Satyr
As far as statistically verefiable studies go I would think the Brokerage statements of the thousands of sucessful traders would be enough.
The problem is that you also need to look at the unsuccessful trades, and you need to subject them to statistical analysis. Several people have posted analyses to me in the past, but all of those were seriously flawed by "data mining". Here is an example of how it works:
one individual contended that by looking at the crossover points between 30 day moving averages and 12 day moving averages it was possible to pick excellent buy/sell points. He proceded to show me chart after chart demonstrating this.
In looking at the charts I discovered that all of the stocks had an RS from IBD of at least 90 or less than 15!. In other words, this man was attempting to prove the efficacy of his method by choosing stocks that had moved significantly in one direction or another in the recent past, and those stocks were all selected after the fact (post hoc). This is the same kind of logical error that O'Neil makes with his CANSLIM approach.
Every statistical study I've seen of TA indicates that its predictions are no better than random. By risk management, I mean managing the beta of the portfolio. Since all serious studies of the market indicate that stock prices are unpredictable, it is pointless to try to calulate such things as support and resistance. They are illusions. Just consider the jargon: when these support and re4sistance levels are broken they are supposed to signal major price movements. But get a bunch of TA afficianados together and ask them what the support and resistance level of any particular stock is and you will not get general agreement. Furthermore, if resistance or support is "broken" and there is no major price movement it is widely interpreted as meaning that that point didn't really represent resistance or support. A human being's capacity for self-decption is boundless, and this vividly demonstrated by technical analysts' rationalizations of market movements.
Some years ago I subjected a number of stock price histories to extensive ARIMA analysis (this is a very sophisticated time-series analysis methodology). All of the analysis indicated that stock price movements were consistent with a "random walk". That is, there is no memory in stock price movements (i.e., no predictive information). This is entirel consistent with the weak form of the efficient market hypothesis.
I began investing in PSFT in 1995, and have an annualized rate of return in excess of 100%! I have added to my position periodically, and have not sold a single share (and thus have not incurred any tax liability). I will continue to hold PSFT for as long as it appears that the company can sustain rapid growth.
Regards,
Paul |
