قیمت گذاری ریسک ارز در بازار سرمایه کانادا

Based on a three-factor international capital asset pricing model, we examine whether the world market, the local market and the currency risks are priced in the Canadian equity market. The analysis presented in this paper is based on data collected from 2003 to 2010. As the dataset also includes the period of global financial crisis, we examine the issue of risk pricing in the full sample as well as in before and after global financial crisis periods. Unlike most existing studies, the empirical results presented in this paper are based on (i) the quasi maximum likelihood estimation (QMLE) based multivariate GARCH-in-Mean specification and (ii) the generalized method of moments (GMM) techniques. Our empirical analysis based on weekly data on 58 largest Canadian firms indicates that the currency as well as the local and the world market risks are priced in the Canadian equity market. This result holds for all exchange currency rates proxies and in all sample periods. We find that the price of the world market, the local market and the currency risks is time-varying and the Canadian equity market is partially segmented.

Finance theory suggests that changes in exchange rates affect the value of assets thereby creating currency risk exposure (Adler and Dumas, 1984 and Jorion, 1990). In addition, within the context of an international capital asset pricing model (ICAPM), it is argued that in the absence of purchasing power parity (PPP), optimal portfolios differ across countries and the return that investors expect on their investment includes a premium for both market and exchange rate risks (De Santis and Gérard, 1998). More recent studies on exchange rate risk pricing (for example Phylaktis and Ravazzolo, 2004, Tai, 2008, Saleem and Vaihekoski, 2008, Saleem and Vaihekoski, 2010, Antell and Vaihekoski, 2007 and Antell and Vaihekoski, 2012) suggest that currency risk pricing is time-variant. Most studies that consider the possibility of time-varying exchange rate risk pricing utilize a modified version of De Santis and Gérard's framework. De Santis and Gérard argue that unconditional models of exchange rate risk pricing are unable to detect the time-varying currency risk. The multivariate GARCH process suggested by De Santis and Gérard is the most common method used to quantify the pricing of currency risk and to test whether risk pricing is time-varying. This method makes use of the variance of each risk factor over time, which is regressed against stock returns conditional on a number of instrumental variables. For example, Saleem and Vaihekoski (2010) allowed for the conditional local influence. They analyzed the size of the risk premia due to time-varying sources of risks and analyzed the potential drivers of risk premium. Antell and Vaihekoski (2012), while extending their earlier work published in 2007, modified De Santis and Gérard's framework. In doing so they relied on Ding and Engle's (2001) covariance stationary specification to test a conditional ICAPM for both Finland and Sweden (in periods of floating and fixed currency rates). The main aim of this paper is to provide a comprehensive analysis of whether or not the exchange rate risk is priced in the Canadian equity market. Canada is a member of G7 (group of seven major industrialized economies) and the North American Free Trade Area (NAFTA). Even before joining NAFTA, Canada was heavily involved in bilateral trade with the US.2 Among the G7 economies, Canada is regarded as the most open (Kia, 2013). Under the floating exchange rate system, the value of Canadian firms is more likely to be affected by fluctuations in foreign currency exchange rates. Through its impact on the expected return on securities, exchange rate fluctuations can be a source of systematic risk on stocks (see Bodnar et al., 1996). The global financial crisis that started in 2008 also affected the Canadian economy. There was a significant decrease in the gross domestic product (GDP), which adversely affected the Canadian imports. Canada also experienced a decline in demand for its exports. However, this crisis had only a short-term effect on Canada's net exports. From 2009, both exports and imports increased gradually and returned to their pre-crisis level soon after (Li and Li, 2013).3 The collapse of Lehman Brothers in September 2008 generated a wave of shocks that affected financial markets across the globe. The failure of this large investment bank exacerbated the liquidity and inter-bank lending problems that resulted in a significant decline in stock market indices in most developed as well developing economies. The foreign exchange markets were also rattled by the higher level of uncertainty caused by the global financial crisis. Due to its strong trading relations with the US, Canadian dollar is almost always significantly affected by fluctuations in the US economy (Melvin and Taylor, 2009). From November 7, 2007 to March 9, 2009, the Canadian dollar depreciated against the currencies of its major trading partners by approximately 28% (Caporale et al., 2013). In summary, the global financial crisis influenced the Canadian economy and therefore the value of the Canadian dollar. Fluctuations in the value of the Canadian dollar can, among other things, affect the expected future cash flows of firms as well as their cost of capital. As a result, changes in the value of the Canadian dollar can affect the values of Canadian firms. Whether or not the exchange rate risk is priced in the Canadian equity market is likely to be of interest to Canadian investors. Depending on their individual characteristics and risk tolerance, fluctuations in the value of the Canadian dollar are likely to have an asymmetric impact on firms and investors. In addition, the trading strategies employed by Canadian investors are unlikely to be identical to those utilized by their counterparts in other developed countries. The Canadian equity market also offers an interesting case study due to the fact that a large number of publicly listed firms in Canada are export oriented. The paucity of Canadian studies in the area of exchange rate risk pricing is noticeable. It is hard to find studies that have focused exclusively on Canada. Except for Samson (2013), most other studies have considered Canada as a part of panel data or multi-country studies.4 But none of these studies provide an indepth analysis of exchange rate risk pricing in the Canadian stock market. Unlike Samson (2013) which, using monthly data from January 1970 to December 2004, mainly focuses on the impact of exchange rate and inflation risk on firm value, this paper examines the issue of exchange rate risk pricing in the Canadian stock market. We use weekly data from January 2003 to December 2010. This allows us to also consider the presence of exchange rate risk pricing in before and after global financial crisis periods. In addition, using the quasi maximum likelihood based GARCH-in-Mean and the generalized method of moments (GMM) estimation procedures, we allow the exchange rate risk pricing to be time-varying. The empirical analysis presented in this paper is based on both an unconditional and a conditional three-factor ICAPM. The three factors are: the world equity market, the Canadian local market index and an exchange rate changes proxy. We use three alternative exchange rate proxies; trade weighted exchange rate index (TWI), the US dollar (USD) and the Euro (EUR). In other words, in addition to using a multilateral exchange rate, we also separately use two proxies for bilateral exchange rates. The sample size used in this study is large – weekly observations on 58 large Canadian firms from 2003 to 2010. Our empirical analysis shows that, in the case of both the multivariate GARCH-in-Mean and the GMM procedures, the exchange rate as well as the world and local market risks are priced in the Canadian equity market. This conclusion holds in all cases – i.e., both for a multilateral and two bilateral exchange rate proxies. Based on the results presented in this paper, it can be argued that Canadian equity returns do include a significant premium for exchange rate fluctuations. In addition, we find that the world, local and currency risks are time-varying. Finally, the Canadian market is found to be partially segmented. The rest of this paper is organized as follows. Section 2 contains a review of the related literature. Methodology is discussed in Section 3. Section 4 contains a description of the data and empirical results. Section 5 concludes the study.

