THIS IS NOT INTENDED TO BE A FINAL WORKING STRATEGY. This is a intriguing pattern that seems to work in European indexes. Maybe can be used as an idea for a complete system or maybe just as filter.
Actually I was trying to build a proorder with a random entry, as a begining for trying some ideas. I thought to use an entry after 2 bars of the last exit (1H TF) with the long/short direction depending of the hour of the day. From 00:00 to 12:00 long and the rest of the day short. To my surprise, the system was wining in DAX and the rest of european indexes, in a way that seems statistical relevant.
It seems that at least the last years (100.000 1H bars) the mornings have been long and the afternoon shorts. I haven´t done a thorough optimization but it seems to work better with the longs from 00:00 to 13:00 (UTC+1), that´s aprox until the USA opening.
Do I miss something here?
Hi TempusFugit, thanks for your post. While reviewing your post to be featured in the library, I stumble upon some questions: did you include spread or not? did you use the tick by tick backtest? I’m not able to reproduce your result so far (please find attached 3 screenshots, first one is without ticks backtest and with no spread at all, the second is without ticks again but with spread included and the third one is a tick by tick backtest with 1.5 points of spread). As you can see there is something wrong somewhere 🙁
Your assumption about a recurring pattern on European index is very interesting though!