動量策略應用於台灣股市

Abstract

本文研究期間從1996年1月至2004年12月以每月資料形成投資組合來進行分析；將研究對象先區分為資訊電子類以及非資訊電子類，以公司的市值在各產業前100大的公司為研究對象來進行分析探討。參考Lee and Swaminathan (2000) 之「動量生命週期」理論，以股價報酬率與周轉率來選擇投資組合，找出何種分類及投資組合可建構出投資績效持續性為較佳之投資策略。 實證結果顯示，週轉率確實可以幫助判斷股票價格動量的大小與方向，但是產業別的影響與報酬率效果類似，皆為不顯著。當以二維來建構投資組合的依據時，在短、中期下低週轉率之投資組合皆顯著優於高週轉率之投資組合，存在週轉率效果；長期時，低週轉率組合的報酬越來越顯著優於高週轉率組合，也就是說長期下台灣股市同樣呈現過度反應的現象。 採用買入低週轉率贏家同時賣出高週轉率輸家組合之早期策略，其績效將遠勝過僅以報酬率為組合建構基礎下之一維簡單動量策略。同時一維簡單動量策略之績效，又比採用買入高週轉率贏家同時賣出低週轉率輸家之晚期策略效果好。

This paper uses the trading data which’s market value are top 100 in electronics industry and non-electronics industry fully quoted on TSEC from 1996 Jan. to 2005 Dec. Following Lee and Swaminathan (2000) in momentum life cyclehypothesis, choices turnover ratio and price return ratio to build investment portfolio. To find out which classification and holding period can get the better performance and sustains. With MLC model, the feasible strategies are formed as follows: buying past winner with low turnover ratio and selling past loser with high turnover ratio at the early-stage (defined as early momentum strategy); buying past winner with high turnover ratio and selling past loser with low turnover ratio at the later-stage (defined as late momentum strategy). The empirical evidence shows that simple momentum strategy is not necessarily profitable less than one year; however, the low turnover ratio winner outperforms the simple momentum strategy. Turnover ratio and past performance is helpful to determine the stock-investing portfolio. For intermediate and short period, the low turnover ratio portfolio outperforms high turnover ratio one. The main results are as follows: the early momentum strategy profits most regardless of time-span. Either early or late momentum strategy outperforms simple price-momentum strategy in short and intermediate term; however, the results will reverse in longer than one year.