.54 27.92 18.68 -10.17 -44.51 -3.01

util -2.70 3.21 -11.27 -39.22 17.75 23.17 23.31 30.06 21.41 -47.74 1.81

This table presents the annual returns (in percent) of euro area sectorial stock markets between 1999 to 2008. The average returns over the 1999-2008 period is calculated from the geometrical average.

17

0. 00

0. 05

R S q u a 0. 10 r e d

S e m i 0. 20 P a r t i a 0. 15 l

0. 25

0. 30

Nam e of

O bser vat i on or Cl ust er

O ilG as Chem i c Ut i l i t Basi cRFoodBe Heal t h Aut om o Const r I ndust Per son Ret ai l Tr avel

Figure 2: Cluster Diagram

Banks I nsur a Fi nanc M edi a Tel eco Techno

18

Table 8: Correlation Matrix

AEX BEL20 CAC40 DAX30 DJESTX50 MIB30 IBEX35 OMXH mro euribor

AEX 1.000∗∗∗ 0.817∗∗∗ 0.916∗∗∗ 0.833∗∗∗ 0.935∗∗∗ 0.843∗∗∗ 0.829∗∗∗ 0.659∗∗∗ −0.047∗∗ −0.049∗∗

BEL20 0.817∗∗∗ 1.000∗∗∗ 0.783∗∗∗ 0.699∗∗∗ 0.780∗∗∗ 0.712∗∗∗ 0.708∗∗∗ 0.488∗∗∗ −0.053∗∗ −0.065∗∗∗

CAC40 0.916∗∗∗ 0.783∗∗∗ 1.000∗∗∗ 0.864∗∗∗ 0.970∗∗∗ 0.886∗∗∗ 0.870∗∗∗ 0.690∗∗∗ −0.052∗∗ −0.048∗∗

DAX30 0.833∗∗∗ 0.699∗∗∗ 0.864∗∗∗ 1.000∗∗∗ 0.921∗∗∗ 0.803∗∗∗ 0.785∗∗∗ 0.616∗∗∗ −0.051∗∗ −0.044∗∗

DJESTX 0.935∗∗∗ 0.780∗∗∗ 0.970∗∗∗ 0.921∗∗∗ 1.000∗∗∗ 0.898∗∗∗ 0.893∗∗∗ 0.731∗∗∗ −0.048∗∗ −0.046∗∗

IBEX35 0.829∗∗∗ 0.708∗∗∗ 0.870∗∗∗ 0.785∗∗∗ 0.893∗∗∗ 0.846∗∗∗ 1.000∗∗∗ 0.642∗∗∗ −0.051∗∗ −0.059∗∗∗

MIB30 0.843∗∗∗ 0.712∗∗∗ 0.886∗∗∗ 0.803∗∗∗ 0.898∗∗∗ 1.000∗∗∗ 0.846∗∗∗ 0.620∗∗∗ −0.053∗∗ −0.056∗∗∗

OMXH 0.659∗∗∗ 0.488∗∗∗ 0.690∗∗∗ 0.616∗∗∗ 0.731∗∗∗ 0.620∗∗∗ 0.642∗∗∗ 1.000∗∗∗ −0.052∗∗ −0.049∗∗

MRO −0.047∗∗ −0.053∗∗ −0.052∗∗ −0.051∗∗ −0.048∗∗ −0.053∗∗ −0.051∗∗ −0.052∗∗ 1.000∗∗∗ 0.932∗∗∗

Euri −0.049∗∗ −0.065∗∗∗ −0.048∗∗ −0.044∗∗ −0.046∗∗ −0.056∗∗∗ −0.059∗∗∗ −0.049∗∗ 0.932∗∗∗ 1.000∗∗∗

This table presents the correlation coefficients of the returns of the AEX Index (Netherlands), the BEL20 Index (Belgium), the CAC40 Index (France), the Performance DAX30 Index (Germany), the IBEX35 Index (Spain), the Milan MIB30 Index (Italy), the OMX Helsinki Index (Finland), the DJEurostoxx50 Index (Eurozone), the ECB main refinancing operations rates (MRO) and Euribor 3 months. (***),(**) and (*) indicate the significant results of levels of 1%, 5% and 10%, respectively.

