Just wondered what peoples opinions were on the validity of strategies that work well, in both IS and OOS, but only on one timeframe (say H1 or D1) and only one instrument (say EURUSD). Changing either, typically results in breakeven or barely profitable results.
Would you trade such strategies ? Why ?
I'm really torn on this subject, because if a strategy works across timeframes and instruments it's likely to be pretty robust, but a jack-of-all-trades is invariably a master of none, so maybe it's best to find the best strategy for EURUSD D1, for example, and not be too concerned if it fails miserably on USDCHF H1 ….
During my trial run I found a couple EA’s that were good on IS & OOS and followed thru to another broker’s live backdata test on that one pair & one TF, that’s what sold me on GB (bought it today), even the thought that anything could be so robust as to go forward well on different period and different broker data was highly in doubt in my mind, surprised me and recovered faith in fx EA world prospects.
I trade strategies that don’t perform well on other currencies, it is really difficult to find one that would run on all currencies without any changes in parameters, etc.
The purpose of testing a strategy on more currencies is testing if the strategy is not curve fitted to the data.
The strategy doesn’t really have to be profitable on the other currency, it is ok if it at least doesn’t go directly to loss.
It is is only losing on other currency it could be that its parameters (SL, PT) are not good for this currency, or it was curve fitted to actual historical data and has no real edge.
Before every live trading I recommend running every EA on the demo first.
You must be logged in to reply to this topic.