과거 수익률을 이용한 거래전략의 성과분석 (A Study on the Profitability of the Trading Strategies Using Past Returns)

서지정보

· 발행기관 : 한국재무학회
· 수록지 정보 : 재무연구 / 25권 / 2호 / 203 ~ 246페이지
· 저자명 : 김상환

초록

외국에서는 과거 3~12개월 동안 주가가 오른 종목을 매입하고 하락한 종목을 매도하는 계속거래전략이 유의적인 수익을 얻는다는 사실이 많은 연구들에서 밝혀지고 있다. 반면 국내 선행연구들은 우리나라에서는 정반대 전략인 반대투자전략이 성과를 보인다는 매우 이해하기 어려운 결과를 제시하고 있다. 본 연구는 우리나라에서 나타난 반대투자전략의 성과는 주식시장이 개인투자자들에 의해 주도되던 외환위기 이전에 국한된 현상임을 보였다. 위기 이후로는 우리나라에서도 외국과 마찬가지로 계속투자 전략이 유의적인 수익을 내는 것으로 나타나고 있다.
거래전략의 성과가 어떤 요인에 의해 발생한 것인가를 분석한 결과 주가의 과잉반응 현상이 크고 유의적으로 나타나 우리나라에서는 투자자의 심리적인 왜곡이 거래전략 성과를 결정하는 중요 원인인 것으로 나타났다. 또한 체계적 위험을 반영한 평균 수익률의 횡단면 분산 또한 거래전략 성과에 중요한 역할을 한 것으로 나타났다. 따라서 거래전략의 성과는 개별주식의 체계적 위험요인이나 행태적인 요인 중 어느 하나에 의해 발생한 것이 아니라 두 가지 요인의 상대적 크기에 의해 나타날 수 있음을 알 수 있다. 체계적 위험을 반영한 평균 수익률의 횡단면분산과 행태적인 요인 모두가 유의적이라 하더라도 전자가 크면 계속거래전략이 수익을 나타내고 행태요인 중 과잉반응현상이 크면 반대거래전략이 수익을 보일 수 있기 때문이다.

영어초록

This paper studies the profitability of the contrarian trading strategies in the Korean stock market in an attempt to provide some insights into why the trading strategies generate the abnormal profits. It has been reported in most foreign literature that the momentum strategies (i.e. buying stocks with high returns over the previous 3 to 12 months and selling stocks with poor returns over the same time period) earn significant profits for the following 3 to 12 months. Contrary to these empirical facts on foreign stock markets, many Korean academic papers reported that contrarian strategies (i.e. buying losers and selling winners) are significantly profitable in the Korean stock market. Yet no research effort has been made to explicate this strange phenomenon in the Korean market. The first objective of this paper, therefore, is to explain why this surprising phenomenon happened in Korea.
Secondly we attempt to determine the sources of the expected profits of the trading strategies that use the past information. In contrast to the previous literature using the circumstantial evidence on the causes of the profitable trading, we directly look into the profits of the trading strategies as suggested by Lo and MacKinlay (1990) and Conrad and Kaul (1998). Adopting this approach we divide profits of trading strategies into two components: one that results from time-series predictability in security returns and the other component that arises due to cross-sectional variation in the mean returns of the securities contained in the portfolio. The significance of the time-series components (namely, over- reaction and lead-lag effect) is interpreted to support the behavioral explanations on the trading profits, while significant cross-sectional variation in the mean returns indicates the risk-based explanations of the traditional asset pricing theory.
We empirically show that contrarian strategies are able to generate profits only in the pre-crisis period. After the Asian crisis, momentum strategies earn significant profits just like in most foreign markets. At first, it may sound puzzling how the completely opposite strategies can remain significantly profitable in the consecutive periods in the same market. But these opposing outcomes in the Korean market do make sense if we consider the substantial structural changes which have taken place in the Korean investors’ behavior. It is suspected that the profitability of contrarian strategies in pre-crisis period is due to the dominant retail investors. For example, the individual investors traded more than 80 percent of the total amount in 1992. They are evidently inclined to buy cheap losers and sell high-priced winners, which leads to the profits of contrarian trading. Ever since the Korean market was opened to foreign investors in the aftermath of the 1997 crisis, the foreign investors and domestic institutional investors have quickly dominated the stock market trading. Their share of the trading amount jumped to 35 percent in 2005 and more than 45 percent in 2008. The dominance of institutional investors is suspected to be the main reason for the profitable momentum strategies. These explanations are also supported by Khil, Kim, and Sohn(2006), who report that the institutional investors and foreign investors show the positive feedback trading behavior while retail investors are chasing the negative feedback trading behavior in the Korean market.
For the robustness check on the J-T strategies, we also test the profitability of contrarian trading strategy using the weighted trading approach. The trading strategy in Jegadeesh and Titman (1993) uses the stocks in the lowest past return decile and in the highest return decile to form the winners and losers. But in the weighted trading strategies, every stock in the sample is traded, with their weight measured relatively to the average return of all stocks. In the whole sample period, the weighted strategy does not generate significant profits. But the contrarian weighted trading did create profits in the pre-crisis period and significant losses in the post-crisis period. This outcome supports the previous findings on the performance of the J-T contrarian strategy.
The analysis of the sources of trading strategies’ profits found that the investors’ over- reaction is a significant factor, implying that the behavioral biases of the traders account for most of the trading strategies’ profits. However, the cross-sectional difference of individual stocks’ mean returns (reflecting the risk of each stock) also contributes to the profitability of trading strategies. This risk-based explanation indicates that the profit-generating trading strategies are compatible with the traditional asset pricing theory. The significance of both over-reaction and risk-based differentials in the Korean market implies that the profitability of trading strategies does not support or deny the behavioral or rational investment theories. If the former factor is larger than the latter, then the momentum strategies will generate profits, while if the over-reaction dominates, then the contrarian profits will be found significant. These results show that trading strategies’ profits can not be explained by either behavioral theory or traditional risk-return theory alone but by both of them. This conclusion is in line with that of Daniel, Hirshleifer and Subrahmanyam (2001), claiming that both investors’ irrational behavior and systematic risks determine the asset prices.

참고자료

· 없음