I have a postive expectancy in both directions

[PhysicsBKK]

Hello all,
I trade multiple strategies o the European and London Opening times. One strategy is a breakout and one is a reversal strategy situated on the highs and lows of the Asian range. I've devised my own market structure guide to root out losers and have developed it to be profitable. In one particular market scenario it is profitable to trade both the reversal (0.5exp) and the breakout strategy (0.66exp). What should I do? Trade both? Trade neither? Pick and choose? This is live and the uncertanity of what to do is frustrating me. This scenraio is not common and happens typically once every few months.
Thanks in advance

 

[Enivid]

I guess it makes sense to trade both. As I understand, you get better edge by expecting one of the trade to reach its TP before another reaches its SL, so you can sometimes profit from both.

 

[PhysicsBKK]

Unfortunately not. I've not explained it well enough.
So today on EurUsd. Asian Range low is 1.3296 (1.33 for sake of ease). My backtesting shows that it is profitable for me to place a Sell stop order at 1.3300 SL 10 pips away 1.3310
TP 40 pips below Entry 1.326
My backtesting also shows me that it is profitable to place a Buy Limit order at 1.33 SL 1.3290 TP 1.3340
Backtesting has been done meticulously over 7.5 years of tick data by hand two seperate times so I trust it as much as anything can be trusted!
To summarise, there are 3 outcomes.1) Both trades lose. 2)Trade A loses, Trade B wins 3)Trade A wins Trade B loses.
I need help concluding what's the best approach and if I can prove it mathematically

Thanks

 

[Enivid]

In that case, you have to pick the trade that has a higher expectancy as they are mutually exclusive.

 

[PhysicsBKK]

Thanks for your reply
Would it not be profitable to trade both or is there a way to calculate it that you are aware off?

 

[Enivid]

Thanks for your reply
Would it not be profitable to trade both or is there a way to calculate it that you are aware off?

Since you cannot win them both, one is a certain loss (-1 expectancy). If it's the reversal one, you have 0.66 expectancy on the breakout, the total is 0.66 - 1 = -0.34. If it's the breakout one, you have 0.50 expectancy on the reversal, the total is 0.50 - 1 = -0.50. If you take just the breakout one, you have your normal 0.66 expectancy.

 

[Enivid]

PS: It might be worth to analyze the combined backtests to look at situations like these. Your expectancy figures are average for the entire periods. They might be very different for situations like you describe.

 

[PhysicsBKK]

Since you cannot win them both, one is a certain loss (-1 expectancy). If it's the reversal one, you have 0.66 expectancy on the breakout, the total is 0.66 - 1 = -0.34. If it's the breakout one, you have 0.50 expectancy on the reversal, the total is 0.50 - 1 = -0.50. If you take just the breakout one, you have your normal 0.66 expectancy.

Thank you. Yes I understand that. Very helpful

Post automatically merged: Yesterday at 13:50

PS: It might be worth to analyze the combined backtests to look at situations like these. Your expectancy figures are average for the entire periods. They might be very different for situations like you describe.

Sorry. Do you mean put the two backtests alongside each other and analyze the combined results?

 

[hayseed]

What should I do? Trade both? Trade neither? Pick and choose? This is live and the uncertanity of what to do is frustrating me. This scenraio is not common and happens typically once every few months.

//----

first, enivid would be far far better at making sense of complicated expectancies or any sort of metric like that than me..... i don't even think in those terms.....
//-----

what seems to help me the most is to invert the problem.....

in other words, should i let a situation that occurs only once every few months involving just 40 pips frustrate me......

if the answer was yes and assuming you can have hedge positions, then open 3 small accounts and trade all 3 versions.....

bet on all horses...... the buylimt in one, sellstop in another and both in the third.....

to me, that would be far more fun.....

being frustrated is so frustrating.........h
//-----

 

[Enivid]

Sorry. Do you mean put the two backtests alongside each other and analyze the combined results?

Yes, you find all such situations using both backtests, see the outcomes, and calculate the expectancies for these situations specifically. But please listen to @hayseed. He is a wise man. If this occurs only once every few months, it's barely worth the hassle.