Need solution. EA keeps freezing MT5 strategy tester 2 or 3 months into test run.
-
Hi. I've been using FXDreema for a few months now, basically just messing around. Currently, I'm busy with a scalping strategy using Heikin Ashi and Awesome Oscillator. The EA uses no SL or TP but rather check profit blocks with close position blocks.
Currently, I have it set up to run at most 8 positions at a time. It runs very well for a month or two, and then it just freezes up the MT5 strategy tester. It does it on both my laptops.
One of the laptops is a monster of a gaming laptop so, I don't think it's the memory. I also don't think it's corrupt history data since other EAs run without interruption.
Here's the project: https://fxdreema.com/shared/68IBqRF1e
Obviously, the check profit blocks have to do a lot of work. Am I doing something wrong, or might it be one of the blocks causing a problem?
I would really like if I can get this EA to work reliably since it's producing a pretty smooth profit graph (at least for about two months before it freezes up the strategy tester). I would appreciate if one of the developers can give it a look to see what might be causing the strategy tester to freeze up.
Thanks in advance and apologies if I'm being stupid.
PS: I'm flying to Thailand in two days so forgive me if I'm not responsive to replies for a few days. I'll probably be on a plane somewhere.
Chris
-
typically when backtesting freezes, its because the EA has run out of equity to keep trading.
Post a screenshot of your journal entry after it ends.
Easy way to test, change your lot size to .01 and increase your starting max balance to 10,000 USD
-
I checked it out and ran it.
0 SL strategies normally crash. They look good for a short time.Your logic is also a little off. Pretend your reading this and you've never seen it before. What goes first? Are you limiting it correctly?
-
@tipsywisdom , thanks but I'm sure that's not it. This equity is just fine. Here's a screenshot of where it's stuck now.
And here is what the graphs currently looks like. You'll note the equity was fine when it froze up.
-
looks like youre using paint for your screenshots....life hack time! CTRL+WNDW+S. Then you can highlight the area you want to show, its copies it for you to paste elsewhere.
-
@tipsywisdom , thanks. I don't really want to use stops but, if that is the problem, I'll try to use S2 and R2 from the pivots in market properties to see if that helps. I'll probably use weekly pivots since I don't really want to even reach them. Mean reversion kind of acts like a stop in this strategy so far on the lower timeframes. I just wish I could run it through a year's data. But right now it's about 1 in the morning here and I probably need some sleep. Thanks for the idea. I'll try to fit stops in somewhere.
-
worst case scenario you put stops and it drops the account to zero. Then maybe you can switch to a different indicator, or price action!
-
speaking of mean reversion, perhaps you can think of a way to detect if price is within a range or "trending." If price is doing one or the other on a larger time frame, this may be more profitable to run this in one of those scenarios vs another.
-
@tipsywisdom , that is why I want it to run until I've blown up the balance on the strategy tester. Then I can go back and look at the higher timeframes to see the bigger picture. It's probably a ranging strategy. But I need to let it run through to see where and how it fails.
Chris
-
Filters like those also typically work better when they are after the conditions.
-
@tipsywisdom , thanks. I managed to fix it using no stops. I set up conditions to basically hedge everything. When it fails to open trades in the opposite direction, it usually takes longer to reach the target but, because of mean reversion in the lower timeframes, it always gets there eventually.
The graph is not always as smooth anymore but I consistently get the same results. Hundreds of runs on different pairs. Never blew up an account so far. So the logic is sound, perhaps just not conventional.
Generally, currencies are ranging, even when they trend. Trending is just currencies ranging wider.
I basically now take profit like this.
No Stop Loss or Take Profit. Just hedging and closing when the sum total reaches a certain amount of pips.
Here's why I like hedging: It's impossible to stop-hunt a trader if the trader has no stops...
What I did is to just terminate the EA when the account grew by 10%. In testing, I used the withdrawal block to emulate withdrawals. So, basically every month you take off your 10% profit and restart the EA from the initial balance value. Most times it takes about 15 days to reach your target.
I think I can sell this thing...
Here's a report to look at. Thanks for the advise.
By the way, Thailand is very nice, even while I'm still under quarantine due to Covid-19.
Chris
-
@chris-6 not bad at all!
I'd still invite you to move split the 1 trade structure into 1 for buy and 1 for sell. Then put both of those green filters just before each by and sell now....to have their own. I'm interested to see how that logic changes it.
-
are you building this on real ticks or 1m ohlc? Testing on one or both as well?
-
@tipsywisdom, I only use every tick and every tick based on real ticks when I build. It's slower but more accurate. Then I run the EA on a demo account for a few months. I'm actually extremely cautious when it comes to real money. So I usually backtest the crap out of any strategy before I use it.
I still have to run this one on a demo account. Only after that will I actually trade with it.
Chris
-
@chris-6 does it trade any different moving those filters?
-
Hello @Chris-6, I came across this thread while troubleshooting my EA.
I'm trying to use a similar strategy and hedging my trades, but I'm having a hard time keeping my DDs low.
I ran the strategy you shared in the original post and I like the way you are closing trades.
It seems that you were able figure out how to keep DDs low, do you mind sharing the conditions you used for closing the trades?
-
@rocketavocado I'm afraid I can't remember. I had to clean up my account and lost the EA. I usually use this guy on YouTube as a quick reference now:
https://www.youtube.com/@BoxxocodeI also usually use three timeframes, for instance 6h, 1h, 10m. I take trades on the lower timeframe, while my higher timeframes are just used to check trend. So, I still keep trend in mind, even while ranging. As far as I remember, I just got in the market while it was ranging and choppy, and I used pending orders.
I relied heavily on one or the other oscillator to check for overbought and oversold (they all basically do the same thing, so you can pick the one you like). It ended up looking more or less like a grid system.
I apologize for not responding earlier. I usually only use FXDreema when I get a new idea.
