Stock Selection - Backtest Time Frame

Stocks selection is critical for trading success.  In order to explore various trading ideas, I will be using backtesting to evaluate whether the ideas have statistically significant edges.  It is important to choose a proper time frame that encompasses various market behaviors so that you can truly test the historical performance of a trading idea.

I am selecting my backtest time frame based on the following criteria:

• Must contain at least one bull market
• Must contain at least one bear market
• Must cover at least 5 years
• Must provide little to no return for a buy-and-hold investor based on the SP500 (SPY)

Based on my available data and the criteria above, I have selected the following time frame for backtesting:

• Start:  September 15, 1998
• Finish:  September 15, 2009
• Total Number of Trading Days = 2870

This time frame spans 11 years (132 months) and contains multiple bull and bear markets.  This period also contains the more recent price behavior of the markets, which is probably more pertinent to the performance of the market over the coming years.  The closing price for SPY on September 15, 1998 was $104.06.  The closing price for SPY on September 15, 2009 was $105.72, a difference of $1.66 per share for a return of only 1.6%.

• Total Number of Months = 132
• Number of Months with Positive Return = 71
• Number of Months with Negative Return = 61
• Percent of Months with Positive Return = 53.78%
• Percent of Months with Negative Return = 46.22%
• Average Return for Positive Months = 3.45%
• Average Return for Negative Months = -3.71%
• Highest One Month Return = 9.93%
• Lowest One Month Return = -16.52%