Serving the Quantitative Finance Community

 
User avatar
flairplay
Topic Author
Posts: 130
Joined: September 26th, 2006, 1:34 pm

Do financial institutions make the best possible use of their quant resources?

October 27th, 2006, 6:25 pm

I sometimes think rather too many quants reckon the maths comes before the phenomenon... Not their fault, as banks hide them away to number crunch without exposing them to the really interesting bits. Sometimes traders are rather shallow and fail to see real issues themselves, or fail to point them out to their quants. I have always thought that smart quants exposed to real life would ask the right questions more often than they do . And I have a suspicion some may turn out better than average traders who get their jobs just because they were mathematically challenged in the first place. The cult of the trader is over blown - there are many monkeys out there (though of course there are many fine ones too).The interesting questions are often not in papers but in real life. Those with a mathematical bent, and with a natural curiosity and interest in the phenomenon, may achieve rather more for their employers than their roles allow them to. Maybe for this reason many quants leave quant work to do something else.I understand financial institutions are not set up for longer term goals, but if they were, the smart way would be to expose quants to the real world as soon as possible, with all its messiness and complexity. In the meatime, quants try to out do each other by invoking more and more abstract mathematics for the sake of it. Model building, especially intuition building, requires simple concepts, not pedantic discussions about sigma algebras. Of course simple concepts are the hardest to come by.It is not easy - but my advice to managers would be to see quants as 1/3 traders, and exotics traders as 1/3 quants. Sadly, that is rarely the case, and quants are rarely given the opportunity to get a feel for the actual world. A lot of expensively trained talent remains under utilised.
Last edited by flairplay on October 26th, 2006, 10:00 pm, edited 1 time in total.
 
User avatar
Lepperbe
Posts: 319
Joined: November 9th, 2002, 10:36 pm

Do financial institutions make the best possible use of their quant resources?

October 27th, 2006, 7:41 pm

excellent point(s).
 
User avatar
NorthernJohn
Posts: 632
Joined: June 2nd, 2003, 9:07 am

Do financial institutions make the best possible use of their quant resources?

October 27th, 2006, 8:12 pm

QuoteOriginally posted by: flairplay I have always thought that smart quants exposed to real life would ask the right questions more often than they do . And I have a suspicion some may turn out better than average traders who get their jobs just because they were mathematically challenged in the first place. The cult of the trader is over blown - there are many monkeys out there (though of course there are many fine ones too).I understand financial institutions are not set up for longer term goals, but if they were, the smart way would be to expose quants to the real world as soon as possible, with all its messiness and complexity. In the meatime, quants try to out do each other by invoking more and more abstract mathematics for the sake of it.Two points I'd make here.The first one is that it is getting a bit tiresome to hear, again and again, people hinting that quants are the more intelligent side of the trader/quant partnership. I just don't see it. It seems to reside in the strange idea that abstract mathematical ability (a particularly "pure" strain of intelligence) implies enough of the other pieces of intelligence to rise to the top in other arenas.And secondly, why the assumption that quants are being kept away from the "action". I really would prefer my quants to sit on the desk with me and the structurers, as part of one team, where ideas flow best, and where everyone can contribute to every discussion.They choose not to. They are happy sitting on a separate floor.
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

Do financial institutions make the best possible use of their quant resources?

October 27th, 2006, 8:31 pm

I've seen this problem in engineering -- it's both easy and comforting to believe in the mathematical model of the system. In the real world, linear systems aren't and neither are constants. I'm not sure about the "simple concepts" idea -- and I mean that I don't know, not that I disagree. Sometimes I think we fail when we try to reduce a complex system to simple concept (and sometimes succeed brilliantly). Sometimes I think we fail when we layer on so much complexity that it obscures a fatal flaw. Einstein's "simple as possible, but no simpler" notion comes to mind. The challenge is that the markets aren't simple -- being the nonlinear combination of all the participants stratagems and behaviors. Yet even as the markets show complex chaotic behavior, they seem to obey some very simple invariants.To me the challenge is in balancing a theoretic versus an empirical approach to markets and balancing a scientific approach to uncovering new knowledge versus a engineering approach to applying that knowledge. I'm sure that some mathematical quants are too wrapped up in their abstractions to see the disconnect between their model's axioms and the real world. I'm also sure that some nonmathematical traders are too wrapped up in the blood and gore of the markets to see the elegant invariant structures. I do agree that combining the two -- the purity of the theoretician and the street smarts of the trader -- is the way to go.
 
