This system is based on the conclusions of this working paper of the National Swiss Bank:

https://www.snb.ch/n/mmr/reference/working_paper_2007_03/source/working_paper_2007_03.n.pdf

They looked for intraday patterns in FX and discovered that: “Domestic currencies appreciate (depreciate) systematically during foreign (domestic) working hours. These time-of-day patterns are statistically and economically highly significant. “

And this happen because… “The prevalence of domestic (foreign) traders demanding the counterpart currency during domestic (foreign) working hours implies a cyclical net positive (negative) imbalance in dealers’ inventory.”

I did a simple bot to check this pattern with Eur/Usd, short in eurpean market hours and long in american market hours. Its results are not brilliant but good enough for me. As far as I can tell looks like a “robust” system: good WalkForward Efficiency and similar results with other pairs/times.

This pattern can be used as a system by itself or maybe like a filter for other systems. Also the thorough study of the paper can lead to a better system, I just wanted to share the “gross” idea.

The system I post include also the parameters optimized for two other pairs (GBP/USD and EUR/AUD), the way it switch optimizations come from a system of our colleague Maz, thanks mate. NO SPREAD CONSIDERED. No Money Management or Position Size.

I have tested it just for a few days in demo and the results are similar to backtesting but considere it more a way of checking the pattern than a final “black box” system.

IMPORTANT NOTE: My timezone is (UTC+00:00) or (CTE-1) ,probably different to most of you, so YOU SHOULD ADJUST THE TIME PARAMETERS because this strategy is time sensitive, very!

Comments and suggestions welcome

fx-homerun-1510053158p8l4c.png FX-MarketEdge-EURUSD-1H.itf