Table 9: Cumulative Abnormal Returns for the Main Euro Area Stock Markets This table presents the cumulative abnormal returns for the AEX Index (Netherlands), the BEL20 Index (Belgium), the CAC40 Index (France), the Performance DAX30 Index (Germany), the IBEX35 Index (Spain), the Milan MIB30 Index (Italy), the OMX Helsinki Index (Finland), the DJEurostoxx50 Index (Eurozone). ∆M RO∗Sign corresponds to the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗degree corresponds to the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign and ∆M RO ∗degree take -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. The cumulated abnormal returns are estimated from the constant mean returns model. Estimation period goes from 244 days to 10 days before the announcement. Event period goes from -2 days to +2 days around the announcement. 0 is the announcement date. (***).(**) and (*) indicate the significant results of levels of 1%. 5% and 10%, respectively.

Stock Markets Indexes AEX BEL20 CAC40 DAX30 DJESTX50 IBEX35 MIB30 OMXH

∆M RO∗ Degree -1 0 +1 −1.78 −0.14 −0.91 −0.64 0.22 0.30 −1.30 −0.27 −0.31 −0.88 −0.12 −0.41 −0.87 −0.12 −0.48 −0.78 0.01 0.44 −2.19 −0.12 −0.57 1.54 0.15 −1.18 N=8 N=140 N=9

-1 −0.02 0.06 0.04 0.04 0.39 −0.20 −0.25 6.02 N=7

∆M RO∗ Sign 0 +1 −0.29 0.19 0.13 1.26 −0.37 0.42 −0.37 0.42 −0.22 0.29 −0.06 1.14∗∗ −0.26 −0.05 0.05 −1.29 N=146 N=4

Table 13: Cumulated Abnormal Returns of Euro Stock Markets according the Economy States.

This table presents the cumu-

lated abnormal returns for the industry Stoxx indexes according the business conditions. ∆M RO∗Sign corresponds to the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗degree corresponds to the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign and ∆M RO∗degree take -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. The cumulated abnormal returns are estimated from the constant mean returns model. Estimation period goes from 244 days to 10 days before the announcement. Event period goes from -2 days to +2 days around the announcement. 0 is the announcement date. The economy states are defined from the methodology of McQueen and Roley (1993). (***).(**) and (*) indicate the significant results of levels of 1%. 5% and 10%, respectively.

Indexes

∆M RO∗Sign

∆M RO∗degree

Stock Markets -1

0

+1

Panel A1: Low Economy State, Non Cyclical Continue on the next page

19

-1

0

+1

∆M RO∗degree

Stock Markets Indexes

-1

∆M RO∗Sign

0

+1

-1

0

+1

Continued from the previous page djestx600o&g

−2.97

0.45

.

−2.71

0.18

.

djestx600chem

1.38

0.10

.

5.12

0.06

.

djestx600food

−0.89

0.51

.

5.16

0.18

.

djestx600health

−1.35

−0.01

.

0.13

−0.16

.

djestx600res

−2.96

−0.39

.

−7.72

−0.39

.

djestx600util

−4.69

−0.62

.

−5.93

−0.89∗

.

0.02

.

−0.07

.

Panel A2: Low Economy State, Cyclical djestx600con

1.89

0.06

.

6.53

djestx600housg

1.59

0.03

.

7.74∗∗

djestx600ind

1.58

−0.04

.

3.27

0.02

.

djestx600retai

−0.27

0.15

.

1.98

0.03

.

djestx600trav

1.61

0.94

.

6.76

0.81

.

Panel A3: Low Economy State, Growth djestx600media

−1.42

−0.06

.

2.62

−0.31

.

djestx600techn

0.56

0.09

.

4.86

−0.03

.

djestx600telec

−0.25

−0.35

.

1.40

−0.41

.

Panel A4: Low Economy State, Financials djestx600bank

−0.99

0.71

.

−2.88

0.66

.

djestx600fins

−1.11

0.75∗

.

2.89

0.47

.

djestx600ins

0.17

0.22

.

−6.12

0.46

.

.

17.82∗

−1.04

.

