How to reverse the order system in FX trading - Expert Advisors and Automated Trading

Stuart Browne 2016.03.22 08:25 #21

Enrico Lambino:

Says who?

Try the millions of traders who have tried to do it before ;-)

It just doesn't work. If it were that easy, there would be literally millions of profitable reverse EAs. Instead, there are literally millions of broke FX traders :D

I have been losing From theory to practice Access struct element using

Marco vd Heijden 2016.03.22 09:20 #22

It only works if you mixed buy and sell order code by accident on a highly profitable strategy while coding.

This has happened to me before.

But not for losing EAs that are simply way off and no where near ever being profitable.

In the first accidental code swap example though, the EA will show a specific losing streak, it will create a runaway cascade event of opening and closing orders, draining the account in mere seconds,

Thus it's not even like a classic losing EA to start with because you will instantly know you made a coding mistake since the account gets drained in seconds.

Advisor managing the work Food for thought and Array sorting

NickF 2016.03.24 03:24 #23

Alain Verleyen:

So what ? In the real world, you can't hope to reverse a strategy and become a winner. Is to so hard to understand ?

But let's assume somebody has found a winning strategy. By reversing the strategy, we can be sure of obtaining the losing strategy.

So, if we consider this losing strategy as the original strategy, then by reversing it we obtain a winning strategy :)

Of course, it's not easy to come across a strategy that is so bad, that by reversing it to get a winning one.

Nick

CatFx50 Could Someone Wins Every Why is it so

Aleksei Beliakov 2016.03.24 05:42 #24

NickF:

But let's assume somebody has found a winning strategy. By reversing the strategy, we can be sure of obtaining the losing strategy.

So, if we consider this losing strategy as the original strategy, then by reversing it we obtain a winning strategy :)

Of course, it's not easy to come across a strategy that is so bad, that by reversing it to get a winning one.

Nick

so true

Marco vd Heijden 2016.03.24 06:28 #25

On the other hand, you EA could be profitable in both directions, so reversing the order process does very little to the profit output of the mechanism...

Some of you should think about that.

Alain Verleyen 2016.03.24 09:35 #26

NickF:

But let's assume somebody has found a winning strategy. By reversing the strategy, we can be sure of obtaining the losing strategy.

No you can't be sure.

Dafiri Irimagha 2016.03.24 23:22 #27

I believe my EA is not profitable. I don't mind some kind of partnership

Doerk Hilger 2016.03.25 08:43 #28

Nice discussion :)

My opinion is, YES, it can, but only IF:

1. It is not a scalper which tries to catch a small amount of pips/points which is close to the spread!

2. It uses fixed SL and TP, no trailing stops.

If the strategy uses fixed SLs and TPs which are n times bigger than the spread, then it´s a actually a very good strategy with just a wrong interpretation, and then the change of SL and TP vice versa makes sense. But the problem is, that the most EAs out there try to catch just very few pips, and this cannot work because the losses come from spread and slippage. The ticks are just modeled in backtesting and if the EA is time sensitive or uses a tight stop loss, it cannot work, because within the tester the ticks are not available in correct order and not with a correct timing. This is also the main reason why results in backtesting differ from results in live.

Doerk

Any rookie question, so Weird results in backtest Strategy Tester Results -

Alain Verleyen 2016.03.25 10:12 #29

It's a recurring discussion, and all the ones who are thinking that a losing strategy will inevitably become a winning one are wrong (even on some optimal conditions is pointed by some contributors).

Here is the main argumentation :

Forum on trading, automated trading systems and testing trading strategies

Food for thought and brainstorming

Alain Verleyen, 2013.03.15 13:59

You seem very obstinate ;-) All responses have already given to you, but maybe not enough clearly evidently. I try once more.:

Main reason that you idea doesn't work is SPREAD. Why ?

Take an example. EURUSD, sl & tp at 40 pips (0.0040), fixed spread 2 pips (0.0002). 10 buy trades : 3 winning, 7 losing.

1. Take a winning one.

BUY at 1.3050 (Ask price, so Bid price is at that moment 1.3048). You win at 40 pips, so at 1.3090 (Bid price, so Ask price is at the moment 1.3092). Stoploss is at 1.3010, we can say that this price is never reach at the moment your Buy trade close with a win.
Inverse that trade by a Sell open at same moment : Open at 1.3048 (Bid), we now this trade don't reach 1.3008 ( as TP) before it reach 1.3088 (SL).
Conclusion : a winning become a losing. It's a fact.

2. Now take a losing buy. Is it become a winner ?

Say open at 1.3155 (Ask price, so Bid is 1.3153), loss at 1.3115 (SL at Bid price, so Ask is 1.3117). It never reach 1.3195 (TP, Bid price, so never reach ask 1.3197).
Reverse that. A sell at 1.3153, so TP is 1.3113 & SL is 1.3193 and these prices are Ask price. We know it never reach 1.3197 (Ask at moment of buy closing) but maybe it reach 1.3193, may be not. Other, we know Ask reach 1.3117 (buy closing at loss) but to win it must reach 1.3113. Reaches it ? You don't know, may be, may be not.
Conclusion : It all depend on the probability to reach SL & TP, we see we can't say anything with certitude about reversed buy losing into sell. With it this is probability ? Another debate.

Hope you are not lost in all that numbers and that help.

EDIT : take some time to write this, and I don't see last posts before.

Here the main topics on the same subject :

https://www.mql5.com/en/forum/8629

https://www.mql5.com/en/forum/9231

https://www.mql5.com/en/forum/16489

https://www.mql5.com/en/forum/36787