We combine the momentum and contrarian approaches, using relative strength of individual stocks in the market, to develop a robust indicator called return percentile. It picks up those stocks which have been performing well over an intermediate-term window, but have underperformed in a short-term window. Using nine-year Indian stocks data, we find that our momentum-contrarian approach provides high returns irrespective of the holding period horizon (monthly, quarterly or annual). The results suggest that the contrarian approach can provide high returns, even in relatively short-term horizon, if applied with robust indicators like return percentile.

Though academics continue to be skeptical about technical analysis, it is quite popular among practitioners. Out of 95 studies conducted by various researchers during 1988 to 2004, 56 found positive results from technical trading strategies, 20 obtained negative results, and 19 indicated mixed results (Park and Irwin, 2007). A technical analyst tries to identify the trend and its reversal, using numerous momentum and contrarian approaches. De Bondt and Thaler (1985) found that lowest-return stocks outperform positive or higher return stocks over the subsequent 3-5 years. They ascribed it to overreaction to information, which causes the ‘losers’ to become underpriced and the ‘winners’ overpriced. Their findings support contrarian approach for long horizon investments. Most of the subsequent findings, like those of Alonso and Rubio (1990) in the Spanish market, Da Costa (1994) in the Brazilian market, Chang et al. (1995) in the Japanese market, Campbell and Limmack (1997) in the UK market, Baytas and Cakici (1999) in seven industrialized countries, Ahmad and Hussain (2001) in the Malaysian market, and Balsara et al. (2008) in the US market also support contrarian approaches. On the other hand, Jegadeesh and Titman (1993) found that the stocks providing high returns in the past perform better for periods of 3-12 months for the US market. This evidence, which indicates investor underreaction, is in favor of the momentum approach for shorter horizon investments. It is supported by researchers like Davidson and Dutia (1989) for the US market, Rouwenhorst (1998) in the international context, Conrad and Kaul (1998) for the US market, Schiereck et al. (1999) for the German market, Kang et al. (2002) for the Chinese market, Griffin et al. (2003) and Hart et al. (2003) for emerging markets, and Forner and Marhuenda (2003) for the Spanish market. The evidence is predominantly in favor of the contrarian approach for the long term and the momentum approach for the short term.

Most of professional security analysts and fund managers exhibit either momentum or contrarian behavior (Morrin et al., 2002; and Menkhoff and Schmidt, 2005), with many of them using moving averages or Relative Strength Indicator (RSI) oscillators for their decision rules (for instance, Wong et al., 2003; and Chong and Ng, 2008). Momentum is indicated by the trend and various crossover rules, whereas contrarian signals are identified with the overbought and oversold positions. However, these methods rely on the relation of current price with the past price of the same stock, which is not a robust indicator of market behavior. We combine the momentum approach (for intermediate term) and contrarian approach (for short term), using relative strength of individual stocks in the market, to develop a more robust indicator that should reduce the instances of whipsaws. We consider a stock to be strong, if its past returns are in a high percentile position compared to other stocks. It is considered to weaken, if the percentile position of its past returns goes down. We find that the contrarian approach provides high returns, even in short-term, with this relative strengthbased method in the Indian stock market.

Applied Finance Journal, Winter Blues, Stock Market Returns, Tunisian Stock Exchange, Trading Strategies, Stock Market Anomalies, Tunisian Stock Market, OLS Regression Method, Clinical Research, Risk Aversion.