Rangebar real trading experience?

New comment
 

Hello

Does anyone have any experience of real trading (ie, real money) with an EA on rangebars? I need some perspective on what Im doing, whether its realistic or not. So far I have:

1. Obtained about 2 years worth of tickdata (tickstory). I have only backtested with 2 years because of strategy tester FXT file size limitation (discussion)

2. Wrote, tested, debugged my own rangebar generator, with help from ppl on this forum. So I can generate a rangebar chart using tickdata. I also have an indicator which will update the chart real-time, so EA's and indys can run on it

3. Written an EA that trades on rangebar chart, and thanks to script you can get here, backtested it.

So Im looking at the backtest results, which are great, and thinking: what have I missed? Because if something looks too good to be true, it usually is, especially in forex. The profit factor is low (1.8), but the equity graph very consistent, risking 3% of account balance. The EA is simple, only uses one indicator. I think the rangebars are what give it the edge. Started forward testing on demo account yesterday.

So what have I missed?? Whats been your experience?

 

HarriMQL4: Does anyone have any experience of real trading (ie, real money) with an EA on rangebars? I need some perspective on what Im doing, whether its realistic or not. So far I have:

1. Obtained about 2 years worth of tickdata (tickstory). I have only backtested with 2 years because of strategy tester FXT file size limitation (discussion)

2. Wrote, tested, debugged my own rangebar generator, with help from ppl on this forum. So I can generate a rangebar chart using tickdata. I also have an indicator which will update the chart real-time, so EA's and indys can run on it

3. Written an EA that trades on rangebar chart, and thanks to script you can get here, backtested it.

So Im looking at the backtest results, which are great, and thinking: what have I missed? Because if something looks too good to be true, it usually is, especially in forex. The profit factor is low (1.8), but the equity graph very consistent, risking 3% of account balance. The EA is simple, only uses one indicator. I think the rangebars are what give it the edge. Started forward testing on demo account yesterday.

So what have I missed?? Whats been your experience?

We cannot say that you've missed something just because your back-tests are too good to be true. You've probably gone through allot of steps getting your expert to this point and we do-not know what you've done. Therefore, I'll say live-test-it for two-weeks and let us know if it maintains the 1.8 profit factor. If the system doesn't place allot of trades then perhaps use a 2-month check-point instead.
 

Thanks for your response ubzen. I understand what you're saying. Just that I've seen people come to this forum looking for help, and they've hit a common/well-known problem with whatever they're doing. So in making the ea Ive tried to take account of the potential pitfalls that I know about. Hence, using Dukascopy tickdata to generate the rangebars. Writing my own generator (because the commercial one I have has got flaws).

A pitfall unique to rangebars is synthetic/fake bars where the price gaps up/down and the gap is greater than size of the pip range. So I generated charts with gaps (EA result was more profitable). But I dont know about any other pitfalls. But I want to do a forward test (demo) before running on live account.

 
HarriMQL4: Thanks for your response ubzen. I understand what you're saying. Just that I've seen people come to this forum looking for help, and they've hit a common/well-known problem with whatever they're doing. So in making the ea Ive tried to take account of the potential pitfalls that I know about. Hence, using Dukascopy tickdata to generate the rangebars. Writing my own generator (because the commercial one I have has got flaws).

A pitfall unique to rangebars is synthetic/fake bars where the price gaps up/down and the gap is greater than size of the pip range. So I generated charts with gaps (EA result was more profitable). But I dont know about any other pitfalls. But I want to do a forward test (demo) before running on live account.

Yeah, I'll strongly recommend a live demo testing before going live. Hopefully two weeks of demo wouldn't be too long. There are usually pitfalls which you're aware of and pitfalls which you don't know exists. The demo testing also gives you time to sort out both that and any technical/programming errors which may arise during trading.
New comment
Reason: