|
An empirical analysis of 130/30 strategies: domestic and international 130/30 strategies add value over long-only strategies Authors: S. Ericson, G. Johnson, and V. Srimurthy Source: The Journal of Alternative Investments Date: Fall 2007 |
Ericson, Johnson, and Srimurthy, three portfolio managers, publish one of the first empirical comparisons of the performance of 130/30 funds and that of long-only strategies. These comparisons, complicated by recent launches and short track records, have hitherto been made from a largely theoretical perspective. To get around these complications, the authors opt for back tests in which the performance of 130/30 funds is simulated by applying quantitative trading rules. The study excludes 130/30 funds using qualitative management, but it remains representative of the 130/30 industry because quantitative funds account for between 60% and 80% of this industry.
The analysis covers the 13-year period from January 1994 to December 2006. Two stock universes are examined, a US domestic universe corresponding approximately to the Russell 1000 Index, and an international universe corresponding approximately to the MSCI EAFE Index. Stocks are divided into quintiles and ranked by alpha, which is calculated with a six-factor model. Among the rules that determine stocks that must be bought and sold, the authors mention: “A sell discipline […] that requires a stock to be liquidated when it falls to the fourth or fifth quintile”. The application of these rules results in portfolios that hold an average of 110 stocks for the long-only strategy, and 143 long positions and 42 short positions for the 130/30 strategy.
First, cumulative pre-fee returns of 130/30 and long-only strategies are calculated over the 13-year period, and they are compared to the Russell 1000 Index and the MSCI EAFE Index. While both strategies outperform their respective benchmarks in the two universes, the 130/30 strategy outperforms the long-only strategy. Excess returns, defined as the annualised compounded return of the strategy minus the annualised compounded return of the benchmark, are 11.27% for the 130/30 strategy and 7.61% for the long-only strategy in the US domestic universe. The ratio of 1.5 between the two returns is also obtained in the international universe. While the performance of the 130/30 strategy is 1.5 times greater than the performance of the long-only strategy, the risk of the first is 1.2 greater than that of the second. In both universes, the 130/30 strategy leads to turnover twice as great as that of the long-only strategy.
Ericson et al. propose a method of measuring the contribution of shorts and longs to the performance of a 130/30 fund. They argue that performance attribution must be evaluated against a benchmark, as in the case of long-only funds. The only difference is the necessity of scaling the excess return. For long positions, the positive excess return is multiplied by 1.3. For short positions, the negative excess return is multiplied by 0.3. For the domestic universe, the average excess return over the 13-year period of 11.3% is explained by a long contribution of 9.1%, a short contribution of 2.1%, and a long-short interaction—generated by periodic rebalancing of longs and shorts—of 0.1%. A calculation of long and short contributions year by year over the entire period permits the conclusion that, whatever the market environment, the positive contribution made either by longs or by shorts exceeds whatever negative contribution may be made by the opposite position. In other words, the short extension makes it possible to obtain greater protection from changing market environments.
