I was spending my afternoon yesterday sitting in my favorite internet cafe ingesting caffeine and playing around with trading statistics (I actually find this fun) when I noticed that my numbers were a little off from what I was seeing on my charts.
I use a statistics package instead of the strategy tester in mt4. In tracking down the reason why the numbers were off a bit, I discovered that while I was calculating an oscillator such as RSI using standard formulas, mt4 (and others maybe?) were smoothing out the RSI before displaying it on the chart by taking a moving average. I suppose so the line looks less jumpy. Mostly they're not off by much, but when there's a big swing, it can be a significant difference. So I had skewed numbers and at least a few trades that I might have acted differently without the smoothing.
Anyone noticed this and have comments? If you were to take a moving average of the oscillator as some systems recommend (fozzy, for example), you would then actually have smoothed it out twice.
Squid
I use a statistics package instead of the strategy tester in mt4. In tracking down the reason why the numbers were off a bit, I discovered that while I was calculating an oscillator such as RSI using standard formulas, mt4 (and others maybe?) were smoothing out the RSI before displaying it on the chart by taking a moving average. I suppose so the line looks less jumpy. Mostly they're not off by much, but when there's a big swing, it can be a significant difference. So I had skewed numbers and at least a few trades that I might have acted differently without the smoothing.
Anyone noticed this and have comments? If you were to take a moving average of the oscillator as some systems recommend (fozzy, for example), you would then actually have smoothed it out twice.
Squid