Position Sizing - How to Add to a Winning Position

Discussion in 'Risk Management' started by VoodooMMI, May 15, 2008.

  1. I like to start a position with 1 contract. That way if the market moves against me to my stop level, I'll flatten my position and only have a loss on 1 contract. However, if the market moves my way, at certain intervals (which is separate question) I will add to my winning position. I'm trying to decide the best way to do this. Some candidates for the number of contracts to add at each level include (1,1,1,1,1,1,1) (1,0,1,1,1,1,1) (1,1,2,3,5,8,13) (1,0,1,1,2,3,5,8).

    How do you add to your winning positions? Have you back-tested your strategy any? If so, what did you find?
  2. I start with X size, then add .5X either when stop can be set to breakeven or when a significant resistance level is broken(this is what usually happens). I then add another .15X everytime a previous wave high is breached (resistance).

    I cant really backtest it because it is discretionary. But one of my prior problem was having winners sized up and keeping losers tiny.

    This is the only way I have kept runners running and continued to hold them with size full on.

    Now I never get out of a position until the market takes me out. Trail your stop, never pull a winner.
  3. Is X your total position or just the size of your initial position. For example if you buy 100, then buy 50 more your current position would be 150. If you were to add .15X is that .15 (100)=15 or .15 (150)=22.5?
  4. MGJ


    "Way of the Turtle" describes a trading system whose pyramiding schedule has at most three add-ons, for a total of four units. Each unit is the same size. In your notation, (1,1,1,1).

    In my own backtesting on my own (non-Way of the Turtle) systems, I don't see huge benefits from the third and fourth pyramided units. However I do find that backtest performance improves when the units are different sizes. If the initial unit is size "1", I make second unit's size "S" and the third unit's size "S squared". For some systems the optimum value of S is greater than 1. (Stunning but true). For other systems the optimum S is less than 1.

    Using your notation: (1, S, S^2).
  5. X = initial position. I just round the .15 off.
  6. You will frequently run into cases where you scale into the trade, only to get stopped out. Not all trades with a positive MFE work out.

    If you trust your setup, just fire. Let position sizing take care of the rest.

  7. Yep, but keep thinking of a way to fix that and you will come across a simple solution.
  8. What solution would that be?

  9. Its not really a solution, its more of a trade off.

    All I do is give up some potential profit for a move that has a higher chance of following through. Now every trade I take has a 70% chance of moving for me. But slightly over 50% of my trades end up break even/small loss as I add to them.

    But I still get to keep the 42.XX% that turn into runners. Keep in mind that only about 30% of those get to a point where I add to them more than one more time.
  10. The first thing to learn about position sizing is that your position size should have *nothing* whatsoever to do with whether your initial entry was a winner, loser, breakeven, or whatever.

    The only rational way to position size is based on an assessment of market action up to the present time. Your own buy or sell decisions in the past have NOTHING to do with what the right position is now.

    Therefore anyone who pursues a policy of adding to a winner, or cutting a loser, solely because it is a winner or loser, is being totally irrational. A position should only be added to if the current price & odds favour a bigger position. A position should only be cut if the current price & odds indicate that the trade sucks.

    More clueless, illogical, irrational nonsensical superstitious bollocks has been written on this than any other subject in the field of trading & the markets.
    #10     May 15, 2008