Why do quants need to know programming?
- Thread starter garfangle
- Start date
Quote from garfangle:
Basically what you are saying is that if one wants to fashion oneself as a quant trader, if one do not currently possess programming skills, he'd better get them to be taken seriously.
V. informative, thx!
Hell yeah!!!
Programming is like lingua franca in the quant world. Not being able to program is like being illiterate. It's like you can't even speak the basic language of the field. The other language is mathematics and quant finance.
It's like the basic language of everything. I think programming should be a prerequisite for ALL high school graduates in the 21st century. I mean you don't need to be a software developer, but having decent to good programming skills are like the basic of modern society in order to create anything of value.
It's very hard to create anything of value otherwise. Without computers, you can't scale anything. You might have the world's greatest trading ideas, but if you can't implement them then it's next to impossible to make any real money. Institutions are interested in scalablity. How big can we run this strategy? Even if you made a million dollars last year trading, it might still NOT interesting enough for them if it's not scalable. A million is pocket change. It can't even pay overhead for them.
The nice thing about prop trading a la ETers is that there are virtually no overhead. So, if you make $1M then it's yours to keep basically. Not that I'm there yet. Or anywhere close. But I'm still working at it. hehe.
good luck dude!
trader99
That is right. The first developer messed up the risk management system because he was a pure hacker. He liked to code for the sake of programming. If you apply for a developer position at a hedge fund, it is very likely that you will be asked some maths and finance questions too. Competition will get tougher and tougher. Just my two cents.........
My two cents: Programming is easy. A typical programming language has about 50 valid constructs (sentenses if you will). Most 2 year olds know more sentences than that.
So the issue is not the programming; it's what to "say", ie the algorithms, models, trading logic. Here a strong academic background (math, physics, CS, EE) is what counts for any quant.
So the issue is not the programming; it's what to "say", ie the algorithms, models, trading logic. Here a strong academic background (math, physics, CS, EE) is what counts for any quant.
Well, I think you mean coding, do you? Programming is a term that refers to tasks more than just typing whatever language constructs in a meaningful or meaningless way.
Techies will be as happy as two year olds if coding is their only jobs. In fact, damn greedy employers now ask more crap like so-called interpersonal skill or something like that. Be an independent trader!!
Techies will be as happy as two year olds if coding is their only jobs. In fact, damn greedy employers now ask more crap like so-called interpersonal skill or something like that. Be an independent trader!!
Anyone coming out of any accredited institution with a degree in math, engineering, business will have some kind of training in programming. Earlier you asked about back in the 80's ... I can only guess that back then it might have been FORTRAN, perhaps PL/1. I graduated in the mid 70's and programming was required both in my undergraduate Quantitative Methods (Business) degree and also in my Operations Research (Engineering College) graduate degree. Prior to my choice to pursue trading when I worked in the corporate world in an engineering role we often had ideas that we developed quick & dirty code for ... just to test out ideas/concepts. It would have been impractical for me to take my ideas to some coder and then have to explain everything to him/her just to build a prototype. I could write my code on the fly ... test it ... tinker with it and continually refine it. If I had something that looked promising we'd clean it up, add some graphical user interface to impress management, demo it and often take it forward as a full fledged project. But to toss the ideas/code back and forth like a volleyball would have been time inefficient and not maximized productivity. We needed domain expertise to develop effective systems and I suspect quants do as well.
All the other reasons already noted are correct, however the most fundamental reason firms don't look at quant candidates who can't program (you're so called "perfect fit" vs. just a good quant) is because they don't have to - they can find people with all the skills. If you can't program software to develop and implement the strategies you're supposed to be coming up with, they don't need you - and as others have said, it's very unlikely to find a really good quant who can't program so there's an implicit corallary that if the guy can't program then he's probably not as good a quant as he claims, so why bother talking to him.
Yep. That's the implicit assumption. I mean, in the grand scheme of intellectual thoughts and academic thinking, programming is NOT that high up in the ladder. I mean if you went to a good school or as smart as you claim you are then you should be able to program. And if you DON'T know how to program then you should be able to learn it pretty fast! For God Sakes, there are books out there that claims to teach you to learn C++,VB, C or whatever in 21 days! haha. And if you are that smart of a quant, then you should be able to read those books and be able to pick it up in less than 21 days. haha.
But the truth is, the pool of people that competes for quant jobs are very talented academically. They have a Ph.D. from a top Ivy League school or a top reputed academic program, they might have even won a Math Olympiad when they were a kid, they can program in 5-6 different languages C++,Perl, VB, Excel, S+, SPSS, Matlab,Mathematica,etc. , have high math ability, and get this SOME even have SECOND Phds in finance. That's how sickening this field as become. I've met quants who have their first phds in math/physics(MIT,Harvard, Princeton), and their second phds in finance(Wharton, the top finance program in the nation) and they have taken graduate courses in computer science(at Carnegie Melon or Caltech). So, they know quant finance theory, really good at math, and can IMPLEMENT it like a pro software developer.
So, there's no lack of quant talent out there. The more you know the better. And it's getting worse every day, because all these really smart academic types hear how much their ex-collegues in hard sciences who left physics and math are making on Wall St and they are coming by the drove. You got guys who are string theorist from Harvard and MIT. Ex math professors sitting on quant trading desks. So, the skill requirements in quant will get MORE DEMANDING not less over time.
The only way for you to NOT be able to program and get hired as quant is to be in quant research. Like the head of quant research. You gotta be sooo BRILLIANT that they will pay you just to think. Fisher Black was head of quant research at Goldman Sachs for years. But before that he was a professor at MIT and got his undergrad and phd at Harvard in math and physics. And he invented the Black-Scholes models and the Black-Derman-Toy fixed income derivatives model. And the successor to Black at GS was Emmanuel Derman,who was a top ranked astrophysics researcher at Bell Labs for a number of years. Derman got his Phd from Cambridge University. So, unless you are brilliant at THAT level, then you probably don't have to program. You can tell your mere mortal local run-the-mill phds type to do your programming, b/c you are the head of research hehe. They got guys who are ex-math professors at Berkeley or renowned in their field who leave academia to go to Wall St and make a few million a year just doing research. Unless you are at that level, then you don't need to program. And get this, and even THEN, those pure research guys can probably still outprogram the average programmer. haha.
good luck!
trader99