Still not chrystal clear, bud? I apologise for my failures as a teacher.
With your permission may I try one (maybe) last time: - in Howard's QTS book he published the FDistribution for the S&P500 daily changes 94 - 06..... it is not normal but for my example it is near enough to it. - you buy at random and exit with +- 1% as your stop/profit, - excluding commission - allowing for slippage you will get everything on Howards graph excluding the prices bounded within +- 1% i.e. your FDist will == the left and right tails of the near normal dist. This is a nearly exact replica of a break even coin toss because: - it is binomial (it comprises two states i.e wins and losses) - the value on each side of the coin has a mean (something like +- 1.5% as a guess). It is stationary: - (it can't ever change unless the S&P does something totally contrary to all past behaviour), - my 10 year history shows S&P daily close is approx 50/50 win/loss +- insignificant variance, - it is bounded by your own rules, - you are highly unlikely to get favourable slippage from your broker (unless it is an abherration) - occassional and random, wild outliers won't change the mean result. How to evaluate the system?: The FDist of the samples is not normal. You certainly can not quantify the performance of the system on one backtest because the liklihood of 'one run' being representative of the system is very low. Knowing coin toss behaviour inside out: I could use BinomialSimulation, produce a CumFreqDist for the possible eq outcomes and then get a probability of risk of ruin. (I probably couldn't be bothered though and would just use some other quick and dirty stats method I learnt when playing Flip!) **************************************************************** YouTube video clip from "The Life Of Brian". Outside a crowd, comprising mostly ignorant peasants, gathers. They clamour at the windows and door, shouting for Brian to come out, "Brian, Brian, ........ Messiah!" Brian's mother goes to the windows and door to slam them shut. "Brian, the Messiah?" "He's not the Messiah, he's just a very naughty boy!" she yells back to the crowd, with finality. --- In [email protected], "Phsst" <ph...@...> wrote: > > And greetings to you Howard... > > I'll get rid of all the baggage of the various previous posts on the old > thread and only focus on one tiny statement you made that seems to do a > very good job of summarizing the gist of the subject: > > "My position is that any system that recognizes an inefficiency, and > trades it profitably, removes some of that inefficiency." > > Technical traders are all about identifying and exploiting repetitive > inefficiencies in the market. Most inefficiencies are short in duration > and therefore not repetitive enough to catch with backtested trading > systems before they vanish. But then some patterns (inefficiencies) > persist long enough to exploit for some period of time (which produces > better results than flipping a coin). > > While (for the most part) I managed to avoid the "Great Recession" of > the past year, I finally had to accept the fact that taking outright > positions in individual stocks had become too risky for my disposiiton > because of the chicanery, hype, mis-direction and outright dishonesty of > Corporate Leaders, Analysts, Talking Heads, Money Managers, etc. > Perhaps Brian's repeated references to "Coin Flips" struck an unwelcome > nerve because of my recent reincarnation into a complete skeptic of free > market management teams and the information they dis-siminate to an > unsuspecting public. > > This does not mean that I do not trade however. > > I've switched to trading Options on the underlying, using various spread > strategies to mitigate as much of the risk as possible associated with > adverse price moves. I can use AB to find lists of candidates for > manually exploring Options strategies, but I no longer search for > strategies which take outright positions in the equities. > > By putting together multi-leg Options positions I can structure some > pretty nice trades that have good pay-off's (when they work) and with > positions that act much like shock absorbers when the crap hits the fan > on the underlying. That does not mean that I don't take losses when > necessary... It just means that my losses no longer have a linear > relationship to adverse moves in the underlying. > > And finally back to your point about finding "inefficiencies", I have > found that the Options Market Makers have their fair share of > mis-pricing and inefficiencies which I've been able to occasionally > derive profits from. > > But I am way off topic. > > Good Luck. >
