آیا سیاست پولی ، نقدینگی بازار سهام را تعیین می کند؟ شواهد جدید از منطقه یورو

The recent financial crisis has been characterized by unprecedented monetary policy interventions of central banks with the intention to stabilize financial markets and the real economy. This paper sheds light on the actual impact of monetary policy on stock liquidity and thereby addresses its role as a determinant of commonality in liquidity. Our results suggest that an expansionary monetary policy of the European Central Bank leads to an increase of aggregate stock market liquidity in the German, French and Italian markets. Furthermore, the effect of monetary policy is significantly stronger for smaller stocks, suggesting a non-linear impact of monetary policy on stock liquidity.

The liquidity of financial markets, defined as “the ease of trading” (Amihud et al., 2005), has recently attracted a lot of attention, as the financial crisis highlighted its role as a precondition for efficient markets. Although central banks all over the world tried to ease financial markets during the recent crisis period by means of massive monetary policy interventions, we know surprisingly little so far about the actual relationship of monetary policy on stock liquidity. Since Amihud and Mendelson (1986) suggested that stock returns are an increasing function of illiquidity, numerous successive studies investigated this relationship. Indeed, the empirical literature generally confirms the theoretical proposition that investors demand higher gross returns as compensation for holding less liquid stocks.1 Another well-established strand of the literature on asset liquidity documents that the liquidity of individual stocks exhibits significant co-movement, which is usually referred to as commonality in liquidity.2 Covariation in the liquidity of stocks implies that the illiquidity risk cannot be diversified and therefore illiquidity should be regarded as a systematic risk factor.3 Furthermore, the observed commonality suggests the assumption that there needs to be at least one common factor that simultaneously determines the liquidity of all stocks in a market, which might be monetary policy.4 The hypothesis we test in this paper is that the monetary policy of central banks is a common determinant of stock liquidity. In particular, we examine the relationship between the European Central Bank's (ECB) monetary policy interventions and the liquidity of German, French and Italian stocks. Interestingly, there are only few relevant theoretical approaches. The inventory paradigm of the market microstructure literature suggests that inventory turnover and inventory risk affect stock market liquidity.5 In a nutshell, this paradigm proposes that stocks are expected to be more liquid if market participants can cheaply finance their holdings and perceive low risk of holding assets. Since monetary policy influences both the costs of financing and the perceived risk of holding securities, it follows that monetary policy should also affect stock market liquidity. Similarly, Brunnermeier and Pedersen (2009) develop a model that addresses the interaction between funding liquidity and asset liquidity. Their model suggests that traders who face capital constraints experience difficulties to meet margin requirements and therefore fail to provide liquidity to the market. The other way around, a deterioration of market liquidity reduces traders' funding liquidity through higher margin requirements. This may lead to a loss spiral and a lower liquidity, higher margin equilibrium. Following this reasoning, an expansionary (restrictive) monetary policy eases (exacerbates) constraints for margin borrowing and thus, facilitates (impedes) the funding liquidity of market participants. Another argument could be that both monetary policy as well as stock market liquidity are closely linked to business cycle movements. Thus, we could expect a considerable impact of monetary policy on stock liquidity, where the real economy might serve as the transmission channel. Few academic studies empirically examine the relationship between monetary policy and aggregate stock liquidity, and their results are to some extent ambiguous. Goyenko and Ukhov (2009) document strong evidence for the U.S. market (NYSE and AMEX) that monetary policy predicts liquidity for the period from 1962 to 2003. A tightening of monetary policy, as indicated by positive shocks to the federal funds rate and negative shocks to non-borrowed reserves, is shown to decrease stock market liquidity. Moreover, the bond market seems to serve as a transmitter that forwards monetary policy shocks to the stock market. On the contrary, Chordia et al. (2005) report only modest predictive power of monetary policy for stock market liquidity. For a sample of NYSE traded stocks they find that an expansionary monetary policy is associated with a contemporaneous increase in aggregated liquidity only during periods of crisis. The authors measure monetary policy by means of net-borrowed reserves and the federal funds rate. Soederberg (2008) studies the influence of 14 macroeconomic variables on the market liquidity of three Scandinavian stock exchanges between 1993 and 2005 and also provides mixed evidence. He finds that the policy rate is able to predict market liquidity on the Copenhagen stock exchange, whereas broad money growth plays a major role on the Oslo stock exchange and short-term interest rates and mutual fund flows predict liquidity on the Stockholm stock exchange. However, no variable is able to forecast liquidity for all three exchanges. Similarly, Fujimoto (2003) studies the relationship between macroeconomic variables and liquidity for NYSE and AMEX stocks. For the period ranging from 1965 to 1982, a positive shock to non-borrowed reserves increases liquidity, whereas an increase in the federal funds rate decreases liquidity. However, for the period from 1983 to 2001, neither shocks to non-borrowed reserves nor to the federal funds rate are able to predict stock market liquidity. We find that an expansionary (contractionary) monetary policy of the ECB leads to an increase (decrease) in the liquidity of stocks, which is in line with the main findings of Goyenko and Ukhov (2009). However, we observe this relationship not only at the macroeconomic level for aggregate liquidity by using vector autoregressive (VAR) models, but also at the microeconomic level for individual stocks by applying panel estimations. Contrary to earlier studies, we are able to report non-linear effects on the individual stock level, i.e. that the effect of monetary policy becomes weaker the higher the market capitalization of the traded stock. Noteworthy, our findings are robust for three different markets (Germany, France, and Italy), seven measures of (il)liquidity (capturing trading activity, price impact and transaction costs) and two variables of monetary policy (base money growth and the Euro Overnight Index Average (EONIA) interest rate). We contribute to the existing literature in three ways. First of all, while previous research focuses primarily on the U.S. stock market and offers to some extent ambiguous results, this study investigates European data. The effect of monetary policy on stock market liquidity might differ between currency areas and across countries, particularly when taking the differences in the statutes and policy aims between central banks into account. We are not aware of any study analyzing in depth the impact of ECB monetary policy. Secondly, we extend the analysis of monetary policy and liquidity to the individual stock level. From a methodological point of view, the application of panel-fixed-effects gives much stronger evidence as some effects could be canceled out at an aggregated level due to (unobserved) heterogeneity among assets. Our panel approach controls implicitly even for unobserved time-invariant characteristics at the individual stock level. To our knowledge, this is the first study applying both panel and VAR models to this specific research question. As shown below, the results in our panel models are not only much more robust than the VAR analysis of the aggregated market, but also offer additional insights regarding individual interaction effects, i.e. the linkage between the effect of monetary policy and the market capitalization of stocks. This leads to the conclusion that unobserved heterogeneity across individual stocks might play an important role, which cannot be captured by an analysis of aggregated liquidity. Finally, we add additional insights by employing in this respect untested, but generally well-acknowledged measures for both monetary policy and asset (il)liquidity. The paper is structured as follows. Section 2 describes the data set and the applied variables, including the measures of monetary policy and (il)liquidity. Empirical results at the macro and micro level with respect to the German stock market are illustrated in detail in 3 and 4, while results from the French and Italian markets are presented as a robustness analysis. Finally, Section 5 summarizes the results and draws some conclusions.

