We all know that past performance is not an indicator of future performance. This does not, however, stop investors from piling into ‘winner’ funds and selling out of funds that have shown disappointing performance. They hope such tactics will improve the performance of their portfolios in the long run.
However, they would do better selling their top performers and buying funds that have underperformed their benchmark, according to a study carried out by Bradford Cornell (California Institute of Technology), Jason Hsu (Anderson Graduate School of Management at UCLA) and David Nanigian (Mihaylo College at California State University Fullerton).
They found that a portfolio consisting of the top-decile performing US equity funds in the previous three years lagged a portfolio made up of average performing funds over the next three-year period by more than 100 basis points. A so-called ‘loser strategy’, comprising the 10% worst performing funds over the previous three years, did even better, underperforming its benchmark by just 0.11%, compared to 2.39% for the ‘winner strategy’. The Sharpe ratio of the loser portfolio is also almost twice as high as the Sharpe ratio of the portfolios comprised of winner funds.
The results also hold when controlled for volatility, size, value and momentum factors in stock returns (see table): the Carhart four-factor alpha is significantly higher for the ‘loser strategy’ than for the ’winner strategy’.
The results of the study, which will be published in The Journal of Portfolio Management, chime with previous research conducted by Expert Investor into European equity fund performance, which found that funds that show top-quartile performance in one year are unlikely to repeat that over the next two years.
 Their paper is called “Does past performance matter in investment manager selection?” (2016)