Reality based coin-tosser method that beats 95% of traders in the world.

Discussion in 'Strategy Building' started by Whisky, Oct 16, 2009.

  1. Yeah, there is one that is fairly well known, but, its not necessarily based on anything emotional; the 10am PST reversal is one given some simple conditions. I suspect this one is due to traders getting trapped or something like that... I think Stocks and Commodities mag did a test on it 6 months or so ago and they probably explain it better. If you modify it slightly, its tradeable, but, you'll have some long DD's. News events and FOMC are usually good times too.

    What I personally look for is anything that happens too fast. Stocks/Futures that rise to quickly or fall too quickly and I toss my coin at those points so to speak. My edge is about 52% across all my systems in real life and I average about 5k-7k trades a year - my most profitable system is at 40% hit rate with 4-1 payoff. Its done very well on the 30 year and the ES this year. Lately I've been moving towards lower hit rate systems as I find them more fun to trade; a homerun versus a law of large numbers kind of thing.

    FYI, the system I posted can actually be modified to produce a decent edge (not good enough for my tastes, hence my posting it here), but, its a good starting point for a newb maybe. Always KISS...

    IMO coin tossing at any price&time based extreme will be profitble... I don't feel the need to prove it, but, lets just say I do quite well with a 52% win rate :D
     
    #471     Oct 22, 2009
  2. P.S. I think your obsession with and use of this word, especially as a mod, makes YOU "the DOUCHE."
     
    #472     Oct 22, 2009
  3. Whisky

    Whisky

    Do you mean 52-48 W:L 1:1 average (4% edge average) or 52% edge?.

    Nice results and very high frequency. Congrats. I bet you do real well.
     
    #473     Oct 22, 2009
  4. Whisky

    Whisky

    He drew first blood. I'll stop when he apologizes or whenever I want. Those are the rules: No holds barred AND I'm the MOD, so I have the edge. That's how I like it. If you think I'm irritating, then try the ES for size.

    You will be called DOUCHE2 from this point on. How do you like that?. Feel free to apologize like a man instead of running to the site owner crying like a baby. I will not backoff even if they ban me.

    So DOUCHE2: Are you ready to make a test changing the EXITS as it was suggested above or you prefer to not participate anymore?
     
    #474     Oct 22, 2009
  5. sosueme

    sosueme

    that is scary, really scary.

    sosueme
     
    #475     Oct 22, 2009
  6. Thanks. I've been trying to take the frequency down a notch since my trading costs are about 23% of gross profits... I'd like to get that down under 10% but maybe that's not realistic unless I go to a shit broker.

    Net across all systems/trades is 51.94% win rate at W:L ratio = 4.150:3.757 (1.104:1). Slightly larger winners and more of em'.
     
    #476     Oct 22, 2009
  7. You will have to prove to yourself and others that your suggestions bring in at least 1 tic per day on average extra profit over the original method rules.


    Other than that :D : actually, the Mersenne Twister RNG (which is very high quality while quite speedy) is implemented in the PHP release I'm using (5.3.0), so all I have to do to get a random number is a simple call, e.g.:

    mt_rand(1,100);

    to get a random number b/w 1 and 100 included.

    IIRC from my previous search (until I found out that I already had it available in my programming envrnmt ^^), there's a lot of stuff on the Net readily available if you google it, including some pseudo-code and maybe even C or C++ code.

    Shouldn't take more than a few mns search, really. ;)
     
    #477     Oct 22, 2009
  8. Whisky

    Whisky

    Wikipedia says it's not so good.
     
    #478     Oct 22, 2009
  9. Whisky

    Whisky

    Are you using some sort of position sizing algorithm in the vein of a % of Optimal F or Kelly or how do you increase your size as you win more, if you do?.

    Some brokers pay you for orderflow if you use limit orders and have a lot of volume.
     
    #479     Oct 22, 2009
  10. I don't think so, matter of fact here's a couple of abstracts:

    in trading simulation we couldn't care less about cryptography. In fact a RNG algo tweaked towards cryptography is most probably not the most efficient for our purpose (randomness vs execution speed).

    and finally, most importantly:

    I'd personnally hardly call that a negative opinion... :)
     
    #480     Oct 22, 2009