یکپارچگی و وابستگی متقابل بازارهای سهام و ارز : چشم انداز استرالیا

This paper examines the integration of the Australian stock market with its two leading trading partners, the US and Japan. In investigating the extent of integration, this study takes into account the interdependence between foreign exchange rates and stock prices, since exchange rates influence international competitiveness of firms, and, via interest rates, the cost of capital. The results indicate that there was a stable long-run relationship among the Australian, US and Japanese markets prior to the Asian crisis but that this relationship disappeared in the post-Asian crisis period. An analysis of the short-run dynamic linkages among markets suggests that, following the Asian crisis, the US influence on the Australian market diminished while the influence of Japan remained at a modest level. Furthermore, the impulse response analysis indicates only a contemporaneous transmission of shocks from one market to other markets. Confidence intervals for impulse responses are estimated using the bootstrap-after-bootstrap method.

This paper investigates the extent of stock market integration between Australia and its two leading trading partners, the US and Japan. The interdependence between stock prices and foreign exchange rates is taken into consideration in evaluating the transmission of stock price shocks across countries. In addition, this study determines whether the extent and nature of stock market integration in the period of the post-Asian crisis differs from that of the pre-Asian crisis. Despite Australia's sound macroeconomic fundamentals, its growing business links with newly industrializing Asian countries and geographical proximity to Asia made it vulnerable to this crisis. In this paper, an attempt is made to identify whether the Asian crisis has changed the extent of Australia's integration with the US and Japanese stock markets. Our interest in this study was spurred by several factors. First, any potential gain from international diversification of a portfolio is inversely related to the extent of stock market integration. A low correlation between returns on national stock indices allows investors to minimise portfolio risk by international diversification. Thus, an analysis of the long-run comovement of national stock prices and their short-run temporal relationships is important for managing an international portfolio. Second, financial deregulation increased the extent of integration between the Australian and US markets (Ragunathan et al., 1999). Moreover, the financial influence of Japan on Pacific-Basin countries following the deregulation of their markets has increased and overtaken that of the US (Phylaktis, 1999). Thus, it is important to examine whether the findings for those Pacific-Basin countries hold for Australia in the post-deregulation period. Third, previous studies generally indicated that intermarket relationships at the time of the stock market crash of October 1987 intensified temporarily for a period around the crash, but then quickly resumed their pre-crash relationships (Roll, 1989). A recent study by Janakiramanan and Lamba (2000) also confirms the finding of a high degree of integration for the brief periods around the October 1987 crash and the 1997 Asian crisis. However, previous studies do not shed light on whether the Australian stock market has become more or less sensitive to the US and Japanese markets in the post-Asian crisis period. At the onset of the Asian crisis, the Federal Treasurer, Peter Costello, proclaimed that the Australian economy was ‘fireproof’, i.e. not vulnerable to crisis in Asian emerging markets. This assertion was based on Australia's sound economic fundamentals and the general premise that the Australian market is primarily influenced by its top trading partners and major international stock markets such as the US and Japan. However, subsequent events indicated that both the Australian dollar and stock prices fell significantly following the crisis in South Korea and Indonesia. If these events did lead to any permanent changes in investors’ perceptions about Australia's linkage with the developed, vis-à-vis emerging, Asian markets, then one would expect a persistent change in Australia's linkage with Japan and the US following the Asian crisis. Fourth, interdependence in stock prices across countries reflects economic integration in the form of trade linkages and foreign direct investment. The widely used dividend discount model suggests that the current share price equals the present value of future cash flows, which depend on the earnings growth of a company. The latter, in turn, partly depends on the macroeconomic conditions of the domestic economy and its major trading partners.2 Thus, comovement of the underlying macroeconomic variables across national economies may lead to comovement of stock prices in these economies.3 As Australia's economic fundamentals are linked to its top two trading partners, the US and Japan, one would naturally expect a reflection of these linkages in stock markets. Fifth, the foreign exchange risk is an important consideration while investing in foreign stocks. The stock market integration literature on this issue can be broadly classified into three categories: (1) studies that ignore exchange rates and use stock indices in terms of local currencies (e.g. Bracker et al., 1999 and Ragunathan et al., 1999); (2) studies that express stock indices in terms of a common currency (e.g. Taylor and Tonks, 1989 and Bekaert and Harvey, 1997); and (3) studies that evaluate whether exchange rate risks are priced in international asset markets (e.g. Dumas and Solnik, 1995).4 The first category of research is based on the presumption that the stock price risk is separable from the foreign exchange risk. The second category of research is driven by the belief that there is a positive equi-proportional relationship between the local currency value of stocks and the exchange rate (value of foreign currency in terms of domestic currency). Put differently, an instantaneous arbitrage condition holds for stock prices, which ensures that if domestic currency depreciates by one percent then investors require a one percent increase in domestic stock prices as compensation for the erosion of the value of their investment portfolios. Our empirical investigation (to be presented in the next section), however, shows that the stock price differential between two countries deviates from that implied by the instantaneous arbitrage condition. Thus, in modelling stock market integration one can neither ignore foreign exchange risks nor impose an instantaneous arbitrage condition. The third category of research overcomes these limitations by evaluating the impact of foreign exchange rates on international stock prices but ignores any potential influence of stock prices on foreign exchange rates. The model used in this paper is flexible enough to capture the dynamic interdependence between stock prices and exchange rates. This study makes the following contributions. First, we examine long-run comovements of stock prices for Australia, Japan and the US and the Australian dollar value of the Japanese yen and US dollar using the cointegration approach. Second, based on the cointegration results from the first step, multivariate time series models are specified to examine dynamic linkages among national stock prices as well as the interactions between stock prices and exchange rates. More specifically, a vector error correction (VEC) model is used for two sub-periods prior to the Asian crisis, while a vector autoregressive (VAR) model in first difference is employed for the post-Asian crisis period. The rationale for selecting different models for various sub-periods is given in the next section. These multivariate dynamic models also allow us to investigate the predictability of financial prices or pricing inefficiency of financial markets.5 Third, the impulse response analysis is conducted to examine how a shock originating in a stock market is transmitted to other stock markets over time. For statistical inference of impulse responses, the bootstrap method (Efron and Tibshirani, 1993) was adopted. It is based on repeated re-sampling of the observed data and has been found to provide a useful small sample alternative to the asymptotic method in many applications of time series econometrics (Li and Maddala, 1996 and Berkowitz and Kilian, 2000). In particular, recent studies by Kilian, 1998a and Kilian, 1998b found that it provides a superior small sample alternative to the asymptotic method for statistical inference in impulse response analysis. The layout of the remainder of this paper is as follows. Section 2 describes the data and investigates comovements in stock prices and exchange rates. Section 3 presents the model of dynamic linkages between national stock markets and discusses the results. Section 4 conducts the impulse response analysis. Section 5 summarises and concludes the paper.

