Can a consistently losing strategy be profitable if I do the opposite of it?

Discussion in 'Strategy Building' started by Amatrue, Oct 25, 2020.

  1. guru

    guru

    A losing strategy likely trades randomly (especially that markets behave unpredictably and therefore randomly in most cases), while an opposite of random is still random. Start with a casino/roulette and bet on red then black, repeat it until your cash is gone. Then see if you’d win doing the opposite.
     
    #21     Oct 27, 2020
    wrbtrader, Amatrue and themickey like this.
  2. Pekelo

    Pekelo

    Not necessary. If the stop loss is too tight in a volatile market, even a winning strategy becomes a losing one.
     
    #22     Oct 27, 2020
    Overnight likes this.
  3. tomorton

    tomorton

    I am trialling such an "anti-signal" trading system on the forex majors and although it is early days it is showing consistency.

    The starting point was to use a simplistic trend direction identifier at the close of the last (London) session and take a basket of trades in that direction through the session prior to the next morning's open, just for the duration of the session. The basket comprises as many of the significant pairs as margin will allow. This seems to show a win rate of about 55% while the market is subject to the usual incidence of trends.

    Now that the majors have become range-bound and unlikely to print any consistent trends, I had expected the system to drop to about 50%. However, reversing the signals is now showing a better performance. It is running at about 75% profitable days and a trade win rate of over 60%. When the anti-signal tactic stops working I will flip back to the with-signal trades. This is the dumbest system I have ever tried but its making money......
     
    #23     Oct 27, 2020
  4. guru

    guru


    Possibly, but even in your example reversing the buys vs sells wouldn’t help.
    So a strategy potentially can be improved but not reversed.
     
    #24     Oct 27, 2020
  5. Amatrue

    Amatrue

    Just backtested my strategy on 1 tick fill resolution from the beginning of the year and the results seem less fake i guess? Also the trades make sense, but there seem to be some anomaly when the largest profitable trade is $1912.5 and largest losing trade is $312.5 when I set take profit and stop loss at $62.5.
    6e1ceb8e408149701d6b427b423465b5.png
     
    #25     Oct 27, 2020
  6. Amatrue

    Amatrue

    I reversed the long/short conditions of my strategy and it gave me the same losing result. How is this possible?
     
    #26     Oct 27, 2020
  7. guru

    guru


    If you’re using compounding then you may be compounding losses. For example start with $1000 and make 10% profit, then lose 10% profit. The end result is $990.
    So you lose $10 just by trading up and down. And this will not change if you reverse the trades and first lose 10% and then make 10% profit. You’ll also end up with $990.
    Try the same strategy without compounding (make $100 then lose $100) and the results may be little different.
     
    #27     Oct 27, 2020
  8. Overnight

    Overnight

    Because yer stops are too tight or your targets are too large. Expand your stops by 200% and reduce your targets by 50%. See how that looks.

    And on the long side, just use no stops. See how that works too.
     
    #28     Oct 27, 2020
  9. userque

    userque

    Look at those actual trades and see what the cause is. Probably one or more of these reasons:

    1. Gap openings if you hold over to the next session.
    2. Trading halts.
    3. Logic/Code issue.
     
    #29     Oct 27, 2020
  10. userque

    userque

    Could be one or more of these reasons:

    1. Commissions
    2. Stop placement
    3. Target placement
    4. Signals based on randomness
     
    #30     Oct 27, 2020