User avatar
flairplay
Topic Author
Posts: 130
Joined: September 26th, 2006, 1:34 pm

Do financial institutions make the best possible use of their quant resources?

October 27th, 2006, 10:35 pm

QuoteOriginally posted by: NorthernJohnQuoteOriginally posted by: flairplay I have always thought that smart quants exposed to real life would ask the right questions more often than they do . And I have a suspicion some may turn out better than average traders who get their jobs just because they were mathematically challenged in the first place. The cult of the trader is over blown - there are many monkeys out there (though of course there are many fine ones too).I understand financial institutions are not set up for longer term goals, but if they were, the smart way would be to expose quants to the real world as soon as possible, with all its messiness and complexity. In the meatime, quants try to out do each other by invoking more and more abstract mathematics for the sake of it.Two points I'd make here.The first one is that it is getting a bit tiresome to hear, again and again, people hinting that quants are the more intelligent side of the trader/quant partnership. I just don't see it. It seems to reside in the strange idea that abstract mathematical ability (a particularly "pure" strain of intelligence) implies enough of the other pieces of intelligence to rise to the top in other arenas.And secondly, why the assumption that quants are being kept away from the "action". I really would prefer my quants to sit on the desk with me and the structurers, as part of one team, where ideas flow best, and where everyone can contribute to every discussion.They choose not to. They are happy sitting on a separate floor.Interesting points.My take on point one: No one group has the patent on intelligence. You can never tell just by training or discipline. I see both misimpressions, that traders are smarter, or that quants are. Even a PhD from a top school tells you little about the individual applicant. Or 10 years at a hedge fund or a bulge bracket firm as a trader.But the point remains:Many quants with the inclination to do real life stuff are kept away because of the path dependency of their careers. Say you are driven and have some numerical talent. Instead of doing a business school MBA, u end up getting a theoretical physics PhD from Berkeley, driven both by the glamour of the field and the challenge. Ground realities about mutiple post docs lead you to finance. You are not necessarily unsuited to trading, or a worse trader than the business graduate who started at 21 in the graduate scheme. But you are pigeon holed. You may be a fair bit of an all rounder in your life, and therefore able to adapt - but trading is not the obvious entry point for you. Maybe you were never the dry individual that some suppose you are.Of course on the other side are the boring quants - the ones who believe the papers they read are the real thing. Models can only ever approximate reality, and writing down this SDE or that will never tell you how non linear real markets are. That the actual fun is non linearity. That the trader is really the non linear "control" hired by management. That ulitimately the fun of trading is the same fun as competitive sport. Instinct, timing, technical skills, anticipation, playing to your strengths, and understanding the interaction between your decision making and that of others. How that is what makes a market.Point two:That is sadly true. There is the quant type who completely avoids the mess and stress of the real world, happy to escape it, and happy to create a parallel world. A fun game is to look a the number of pointless quanty results in many papers going around which are either trivial to traders, or which satisfy Pauli's famous epithet of "not even wrong." How many of these results can a trader guess even without reading the abstract? Are there truly very many surprising results out in the quant literature?I think finance is a very young field, still immature, where many senior board members and the like are still not overly knowledgable and do not make the best use of the talent they hire. Finance is an applied field, with some elements of science, some art or better still artisanship, and many of competitive sport. The best practioners are those who combine all the above elements, and if you automatically rule out tehnically trained people, then you are imposing a narrow view of the world.In contrast, if you were to apply to NASA with an Engineeing PhD you need not automatically become a researcher, you may be the one in the astronaut program. In fact, your training would not be held against you - perhaps even seen as a plus.One of the analogies I often use is that while a butcher may be good with a knife, it does not necessarily make him a fine neurosurgeon. Years of hard training in medical school does not mean that a butcher, with his instincts for cutting a lean peace of meat, would necessarily make a better surgeon than his rigorously trained academic counterpart. Long as the medical graduate recognises that his thoretical knowledge is only a tool, that he must learn the practical aspects of neuorsurgery, he has no obvious disadvantages compared to a butcher - except a later start in the job market. Of course many quants suffer from hubris and make the mistake of not learning about the real world. Ultimately that is lack of common sense and general intelligence.But I have seen often of the other fallacy as well - the lore of trader as king. It has slowly disappeared as it is seen as those with instinct and timing, openness and skill, make the better traders, and technical skill - just like good technique acquired from years of practise in sport - can be a handy advantage if sensibly used. Still many a prospective neurosurgeon in finance is managed by a butcher without much appreciation of the finer points.I think as the field matures further, both instinct and skill - along with technique - will be seen as desirable attributes. We trade in ever more esoteric parameters and patterns, and not being able to understand the product will be a handicap. Just as it is in neurosurgery.Skill is never just instinct - it is talent honed by practise and hard work, acquisition of technique and ability to use it when it matters. So many supposed "instinct" traders with poor skill have got burned over the last 15 years that it is clear that skill is not just instinct. Increasingly skill in traders will be seen to be a combination of instict and technique - and to trade competitively you wont be able to manage without one or the other.So many traders I have seen delude themselves that they dont need understanding and knowledge - almost always with predictable disastrous results. And too many quants ignore that knowledge - especially idealised and limited - is only just that. A derivatives trader 30 years from now will have to have both - almost like a good test pilot. Rather rambling - and not sure I have articulated this too well.
 
