دینامیک انتظارات بازار سهام بین المللی

This paper develops a noisy rational expectations model of the way in which international investors adjust their expectations of asset payoffs in a given country in response not only to public information signals but also to private information signals whose precision differs across investors. The model predicts that the perceptions of investors in one country about the future market returns in another country are related differently to realized past returns depending on their informational disadvantage relative to other investors: the greater is that informational disadvantage, the greater is the change in perception associated with returns. The predictions are confirmed by monthly survey data of institutional money managers investing in developed markets from 1995 to 2000.

There is now extensive evidence that international flows of equity capital are positively correlated with the returns on the markets of the destination countries. Brennan and Cao (1997) have shown that this positive correlation can be due to an asymmetry of information between foreign and domestic investors. According to this asymmetric information theory, the expectations of less well-informed foreigners respond more elastically to economic news than do the expectations of better-informed domestic investors. Good news for example, which is associated with positive stock returns, leads foreigners to purchase stock from domestic investors, creating a positive association between foreigners’ purchases of domestic stocks and domestic stock returns. In this paper we develop the implications of the asymmetric information theory for changes in the degree of bullishness reported by domestic and foreign investors. Consistent with the theory, we find that there is a strong tendency for foreign institutional investors to become more bullish about a given national market following a positive return on that market. Thus, our evidence provides further support for the hypothesis that information asymmetry is an important determinant of international capital flows. The framework for our analysis is a dynamic version of the multi-asset noisy rational expectations model of Admati (1985), which we use to analyze the determinants of the proportion of investors in country m who are bullish and bearish about the equity market in country k. The basic assumption underlying the model is that domestic investors are better informed about the payoffs on domestic assets than are foreign investors. This causes the beliefs of foreign investors to be more sensitive to new public information than are the beliefs of domestic investors. As a result, foreign investors buy from domestic investors when there is good news and asset prices rise, and sell to domestic investors when there is bad news and asset prices fall, which can account for the observed trend-following behavior of foreign investors. In order to focus attention on the role of information asymmetry, exchange rate risk and interest rate differentials are ignored, and the analysis is conducted in a model with many trading periods but only a single terminal consumption period. The implications of the model are tested using survey data on the perceptions of institutional investors about the future returns on the stock markets of various countries. The primary data come from the Merrill Lynch monthly Fund Manager Survey, which is completed by around 250 large institutional managers around the world. We show that institutional investors tend to be more bullish about the stock market of their country of residence. More importantly, we show that foreign investors become more bullish (bearish) about a country as the returns of that country's market portfolio increase (decrease), while this is not true of domestic investors. This is consistent with foreign, but not domestic, investors being at an informational disadvantage relative to all investors in a given market. Although the very notion of informational asymmetries among institutional investors in different countries may seem suspect in the world of modern communications, recent work by Coval and Moskowitz, 1999a and Coval and Moskowitz, 1999b shows that physical propinquity, even within the same country, is important for the portfolio composition and profits of institutional investors (Grinblatt and Keloharju, 2001), and Hau (2001) shows that the profits of professional traders in German stocks are higher for those located in Germany or in German-speaking countries. These papers are part of a growing literature on differential information endowments in financial markets and the role of distance in portfolio composition.1Brennan and Cao (1997) and Brennan and Aranda (1999) use a framework similar to the one presented here to predict trend-following behavior by foreign investors in international equity and debt markets respectively, and provide evidence of trend following by foreigners in certain markets. Further evidence of trend-following behavior by foreigners has been reported by Karolyi (2002), Froot et al. (2001), Choe et al., 1999 and Choe et al., 2001, Kim and Wei (2002), Grinblatt and Keloharju (2000), Griffin et al. (2003), Richards (2004), Edison and Warnock (2003), and Dahlquist and Robertsson (2004). On the other hand, Hamao and Mei (2001) argue that foreign investors in Japan tend to be long-term contrarians, and Grinblatt and Keloharju (2000) argue that foreign institutional investors are the most sophisticated class of investors in Finland, despite the fact that they pursue momentum strategies. Interpretation of the implications of this body of evidence for the differential information hypothesis is complicated by the fact that different studies use data at different frequencies, ranging from transactions data to quarterly flow data. Moreover, while the theoretical model of information asymmetry predicts a contemporaneous relation between flows and returns, realistic information and decision-making lags suggest that flows from less well-informed foreign investors might well lag returns, at least at high frequencies. Many of the empirical studies report only the results of regressing flows on lagged returns and interpret their findings in terms of “trend-chasing” behavior or “momentum investing” without asking why it is foreign investors who are prone to such behavior. Even a finding of a contemporaneous relation between flows and returns may be difficult to interpret in terms of information asymmetry since such a relation could be caused either by the price impact of foreign capital flows or by the use of portfolio insurance strategies by foreign investors; Leland (1980) discusses the conditions under which investors will find it optimal to follow portfolio insurance strategies. Therefore, in this paper, instead of examining the relation between flows and returns, we examine the relation between reports of investor perceptions and returns. Understanding how the expectations of different classes of investor are formed and change over time is important for a better understanding of the portfolio choice and dynamic trading strategies of investors and firms in international financial markets. Despite the continuing home bias in investor portfolios, during the last two decades holdings of foreign equities have become increasingly important, cross-border equity flows have increased dramatically, and international cross-listing of shares has become common ( Pagano et al., 2002). The remainder of the paper proceeds as follows. Section 2 extends the model first developed by Brennan and Cao (1997) to an economy with a large number of trading sessions. Its implications for the behavior of the fraction of investors in a given country who are bullish about the stock market of another country are developed in Section 3. Section 4 describes the data and Section 5 reports the empirical findings, which imply that foreign institutional investors in a given country's equity market are at an informational disadvantage relative to the average investor in that market. The evidence is particularly strong for the Japanese market and for Japanese institutions in foreign markets.