Recent studies on the integration of the Australian stock market with major international markets (e.g., Ragunathan et al., 1999 and Janakiramanan and Lamba, 2000) ignore the interdependence of foreign exchange and stock markets. This paper takes into account the correlation between exchange rates and stock prices and investigates whether, and to what extent, the Asian crisis influenced both the long- and short-run intermarket relationships. More specifically, this study makes the following contributions. First, it examines the long-run relationship of stock prices for Australia, Japan and the US, and the Australian dollar value of the yen and US dollar using the cointegration approach. A cointegrating relationship among the variables in the system is found prior to the Asian crisis but no such relationship exists in the post-Asian crisis period. This result seems to indicate that prior to the Asian crisis financial prices in these countries were driven by a common international factor but after the Asian crisis the country-specific factors have become more important than the international factors, leading to lack of comovement in financial prices. Second, it employs multivariate time series models to analyse the dynamics of the system. The results indicate that in the pre-crisis period, the Australian stock market was primarily led by the US stock market (as opposed to Japan), however, in the post-crisis period, the Australian stock market has become more dependent upon its own past and less on the US stock market. Third, this paper observes that stock returns are led by foreign exchange rates but that the former do not significantly influence the latter. Finally, an impulse response analysis was conducted to investigate the transmission of a price shock from one market to other markets. The results reveal a quick transmission of shocks among the stock markets both before and after the Asian crisis. Thus, in spite of some evidence of lack of cointegration (long-run comovements) in financial prices in the post-Asian crisis period, the short-term linkages among markets have remained strong.