Here are some key data from the backtesting report:
Period: 01.01-2001-01.01-2012 (11 years)
Profitfactor: 4.27
Max drawdown 3,79%
Modellingquality: 90.00%
Profit trades: 50,44% averege profit 4851
Loss trades: 49,56% averege loss 1155
This seems too good to be true doens’t it? What do you think?
And can anyone answer me how reliable is such a backtesting, or how reliable a bactesting in general? :33:
We as trading company get a lot of mails every day with good backtests. Please don´t trust them, you never see any spread.
You should make a 99% backtest on your own, and if it looks good start a demo before you go live!
I have made 2 EA’s thus far. The first one if backtested from the begging of this year until now has won 100% of it’s trades. I am currently forward testing it to confirm that it still wins. So far it’s been doing fantastic. Only four trades since my orignal forward testing has begun; I started it may 28th, (yes it’s a practice account) but like the others have said, the forward testing results are what is going to prove whether the EA works or not. Even though these first 4 trades are all profitable, I’m going to continue to watch this EA’s every move for months. I hope it makes at least 20 more trades in a row that are profitable. At that point I feel like I am on to something.
If you where to backtest it from 1971 until now it would win 92% of it’s trades. But winning average really doesn’t mean much either, it’s your profit to loss that really counts
If you have mt5 I will give you a copy of this ea free of charge. If you decide to use it after you forward test it let me know. Make a million dollars for me and hook me up with a little dough xD.
Sure thing. I will make sure to post a new thread about it solely dedicated to forward testing. I already have a thread about it on here called “more or less profit for the bot competition?” be sure to check that out if your looking to see what its back testing results where. I look forward to working with you at fine tuning this EA and comparing results on the forward test. Bear in mind it does goes a period of a few to several days time where it does not place any trades. It hasn’t placed a trade on my practice account in 2 days thus far since the start of the forward test. The day i started it placed four trades so its kinda acceptable for it to go a few more days without placing any. But we’ll see how he runs. Add me to your friends and I will gladly pm you my email so that I can send him over to you.
Don’t pay any money for an EA until you have tried forward testing a few free ones
I would live to test it out. I tried to PM you but I am new here, and it won’t let me.
My email address is: ezra (at) straypoetry (dot) com
Thanks for all your hard work!
Hello, I would be glad to backtest it. I have oanda data (from tick) from 2004 onwards. I can provide an accurate report, including profit distributions, montecarlo simulation and worst case scenarios.
To perform a rigorous evaluation there are only two requirements from my side:
the timeframe must be 30 min or larger
sl and tp must be normally not smaller than 10-15 pips
That just to provide a backtesting in mt4 that makes sense.
Let me know if you’re interested.
In any case, the results look a bit suspect (at least without further info). I would check if this backtesting doesn’t exploit simply some flaws in mt4, providing false results.
Sounds great Lorenz. From what I have already found, if you backtest it on a different year the chart that you attach the EA to will need to be adjusted so that profit can be made.
I have mt4 but the original ea was made for mt5. The issue with mt4 is that it does not have the 2 hour chart that the ea requires to make trades. I have a mt4 version but it’s technically a completely different EA. I’m really looking for testers on the mt5 version. I know very few have a copy of mt5 or even use it for a live account, but a practice account is easy to make and mt5 is free.
Email me at theleg0nd at aol if you can test it on mt5.
90% data is simply not good enough these days[/QUOTE]
IMHO, the MT4 backtester has such kind of limitations that the use tick data (or 99% Quality) is almost pointless and doesn’t give you any significant advantage.
It is important to keep in mind that the MT4 backtester’s drawdown figure is based on balance and not on equity. This is not a small distinction because many Martingale based strategies show very small balance drawdown, but the risk of ruin tends to be near certain, and the % equity drawdown on normal open losing trades tends to be very high with lots of small winners, until the one trade that kills the account.
