First post alert, please be gentle!
Hi all, as a one year newcomer to forex I thought I'd be brave and after lurking here just reading and learning I thought it time to share my first attempt at a system. I'm sure this idea is not new but the entire thing is something I have devised on my own with the idea of developing something that works for me and within my comfort zone as a rookie. I have much more work I'd like to explore with this but it does seem to be working so far, and I guess as a newbie i'm looking for advice on how to continue improving it and ideas to take it further (apologies if this system is so simple that you want it in the newbie section!).
The basis of this system is that I wanted to design something simple that was easy to accurately backtest without any concerns about indicators giving me false information. I also wanted it to be easy to trade mechanically with no room for interpretation and hence no excuse for me to break discipline and try something wild!
Its perhaps a scalping system, the idea is to take the momentum and trade direction from the london morning session and try to take just 10 pips from this move so here it is:
Make one trade a day at 13:00 London time (GMT+1 in summer, GMT+0 in winter).
Only pair tested (so far) is EUR/USD
If price at 13:00 is greater than at 09:00 then buy
If price at 13:00 is less than at 09:00 then sell
It really is that simple. I am not a programmer (yet... its all on my forex study list!) and I'm sure this could be automated in a second by someone in the know. I have however backtested this between march 26th this year through to yesterday giving me just over 7 months worth. I have been live trading it too for about a month with a 10% return so far. I am ok with excel but do not know visual basic so although I have historical EUR/USD data, the only way I could backtest this was to go through the charts manually on my criminal (IG Index) with a paper and pen. I would like more backtesting but the IG charts only go as far back as 26th March 2012: Yes this took me a long time but I knew it was something I could do with a system this simple and to my mind it should be a very accurate backtest albeit only 7 months worth.
Attached is a summary of results of the backtest, for those that are interested. During my backtest analysis I like to assign a fictional account balance so I can see how a sytstem will pan out in real £ terms... (In my backtest the account started with a fictional £2000.)
In short based on a "% risk, a TP 10 SL 10 gave 66% win rate and 185% return had it been traded from 26th March to Yesterday 13th November with just a 6% drawdown.
TP of 15 and SL of 10 performed best: 54% winrate, 208% return and 12% DD
I have also backtest using a SL of 15 and a SL based on the 4hr ATR but these did not perform as well as SL10. I have also tried filtering trades based on the nanningbob 10.2 TMA slope, i.e. a trade only taken if the TMA slope direction agrees with the 0900-1300 signal... I tested filtering by several TMA levels (>0.2, >0.4, >0.6, >0.8). The naningbob filter improved the win rate but as it restricted the amount of trades taken the return rate is not as good as just trading everyday with a stop of 10.
Ideally I'd like to trade this system on several pairs, but my lame method of backtesting is taking some serious man-hours. I have done the odd month on other pairs and I think it could have success although the SL and TP levels will need adjusting I think. Any advice on how I can speed up my testing and analysis of this system on other pairs would be wonderful!!
Thanks for reading - please dont laugh!
. May be it's due to lack of tick data as I only have M1 OHLC...