Using weekly data from 2003 to 2010, this paper provides a comprehensive analysis of the issue of exchange rate pricing in the Canadian equity market. We consider the possibility of exchange rate risk pricing in the full as well as before and after global financial crisis periods. The empirical analysis presented in this paper is based on a three-factor international capital asset pricing model. The three factors are (i) local market risk, (ii) the world market risk and (iii) the exchange rate risk. The exchange rate risk is approximated by a multilateral and two bilateral exchange rates. Trade weighted index (TWI) is used as a measure of the multilateral exchange rate, whereas the US dollar (USD) and the Euro (EUR) are used as bilateral exchange rate proxies. Unlike most existing studies, we first examine the issue of the pricing of these risk factors by means of a multivariate GARCH-in-Mean procedure, which involves the use of the quasi maximum likelihood estimation (QMLE) technique. In addition, we also examine the issue of risk pricing and the presence of time-varying risk by means of the generalized methods of moments (GMM). The GMM estimation involves a set of instrumental variables that are used to approximate the information set. We use several important variables as risk factors. Conditional on the set of the most recent available information, the expected return of firm is regressed on the three factors. Our empirical analysis suggests that the local market, the world market and the currency risks are priced in the Canadian equity market. This result also holds in the pre and the post-global financial crisis sub-periods. The general result concerning the pricing of the exchange rate risk in Canadian stock market presented in this paper is consistent with prior studies on Canada, such as Vassalou (2000), Moore and Wang (2013) and Samson (2013). The empirical results presented in this paper also suggests that (i) partially segmented asset pricing models are relevant for economies like Canada and (ii) all risk pricing factors, including the local market risk, in the Canadian equity market are time-varying. We find that the Canadian stock market is partially segmented. The regulatory restrictions adapted by the Canadian Government in the aftermath of the global financial crisis can be viewed as a source of segmentation (Melvin and Taylor, 2009). Finally, the issue of time-varying risk in the Canadian stock market requires further investigation. Furthermore evidence of segmentation presented in this paper highlights the need for modeling the integration of the Canadian market with other markets by means of a dynamic process.