تاثیر شوک سیاست پولی بر قیمت سهام: شواهد از کانادا و ایالات متحده

Using structural VAR models with short-run restrictions appropriate for Canada and the United States, we empirically examine whether trade and financial market openness matter for the impact on and transmission to stock prices of monetary policy shocks. We find that, in Canada, the immediate response of stock prices to a domestic contractionary monetary policy shock is small and the dynamic response is brief, whereas in the United States, the immediate response of stock prices to a similar shock is relatively large and the dynamic response is relatively prolonged. We find that these differences are largely driven by differences in financial market openness and hence different dynamic responses of monetary policy shocks between the two countries that we model in this paper.

Stock markets are notoriously sensitive to unexpected changes in monetary policy. But this sensitivity may vary across different economies. In this paper we investigate whether the response of stock markets to changes in monetary policy differ significantly between a small open economy (Canada) and a large and relatively closed economy (the United States). There are two reasons to hypothesize that in small open economies a contractionary domestic monetary policy shock would have a smaller negative influence on stock prices and that domestic monetary policy shocks have a relatively smaller contribution to overall stock price volatility. First, in small open economies domestic monetary policy takes the world interest rate as given, and as such has a relatively limited influence on this discount rate that matters for domestic stock prices. At the same time, shocks originating from the rest of the world have a larger impact on stock prices in small open economies. This is in part due to the fact that small open economies operate in international financial markets and unexpected capital inflows and outflows tend to be a larger share of their GDPs, and in part because small open economies tend to have less diversified economic structures than large and relatively closed economies do.1 Second, in small open economies the exchange rate is an important component of propagation mechanisms, especially compared to in large economies. International trade is typically a larger share of GDP in small economies than in large and relatively closed economies. Thus, the degree of trade openness may influence the impact and transmission of domestic monetary policy shocks on domestic asset prices to the extent that monetary policy affects the exchange rate. 2 Given the above considerations and the fact that Canada is a quintessential small open economy, and the United States is a large and relatively closed economy, in this paper we address two important and related empirical issues in the context of Canada and the United States: whether trade openness and financial market openness matter for the monetary policy shocks’ impact on and transmission to stock prices. Our empirical results based on the comparison between Canada and the United States suggest that indeed openness matters significantly in terms of the overall response of stock prices to unanticipated changes in monetary policy. Using structural VAR models with short-run restrictions, we find that in Canada the immediate response of stock prices to a domestic contractionary monetary policy shock is small and that the dynamic response is brief. By contrast, in the United States, the immediate response of stock prices to a domestic contractionary monetary policy shock is relatively large and the dynamic response is relatively prolonged. We also provide an economic explanation of these differences, which are largely driven by differences in financial market openness and hence different dynamic responses of monetary policy shocks between the two countries that we model in this paper. Our findings underscore for Canada the relative importance of financial market openness for stock prices. We find that unanticipated changes in the U.S. federal funds rate significantly affect the forecast error variance of the Canadian stock prices. On the other hand, we find that the overall impact of external demand shocks on Canadian stock prices is relatively small. One interpretation of this finding is that the floating exchange rate regime provides a cushion for the transmission of the external demand shocks to the real sector in Canada. While monetary policy primarily influences aggregate demand, the stock market serves as an important channel of the monetary transmission mechanism. Stock prices influence financial wealth, and hence affect consumption, investment and labor supply decisions (Poterba, 2000 and Lettau and Ludvigson, 2004). Moreover, in both Canada and the United States, the share of financial assets in household wealth grew rapidly during the time period we examine in this paper.3 Thus, it is increasingly important to understand whether, and how, monetary policy affects stock prices over time in the context of trade and financial market openness. Despite the growing prominence of the link between asset prices and aggregate demand, most empirical open economy models continue to rely on the conventional Mundell–Fleming framework. In this framework, given that households divide a fixed amount of wealth between a bond portfolio and domestic money holdings, one can write down the equilibrium conditions for either the bond market or the money market, which typically is the one specified in the empirical models. The bond portfolio, on the other hand, consists of domestic and foreign bonds, which are perfect substitutes. Consequently, empirical models appeal to the uncovered interest parity condition to link interest rates for domestic and foreign bonds to the exchange rate. In our analysis, for the Canadian economy, we adopt the basic features of the Mundell–Fleming-type small open economy model. However, we extend this framework by including stocks in the menu of assets in the domestic portfolio. We allow stock prices to depend on real and nominal variables, and allow stock prices to change in response to portfolio shocks. We model the U.S. economy following the lead of Christiano et al., 1996 and Christiano et al., 2000, but also augment their model by including stocks as a component of wealth. From a methodological standpoint, the core of our empirical analysis is the identification of the impact of monetary policy shocks on stock prices after taking into account the important interactions among main macroeconomic variables. To examine this impact, we use structural vector autoregressive models (VARs), which impose restrictions on the contemporaneous (“short-run”) effects of shocks upon certain variables included in the model ( Bernanke, 1986 and Sims, 1986). There are several important advantages to estimating the interaction between monetary policy and stock prices using short-run restrictions: it allows us—within a unified empirical framework—to impose (arguably) more plausible restrictions based on economic and structural considerations, to use statistical model selection techniques in conjunction with macroeconomic theory, and to identify structural shocks, including unanticipated changes in monetary policy. Earlier studies also provide a useful guidance on this choice. Kim and Roubini (2000) argue that structural VARs resolve a number of anomalies detected in the empirical small open economy recursive VAR models. Christiano et al. (2006) argue that structural VARs with short-run restrictions yield remarkably sharp inference in the context of response analysis with structural shocks. 4 The structural differences between Canada and the United States, and the macroeconomic dependence of the small open Canadian economy on the large and relatively closed U.S. economy are the distinguishing features of our analysis. It is perhaps surprising that the existing literature has not fully explored how these differences affect the link between stock prices and monetary policy shocks. In a framework that is closest to ours, Lastrapes (1998) uses the same structural model with long-run restrictions for eight countries (G7 plus the Netherlands), but does not control for their potential differences in economic structure, and does not allow for macroeconomic interdependence. 5 Consequently, our analysis builds on a framework that explicitly models the structural differences between Canada and the U.S., and that accounts for the differences in their responses of stock prices to monetary policy shocks in an internally consistent way. We organize the remainder of the paper as follows. In Section 2, we provide a brief review of the literature on stock prices and monetary policy shocks, and discuss the relevant monetary policy instruments in Canada and the U.S. In Section 3, we present our baseline VAR models, and discuss the identification strategy. In Section 4, we present and discuss our key empirical findings. In Section 5, we evaluate the robustness of our results to alternative identification restrictions. We conclude in Section 6.