In this paper we have developed new implications of the hypotheses that domestic institutional investors enjoy an informational advantage and foreign institutional investors an informational disadvantage in a country's equity market relative to the average investor in that market. In particular, we have shown that, under certain auxiliary assumptions, the hypothesis that domestic institutional investors have an informational advantage implies that the change in the fraction of domestic institutional investors who are bullish about the domestic market will be negatively related to domestic market returns. The hypothesis that foreign institutional investors have an informational disadvantage implies that the change in the fraction of institutional investors in a given foreign country who are bullish about a particular market will be positively related to that market return. We tested these hypotheses using survey data on institutional investor views on four different equity markets. There is strong evidence that, consistent with the foreign (institutional investor) information disadvantage hypothesis, the fraction of bullish foreign institutional investors tends to increase following a rise in the domestic equity market. This is most pronounced for investors who are domiciled in Japan, and least pronounced for investors who are domiciled in the U.S. It is also most pronounced for foreign institutional perceptions with respect to the Japanese market and least pronounced for perceptions with respect to the U.S. market. There is also evidence of a domestic (institutional investor) information advantage for institutions domiciled in the U.K. and the U.S., in that for both countries the proportion of investors who are bullish about the domestic market declines following a positive return on that market; there is no evidence of such a domestic information advantage for E.U. or Japanese institutions. These findings are robust to the inclusion of control variables to capture the possibility that changes in bullishness are influenced by lagged market returns or lagged changes in bullishness, by changes in bullishness among institutions domiciled in the country of the market concerned, or by returns on the investor's home market. The evidence of a foreign institutional investor disadvantage is consistent with the arguments made by Brennan and Cao (1997) that international portfolio flows of equity capital are driven, at least in part, by informational considerations, and while there is only limited data on inter-country portfolio flows, changes in perception are shown to be significantly related to reported aggregate monthly portfolio flows from the U.S. to both the E.U. and Japan.