ATR issue

Hi,

ATR indicator from SQ count SL/TP differently at 2 real accounts at 1 broker. EA is the same. But at one account is SL based on ATR counted 105 pips, on the other account (same data, same broker, same EA, same setting) for example 106.7. 

Looks like little, but as little as 1.7 pips can decide if you earn or lost. so one account will be after one trade 105 pips in profit, another one will touch SL 50 pips. The difference than between those 2 real account is 155 pips. Same EA. Different results. So for now, one account is going up, one account is going down. 

My Q is, is it because the indicator is not so exact? Its imported from SQ. Or this is, because very very little difference in data between 2 real accounts means bigger difference on ATR indicator? Or is ATR indicator “lazy” and is not updating so often? may be it needs “wake up call” every few seconds. Any idea?

Thank you for help.

Patrick

Hi Patrick,

I use 2 accounts at the same broker as well (actually 3), all orders (5 EAs on each account with the exact same settings) in terms of SL / TP (ATR based) do match 100% between all 3 accounts. Are you sure the accounts are EXACTLY the same in terms of spread, commission, stop level, etc? That can be the only explanation.

Hi geektrader,

yes its the same account – Pepperstone / Razor (ECN)

you mean SL and TP is always same for 100%? 

it is like:  Stop Loss = (3.28 * ATR(48)) pips;

              Profit Target = (4.81 * ATR(14)) pips; an the stop loss is different between same strategy, same MM (risk in $ for trade – only once it is for example 100$ and once it is risk 10$, but in both cases it should be same amount in pips, only size in lots should be different…)

i will whatch again for few days/weeks and try to find out whats happening.

Hi Patrick,

yes, the values in pips should be the same even if the lot-size is different, it won´t matter. Can you try the following just to make sure all settings are 100% the same between both accounts?: Close MT4 and copy the MT4 folder of one ready setup MT4 instance (with all EAs attached to the charts) to another new folder and then start this instance as well and the only thing you do is to login to your other Razor account? This way we can be sure everything is setup EXACTLY the same. Of course you´d have to start both in /portable mode via commandline since the data-folder will not be copied otherwise.

i think i know the answer. could you please, geektrader, look into your history if you were so good and nice :-),or watch for that when you check your trades,

when the open price is different (ussually when news aor high volatility), than SL and TP based on ATR is different. When the broker is real ECN, than price is changing every second and 2 ppips difference in open price afford the TP and SL counted by ATR. So SL is than in my case 2.5 pips different from another account and +difference 2 pips in entry price its 4.5 pips difference between 2 acount, which decided that 1 account is in loss and one account is in profit.

looks like small issue, but these small things are making difference between profit and loss…. 

File: ATRissue.png

Yes, of course if you get a different entry price due to slippage, the SL is adjusted to the REAL open price, that´s just normal. If your strategy is SO sensitive to that, then it´s not robust enough and you should not trade it at all honestly.

Geektrader, it not depends on strategy, when the movement is high no matter how good is the strategy..its H1 strategy at EURJPY. Check in future, i dont trust it will never happen to you or someone other. My real is going up, i just wanted to say i found out that no matter how good backtest or values the strategy has, real market is “the worst exam”.

Odesláno z mého Lenovo A536 pomocí Tapatalk

Yes I know, it´s because of slippage, that´s normal of course as SQ needs to adjust the ATR StopLoss to your entry price, and if that slipped, it has to be done like that. Because in a worst case, the value for the ATR SL would already be surpassed by the market if extreme slippage on the entry happened for example.

And yes, real market is always the hardest, that´s why you should backtest with realistic slippage values and run robustness tests on randomized data and only trade the strategy if it passes all that.

Fortunately it happens only time to time. Other trades are exact. But is +-90% ok, other is different. But there is harmony in the direction of equity in long term and this is important. This is probably forex, nothing is exact.

I do realy hard RT tests, only WFM+WFA i am not using, because i saw when i change something little bit the result is totally different so i am still not sure how to do that (i read all articles about it, but i am still not sure that i use the right setting…)