شاخص های سیاست های پولی ایالات متحده و بازده سهام بین المللی: 1970-2001

It is documented in the literature that U.S. and many international stock returns series are sensitive to U.S. monetary policy. Using monthly data, this empirical study examines the short-term sensitivity of six international stock indices (the Standard & Poor 500 [S&P] Stock Index, the Morgan Stanley Capital International [MSCI] European Stock Index, the MSCI Pacific Stock Index, and three MSCI country stock indices: Germany, Japan, and the United Kingdom) to two major groups of U.S. monetary policy indicators. These two groups, which have been suggested by recent research to influence stock returns, are based on the U.S. discount rate and the federal funds rate. The first group focuses on two binary variables designed to indicate the stance in monetary policy. The second group of monetary indicators involves the federal funds rate and includes the average federal funds rate, the change in the federal funds rate, and the spread of the federal funds rate to 10-year Treasury note yield. Dividing the sample period (1970–2001) into three monetary operating regimes, we find that not all policy indicators influence international stock returns during all U.S. monetary operating periods or regimes. Our results imply that the operating procedure and/or target vehicle used by the Federal Reserve Board (Fed) influences the efficacy of the policy indicator. We suggest caution in using any monetary policy variable to explain and possibly forecast U.S. and international stock returns in all monetary conditions.

Recent research suggests that certain groups of monetary policy indicators or variables based on the federal funds rate or the discount rate have the ability to help explain and/or possibly forecast U.S. and international stock returns. Since equities represent claims on future profits of firms, which, in turn, are generated from future economic output, changes in monetary policy to raise or lower interest rates should measurably impact stock returns. The standard discounted cash flow model posits that increasing interest rates should adversely affect stock returns while decreasing interest rates should positively impact stock returns. Changes in interest rates will be felt directly through the discount rate used in the model. An additional indirect impact will flow from the anticipated increase or decrease of cash flows due to changed economic activity. The focus of this study is to empirically examine the sensitivity of several international stock total return series to Federal Reserve Board (Fed) monetary policy as reflected by two groups of monetary indicators controlled by the Fed: the federal funds rate and the U.S. discount rate. We are particularly interested if the influences on stock returns by these monetary indicators are robust to different historical and current operating procedures and/or target variables used by the Fed to control the money supply. Bernanke and Blinder (1992, p. 902) suggest that monetary policy affects aggregate demand through the demand for bank credit. They state, “…we entertain the idea that the federal funds rate (or the spread between the funds rate and some alternative open-market rate) is an indicator of Federal Reserve policy.” Extending this notion, Patelis (1997) and Thorbecke (1997) analyze the ability of several federal funds variables to forecast stock returns. Applying VAR methodology to stock return data supplied by CRSP for the sample period 1953–1990, Thorbecke (1997) finds a statistically significant negative relationship between changes in the federal funds rate and industry and size portfolios. Thorbecke (p. 648) also presents additional evidence from an event study that “…there is a statistically significant negative relation between policy-induced changes in the funds rate and changes in the DJIA (Dow Jones Industrial Average) and the DJCA (Dow Jones Composite Average).” With a multifactor VAR model, Patelis (1997) examines the impact of the federal funds rate and the spread of the federal funds rate to the 10-year Treasury note rate on monthly NYSE value-weighted excess stock returns obtained from CRSP. The sample period ranges from January 1962 to November 1994 with analysis focused on monthly, quarterly, annual, and biennial horizons. Patelis finds the federal funds rate and the federal funds spread to be highly significant (P values of .000 and .017, respectively). In a different approach to measuring monetary policy, Jensen, Mercer, and Johnson (1996) use a binary variable to indicate the direction of monetary policy. They define monetary policy as either expansive or restrictive based on the direction of change in the discount rate. Using time series regressions on monthly and quarterly data over a sample period from February 1954 through December 1991, they find that security prices are sensitive to macroeconomic variables as a function of the Fed monetary policy. For ease of description, we label this binary variable approach to describing monetary policy as a discount rate regime (DRR). Using t tests, F tests, and time series regressions, Conover et al., 1999a and Conover et al., 1999b and Johnson, Beutow, and Jensen (1999) extend the use of the DRR policy variable to analyze monthly international stock and mutual fund returns. Using a sample period that extends from January 1970 through December 1995, Conover et al., 1999a and Conover et al., 1999b find that stock returns are generally higher during periods of expansive U.S. as well as local monetary policy. In Johnson et al. (1999), the sample period runs from January 1976 through September 1998. Using the DRR methodology over the period from January 1970 to December 2001, we count nine distinct periods of increasing and eight decreasing discount rates. In addition, during this same period, we observe the Fed using several distinct operating procedures and/or target variables to implement monetary policy. To describe a period when the Fed uses a procedure or variable to control the money supply, Ogden (1990) suggests the notion of a monetary operating regime. He defines a monetary operating regime as a period when the Fed uses a distinct set of procedures and/or target variables to implement the desired monetary policy (restrictive or expansive). With this definition, a monetary operating regime or period is thus differentiated from monetary policy or stance, and together they constitute the current monetary condition. We examine two questions in this study. First, are U.S. and international stock index returns consistently sensitive to the two groups of monetary policy indicators in light of the several historical monetary operating regimes evident from 1970 to 2001? Second, are any of these monetary policy variables statistically important during the current monetary operating regime of specifically targeting the federal funds rate and thus can be used as a current indicator of monetary policy? This study extends previous research in several ways. First, it provides a side-by-side comparison of the two groups of monetary indicators that are most prevalent in current research. Second, this study examines the impact of these various monetary indicators across several distinct monetary operating regimes. Third, this study extends analysis through December 2001. Section 2 outlines the definition of a monetary operating regime, while Section 3 describes data and methodology used this analysis. We present our empirical results in Section 4 and conclude our analysis in Section 5.

While both binary monetary policy indicators are statistically significant when examined across the entire sample period, we find spotty results of their statistical influence on international stock returns by monetary operating period. In the current monetary operating regime of targeting the federal funds rate, we do not find Europe, Germany, and UK stock returns to be sensitive to the DRR variable of Conover et al., 1999a and Conover et al., 1999b and Jensen et al. (1996) either singularly or in a regression framework with an economic activity control variable. In the regression analysis, we find low statistical sensitivity of U.S. stock returns to DRR and no indication of sensitivity when returns are sorted into expansive and restrictive periods. We do find Pacific and Japan stock returns strongly sensitive to DRR in the regression analysis. Most likely, this is due to the U.S. being the largest trading partner of Japan. However, we are concerned about the efficacy of DRR as a robust indicator of monetary policy to explain stock returns in all monetary conditions and especially when the Fed uses a targeted federal funds rate. We find that stock returns seem more sensitive to our hybrid MPR variable in the regression studies especially during Period 3. This suggests possible use of the targeted federal funds rate as an indicator of monetary stance in regression analysis in addition to economic activity variables. We also find that monetary indicators based on the monthly average federal funds rate produce spotty results by monetary operating periods. Perhaps, the most consistent indicator to impact stock returns is the first difference in the federal funds rate. Some of this result may be due to short-term frequency of the analysis. We conclude that Fed operating procedures and/or target variables impact the sensitivity of international stock returns to these five historical monetary policy indicators. The results suggest caution in using any monetary policy variable to explain and possibly forecast U.S. and international stock returns in all monetary conditions.