روابط تعدیل قیمت و منجر به وقفه بین بازده های بازار سهام: شواهد ساختاری از بازار سهام تایوان

This paper investigates the price adjustment and lead-lag relations between returns on five size-based portfolios in the Taiwan stock market. It finds evidence that the price adjustment of small-stock portfolios is not slower than that of large-stock portfolios. Additionally, limited evidence supports a positive leading role of large-stock portfolio returns over small-stock portfolio returns. These two findings are substantially different from the results of previous research on developed markets.

Economists have spent many years analyzing security markets in order to understand the short-term movements of security returns. Up to now, there is substantial evidence that short-horizon security returns and their variations are predictable. Conrad and Kaul (1989) demonstrate that time variation in expected return accounts for large proportions, in excess of 25% for small firms, of the variance of both weekly and monthly portfolio returns. Fama and French (1988) report that 25–45% of the variations of 3- to 5-year stock returns are predictable from past returns. All the evidence poses a serious challenge to the long-held view that stock prices follow a random walk. It has been well known that individual stocks exhibit negative autocorrelations in short-horizon returns, e.g., Fama and French (1988) and Jegadeesh (1990). Moreover, empirical studies observe the profitability of contrarian strategies, i.e., selling winners and buying losers. Lo and MacKinlay, 1988 and Lo and MacKinlay, 1990 argue the contrarian profits result mainly from the existence of asymmetric cross- (instead of own-) autocorrelation of stock returns. The authors, introducing the lead-lag relations, discover that the lagged returns on U.S. large-stock portfolios are correlated with the current returns on U.S. small-stock portfolios, but the lagged returns on small-stock portfolios are not correlated with the current returns on large-stock portfolios. This type of asymmetric cross-autocorrelations suggests a strong leading role of large-stock returns over small-stock returns that cannot be fully explained by non-synchronous trading. Yet, not all economists agree on the importance of the lead-lag relations between stock returns. Jegadeesh and Titman (1995) and Hameed (1997) conclude that most cross-autocorrelations are due to stock-price overreaction and a very small fraction can be attributed to lead-lag relations. Boudoukh et al. (1994) summarize that a group of economists, called loyalists, view that large autocorrelations at short horizons are not due to fundamentals. Instead, they arise from the market frictions (e.g., non-synchronous trading, price discreteness or bid-ask spread), institutional structures or microstructure effects. The purpose of this paper is to study the price adjustment and lead-lag relations between size-based portfolio returns in the Taiwan stock market where the market microstructures are considerably different from mostly developed markets. In sum, this paper observes new microstructure evidence for the Taiwan stock market under mainly a reduced-form specification. The evidence can be regarded as a supplement and extension of previous research. Market microstructures, according to previous literature, frequently play a key role in explaining the lead-lag relations. Boudoukh et al. (1994), for instance, claim that Lo and MacKinlay, 1988 and Lo and MacKinlay, 1990 underestimate the non-synchronous-trading effect that induces the autocorrelations of (particularly small-) stock returns. Their analyses suggest that institutions and factors are the most likely sources of the autocorrelation patterns. Mech (1993) and Badrinath et al. (1995) focus on other market microstructures, such as transaction costs and institutional ownership, to explain the observed cross-autocorrelations. According to Mech (1993) and Hameed (1997), the difference in stock-price adjustment to information may be one of the primary causes of lead-lag relations. Prices of large firms, usually suffering less from a variety of market frictions such as the non-synchronous trading and transaction-cost problems, respond to (macro) news faster than the prices of small firms and, thus, lead the prices of small firms. One of other market frictions related to the stock-price adjustment to information is the price limit imposed on stock trading. Chang et al. (1995), studying this issue in the Tokyo stock exchange, suggest that price limits tend to induce autocorrelations of daily returns stronger than those of weekly returns and much stronger than those of monthly returns. Kim and Pitman (2000) find that volatile stocks, actively traded stocks, and small stocks hit price limits more often than other stocks. Kim and Rhee (1997) support that price limits prevent prices from efficiently reaching their equilibrium level (delayed price discovery hypothesis). Due to the differences in the market microstructures, many hypotheses held in mostly developed markets may not be held in Taiwan. For example, McQueen et al. (1996) assert that the returns on the largest-stock portfolio on average react to news sufficiently fast and can proxy for macroeconomic news, because the included stocks are usually traded frequently and incur low transaction costs. Yet, we observe that the less severe non-synchronous-trading and transaction-cost problems in Taiwan likely make the price adjustment of small-stock portfolios not slower than that of large-stock portfolios. Furthermore, the imposition of price limits could add an additional consideration that price limit can aggravate the delayed reaction for stocks responding to news and likely generate extra difficulty in analyzing the lead-lag relations. We, first, apply reduced-form models, often employed by previous papers to study the lead-lag relations in developed markets, and, then, propose revisions to accommodate the unusual microstructures in Taiwan. In Appendix A, furthermore, we explain the possible difficulty in interpreting economically the results derived from reduced-form models without considering the contemporaneous correlations between stock returns. Taking this issue into account, we modify the applied reduced-form models and re-estimate the lead-lag relations. As a conclusion, first, there exist strong contemporaneous correlations between size-based portfolio returns in the Taiwan stock market. Second, we find evidence that the price adjustment of small-stock portfolios is not slower than that of large-stock portfolios. Third, the cross-autocorrelations between the large-stock portfolio returns and other small-stock portfolio returns are mostly negative. Hence, a positive leading role of large-stock portfolio returns over small-stock portfolio returns does not exist in Taiwan, substantially different from the results of previous research on developed markets. The rest of this paper proceeds as follows. In Section 2, we describe the microstructures in the Taiwan stock market different from those in developed markets. In Section 3, we introduce our data and their summary statistics. Section 4 emphasizes the issues of this research and the models and then discusses our results. We summarize this paper in Section 5.

In this paper, we have studied the price adjustment and lead-lag relations in the Taiwan Stock Market where the market microstructures are different from mostly developed markets. First, most stocks were actively traded, regardless of their sizes. Second, transaction cost was symmetric and relatively low for all stocks. Third, the market was full of small, uninformed, and short-term investors. Fourth, price limits were imposed on all traded stocks. Because of those microstructures we find new evidence for the lead-lag relations between size-based portfolio returns in the Taiwan stock market. Comparing our results to those of previous papers focusing on similar issues in mostly developed markets, we observe two major differences. First, in the Taiwan stock market, we find evidence that the price adjustment of small-stock portfolios is not slower than that of large-stock portfolios. Second, we find no evidence for a positive leading role of large-stock portfolio returns over small-stock portfolio returns in Taiwan. The two differences may result from the discrepancies in market microstructures between the Taiwan stock market and other developed markets.