Difficult, I subscribe to the self-fulfilling-prophecy theory a lot of people around here are propagating. If everyone's watching their indicators, they'll all buy at once and thus accelerate a rallye! But I also think that moderate indicators like MACD do work! Even if people aren't looking at them while making trading decisions!
But it doesn't really make sense that TA in general should be a self-fulfilling prophecy. In general terms: If TA predicts that the price will be low at time t1 and high at time t2, then anyone who believes in that particular prediction will plan to buy at t1 and sell at t2. The effect will be to drive up the price at t1 and to drive down the price at t2. IOW, the
reverse of what a self-fulfilling prophecy would do. (Self-preventing? self-inhibiting? ... )
From the other hand, it can explain trends existence: if price grows and everybody believe it will continue to grow for some time, everybody will buy, to ride the trend, and the price will keep growing.
TA is at heart a follow-the-herd mentality. As opposed to fundamental analysis, which is a lead-the-herd mentality. So you are right: if the price starts to go up, then the typical TA reaction is to say that "the uptrend is confirmed" which means it's time to buy. And vice versa.
But I propose the following theorem:
It is impossible for all three of the following to be true regarding a particular TA pattern:
1) The pattern makes well-defined predictions of price movements.
2) The pattern is widely recognized by lots of traders who act on it.
3) The pattern actually plays out as predicted.
My argument is basically what I said in my earlier post, which is that if 1) and 2) are true, then they will cause the opposite of 3) to be true. iow, the more well defined it is, and the more widely recognized it is, then the more untrue it becomes.
I agree with you on it. With trends it probably looks like entering/exiting too late.
So, my current stance on TA is:
1) Patterns do exist.
Proof: Statistical evidences (fat tails etc) and my poll
that shows that belief in TA is (almost) not correlated with self-deception.
2) However, an average trader cannot profit from it.
Proof: trading is zero-sum game, so majority loses. If this majority uses some TA, majority of TA users cannot profit from it.
3) However, an average trader wrongly believes he can profit from it.
Proof: he trades using TA.
So, at which conditions you can profit from TA? I can think of three cases:
-You just get lucky.
-Average trader is using TA, but you are much better in TA than average trader (you have a magic monkey, for example
).
-Average trader is not using TA (influx of noobs), while you are using those TA's instruments that cannot be explained just by belief in TA (like, you use MACD, but not fibs).
