Sle, at least, there's one point we agree on. I think the press likes to focus on quant HFs that blow like LTCM,etc. It makes for good stories for people to hear. But I would BET $ that NON-QUANTS as a WHOLE have lost many multiples of quants, just because there are more of them. Just look at your typical mutual fund managers. The Janus and tech fund of the 90s have lost 50-70% of their asset. They have lost in the TRILLIONS! Yes, TRILLIONS. But no one said anything, because the public has been tricked into thinking "oh, it's my long term investment. it will come back." It might or might not come back. But unrealized loss is the same thing as realized loss for a real trader. How come no one is saying anything about those equity fund managers huh? I read in a recent Money, Bloomberg or one of those popular finance magazines about a Lincoln Asset Management(a money manager). They had $20B in 2000 peak. And now they have $1B. That means they lost 95%!!! 95% FREAKIN' percent of their asset! That made the LTCM's loss of a few billions look like nothing. How come the public is not enraged by that? And we are not even talking about the untold billions and billions lost by traditional funds either. The public is fooled as always. Hey, my point is that having a good strategy, good risk mgmt, good money mgmt, etc. are all necessary. REGARDLESS of methods used. I'm not championing QA, TA, FA,etc. I used a combo of all three with more emphasis on TA and QA. But before people start bashing on quants as gamblers you gotta know that traditional long term investors are the biggest gamblers of them all! They just sit in the position into the bitter end or bail out at the bottom or just hope in 20yrs things will come back. As Jesse Livermore said,"The biggest gamblers in the markets are investors. IF they are wrong, then they lose it all"
NT is yet another personality that is worth dicussing in depth, especially considering that all of his money was made by being long vol during a particular market crisis. And after that he writes books like "Fooled by Randomness". I personally think is is stuck up (and if you read his books it is apparent) and not worth much as a mathematician, despite all his claims.
Well, it really depends on what you trade - I would really like to see some street-wise kid trading CDO's . As for models - my motto is always "give me a simple model i can understand, not a complex model i can trust" A good trader has to understand the model - but if the model is really tricky (like HJM, for example), a trader has to be trained in the right stuff.
Ok, what are tradestation users doing? They aren't quants I would have to guess. I use Tradestation and I would have to say it is largely tech analysis oriented. TRAD, in one of their quarterly financial statements said that their average account was earning 7% per annum and that was a while back and in a really bearish market where the indexes were falling. Now if a portion of those people are basically learning, then even including newbies, tech traders are making money. I don't know what quants are making or losing annually on the average but the average fund loses compared to the indexes, so I can rank tech analysis oriented traders higher than fund managers and quant type funds I can't rank, I don't know where to get that info. As far as funds losing trillions, they did not lose it in a zero sum game like quants do, they just lost it because the price of equities fell across the board for a while, not much to compare there.
To really understand the opposition of quants vs ta one must remember this - I copy and paste from another thread - by Michael Young from Reef fund who wrote a paper before LTCM's debacle - untitled "why diversification doesn't work" - but I don't quote him for that but for his historical perspective in finance research - and who reminds that "there are three major schools of investment: Fundamental, Technical and Quant. Until the early 1970s, the Fundamental and technical schools supplied the ideas and methods that sophisticated investors used to make decisions. But recently, the Quant school has captured the bulk of academic and professional attention." I could also quote famous Hedge Fund Manager Peter Berstein who wrote "the Improbable Origin of Wall Street" where one can see that it is since Samuelson that TA have lost their prestige. But since LTCM debacles and the last Nobel Prize for Behavorial School they have regained some prestige but the quants have far from having lost theirs although they have been titled like "when genious failed" refering to their inabilities to take into account some realities about market's psychology according to the opposite school.
I do not think there is any opposition - comparing quants to non-quants in general is like comparing cars to planes. I do not think equity trading is that different depending on wether you are looking at head and shoulders pattern of some sort or at some index correlation and naming it some cute greek letter [disclamer: i have never done either]. On the other hand, I do not think it is possible to trade instruments that are more complex then vanilla options without some quantitative training.
If you agree with his assertion of non-markovian nature of market, I don't. As I said in quantum thread "some people talks about quantumic effect like other of fractals in market : as I said Elliott has never demonstrated fractality in market he only POSTULATED if not so it is a long time that scientists and economists would recognise Elliottism but they don't; it is the same thing for Quantum effect in market it has never been DEMONSTRATED ONLY POSTULATED. And the reason is very simple: as long as there isn't any true fundamental law of market that is found - that some can even doubt that it could exist - it is nearly impossible to distinguish from a random generating process - to simulate jumps in stock market one can add some poisson law process for example to "explain" the jumps in price but it is of course stochastic and not at all deterministic values so that quantum effect can be denied to exist." There is the same kind of reason for being in favor of non-markovian process : it is only BY DEFAULT OF PROOF - as when you have a suspect of crime you can't put him in jail as long as you don't have a sure proof so that you declare him "innocent" . That is to say the market is "innocent" from having a markovian nature only by default of proof as in the criminal metaphore. But I have found a fundamental deterministic law and can prove that the market has a markovian nature. In fact even academics like Mandelbrott recognise that the market must have a markovian nature since they talk about long term memory effect - notwithstanding short term memory effect that has been recognised officially so that they had to disconnect Market Efficiency Hypothesis - which is the pillar of Modern Finance that's why it needs to be defended at any price by academics gainst all barbarians attacks - from brownian motion and connect it with martingale process hypothesis which was less restrictive than brownian motion hypothesis - but the problem of market efficiency which is tightly linked to the proof of non markovian nature of market concerns long term horizon. And since about the 1995's at least even on long term horizon there are now official debates among academics questioning the markovian nature of market and so efficiency. So will they have to disconnect again the Efficiency Hypothesis from non-markovian nature as they have done in the past by disconnecting efficiency from brownian motion : I say yes for sure and the market will then stay "efficient" .