In this study, we evaluated the economic significance of the stock prices in the transmission of domestic monetary policy shocks in Canada and the United States by incorporating stock prices into empirical monetary business-cycle models featuring open and closed economies, respectively. We relied on macroeconomic theories to impose short-run restrictions on the structural VAR models and to identify impulse responses, which provide valuable economic insights. We found that in response to an unanticipated 25 basis points increase in the instrument interest rate, stock prices in the U.S. decline by about 4 percent within seventeen months after the shock, and in Canada they only decline by about 0.8 percent within merely four months after the shock. These differences are largely attributable to the different dynamic responses of domestic short-term interest rates to monetary policy shocks. In Canada, the interest rate response is rapid, but not very persistent, whereas in the United States the response is prolonged. In the model, we paid attention to the differences in trade and financial market openness between these two economies, especially through external demand and monetary policy shocks. We found that monetary policy shocks in the United States have significant impact on the Canadian stock prices and contribute substantially to their volatility. We note that the contribution of external demand shocks to Canadian stock price volatility is very small. Our results, therefore, suggest that incorporating wealth effects into empirical open economy monetary business-cycle models is important in understanding the transmission of monetary policy shocks. This analysis did not include real estate and other forms of wealth. Incorporating these refinements, and examining whether they have substantive influence on our findings from VAR models with short-run restrictions are left for future research.