本文旨在探討投資人在避險和投機部位回饋交易的行為,本文以台灣期貨交易所(TAIFEX)提供獨特的資料庫,區分自然人、證券自營商、期貨自營商和外資,並使用向量自我廻歸模型(Vector Autoregression model,VAR) Granger因果檢定(Causality Test)來進行分析。研究結果發現,自然人不具有回饋的交易行為。期貨自營商為正向回饋交易者,外資為負向回饋交易者。此結果與Lin et al. (2005) 和Cheng et al. (2007) 的發現有所不同。Lin et al. (2005) 發現自然人是負向回饋交易者,期貨自營商、外資及證券自營商是正向回饋交易者。Cheng et al. (2007) 發現自然人與在穩定性分析中,除自然人與期貨自營商外,多空頭市場對投資人回饋交易行為有顯著的影響;除期貨自營商外,到期日效果以短天期效果對投資人回饋交易行為有顯著的影響;星期效果對投資人回饋交易行為並沒有影響;除期貨自營商與證券自營商外,在高流動性效果對投資人回饋交易行為最為顯著;自然人在高深度效果回饋交易行為最為顯著,外資則在低深度效果較為明顯。在此研究期間,自然人一直是虧損的狀態,而在期貨市場中獲利最大的是期貨自營商,其次是外資,接著是證券自營商。
This paper examines a feedback trading behavior for hedging and speculative positions by trader types. By analyzing a unique data set provided by Taiwan Futures Exchange (TAIFEX), we adopt VAR model and discuss the lead-lag relation between return and the hedging and speculating positions by trader types. Although no evidence shows that individual traders are feedback traders, futures dealer is a positive feedback trader and foreign investor is negative feedback trader. The results are inconsistent with findings of Lin et al. (2005) and Cheng et al. (2007). Lin et al. (2005) find individual trader is negative feedback trader, but futures dealer, foreign investor, and security dealer are positive feedback traders. Individual traders have opposite directions on feedback trading behaviors across up and down markets. In addition, maturity effects significantly affect feedback trading behavior by trader types except for future dealer, especially, in short maturity. However, day-of-the-week effects are not related by feedback trading behavior. Then, only for individual and foreign traders, liquidity effect and depth effect influence feedback trading behaviors. Finally, in this study, individual traders have loss money, while the largest profit in the futures market is the futures dealer, followed by foreign investor, followed by security dealer.