Making of a method

Discussion in 'Journals' started by game, Apr 15, 2013.

  1. Hi Tonkadad,

    I use a volatility-based stop-loss that is determined by how big the price bars have been recently at the time that I put the trade on. I guess there's different ways to do it, but the one I settled on is to use a fraction of the average true range at the time that the signal bar has printed.

    So, for example in futures trading right now my favorite system uses a stop-loss that is 0.7 x the current 12 bar ATR. This fraction was chosen by experiment, because I've found that the full stop-loss is hit on less than or equal to ~ 20% of trades.

    Since March I've been testing a system for trading currency futures that uses a stop-loss of 0.5 x ATR and also takes profits, via programmed limit order, when price either moves forward by 0.5 x ATR or by 1.0 x ATR. The 0.5 x ATR profit-taking method had a 70% win rate in backtesting, but it isn't doing as well in real world testing. That seems to be a common theme in my backtesting and forward testing...
     
    #1381     May 12, 2014
  2. ST thank you for your detailed response, look forward to more of your contributions.
     
    #1382     May 12, 2014
  3. game

    game

  4. game

    game

  5. game

    game

  6. game

    game

  7. game

    game

  8. game

    game

  9. game

    game

  10. game

    game