User avatar
NorthernJohn
Posts: 632
Joined: June 2nd, 2003, 9:07 am

Do financial institutions make the best possible use of their quant resources?

October 27th, 2006, 11:45 pm

QuoteOriginally posted by: flairplayMany quants with the inclination to do real life stuff are kept away because of the path dependency of their careers. Say you are driven and have some numerical talent. Instead of doing a business school MBA, u end up getting a theoretical physics PhD from Berkeley, driven both by the glamour of the field and the challenge. Ground realities about mutiple post docs lead you to finance. You are not necessarily unsuited to trading, or a worse trader than the business graduate who started at 21 in the graduate scheme. But you are pigeon holed. You may be a fair bit of an all rounder in your life, and therefore able to adapt - but trading is not the obvious entry point for you. Maybe you were never the dry individual that some suppose you are.It is the idea that this person, who you describe, will be railroaded into being a quant that I disagree with.This is exactly the path I took. Talented physicist, who always assumed that I would always be in science. Very good first degree, a doctorate in theoretical physics, and a year in research, during which I realised how bad that world could be.So, what did I do?I applied to some banks, did the graduate program, and, at the end of it, became a trader.This story is repeated again and again in people I know.
 
User avatar
flairplay
Topic Author
Posts: 130
Joined: September 26th, 2006, 1:34 pm

Do financial institutions make the best possible use of their quant resources?

October 28th, 2006, 1:12 am

QuoteOriginally posted by: NorthernJohnQuoteOriginally posted by: flairplayMany quants with the inclination to do real life stuff are kept away because of the path dependency of their careers. Say you are driven and have some numerical talent. Instead of doing a business school MBA, u end up getting a theoretical physics PhD from Berkeley, driven both by the glamour of the field and the challenge. Ground realities about mutiple post docs lead you to finance. You are not necessarily unsuited to trading, or a worse trader than the business graduate who started at 21 in the graduate scheme. But you are pigeon holed. You may be a fair bit of an all rounder in your life, and therefore able to adapt - but trading is not the obvious entry point for you. Maybe you were never the dry individual that some suppose you are.It is the idea that this person, who you describe, will be railroaded into being a quant that I disagree with.This is exactly the path I took. Talented physicist, who always assumed that I would always be in science. Very good first degree, a doctorate in theoretical physics, and a year in research, during which I realised how bad that world could be.So, what did I do?I applied to some banks, did the graduate program, and, at the end of it, became a trader.This story is repeated again and again in people I know.Interesting. I doubt many are rail roaded, but at least in the early 1990's, many were naturally hired into quant roles - often not through the graduate program. Graduate programs tend to give you enough of a broad exposure to make you choose your mind.Those who get hired straight into one specialised area tend to miss making choices. Maybe things are a changing.....
 
