You need no special software for that when you have a decent trade software that supports statistics plots. A simple script should do.
This is a small lite-C script that can test weekly seasonality:
Inserted Code
function run()
{
BarPeriod = 60;
set(PLOTNOW);
PlotWidth = 2000;
PlotScale = 12;
int h = hour() + 24*(dow()-1);
var value = (price(0)-price(h))/PIP;
plotBar("Price",h,hour(0),value,AVG|BARS,RED);
plotBar("StdDev",h,hour(0),value,DEV|BARS,BLUE);
}
This script plots a week statistic with the hourly mean and standard deviation of the difference of the current price to the price of the begin of the week. This is the result for EUR/USD:
The red bars are the price, the blue bars are the standard deviation. You can see a weak seasonal effect Monday and Tuesday. The price tends to go down Monday, then tends to go all the way back and moves in opposite direction by Tuesday. It should be possible to exploit this effect in a strategy, however the standard deviation is much higher than the price difference, so it would be not easy.
You need the free Zorro software for running that script and experimenting with seasonality.
