Monday, 30 April 2012

Risk measures for the previous XIU post

Here are some risk measures for the XIU trading strategies presented in my last post. I considered three investing strategies: buy and hold, a MA(10) trend following strategy, a seasonal switch.The MA(10) trend following strategy produces the highest average annual return (6.85%) and the second largest standard deviation. Based on Sharpe ratios, the MA(10) trend following strategy is preferred. The seasonal strategy has the lowest downside risk while buy and hold has the highest downside risk. Downside risk is measured by semistandard deviations with a benchmark of 0.

b&h ma10 seasonal
mean 3.76 6.85 6.00
stdev 15.79 10.62 10.12
sharpe 0.08 0.63 0.56
downside 11.76 7.30 6.83


  1. The results go against the conventional wisdom that trend following strategies perform better in trending markets as opposed to volatile markets as they may subject the trend follower to whipsawing. The fact that it has outperformed over a 10 year period gives credence to the strategy and moreover, it appears the long run moving average perhaps neutralizes the issue of a high signal to noise ratio that may ultimately give an investor/trader the wrong trading signal with a shorter moving average. Mean reversion leading to sideways market activity has been the Achilles heel of such strategies in the past; have they not included dividends in their results?

    1. Using a ma(10) with monthly data greatly reduces the whipsaw effect that one experiences when using daily data. A ma(10) with monthly data seems to work reasonably well with most broad based equity indexes.