User avatar
TraderJoe
Posts: 11048
Joined: February 1st, 2005, 11:21 pm

Do financial institutions make the best possible use of their quant resources?

October 28th, 2006, 11:49 am

QuoteOriginally posted by: flairplayI sometimes think rather too many quants reckon the maths comes before the phenomenon... Not their fault, as banks hide them away to number crunch without exposing them to the really interesting bits. Sometimes traders are rather shallow and fail to see real issues themselves, or fail to point them out to their quants. I have always thought that smart quants exposed to real life would ask the right questions more often than they do . And I have a suspicion some may turn out better than average traders who get their jobs just because they were mathematically challenged in the first place. The cult of the trader is over blown - there are many monkeys out there (though of course there are many fine ones too).The interesting questions are often not in papers but in real life. Those with a mathematical bent, and with a natural curiosity and interest in the phenomenon, may achieve rather more for their employers than their roles allow them to. Maybe for this reason many quants leave quant work to do something else.I understand financial institutions are not set up for longer term goals, but if they were, the smart way would be to expose quants to the real world as soon as possible, with all its messiness and complexity. In the meatime, quants try to out do each other by invoking more and more abstract mathematics for the sake of it. Model building, especially intuition building, requires simple concepts, not pedantic discussions about sigma algebras. Of course simple concepts are the hardest to come by.It is not easy - but my advice to managers would be to see quants as 1/3 traders, and exotics traders as 1/3 quants. Sadly, that is rarely the case, and quants are rarely given the opportunity to get a feel for the actual world. A lot of expensively trained talent remains under utilised.Read "When Genius Failed" to see what happens when quants give themselves too much freedom , and "The Smartest Guys in the Room" to see what happens when the traders are allowed to take over... Unsettling reading both. Given the profits the IB's have been making I'd say they have the balance about right and that you're just talking shit.
 
User avatar
flairplay
Topic Author
Posts: 130
Joined: September 26th, 2006, 1:34 pm

Do financial institutions make the best possible use of their quant resources?

October 28th, 2006, 12:01 pm

QuoteOriginally posted by: TraderJoeQuoteOriginally posted by: flairplayI sometimes think rather too many quants reckon the maths comes before the phenomenon... Not their fault, as banks hide them away to number crunch without exposing them to the really interesting bits. Sometimes traders are rather shallow and fail to see real issues themselves, or fail to point them out to their quants. I have always thought that smart quants exposed to real life would ask the right questions more often than they do . And I have a suspicion some may turn out better than average traders who get their jobs just because they were mathematically challenged in the first place. The cult of the trader is over blown - there are many monkeys out there (though of course there are many fine ones too).The interesting questions are often not in papers but in real life. Those with a mathematical bent, and with a natural curiosity and interest in the phenomenon, may achieve rather more for their employers than their roles allow them to. Maybe for this reason many quants leave quant work to do something else.I understand financial institutions are not set up for longer term goals, but if they were, the smart way would be to expose quants to the real world as soon as possible, with all its messiness and complexity. In the meatime, quants try to out do each other by invoking more and more abstract mathematics for the sake of it. Model building, especially intuition building, requires simple concepts, not pedantic discussions about sigma algebras. Of course simple concepts are the hardest to come by.It is not easy - but my advice to managers would be to see quants as 1/3 traders, and exotics traders as 1/3 quants. Sadly, that is rarely the case, and quants are rarely given the opportunity to get a feel for the actual world. A lot of expensively trained talent remains under utilised.Read "When Genius Failed" to see what happens when quants give themselves too much freedom , and "The Smartest Guys in the Room" to see what happens when the traders are allowed to take over... Unsettling reading both. Given the profits the IB's have been making I'd say they have the balance about right and that you're just talking shit.I dont really like reading finance books to be honest - they come with so much baggage about personality cults etc.Why Investment Banks make money is a different and altogether separate matter. I would be hard pressed to name many banks that dont make bumper profits in equity derivatives these days for example. To say because banks make money means that they utilise their resources well is a big extrapolation. Is the balance right, just because retail investors pay 300bp margin on reverse convertibles with a down barrier? Maybe. As well, the two examples you mention, referring to the two books, actually seem to reinforce my point about getting the mix right. You think the mix is right - I think sometimes it isn't and that going forward it will get better. It already has from the early 1990's.Also, kindly refrain from silly language - if you disagree with the point, you dont have to say "you are talking xxxx".
 