Panel A5: Low Economy State, Automobile djestx600auto

7.32

−1.29

Panel B1: Medium Economy State, Non Cyclical djestx600o&g

−0.95

0.19

0.47

−4.81

0.27

−0.43

djestx600chem

0.10

−0.03

1.07

−5.72∗

0.09

0.69

djestx600food

2.73

−0.14

−0.68

3.71

−0.02

−1.83

djestx600health

0.58

−0.21

1.06

3.40

−0.16

0.19

djestx600res

−1.48

1.41

−8.32

djestx600util

1.61

−0.40

3.21

0.09

−1.50

−0.02

−2.43

0.20

−0.54

0.50∗∗ 0.02

0.55∗∗

1.51

Panel B2: Medium Economy State, Cyclical djestx600con

−0.17

0.16

Continue on the next page

20

∆M RO∗degree

Stock Markets Indexes

-1

∆M RO∗Sign

0

+1

-1

0

+1

−0.56∗

0.29

−5.76∗∗

−0.55∗

0.26

0.12∗

2.43

−2.06

−1.10

1.77

−0.22

−1.68

1.47

−1.23

−0.66

1.33

Continued from the previous page djestx600housg

−2.40∗

djestx600ind

−0.07

djestx600retai

0.17

djestx600trav

−0.65

−0.30 0.45∗

0.22∗∗ −0.32 0.43∗

1.54 −1.07 −0.37

Panel B3: Medium Economy State, Growth djestx600media

−0.75

−0.70

djestx600techn

0.49

0.27

2.54∗

1.62

0.29

2.63

4.44∗∗

0.20

0.69

7.68∗∗

0.26

0.86

djestx600telec

Panel B4: Medium Economy State, Financials djestx600bank

−2.02

−0.01

−1.18

−5.48∗∗

0.00

−1.74

djestx600fins

0.42

0.05

−0.69

−0.37

0.06

−0.71

djestx600ins

−0.38

0.10

−2.82∗∗∗

2.62

−0.08

−1.17

−11.60∗∗∗

−0.19

−0.19

Panel B5: Medium Economy State, Automobile djestx600auto

−3.57∗

−0.16

−0.75

Panel C1: High Economy State, Non Cyclical 0.54∗

1.90

−0.48

0.43

0.37

1.65

djestx600food

−0.26

0.22

djestx600health

0.38 3.07∗∗

djestx600o&g

−0.48

djestx600chem

djestx600res djestx600util

−2.77

0.56∗

2.06

0.43

0.37

2.29

2.09

−0.26

0.29

1.35

0.93∗∗

0.49

0.38

0.91∗∗

0.95

0.22

2.100

3.07∗∗

0.25

3.59∗

0.04

1.29

−2.77

0.04

1.67

Panel C2: High Economy State, Cyclical djestx600con

0.48

0.08

2.03∗

0.48

0.12

2.15

djestx600housg

−1.17

0.05

−1.85∗

−1.17

0.08

−3.13∗∗∗

djestx600ind

0.81

−0.11

0.14

0.81

−0.08

−0.20

djestx600retai

0.07

0.36

0.12

0.07

0.35

0.08

djestx600trav

1.05

0.73∗

0.23

1.05

0.71∗

0.29

−0.10

−0.58

−0.32

Panel C3: High Economy State, Growth djestx600media

−0.10

−0.57

−0.56

djestx600techn

0.86

−0.47

−4.59∗∗

0.86

−0.44

−6.71∗∗∗

djestx600telec

0.81

−0.73

0.00

0.81

−0.76

0.97

Continue on the next page

21

∆M RO∗degree

Stock Markets Indexes

-1

∆M RO∗Sign

0

+1

-1

0

+1

Continued from the previous page Panel C4: High Economy State, Financials djestx600bank

0.31

0.22

1.41∗

0.31

0.26

1.08∗

djestx600fins

1.61

0.15

1.39

1.61

0.20

0.70

djestx600ins

−0.46

−0.11

1.42

−0.46

−0.09

1.72

1.39

−0.55

0.39

1.00

Panel C5: High Economy State, Automobile djestx600auto

−0.55

0.35

22

Table 10: Abnormal Returns for the Euro Area Stock Markets This table presents the abnormal returns for the AEX Index (Netherlands), the BEL20 Index (Belgium), the CAC40 Index (France), the Performance DAX30 Index (Germany), the IBEX35 Index (Spain), the Milan MIB30 Index (Italy), the OMX Helsinki Index (Finland), the DJEurostoxx50 Index (Eurozone). ∆M RO∗Sign corresponds to the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗degree corresponds to the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign and ∆M RO∗degree take -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. The cumulated abnormal returns are estimated from the constant mean returns model. Estimation period goes from 244 days to 10 days before the announcement. Event period goes from -2 days to +2 days around the announcement. 0 is the announcement date. (***).(**) and (*) indicate the significant results of levels of 1%. 5% and 10%, respectively. ∆M RO∗degree

∆M RO∗ -2 Panel A: AEX -1 −0.73 0 −0.22∗∗ 1 0.19 Panel B: BEL20 -1 −0.39 0 −0.12 1 0.22 Panel C: CAC40 -1 −0.67 0 −0.27∗∗ 1 0.38 Panel D: DAX30 -1 −0.24 0 −0.26∗∗ 1 0.63∗ Panel E: DJESTX -1 −0.30 0 −0.27∗∗∗ 1 0.48 Panel F: IBEX35 -1 0.45 0 −0.23∗∗ 1 0.64 Panel G: MIB30 -1 −0.67 0 −0.30∗∗∗ 1 0.18 Panel H: OMXH -1 −0.35 0 −0.33∗∗ 1 0.18

