Jack - That was probably the best explanation I have seen you give recently on your trading philosophy. But eloquence does not necessarily equate to truth. As I have been telling you for years, you really need to take a day off from YOUR orthodoxy and watch a one tick chart all day. Price most often leads volume, and often volume will even cause price to react in the opposite direction. Price leads the crowd around by the nose like a swineherd leads a hog to slaughter. Best regards. - Mike
My backtest used Spydertrader's chartscript code to quantify the "P,V boolean relation." If anyone has a more accurate way, I'm all ears. But it's really a very simple concept that I doubt Jack thought would be backtested. P.S. if you check the newsgroup archives, Jack strung along a guy from Harvard for over a year on this until it was concluded that it's boolean ballyhoo. Nothing of Jack's tests out that I'm aware of. That's one of the reasons why I think he's such a master of obfuscation... to try to keep from being exposed and to keep the naive begging for more.
Thanks for the feedback. JH's play on guiding newbies reminds me the international comedians show last night. It seems he'd be quite unhappy if not every trader becomes millionaire in a fairly short time.
Lots of TA doesn't work simply because too many people know about, plain and simple. The market, as discounting mechanism, reflects ALL trades, whether initiated through FA or TA, so "sell the news" phenomenom exists for TA as well. Typical chart patterns used to work far better when they needed to be drawn by hand; they don't work so nearly as well when everyone and their mother can pull up a chart in 2 seconds. So there is good TA and "old" TA; buying something because it shows a "classic" chart pattern has about the same chance of succeeding as someone who just buys a stock because earnings released were "good". I'm sure the same applies for QA as well.
Looks like this system would work if you took short positions, rather than long positions (ie turn graph upside down).
TA is subjective. If not go ahead and try writing a code that defines a Reversal Patterns (Head and Shoulder, x2/x3 Bottom, Cup & Handle... etc.) and Continuation Patterns (Flags, Triangles and etc.). Let's see if we all can agree on the indications. How about duplicating a technical discretionary trader's mind into a computer? This may seem irrelevant but if TA was "fully" objective then it's trading decisions can also be duplicated... It cannot be duplicated because it's subjective. Now, does TA work? Yes. Can TA be applied to an automated trading system? Yes. Does it matter if people know what kinda of TA you use? No. Trading system / Actual Trades? Yes. Liquidity and slippage. Just because I can't play golf, it doesn't mean other can't. Just because I can trade, doesn't mean you can, in reverse...
Interesting on it's own but it doesn't discuss whether a trading rule can be developed that might capture some of the conditional returns they analyze.