User avatar
AlphaNumericus
Posts: 258
Joined: December 25th, 2004, 9:17 pm

Do financial institutions make the best possible use of their quant resources?

October 28th, 2006, 12:29 pm

QuoteOriginally posted by: flairplayI dont really like reading finance books to be honest - they come with so much baggage about personality cults etc.Why Investment Banks make money is a different and altogether separate matter. I would be hard pressed to name many banks that dont make bumper profits in equity derivatives these days for example. To say because banks make money means that they utilise their resources well is a big extrapolation. Is the balance right, just because retail investors pay 300bp margin on reverse convertibles with a down barrier? Maybe. As well, the two examples you mention, referring to the two books, actually seem to reinforce my point about getting the mix right. You think the mix is right - I think sometimes it isn't and that going forward it will get better. It already has from the early 1990's.Also, kindly refrain from silly language - if you disagree with the point, you dont have to say "you are talking xxxx".Careful, that guy has a history of threatening to have people fired from their jobs for disagreeing with him here I don't agree that When Genius Failed, or The Smartest Guys in the Room, or Liar's Poker are "finance" books. Their meant to be popular books and to enertain the general audience - the people who don't need to understand the difference between stocks and bonds to buy these books. Derman's My Life as a Quant describes the relationship between traders and quants at Goldman Sachs (a very successful shop, mind you) in a much more interesting and detailed manner...
 
User avatar
twofish
Posts: 4944
Joined: February 18th, 2005, 6:51 pm

Do financial institutions make the best possible use of their quant resources?

October 30th, 2006, 5:01 pm

QuoteOriginally posted by: flairplayNot their fault, as banks hide them away to number crunch without exposing them to the really interesting bits.There was one bank that I interviewed for, and I was struck at how shy the interviewers were. One thing about mathematics is that it is one area in which you can be very shy and still do very, very well, so I suspect that some of the reason at least some quants don't get exposed to the action is that it really makes them unconfortable, and they perfer number crunching. I've found that physicists are much less shy than mathematicians. I suppose the reason is in mathematics you can literally go off into your own world, whereas in physics you have to interact with data which means interacting with people.QuoteThe interesting questions are often not in papers but in real life. Those with a mathematical bent, and with a natural curiosity and interest in the phenomenon, may achieve rather more for their employers than their roles allow them to. Maybe for this reason many quants leave quant work to do something else.I've found that there is a tension between the problems I think are interesting, and the ones which my employer thinks are interesting. The problems I find interesting are those which are mathematically challenging and those with "high impact" (understanding type II supernova collapse or the Chinese economy). These may or may not be ones which are currently profitable.It's not necessarily a bad thing. My problem is that I find too many things interesting, and one of the nice things about business is that it does give me some much needed discipline. Also the fact that I find lots of things interesting, means that there is a pretty good chance that given a problem, I can find some way of turning it into something interesting.I understand financial institutions are not set up for longer term goals, but if they were, the smart way would be to expose quants to the real world as soon as possible, with all its messiness and complexity. In the meatime, quants try to out do each other by invoking more and more abstract mathematics for the sake of it. Model building, especially intuition building, requires simple concepts, not pedantic discussions about sigma algebras. Of course simple concepts are the hardest to come by.It is not easy - but my advice to managers would be to see quants as 1/3 traders, and exotics traders as 1/3 quants. Sadly, that is rarely the case, and quants are rarely given the opportunity to get a feel for the actual world. A lot of expensively trained talent remains under utilised.