Market efficiency in KOSDAQ: A volatility comparison between main boards and new markets using a permanent and transitory component model

In this paper, we examine the market efficiency in KOSDAQ in the context of the relative market-macrostructure of main board and new market, focusing on volatility which is a market-microstructure variable. For this, we apply the Engle and Lee (1999) component model to the indices representing the main boards and new markets of Korea along with the U.S. and the U.K. stock markets, where the component model decomposes volatility into a permanent and a transitory component. In order to compare the estimates of each volatility component in main board and new market of a country, we develop new test statistics. We use the following indices for analyses: the KOSPI200 stock index and the KOSDAQ50 stock index, the S&P 500 index and the Nasdaq 100 index, the FTSE 100 index and the FTSE AIM all-shares index. We use intraday data exclusively for the KOSPI200 stock index (Jan. 2000 to Dec. 2005) and the KOSDAQ50 stock index (Nov. 2000 to Dec. 2005). We also use daily data for all related indices from Jan. 1990 to Dec. 2005, except for the KOSDAQ50 stock index (Jan. 2000 to Dec. 2005) and the FTSE AIM all-shares index (Jan. 1996 to Dec. 2605) due to limited data availability. Our paper is the first paper which evaluates the market efficiency in KOSDAQ by taking into account its economic status in the Korean stock markets as well as comparing KOSDAQ to Nasdaq and AIM, the most successful new markets in the world with similar economic status to that of KOSDAQ in the stock markets of their respective countries. We obtain the following results. <Results using intraday data> Both components are statistically significantly persistent, longer than a trading day, for both the KOSPI200 stock index and the KOSDAQ50 stock index for all return periods except the fifteen-minute return. Unexpectedly, the transitory component of the KOSDAQ50 stock index is less persistent than that of the KOSPI200 stock index. The relative magnitudes, the ratio of the transitory component to volatility, in KOSDAQ are statistically significantly larger than in the former Korea Stock Exchange (hereafter the KOSPI market) for the fifteen-minute and the one-hour returns. However, the difference in the relative magnitude between two markets is much smaller and is statistically insignificant for the returns in longer periods. <Results using daily data> The persistence of permanent and transitory components in the U.S. and the U.K. main boards and new markets is very similar to that in Korean main board and new market for which we have used both intraday and daily data. The only exception is a transitory component of FTSE AIM all-shares index, whose persistence is very short. The relative magnitudes in both KOSDAQ and the KOSPI market are fairly high compared to those in the U.S. and the U.K., but the difference in their relative magnitudes is not statistically significant. Since the IT bubble burst in 2001, however, the relative magnitude in KOSDAQ has become larger than that in the KOSPI market. On the other hand, both differences of relative magnitudes between main board and new market in the U.S. and the U.K are statistically significant. For the U.S., the relative magnitude in the Nasdaq 100 index is surprisingly smaller than in the S&P 500 index. For the U.K., as expected, the relative magnitude in the FTSE 100 index is smaller than in the FTSE AIM all-shares index. We draw the following conclusions from the above results: In general, the market efficiency in KOSDAQ is not inferior in any way to that in KOSPI market. Recently, however, the KOSDAQ has been relatively less efficient than the KOSPI market. In addition, in the context of macrostructure, KOSDAQ is comparable with Nasdaq and AIM. In other words, given the market efficiency of main board in Korea, U.S., and U.K. stock markets, the market efficiency of KOSDAQ is as great as that of Nasdaq, but it is larger than the market efficiency of AIM.