Opening Gaps in SPY/S&P 500, part 2

The opening gap strategy I published last week showed promising returns. However, can we trust the data? The test was performed on EOD data from Yahoo! (open, high, low and close per day).

Today I have downloaded 30 mins data from IQFeed. The test period is the same: 1. January 2010 until 1st of July 2012. All the other parameters are the same. The result is still quite good:


In total there is 89 fills, 21 less than on the EOD test.  Average gain for long is 0.22% and 0.08% for short.

I have no idea why the numbers are so different. This proves that backtesting is just an indication, you need to test live to get a grasp of the strategy.