سیاست های پولی و انتقال شوک های نفتی

This paper provides evidence on the role played by monetary policy in the transmission of oil shocks to the US economy. We show that for the period since 1986, oil shocks have had a negative effect on stock returns, regardless of whether the oil shock is defined as the percentage change in the price of oil or a nonlinear transformation of that series. We then demonstrate that there is no relationship between the reaction of individual stock prices to oil shocks and to monetary policy shocks. This implies that oil shocks do have effects on the economy beyond their effect on monetary policy. We conclude that systematic monetary policy is not as effective as suggested in some previous studies.

Monetary policy can affect the economy in two ways. The first, which has been the focus of most of the literature in the past two decades, is through shocks to the Federal Reserve’s policy instrument, typically the federal funds rate. The primary challenge when measuring the effect of monetary policy shocks is providing a plausible set of assumptions sufficient to identify exogenous movements in the policy instrument. Alternatively, the Federal Reserve can affect the economy through its choice of policy rule (see e.g., Sims and Zha, 2006, Bernanke et al., 1997, Cochrane, 1998, Sims, 1999, Hoover and Jorda, 2001, Hamilton and Herrera, 2004 and Bernanke et al., 2004 and the references contained therein). Woodford (2003, p. 7) takes the view that it is only the effect of systematic monetary policy that is of interest, as “under almost any view it is desirable to eliminate such shocks…to the greatest extent possible.” Measuring the effects of systematic monetary policy is difficult. BGW tested the hypothesis that it is the response of monetary policy to oil shocks, rather than the direct effects of oil shocks on production, that causes recessions. In an attempt to control inflation, the Federal Reserve pursues a contractionary monetary policy after an oil shock, and it is this tightening of monetary policy that causes output growth to slow.1 If the Federal Reserve had instead pursued a policy of not responding to oil shocks, according to this argument, the post-World War II recessions either would not have occurred or would have been less severe. The difficulty in evaluating this hypothesis empirically is that it requires an estimate of the path output would have followed under a policy regime different from the one that was actually in place. BGW estimated a structural VAR model for seven variables, including output and monetary policy. Impulse response functions for GDP after an oil shock were computed in the usual fashion. They then changed the equation describing the Federal Reserve’s reaction function to “shut off” the response of monetary policy to oil shocks, and computed new impulse response functions for output, assuming the coefficients of the other equations would not have been affected.2 The difference in impulse response functions is interpreted as an estimate of the effect of systematic monetary policy on the economy. They concluded that oil shocks may not have had any effect on output under the alternative policy rule, implying that systematic monetary policy is important, as it can even push an economy experiencing normal growth into a recession. Several authors have discussed in detail the shortcomings of the approach taken by BGW (see e.g. Sims, 1997, Hamilton and Herrera, 2004, Bernanke et al., 2004 and Sims, 1997 argues that the alternative policy rule imposed by BGW implies “unsustainably explosive behavior of prices”. Hamilton and Herrera (2004) are skeptical about BGW’s conclusions about the importance of systematic monetary policy for two reasons.3 First, they show that the monetary policy actions required to implement BGW’s alternative policy rule are implausibly large, as much as a 900 basis point reduction in the federal funds rate. Second, they question the specification of the estimated VAR model, demonstrating that monetary policy is far less powerful when more lags of the variables are included. Bernanke et al. (2004) suggest that it may be necessary to move toward a more structural approach in order to make further progress. This paper provides evidence on the role that monetary policy plays in the transmission of oil shocks to the economy, but using a methodology that is not subject to the criticisms of BGW.4 If oil shocks affect the economy only indirectly through monetary policy, then only interest-rate sensitive stocks will respond to fluctuations in the price of oil. Bernanke and Kuttner (2005) find variation in the responses of different industries to monetary policy shocks, with some industries reacting strongly to an unexpected monetary policy tightening, and others showing no reaction at all. The hypothesis that oil shocks do not have direct effects on the economy implies that oil shocks are nothing more than an indicator of the future direction of monetary policy. Hence, if that hypothesis is correct, there should be a similar pattern in the response of stock prices to oil shocks and to monetary policy shocks across industries. It is straightforward to test this hypothesis without making any strong assumptions. It is clear from the large literature on the oil price–GDP relationship that the choice of oil shock measure is important (see Hamilton, 2003 for a review). For this reason we begin by estimating the response of stock prices to different oil shock measures, to determine the correct specification of the relationship between oil price changes and stock returns. The rest of the paper is as follows. Section 2 provides estimates of the response of stock prices to linear and nonlinear measures of oil shocks. Section 3 compares the response of stock returns to both monetary policy shocks and oil shocks. We then investigate some explanations for our findings and discuss the implications for the importance of the Federal Reserve’s policy rule. Section 4 provides concluding comments.

This paper was motivated by the important work of Bernanke et al. (1997). They found that, “…a non-responsive monetary policy suffices to eliminate most of the output effects of an oil price shock…”.13 After allowing for changes in the underlying assumptions, their point estimates still suggest that monetary policy is responsible for two-thirds to three-fourths of the response of output following an oil shock. However, in commenting on their paper, Sims (1997) writes that, “It remains possible for a skeptic to maintain the view that the effects of both systematic and random shifts in monetary policy are negligibly small.” The finding that oil shocks have little or no direct effect on the economy is important both for understanding oil shocks and monetary policy, but is by no means a settled issue. The present paper complements the work of BGW by employing a different empirical methodology. We find that stock prices exhibit a significant response to oil shocks, whether the oil shock is defined simply as the change in the price of oil, or as a nonlinear transformation of oil prices, such as the net oil price increase introduced by Hamilton (1996). Large oil shocks explain as much as 28% of the same-day variation in individual stock prices. Our primary goal has been to determine whether the response of stock prices following an oil shock is due to the direct effects of oil prices on the economy, or due to the Federal Reserve’s reaction to oil shocks. Our results are, in fact, consistent with monetary policy playing no role in the transmission of oil shocks to the economy. It appears that oil shocks do have a substantial effect on the economy even in time periods when the Federal Reserve does not react to oil shocks.