Degree of market imperfections: evidence from four Asian index futures markets

Abstract

The degree of market imperfections has important implications for the behaviour of stock index futures. This work extends the evidence regarding the degree of market imperfections in three ways. First, this work represents the first attempt to compare the market imperfections of four Asian index futures markets. As anticipated, the degrees of market imperfections are significantly higher in the emerging markets (Korea and Taiwan) compared to the developed markets (Japan and Hong Kong). Second, this work compares the relative performance of three alternative volatility estimators in estimating the degree of market imperfections: the bivariate error correction GARCH(1,1) model, the exponentially weighted moving average (EWMA) and the power EWMA. The comparison results provide support for the conclusion that among the volatility estimators examined, the bivariate error correction GARCH(1,1) model performs the best. Third, this work tests the stationarity of the degrees of market imperfections. The empirical results demonstrate that the degrees of market imperfections for all four Asian markets are stationary, suggesting that the estimates of the degrees of imperfections obtained from ex post data are useful to investors on an ex ante basis.

Acknowledgement

The author would like to thank National Science Council of the Republic of China, Taiwan for financially supporting this research.

Notes

¹Two primary market imperfections are clarified: market imperfections between markets and market imperfections within a market. The former represents the imperfections involved when implementing activities (e.g., arbitrage) between two or more markets. The latter comprises the imperfections involved when implementing activities within a market. A model for valuing the degree of market imperfections is developed on the basis of the linkage of two markets (that is, the spot and futures markets).

²Under the Morgan Stanley Capital International (MSCI) taxonomy, Japan and Hong Kong are categorized as ‘developed’ markets. Korea and Taiwan are classified as ‘emerging’ markets.

³Writing the cost of carry model in return form, Brooks et al. (1999) demonstrated that the futures and index returns are perfectly and positively contemporaneously correlated.

⁴Empirical findings for various financial markets suggest that the bivariate error correction GARCH(1,1) model can also improve hedging performance [e.g., Kroner and Sultan (1993); Park and Switzer (1995); Yang and Awokuse (2003); and Floros and Vougas (2004)].

⁵The regression results show that all of the estimated coefficients (β) on the degree of market imperfections are significantly positive at the 1% level, indicating a significantly positive relationship between the degree of market imperfections and the absolute error across all estimators for all four Asian markets. Furthermore, the regression results are omitted to save space.

⁶In Japan, cash dividends are clustered at the end of March and September. In Korea, most of the companies pay dividends at the end of the year. As for the Taiwanese shares, cash dividends for the underlying component stocks are mostly paid only once per year, and are concentrated in July.