[DcifrThs]

[ QUOTE ]
The are a few problems with TA (all of which are really the same thing):

1) There are a variety of definitions of TA, many of which contradict each other.

[/ QUOTE ]

this is the one i have the most trouble with. wtf is TA? lol.

[ QUOTE ]


2) No one can ever explain why TA (regardless of definition) should work, only that "it just works". Which by the way, is a viewpoint sorely lacking in evidence.

[/ QUOTE ]

well on some definitions of TA there is definitely documentation and studies that prove it works over time but can't be expected to in the future. the reason is because it has been documented to have worked over time LDO.

my favorite viewpoint on the subject is from Benoit Mandelbrot. he is arguably one of the smartest people of the century (20th) and certainly contributed a ton to mathematical and statistical undertakings.

he wrote that with is methodology of replicating price charts, he fooled many technical analysts who immediately showed him supports, reversals etc. etc. etc. thinking that the price charts were actual market data instead of a totally made up random simulation.

he refused to believe that TA is worth while simply because he could so easily fool its practitioners.

he did concede though, and i do as well, that TA seems to work in some instances and there is one logical reason i can think of: one unchanging factor of TA seems to be human reaction to price movements. there are psychological factors that imo cause humans to act a certain way and TA uses proxies (unknowingly possibly) of those traits in its practitioning.

[ QUOTE ]
3) As mentioned, it is kind of curious that all of the long term great investors practiced some variation of FA.

[/ QUOTE ]

yet some great hedge funds use TA and are killing it (here you have to include quantitative strategies as TA since they are based on arguably non- fundamentel factors) ... i.e. renaissance, shaw etc.

it doesn't take mathematicians to do fundamental analysis but simons only hires the absolute creme de la creme from the math/stat world (for the most part he doesn't even look at US PhDs...which says something about our education in that subject). so i think he's pretty much a TA type, but again, goes back to wtf is TA???

great post btw spider.

Barron