The Comparison of Optimal Price Formation Processes for Different Trading Performance Futures Contracts

• Main Authors: Mei-Hsing Cheng, 鄭美幸
• Other Authors: Hsin-Hong Kang
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/33922699591076749229

博士 === 國立成功大學 === 企業管理學系碩博士班 === 94 === 　The Taiwan futures market altered its price formation process of middle trading session from call auction to continuous auction on July 29, 2002. This dissertation investigates the effects of price formation process on the futures market quality by analyzing the liquidity ratio and P-GARCH (1, 1) model. Also, the optimal price formation processes for futures contracts with which trading performance are different are compared. Meanwhile, we use the method of Bessmbinder and Seguin (1993) to examine the volatility-volume relations for the Taiwan futures market, and investigate the effects of price formation process on the volatility-volume relations. 　The paper shows that for Taiwan Stock Index Futures (TX) liquidity and information efficiency are improved significantly, and the intraday volatility reduced, not significantly though, after transferring into the continuous auction. For the other two futures contracts, The Electronic Sector Index (TE) and the Banking and Insurance Sector Index (TF), continuous auction is helpful to increase liquidity and decrease volatility, but it is harmful to information efficiency. The results indicate that the continuous auction is optimal for the highly traded futures contracts, yet it may not for the thinly traded ones. 　Consistent with previous studies, we find that the trading volume is positively related to daily returns volatility, and the open interest is negatively related to volatility. The effects of unexpected components of trading volume and open interest on volatility are larger than those of expected components. The effect of price formation process on volatility-volume relations is not notable, which may because that the new price formation process is only applied to the middle trading session and remains the closing session unchanged. 　Besides, I found that the relations between volatility and trading activities are much stronger for TE and TF than for TX. Furthermore, under continuous auction the effects of expected trading volume and unexpected open interest on volatility and the magnitude of an increase in unexpected open interest lessens the impact of an unexpected trading volume on volatility is larger, and the ratio of unexpected volume effect on volatility to expected volume effect on volatility is smaller for TX, and all are reversed for TE and TF. The evidences show that an unknown linkage between the market performance and the volatility-volume relations exists.