It's OK. We all know the OP's original statement is a falsehood generated from legions of failed traders who choose to make excuses (via a process known as mental masturbation) as to why someting cannot be done, rather than attempt to do it. He doesn't deserve your anger or disrespect, but rather your compassion and the patient kindness you would show anyone else who has a challenging handicap, because he will never gain the intellectual knowledge or do the real emotional and psychological work required to actually succeed at trading. Good trading, Jimmy Jam
Exactly. Transactions costs alone would, given a 50/50 outcome on every trade, bankrupt every scalper within a year or two. Simply to be breakeven after 12 years scalping is sign of some predictability in price movements. Profitability over such a period is an extreme outlier, even in a sample of 100 million traders. The statistical significance of a 12 year scalping track record is greater than that of the correlation between smoking and lung cancer, obesity and heart attacks, and many other accepted statistical correlations in the scientific community. Another problem with the "survivorship bias" and "lucky fool" idea is that you could equally apply such concepts to other competitive fields like professional sports, chess grandmasters etc. Given the number of people playing basketball and chess, clearly the achievements of Michael Jordan and Garry Kasparov were just luck
Try significance testing. Then apply it to the record of any successful high-frequency trader, not forgetting to incorporate commissions and slippage. Calculate the probability that profitability over a period of years, given this enormous transactions costs overhead, was down to random chance. A statistician did a standard significance test on the results of George Soros from the founding of his fund up to the mid 80s (when he stopped running it himself). The result was that the chance of his performance occuring by random chance was over 400 million to 1. How many hedge fund managers were there in the world from 1969 to 1985? Maybe a few hundred? A thousand or two at most? A trained statistician found that the results were significant to a degree far higher than most scientifically accepted correlations. What are your thoughts about that?
Cutten, Do you have a link to that study of Soros ? Alpha : http://www.investopedia.com/terms/a/alpha.asp
No it isn't. Come on, at least get your data right. Firstly, it is tick returns that you need to look at, that is the true data, not daily returns (which ignores intraday highs & lows - a market that fell 50% than rallied back to unchanged would be a 0% change day on daily data, for example). Second, just run a quick test on the movement on Black Monday October 1987, and tell me what was the chance of that occuring, based on a normal distribution. No one at the cutting edge of financial academia still thinks prices are normally distributed. You don't even know your theory - and this is academics, not even practitioners!
Unfortunately no, but it was mentioned in the foreward Paul Tudor Jones wrote to "Alchemy of Finance". Given the fact that they traded frequently, paying 1970s transactions costs (i.e. extremely high by today's standards) and compounded almost 40% per annum from 1969 to 1981, whilst the Dow stayed flat, I don't find the probability that surprising.
Out of interest, why are you in an S&P index fund? After all, the S&P has only had a statistically small sample number of years from which to draw inferences from the data. Isn't it possible that the performance of the S&P so far has been luck?
some points... it is correct that ASSUMING (Like many academics have done) a normal distribution curve is problematic (and i say this even as an advocate of market profile). it's also a flaw in the black-scholes formula...