Deciding when to take profits (I let most of my winners turn into losers)

Discussion in 'Strategy Development' started by IronFist, Apr 30, 2008.

  1. Ok this is driving me nuts.

    I have a methodology that I've been working with on the YM and NQ. It produces large winners, especially in the first and last hours of trading each day, but then in between, or even sometimes during those "good" hours," I get chopped to hell and I end up negative for the day (OEC demo account).

    The problem is I don't know when one of my signals is going to pan out and be a big winner, or when it's just BS.

    Say in a given day I get 30 signals. Three might be winners (+$300 each) and 28 might be losers (-$30 each). So that's $600 - $840 = fail. I try to cut my losers early but even with my winners there's always some chop before they take off.

    So my big winners are big (for me, trading 1 lot at a time), but sometimes they'll get to +$80 or +$150 or so and then reverse. I don't sell until I get my sell signal. Sometimes they pick back up and end up +$200 or +$300, but sometimes they go back down and my sell signal triggers around +$5 or $10. Then I get pissed at myself when at one point I was up $100 and I sold for $10 profit.

    So does anyone have any cool algorithms for solving this problem?

    ...or for reducing the frequency of my losers? I've tried stop and reverse orders (so buy 1 contract, and then sell 2 (which puts me short 1), and then buy 2 (which puts me long 1), etc)., but then I end up getting a bunch of $+20 trades that turn into losers. If I had scalped all those $20 trades I'd be positive at the end of the day, but I'm always waiting for them to turn into that one huge trade so I hold it and it turns into a loss, I reverse it, +10 or 20, then it turns into a loser where I reverse again, +10 or so, then it turns into a loser, etc.

    My first day on this demo I was +$800 on 3 trades at 1 contract each (two on YM and one on NQ) after the first two hours of trading. I thought "awesome, this is easy". And almost every day since then, employing the same methodology, I've ended up negative.

    Here hang on, I will attach some charts of what I'm talking about.
  2. magicz


    I think you're using lagging indicators for your signals. this is not a good way to go when there is a lot of volatility. better when the underlying is trending.
  3. apply the POEM in my today's thread "11:11AM ...".
  4. The other day:


    Today so far:

  5. The only indicator I'm using is a 60EMA. Go long when the price goes above it, go short when the price goes below it.

    It looked like it would work so beautifully in looking at charts. I didn't know the chop would be so bad.

    Seriously. I thought I was on to something because I wasn't using 20 different indicators like most beginning traders do.

  6. You have rediscovered the basic problem with trend catching systems - works great for trends and doesn't work in chop.
  7. is it possible to identify when it is or is going to be chop and just sit out on the sidelines, then?

    Or is this the reason successful traders scalp; so trends dont matter?
  8. Chris311


    trends matter but if there was a easy solution to your problem then we would all be retired... Have you tried using a much shorter MA for your sells and stick with the longer for your buys? May get whipsawed more but should be able to lock in gains a little quicjker before you dump all the profit. Plus if it is only a short retracement you can get back in when it crosses your long MA again.
  9. Well a few ideas

    Take profit a the previous days high or low

    Use a projection of the last swing high/swing low

    Just before round numbers

    Time of day can be helpful, often runs end around noon eastern, depending on what you are trading, US equity products often reverse at 3:30pm

    You can often get better fills with order that end in "9" or "1" than round numbers and on some products 8/2 works even better

    hope this helps
  10. ssbc19


    Do you actually trade this or is it a program?

    If all you had to do to make money is write a program to buy when price crosses above 60ema and sell when it crosses below we would all be rich.

    And why not just program it because a computer would be a lot faster with the trades.

    I think if you incorporated some analysis of the tick charts you could find better entries and exits
    • ym.png
      File size:
      68.3 KB
    #10     Apr 30, 2008