You are totally INCORRECT! The first waves of quants in the early 90s to mid 90s were phds in physics hired "right out of particle accelerator" labs. Now, employers are a bit more picky and asking for financial experience. But at the really really hardcore quant trading desk/hedge funds, they do NOT ask for finance background. In fact, at DE Shaw and Renaissance having some financial background might COUNT AGAINST you. They want really really smart people who has not been tainted with finance knowledge. Just great ability to think. That's it. But for the average wall st ibank, they like a phd in math,physics, econ,finance, etc. PLUS strong software skills PLUS finance knowledge. But to say derivatives has nothing to do with physics is a yes and a no. The Black-Scholes option pricing equation is nothing more than a heat diffusion stochastic partial differential equation with the variables changed to reflect volatility, time to expiration, strike price,etc. In fact, Fisher Black, had NO finance training when he went to finance and came up with the option pricing model. Black did his undergrad and phd in math&physics at Harvard University. Robert Merton, who won the Nobel Prize for options as well, had a mathematical engineering background and did NOT take a single economics course before he enrolled in the MIT econ phd program. Physics and finance have had a long love affair in terms of cross fertizilation and borrowed analytical techniques: PDE, Brownian motion, Monte Carlo, etc. used to day in finance were all from physics domain. so, don't say it does NOT apply!
Actually, D.E. Shaw almost went bankrupt in 1998. It is(or was) true that some managing directors there never enter or execute a single trade. However, I heard that Renaissance have been doing well. http://www.elitetrader.com/vb/showthread.php?s=&postid=200371&highlight=shaw#post200371 http://www1.excite.com/home/careers/company_profile/0,15623,811,00.html
I'm not sure having a hard science/math background was WHY LTCM blew up. They blew up because of HUMAN ARROGANCE and HUBRIS. Has little to do with their pure academic training. There are probably countless other non-acadmic/non-math/non-physics traders who blew up many times. Just listen carefully here on ET! And in Market Wizard. Blowing up is just poor risk and money management and overleverage. Blowing up is a human nature thing. Though, I must admit, the type of thinking that LTCM embraced that markets will revert back to normality is particularly dangerous during catastrophe. misc
Interesting to hear people who (claim they) have multiple advanced degrees miss the point so completely regarding LTCM. Those that are interested should try to get a copy of Nova's program on the subject. They researched the subject thoroughly and in the segment, interviewed Merton, and other principals in order to develop an accurate picture of what really happened. As it turns out, what they "failed" to understand was how differently markets act when everyone is trying to get to the exits at the same time. Dynamic hedging works fine in theory but in an emergency, you still have to have someone to sell to. VAR (value at risk) suffers from the same problem. Hedging has become a complex subject that is discussed in detail at sites like (Paul) Wilmot's. Alternatively folks with the background to understand it can read Philippe Jorion's books on VAR or Neil Chriss books. Because they have had big problems managing derivative risk exposure, JP Morgan developed their own system called RiskMetrics. The training catalog deals with some of the same subjects. Clients with a wrap account might be able to get access to that info. Good luck. Steve46
Can someone explain to me what does the trader in big firms do? Do they basically trade using financial models developed by quant-developer or are they free to employ whatever methods they deem appropriate? And exactly what does these quant-developers do? Developing financial models with no financial backgrounds or trading experience? And after all their hardwork, can they derive an edge? If they fail to deliver any good models because there is no holy grail, how can they keep their job? Thanks
They mitigate risk, that is the objective as the house makes their profit creating and selling the product and the product has to be defended to be successful.
Loosenup: If we are talking about derivatives, the trader is subject to exposure limits. He/she gets a daily sheet that maps out exposure. and lists inventory. This is your "book". The general program is written and/or developed by the quant (the guy with the Phd) then the traders execute within the limits of the program. "Within the limits" means that they try to keep exposure to the greeks (delta/vega for instance) within a certain allowable boundaries. At the end of the day, your positions are reviewed by your manager and you had better be within limits. Losses are expected periodically. At the end of the year the bonus is calc'd from the bottom line. Hope this helps. Steve46
Hi: I am getting some PM's regarding LTCM and so I want to TRY to clear up the issue for those who are interested. Following is an excerpt from an article on the subject: "Although it is commonly thought that the trading positions taken by LTCM were predominately convergence trades, this is in fact not true. As LTCM's capital base grew the need for additional returns on that expanded capital led it to undertake other trading strategies. Although these trading strategies were non-market directional, i.e. they were not dependent on overall interest rates or stock prices going up (or down), they were not convergence trades as such. By 1998 LTCM had extremely large positions in areas such as merger arbitrage (profitable when the mergers were consumated, unprofitable if they were not) and S&P500 options (net short long term S&P vol). In fact some market participants believed that LTCM had been the primary supplier of S&P500 gamma which had been in demand by US insurance companies selling equity indexed annuities products for the prior two years. The profits from these trading strategies were generally not correlated with each other and thus normally LTCM's highly leveraged portfolio benefitted from diversification. However the general flight to quality in the late summer of 1998 led to a marketwide repricing of all risk and these positions then did all move in the same direction. As the correlation of LTCM's positions increased the diversified aspect of LTCM's portfolio vanished and large losses to its equity value occurred. Thus the primary lesson of 1998 and the collapse of LTCM for Value At Risk (VAR) users is not a liquidity one, but something more fundamental. Namely that the underlying covariance matrix used in VAR analysis is not static but changes over time." "A Deeper Understanding of Risks Taken by LTCM" 2000 in Literature. I am guessing that some folks will not really understand the logic, but there it is for those that do. Regards, Steve46
LTCM is still surrounded by mystique and inspires a general feeling of awe for many in this industry. With legends like Merriwhether, Scholes and Merton on the staff, its not difficult to see why. As to why they really went broke, I think it is a combination of factors. But the general lesson to be learnt from LTCM, IMHO, is that you need real traders, people who realise that something is not right and that the model is not working. Cut your losses, get out, sort out your stuff and get back in when your edge is back (or you realise that under such extreme market conditions correlations tend to go to 1). After all, the quant models are just that. Models of the real world, and we are not talking about chaos theory weather systems either, but people and mass psychology. I did see the Nova program "Trillion Dollar Bet". It was interesting to hear Scholes questioning the inputs to the models.... Getting back to the subject. For all who now want to go and do PhDs. Getting a PhD takes an enormous amout of work and dedication. You basically have to love what you do in order to slave away for years on a scholarship. If your main reason is to do it for Quant work, then please think again. You might be better off doing an MBA, or even a MSc in FinMath. I believe Chicago and Columbia have excellent FinMath MSc programs (Derman left GS and went to teach at Columbia). It should basically be a case of, you want to do a PhD no matter what. You just have to do it, regardless of whether you end up as a Quant. As times are changing, my recommendadtion to people who still want to pursue this path would be to get into a Finance PhD program with a strong mathematical perspective. Unless you are the best of the best, getting a PhD in Physics, Engineering or Applied Maths will in my view make it difficult for you to get into this area without prior finance experience. At least with a Finance PhD (with some serious maths) your options are open, from trading, to Quant, to Fin Eng, etc. Oh yeah, and learn to Code ! Just something for people to consider. Peace and good luck.