To SCALE or Not to SCALE a profit?

Discussion in 'Trading' started by Trend Fader, Oct 10, 2002.

  1. You are right but for the wrong reason :D This approach might have phsychological side benefits. But those aren't the main reason to do it. The main reason is it reduces the volatility. In theory it is irrelevant if you can tap on additional resources. In practice, the volatility of your account size matters.
    You can also look at it from a different angle, although it may not seem relevant at first. But bear with me for a sec.
    The main reason academics (and practitioners alike) argue one should diversity is to reduce the volatility. In this case, you are looking at diversification <b>over time</b> as opposed to the usual <b>accross assets</b> look. In other words, if you have a strategy where you take a position in some security and close it later to roll over into the next position, you might on average have a profit of some percentage. BUT, that doesn't mean you'll ever get there b/c the volatility of outcomes may take you out. By scaling in/out you reduce the variability of final outcomes and by making the distribution tighter are more apt to get that average return.
     
    #31     Oct 11, 2002
  2. Where are the Wealthlab and TradeStation guys when you need them? Somebody model a simple breakout system and compare exiting techniques. Just to be fair better do the same thing with a reversal system since it might work differently.
     
    #32     Oct 11, 2002
  3. No man, you misunderstoon me. I wasn't saying the thread was BS. I was just saying the answer seemed quite obvious even from the very first post. In other words, to me an intellectual discussion is interesting and stimulating when there are potential pros and cons to both sides. In this case, it's somewhat skewed to the extent that there aren't many benefits to not doing it except if you get REALLY REALLY lucky and get a sequence of high positive tail outcomes and then QUIT and go spend the dough for the rest of you life. B/c if you don't quit, you'll see it shrink as soon as you hit some of the outcomes from the other tail.
    By scaling in/out you remain closer to the mean/median/mode, which is where I believe avery trader should be shooting for (assuming they do have a strategy that gives them a postive mean :D). Otherwise they are just gambling.
    To me, trading is trying to get that everage return, and reduce the volatility as much as possible.
     
    #33     Oct 11, 2002
  4. And how exactly is that???? If you have a strategy that predicts where the price will be on average over the next, say ten minutes, just b/c it's not there yet, doens't mean it won't be. Heck, even if you are just randomly opening positions and it goes against you, I don't see why it would be more likely to keep going against you than not.
     
    #34     Oct 11, 2002
  5. bone

    bone

    You go, girl.
     
    #35     Oct 11, 2002
  6. birddog

    birddog

    Well, I don't know what is the correct answer, but I know what works for me.

    I trade in increments of 5000 shares building up to 20,000 shares. When I put in the first trade for 5,000 shares and it goes against me I am out for 1-3 cents loss. If the trade goes my way (4-5 cents ahead) then I usually add another 5,000 shares. If it goes against me I get out of the whole position and start again. If the trade goes ahead a little more then I usually close out half and then reenter again on a retracement (which doesn't always happen especially on shorts). If it continues I usually close it out as well after another 4-5 cents. This way I am ahead by 8 cents on 10,000 shares (or more) and I am out 2-3 cents on 5000 shares.
     
    #36     Oct 11, 2002
  7. bone

    bone

    Yeah Vlad, I'm bullish interest-rate futures for a ten-minute span, like in your beautiful one-lot wonder scenario. It's Friday and nobody wants to go home short interest-rates over a three day weekend. What about Iraq and terrorism? I'm long out my ass. Three hundred CBOT Ten-Year Note futures. $4,678.50 per one-half tick.

    Hey, out of the blue Merrill-Lynch comes out today and says this is a great place to buy GE and IBM. S&P, Dow, and NASDAQ futures take off to the upside. Over 300 points on the Dow.

    I think I'll scale out of this apparent loser.

    Vlad says that's what I should do.

    Idiot.

    Amateur.

    You have no fucking idea.

    Cost me $15,000 to get out. All at once. Hit the entire Bid. If I chose to scale out, would have cost me about $250,000.
     
    #37     Oct 11, 2002
  8. :p :D -- man you are brutal bro! I love it
     
    #38     Oct 11, 2002
  9. nitro

    nitro

    I have modeled this in TS. My conclusions are inevitable...you can scale into a trade, and that doesn't seem to affect the result terribly on the short side. However, it seems to on the longside - you want to get out "en-masse" on a long trade that goes against you. [BTW, there are some hidden variables here that I must disclose - this assumes symmetric stop loss pain - perhaps an oversimplification]

    FWIW, I always go in full force and get out full force. It doesn't mean I won't add to a trade, but that is because I would see it as a new trade, not as scaling into an established one. The fact that there is already a position on with "identical" entry parameters is a coincidence for me.

    nitro
     
    #39     Oct 11, 2002

  10. I disagree. Assuming you are a simple breakout trader or buy the pullbacks in the direction of the primary trend. Look back at your most profitable trades.. I bet you that more than %60 will be in profit right from the get go. This is a fact.. its true in relation to my performance.. and I bet its true in relation to most traders. THe point is.. if you are at a loss from the get go by scaling out the loser.. you increase necessary exposure. When trading markets you always have to assume that when a market is set in motion it will continue until proven otherwise.

    If I am at loss.. I think the market will go down forever.. and if I am at profit.. it will go up forever. Of course it will wiggle and waggle, but usually the primary stop loss is the place outside the wiggle room. So if it penetrates the stop loss.. it has a very high chance of continuing to go down.

    --Mike
     
    #40     Oct 11, 2002