I'm about 70 pages into (what I'm told is) the Bible in this area, Technical Analysis of Stock Trends (9th edition).
I can definitely understand its "classic" status. It is written authoritatively and defines terms carefully upfront. It has plenty of charts to show you exactly how to apply the concepts, and devotes space to objections. So as a neoclassically-trained economist, and one who has studied modern finance a decent bit, I was pleasantly surprised by the book's rigor.
On the other hand, it seems there is a basic problem right at the start. Now admittedly this is more of a philosophical issue; after all, they list impressive tables showing the results from employing basic Dow Theory from 1890 through the present. (N.B. here I'm a little skeptical because the book admits that actual Dow theorists can disagree in the heat of the moment about whether the market has reversed course, etc. So I'm not sure how much "art" went into the historical backtesting to call decision points. But even so, the results are pretty amazing; even if they're exaggerated by a factor of 3 there's still something there.)
Anyhow, the problem is this: Classical Dow Theory has as its primary tenet--and by the way, this amazed me; nothing is new in the social sciences!--that the stock market price right now fully discounts all available knowledge about the stock in question. So that's why (and here I'm paraphrasing a bit) the "fundamentalist" trader really can't hope to beat the market in the long run; other people can read those reports, look at the balance sheet, etc. That's why the true believer in Dow Theory thinks you need to look at the "market action" itself, in order to ride trends, if you hope to beat the crowd.
I have no problem with this point; in fact, as I alluded to in the interjection above, I was surprised at such an early statement (did Dow himself explicitly say this back before 1900?!) of the efficient markets hypothesis.
But the obvious question, of course, is why don't the technical analysts apply that same critique to themselves? How can it possibly be the case that drawing support and resistance lines, etc., can give someone an advantage in the market? How can it possibly be the case that the expert chartists know that we're in an uptrend, and that it would be really smart to buy stock ABC at such-and-such time?
To put it in other words, why doesn't the current stock price reflect all available information, including the information derived from the stock price's history?
Again let me reiterate, I'm not here writing off technical analysis. At the very least, if a big chunk of stock speculators believe in it, then every investor should at least be familiar with the terminology, etc.
All I'm doing in this post is questioning why someone who starts off with "the stock price reflects all available knowledge" would then spend time on a publicly available toolbox of techniques.