-1

0

∆M RO∗Sign

Events 1

2

-2

-1

0

1

2

−0.96 0.06 −0.32

1.09 −0.09 0.38

2.16∗ 0.02 0.25

−1.88 −0.02 −0.66

−0.54 0.04 0.59

1.53∗ −0.00 −0.21

−0.88 0.05 0.32

−0.53 0.09 −0.56

0.84 −0.11 0.28

−0.34 0.13 −0.25

−1.02 −0.02 −0.57

−0.42 −0.25∗∗∗ 0.53

0.13 0.19∗∗ −0.08

−0.02 0.04 0.28

−0.13 0.07 −0.41

−0.23 0.04 0.29

−0.19 −0.13 0.31

−0.23 −0.01 −0.59

0.92 −0.06 0.08

−0.29 0.08 0.05

−1.03∗ −0.01 −0.22

−0.57 −0.29∗∗∗ 0.76∗

−0.19 −0.04 −0.26

1.01 −0.03 −0.07

1.76 −0.00 0.32

−1.97 −0.01 −0.33

0.31 0.04 −0.77

0.88 −0.08 0.13

−0.50 0.17 −0.04

−1.32∗∗ 0.02 −0.37

−0.12 −0.27∗∗ 0.86∗

−0.68 0.06 −0.68

1.04 −0.07 0.38

1.29 0.10 −0.07

−2.27∗∗ −0.00 −0.22

−0.28 0.06 −0.72

1.28 −0.02 0.09

−0.45 0.12 −0.02

−1.13∗∗ −0.00 −0.31

−0.15 −0.28∗∗∗ 0.81∗∗

−0.58 0.04 −0.52

1.68 0.00 0.10

1.60 0.03 0.30

−2.15∗ −0.01 −0.40

−0.31 0.13 −0.40

1.24 −0.02 0.11

−0.69 0.17∗∗ 0.17

−1.46∗ −0.03 −0.08

0.25 −0.22∗∗ 1.16∗∗

−0.36 0.10 −0.26

1.34 0.02 −0.08

1.06 0.09∗ 0.35

−2.48∗ −0.05 −0.02

−0.39 0.09 −0.54

0.41 −0.01 −0.16

−0.80 0.13∗ −0.01

−0.75 −0.02 −0.05

−0.26 −0.33∗∗∗ 0.47

−0.58 0.07 −0.57

0.32 −0.01 0.03

1.39 0.03 0.10

−1.12 −0.03 −0.09

0.20 0.27 0.02

−0.58 −0.00 −0.28

−0.26 −0.35∗∗ 0.56

−1.40 −0.08 −0.93

5.20 0.31 −0.57

2.55 0.18 0.55

−0.08 −0.01 −0.89

−1.10 −0.03 −1.14

3.38∗ 0.25 0.04

0.13 0.17∗ 0.26

Figure 3: Moving Correlation Between DJEurostoxx50 and Euribor 1

0.8

0.6

0.4

0.2

0 2000

2001

2003

2004

-0.2

-0.4

-0.6

-0.8

23

2005

2007

2008

Table 11: Cumulated Abnormal Returns of Main Euro Stock Markets according the Economy States This table presents the cumulated abnormal returns for the AEX Index (Netherlands), the BEL20 Index (Belgium), the CAC40 Index (France), the Performance DAX30 Index (Germany), the IBEX35 Index (Spain), the Milan MIB30 Index (Italy), the OMX Helsinki Index (Finland), the DJEurostoxx50 Index(Eurozone) according the business conditions. ∆M RO∗S ign corresponds to the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗degree corresponds to the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign and ∆M RO∗degree take -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. The cumulated abnormal returns are estimated from the constant mean returns model. Estimation period goes from 244 days to 10 days before the announcement. Event period goes from -2 days to +2 days around the announcement. 0 is the announcement date. The economy states are defined from the methodology of McQueen and Roley (1993). (***).(**) and (*) indicate the significant results of levels of 1%. 5% and 10%, respectively. Stock Markets ∆M RO∗degree Indexes -1 0 +1 Panel A: Low Economy State AEX −4.76 0.82 −10.67∗∗∗ BEL20 −1.89 0.37 . CAC40 −2.81 0.40 . DAX30 −3.39 1.10 . DJESTX50 −2.22 −0.28 . IBEX35 −1.33 0.34 −2.84 MIB30 −4.24 −0.31 . OMXH −1.36 −0.80 3.61 Panel B: Medium Economy State AEX 0.44 −0.33 0.05 BEL20 2.09∗ 0.13 −0.96 CAC40 −1.17 −0.30 −0.42 DAX30 −0.57 −0.20 −0.48 DJESTX50 −0.87 0.06 −0.64 IBEX35 −0.23 0.20 1.33 MIB30 −2.70 0.05 −1.25 OMXH 2.08 0.48 −1.55 Panel C: High Economy State AEX 1.98 −0.39 0.49 BEL20 0.12 0.31 2.61 CAC40 0.76 −0.66 −0.47 DAX30 −0.38 −0.54 −0.57 DJESTX50 1.15 −0.37 −0.31 IBEX35 . −0.49 0.52 MIB30 1.63 −0.39 0.51 OMXH 11.50 −0.08 −8.54∗

