Can I get a few thoughts on these backtest results? - page 2
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
You really want to get rid of your mismatched chart issues and have your EA up to 90% modeling quality. I can tell you that when I have only a few mismatched data on my EA, the results are extremely different than when it has good data.
You really want to get rid of your mismatched chart issues and have your EA up to 90% modeling quality. I can tell you that when I have only a few mismatched data on my EA, the results are extremely different than when it has good data.
Thanks for reminding me to fix that. Another poster sent me a link, and suggested I do it as well, I let it slip my mind. It did make a difference in the total return, Not alot compared to the return this EA seems to post in test mode. But it will be nice to do it correctly from the start next time when I test this EA on other pairs and and also test a different EA I am working that probably won't have the same type of curve and return as this one.
Do me a favor, run it from 2009.07.08 09:05 - 2010.06.29 23:55 and post the curve here. Thanks in advance ;)
Interesting, results were similar for 2008-2009 as well. Well at least my risk fuctions seem to adjust correcly, they let me loss my money slowly.
Is there a general length in time or total number of trades, for a good backtest that people use here? Prior to running these tests here I have always used forex tester to my backtesting manually.
....Is there a general length in time or total number of trades, for a good backtest that people use here?....
Nope, in my dear opinion: the more the better. At some point you have to say good enough and roll with what you've got. This does not mean you should optimize 10-Years worth of data. Been there ... done that. Now-a-days, when I create a new strategy, I test it on 3-months and move to 1-year. Then I go year-by-year backwards. Then I go currency by currency. At some point it breaks. I take note of where it didn't work and try to figure out why and move on.
The more you play with the tester the more feel you'll get for different strategies and what works on what market behavior. Of-course you'll have to do all this on Live-demo and Pennies accounts b4 committing your retirement account to it. Generally if it's showing bad results on 8 out of 10 years, especially if those bad periods are in recent times, you'll need a real good excuse to trade it on real money. However, if its vice-versa on the good side and does well on multiple currencies. Then I focus on Fundamentals as the breaking factor. At least that's the plan thus far. I'm just learning like you ..... time will tell.
For manual strategies, if you're new then give it about 1-Year results (3-months if you have some experience). For expert advisors, give it about 3-months but thats after you've done detail back-tests and understand every angle of the system and know that the code is bug free.
Here's my current EA from 2009.07.08 09:05 - 2010.06.29 23:55
And here it is from 2011-01-01 to Present (3.82 profit factor, 52% profit trades)
* Starting with a $10k balance, using 30pip stop losses and money management strategies.
I'm not afraid to show it. I'm a newbie, that was my first post ever. :-) I was just trying to relate to the person who posted the initial question. Here's the report from the last 6 months (the 2nd graph picture)
Absolute drawdown was only $34, but I imagine that is because of the time I started the tester. It was probably sheer luck, and of course, as you saw starting in the timeframe you suggested (which was probably a horrifically volatile time in the market) proved that there were bigger losses. Anyways... here you go. :-)
**Also, the 6 month gap is because you specifically asked for that timeframe, and I had been doing the majority of my testing on jan/2011 to present, which is what I showed here. I could post those as well, but I'm far from ready to start saying my system is awesome, and I'm sorry if you got that impression from my original post. Again, I'm just trying to relate to the original posters question. I actually am glad you gave us that timeframe to look at, because it is a real bear to get through and will help in future backtesting!
Your test is simply awesome as far as stats go; I don't care what anyone else says. It's not Long or Short only. Relative Draw-down is almost in single digits. Lots sizing seems on point. It breaks-even during the bad years. Average # of trades per year would make good compounding.
The only other thing I can recommend is running the system more on other periods and currencies which it haven't been optimized for. Then Live-test that baby. I like this system, it must be a trend follower which plays by all the rules. Something even I haven't been able to master.
Thanks! I do indeed follow the trends. I ended up making my own indicator that shows 2 other timeframes trend and also a signal line that gives me a pretty good indication of when it's in a rangebound/sideways state. I like to trade the 30minute chart, so I set my indicator for the next 2 highest timeframes (1hour and 4hour).
So, on my indicator, above the 0 line is the TimeFrame1 (1hour) and below the 0 line is TimeFrame2 (4hour). If the blue line goes below the zero line, as you see here, it's a rangebound/sideways market and I stay out. Also, I only enter a position when both the TimeFrame1 and TimeFrame2 are matching. If they are both green, it's a Long, if they are both red, it's a Short. Of course, I lose some of the price movement as I am waiting on the indicators to catch up, but I find that it's better to wait for a higher probability trade, than to take a trade too early and hit your stoploss. This is just one of several indicators I use to enter and take add-on positions. I'm sure there are indicators out there that can tell me the same thing, I just got tired of looking for one and decided to make my own. :-)
* The trades you see are another money management idea I had. Basically, in my strategy, the entry point is the most critical part. So, I split it into 2 separate equal trades. 1 trade has a take profit of 2x the stoploss, and the other trade has no take profit, I close trades based off of indicators and trend reversing. So, what happened in both of these trades is that there were 2 entry positions. Both taking 2 trades. Both of the 1st trades hit the take profit and the 2nd trades went up even higher until my other indicators closed them out. The idea behind this is that more times than not, I have found that even when I make a bad trade, the price will hit 2x my stoploss in the direction I intended before it spikes back down to hit my stoploss. So, by making 2 equal trades, and allowing one to take profit, if the price does go back on me and hit my stoploss on the 2nd trade then I am at a break even point. If the price just immediately goes south on me and both trades hit the stoploss, then I haven't lost any more than I was willing to risk to begin with. :-) This may not be new to anyone, but it was a game changer for me. lol