corso, i understand what you are saying, and yes, there is validity to your statement in some aspects of the mentioned disciplines. best, surfer
fortunately or unfortunately, as the case may be, many egotistical quant finance clowns are blissfully unaware of the latest quanatative studies on TA. for example, this study done at MIT in 2000 indicates a basis for so called technical patterns. http://web.mit.edu/alo/www/Papers/1705-1765.pdf best, surfer
if anyone is a clown, that's andy lo. nobody on the street takes him seriously. there are plenty of people who claim things like that, but so far there has been little proof other then a few papers. To put it bluntly - while nobody will ever claim that markets are truly efficient, Andy Lo keeps touting his gausian kr-s at every conference. Did anyone make any money on it? nah. andy is mixing up a number of things that good traders have gotten right long time ago - in particular, that TA and scalping are different things. But then again, he is an MIT professor, and not a trader.
Who told you andy lo is a clown?! If consulting for Goldman Sachs and other major Wall St firms for a hefty fees is a clown then Andy Lo is #1 clown. He's taken pretty seriously on the Street. His book Econometrics of Financial Markets is like the quant Bible. He's running a $300M quant hedge fund funded by Deutsche Bank. How much money are you running?
Well, there are plently of people including many other clowns (myself included) who did or do that. I was saying he is a clown for other reasons - he has a single idea that he was promoting his entire life and the mathematical apparatus he developed for that idea is vastly inadequate. While I do agree with his assertion about non-markovian nature of equite markets, I think attempts to tackle it with purely statistical apparatus are futile. To summarize - nobody told me, I read his papers and while some of the ideas are valid (like the data-snooping biases), he by no means can be placed in the same list as Jarrow, Ross and many others (that's what IAFE tried to do by electing him "Financial Engineer of the Year"). What I really do dislike about him is his relentless self-promotion, that is why i describe him as a clown. It is not. While there is no single quant bible (Hull is probably the closest candidate, with Musiela and Rutkowski a close second for the next level), his book by no means qualifies. Nevertheless, I do own it. First of all, it is not Deutsche Bank, it is DeAM, which is their asset management group. Second of all, he is not "running" a fun, there is a trader that is responsible for day-to-day operation, Jack Russo (if I am not mistaken). AL is more of a strategical advisor to the fund, but that does not make his research any more valid. Question "how much are you running" is not applicable to people on the sell side, especially in IR derivatives. To understand the scale of any IR derivatives operation on the sell side, think about the following fact - average notional on a vanilla interest rate swap is 10M (ranging from 1M to more then a few bln). A reasonably sized book would contain 5-6 thousand of these OTC deals. It is a different game with different rules. Did I answer all of your questions?
By no means, quant finance is not opposed to market timing. If you look at many FI funds, they are doing exactly that with the shape of the yeild curve. So far I have seen exactly the opposite - it is the technical traders that advertise their prowess. Have you ever seen a book "Make easy money with quant finance"? Or "quantitative trading techniques for dummies"? As for loosing money - what happend to all the great equity trend followers that made millions during the 90's? You apparently do not realize two things - quantitative finance in large is concerned with derivatives and derivatives are a zero-sum game. Somebody's loss is allways someone else's gain. What actually is being traded is risk. Secondlu, you need to get a handle on how much money is being piped through quant finance - obviously, you would be running large risks. And yes, there is no excuse for LTCM's debacle, but Barings, Daiwa, and many other non-quant related disasters show that money can be lost in many various ways. For instance, here is a short list of major non-quant related problems of the top of my head, just to give you the scale of things: Mexico, ~1995 - 20 banks went bust over the peso problem. Scale of loss - approx 72 billion, that's like 15 percent of Mexico's GDP Bank Negara (central bank of Malasia), 1992-93 - lost some 5bn over the collaps of the EMS - they bet that GBP will stay in and it did not.
They are both useful to a certain extent, why wouldn't? I really don't think these two approaches are mutually exclusive. Both approahces are still young today in terms of their development history. Possibly to avoid immature conclusions, we'd better just wait and see.
quant traders generally consider technical traders as "noisy traders" but it seems that quant traders can also belong to another category of "noisy" tarders if one reads this from Nassim Taleb interview in derivatives strategies magazine http://www.derivativesstrategy.com/magazine/archive/1997/1296qa.asp DS: You're not ready to give up on all quantitative techniques. You were trained as an econometrician. You don't make wild speculative bets and I assume you try to hire traders who have some kind of quantitative skills. NT: I have the following problem. Anytime I take a street-smart kid with a strong Brooklyn accent and train him or her in quant methods, I develop a wonderful quant trader who knows how to squeeze the sitting ducks. When you take extremely quantitative trainees, particularly from the physical sciences, and try to make them arbitrage traders, <FONT COLOR=RED>they freak out and become pure gamblers</FONT>. They can't see the edge, and they become the sitting ducks. The world has too much texture, more than they can squeeze into the framework they're used to. I see a huge incidence of pure speculative gambling on the part of these people who are hired on the strength of their knowledge of quantitative methods.
I agree with him: for example for my model I have hired a physician engineer to program the stuff, well he had done a marvelous job as for programer but never I will let him trade - I have tried then I judged that he hasn't the good profile . Nevertheless let's not generalise. Let's say that Physics world is far different from Trading worlds so that quant trainees who mostly come from Physics world have some difficulty to realise that market price is not really like atom orbits although there can be some analogy in details it is different: the evil always hides in details