24

-1

∆M RO∗Sign 0 +1

−6.71 −4.04 −3.16 . −3.22 −1.95 −2.42 4.31

−0.06 0.16 0.13 0.87 −0.39 0.03 −0.71 −1.10

. . . . . 0.32 . 3.61

2.67 2.09 1.79 −1.12 2.50 0.38 −1.86 4.13

−0.34 0.05 −0.41 −0.26 −0.08 0.17 −0.09 0.42

0.05 0.07 1.21∗ 0.41 0.53 1.33 −0.38 −1.92

1.98 0.12 0.76 −0.38 1.15 . 1.63 11.50

−0.35 0.31 −0.66 −0.54 −0.38 −0.50 −0.36 −0.08

0.23 2.61 −0.47 −0.57 0.01 1.31 0.17 −8.54∗

Table 12: Cumulative Abnormal Returns for the Euro Area Stock Markets This table presents the cumulative abnormal returns for the noncyclical, cyclical, growth, financial and automobile stock groups. The definition of these groupes are from a cluster analysis of supersectors of Stoxx. ∆M RO∗Sign corresponds to the sign of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗degree corresponds to the degree of unexpected announcements of the ECB main refinancing operations rates. ∆M RO∗Sign and ∆M RO ∗degree take -1 for an unexpected decrease, 0 for a statu-quo and +1 for an unexpected increase. The cumulated abnormal returns are estimated from the constant mean returns model. Estimation period goes from 244 days to 10 days before the announcement. Event period goes from -2 days to +2 days around the announcement. 0 is the announcement date. (***).(**) and (*) indicate the significant results of levels of 1%. 5% and 10%, respectively. Stock Markets ∆M RO∗degree Indexes -1 0 Panel A: Non Cyclical djestx600o&g −1.59 0.34∗ djestx600chem 0.66 0.11 djestx600res −0.90 0.28∗ djestx600food 0.62 0.07 djestx600health −0.19 0.16 djestx600util −1.85 −0.09 Weighted Average −0.80 0.11 Panel B: Cyclical djestx600con 0.77 0.12 djestx600housg −0.60 −0.28 djestx600ind 0.77 0.03 djestx600retai −0.02 −0.03 djestx600trav 0.63 0.63∗∗∗ Weighted Average 0.40 0.03 Panel C: Growth djestx600media −0.84 −0.55∗ djestx600techn 0.61 0.02 djestx600telec 1.77∗∗ −0.17 Weighted Average 0.72 −0.14 Panel D: Financial djestx600bank −1.05 0.18 djestx600fins 0.14 0.20 djestx600ins −0.19 0.06 Weighted Average −0.73 0.15 Panel E: Automobile djestx600auto 1.27 −0.22 N=8 N=140

25

+1

-1

∆M RO∗Sign 0

+1

1.02 1.29 2.13 0.57 0.91 0.54 0.91

−2.12 0.06 −2.47 2.09 1.07 −2.07 −0.82

0.34∗ 0.16 0.30∗∗ 0.10 0.15∗ −0.10 0.12

0.42 1.14 2.30 −0.42 0.55 0.12 0.46

0.89 −0.64 1.49 −0.53 −0.02 0.58

1.27 −0.09 0.71 0.97 1.80 0.80

0.15 −0.29 0.10 −0.07 0.59∗∗∗ 0.05

0.62 −1.08∗∗ 0.86 −0.49 −0.09 0.20

0.62 −0.29 0.29 0.05

0.30 2.05 2.67 2.06

−0.58∗∗ 0.02 −0.16 −0.18

0.62 −0.85 0.75 0.23

−0.08 0.16 −1.10 −0.36

−1.93 1.43 −1.10 −1.48

0.19 0.17 0.02 0.14

−0.57 −0.10 −0.08 −0.40

−0.02 N=9

1.28 N=7

−0.18 N=146

0.17 N=4