There are 3 implicit concepts that are involved in the question: 1°) The concept of Risk of Ruin 2°) The concept of Martingale/Sub-Martingale/Super-Martingale which is the mathematical concept behind usual Fractional betting Rules in Trading or Gambling 3°) The concept of Independance The question is not really "Are trades independant", it is "Does the usual heard Money Management Rule imply that trades are dependant ? " Hey remember that your professors should have given 0 to your exam if you answer to another question than the one really asked . What are these 3 concepts: 1°) Concept of Ruin: see http://www.elitetrader.com/vb/showthread.php?s=&postid=457222&highlight=Risk+and+ruin#post457222 The definition of risk of ruin is nothing arbitrary and exists since prob has been "invented" with Pascal. There are two variants as far as I know: - the first is to suppose an initial capital (account) for the player and the banker - the second is to suppose an initial capital (same as above) and a capital objective which plays the equivalent role to the banker's account. In my prob faqs I didn't precise that it is the second variant that I gave : Risk of ruin = 1-P[K]=1 - {1 - (q/p)^K0}/{1 - (q/p)^K}. p is the probability of winning 1$ at each bet and q = (1-p) the probability of losing 1$ at each bet, K0 is your initial Capital and K the fortune you would like to reach. The faqs is not finished yet. I only gave the basic case, but it is necessary to understand the basic case before the more general case (if you wana risk Maxrisk of your account it's just substitution of K0 by 1/Maxrisk, if win/loss is not symetrical then Z = winpct / 100 * ( AvgWin) - ( 1 - winpct /100 ) * abs( AvgLoss); the average win/loss should be considered). When I will have time I will complete the FAQs. ------------ 2°) Concept of martingale: http://www.elitetrader.com/vb/showt...e=6&highlight=martingale and sub&pagenumber=2 In a coin flipping game the expectancy will stay the same whatever "strategy" - sequence of variation of leverage including the so-called ANTIMARTINGALE RULE - that is to say ZERO if the coin is fair (which is reasonable to assume for a casino). Mathematically something that can give only ZERO EXPECTANCY whatever strategy is used is called MARTINGALE. Something that gives negative expectancy is SUPERMARTINGALE and positive expectancy is SUBMARTINGALE. I will detail later : as I said here http://www.elitetrader.com/vb/showt...=6&pagenumber=6 I will make a thread dedicated to martingale. Stock market is not exactly like flipping coin in a casino : it has advantage over the casino (that's why I never go to a casino whereas I am in stock market : casino is a martingale/supermartingale except if the casino is idiot whereas stock market can be a submartingale). 3°) Independancy which is defined by Bayes Axiom: http://www.elitetrader.com/vb/showthread.php?s=&postid=462625&highlight=Bayes#post462625 p(A and B) = p(B) * p(A/B) (also = p(A) * p(B/A)
From above the answer is obvious: no as martingale rule it doesn't imply that the trades are dependant. Such martingale rule comes from gambling and as everybody knows independancy of roulette or any other casino game is practically assured. Now it can be true that in trading , trades can be dependant but the dependancy of trades doesn't come from the reason above per se.
no, it' s about 1 in 250. that's roughly 3 positions a trading day, which is probably on the low side for the ET crowd. from what i've ETers are at least double that. at which point you're talking about 8 straight losers happening every 8 weeks or so. but that's only on average, in reality things will be both better and worse than that.