This study sheds light on the role of monetary policy as a potential determinant of stock liquidity. We examine whether an expansionary (restrictive) monetary policy of the ECB increases (decreases) the liquidity of stocks both at the macro and micro level for the Xetra trading system, Euronext Paris and Milan stock exchange. In order to measure (il)liquidity we employ seven variablesthatcapturetheaspectstradingactivity,priceimpactandtransactioncosts.ThemonetarypolicyoftheECBisapproximatedeitherby the twelve-month growth rate of the monetary base or by the EONIA. Our findings can be summarized as follows. Firstly, to examine the relationship between monetary policy and aggregated market liquidity we use VAR models in order to take potential endogeneities into account. The Granger-causality tests favor the conclusion that the central bank policy causes stock market liquidity, while evidence for a reversed relationship is rather weak. These observations are consistent with the fact that the ECB clearly focuses on price stability, thereby being less activist with regard to other objectives. The estimatedimpulse response functi ons confirm that an expansionary monetary policy entails more liquid stock markets. Most signs of the responses of the aggregated market (il)liquidity measures after a shock in the monetary policy variables are well in line with our hypotheses, though those for a shock to the EONIA are in general not statistically significant. Secondly, we complement the macro analysis by means of panel estimations with stock-fixed effects a nd find that an expansionary (restrictive) monetary policy also significantly leads to an increase (decrease) in the liquidity of individual stocks. Interestingly, we find that individual stock characteristics like market capitalization, which have not been considered by prior research, indeed play a role in the relationship. In general, smaller firms seem to be more responsive to liquidity effects of monetary policy. This observation is in line with a scenario in which monetary policyaffectsfundingcostsoffirms.Especiallyinthecontextofbank-dependentsystemslikethoseoftheeuroarea,monetarypolicymay impact stock fundamentals through an increase in the cost of bank financing. Our results thus highlight the importance of considering severalmeasurescapturingdifferentaspectsoftheconceptofliquidityandmonetarypolicyvariables.Moreover,monetaryinterventions of central banks should be considered as a determinant of individual stock liquidity, which may help to explain observed commonality in liquidity, as well as variations in liquidity at the aggregated-market level. Our study leaves several doors open to further research. From a methodological point of view, the issue of cross-sectional homogeneity of the slope parameters in the panel models has to be investigated and tested in future work, as assuming fixed effects and including an interaction of monetary policy and market capitalization of stocks in our panel estimations might not be sufficient to account for all potential forms of cross-sectional heterogeneity. Furthermore, a content-related extension could, for instance, take the bond market into consideration. The impact of central bank policies on bond markets seems rather obvious due to open market operations and other monetary policy instruments. Moreover, as suggested by Keynesian arguments, the final effect of monetary policy on liquidity depends on the relative attractiveness of other asset markets (i.e. the bond market). A tightening (easing) of the monetary policy would for instance make bonds relatively more attractive compared to equities and part of the effect of monetary policy on stock market liquidity would be channeled through the bond market (flight-to-quality or flight-to-liquidity episodes). Noteworthy, existing literature supports this conclusion, as shown by Goyenko and Ukhov (2009) .In this respect, it should be noted that this effect does not change the causation direction observed in our study (from monetary policy to stock market liquidity), but solely concerns the transmission mechanism of monetary policy shocks to the stock market (potentially also through the bond market). Moreover, studying cross-market information could also turn out to be an interesting research question. Specifically, information across countries could play a role in determining to what extent co-movements towards low-liquidity levels across countries (for example in periods of global crisis) would determine the conduct of common monetary policy in the euro area. Such an extension would add information about a potential reverse causality in those periods