# A Few Questions for Twofish-Quant.

1. Nov 13, 2009

### doinok

Hey Twofish-Quant,

I have a few questions for you and I posted them here so others can benefit from reading your responses.

I'm currently an undergraduate majoring in Computational Physics at a state school. Taking upper division Physics classes this year has made me realize that I like learning about computational math and numerical analysis more than the Physics material. I've become disenchanted with Physics as a career over the past semester, and I'm almost certain that I want to go into industry. My goal is to have a job that is intellectually challenging and has great pay.

Before reading your posts, my plan was to finish my undergrad in Physics and then get a Master's in Mechanical Engineering, focusing on computational fluid dynamics. Your description of Quantitative Finance has got me rethinking my plans, though. Programming is a lot of fun and I like learning about numerical methods, and it sounds like your job is my dream job: good pay, doing what I like, but...

But I want to make sure that I'm not deluding myself in thinking that I have a chance at getting one of these jobs, so I have a few questions. (And I appreciate you even reading this, so feel free to answer only some of, or none of, my questions.)

What does Wall Street's ideal Quant candidate look like? Are they partial to Physics Phds, or are they looking for anyone with an advanced degree in a technical field who's done a lot of numerical programming? Do they hire engineers? If so, would a Master's in Mechanical cut it?

I also hope that I'm not over-estimating my intelligence and ability to do something like this. What is the minimum intelligence that would suffice? 97th percentile? 99th? Are you and your colleagues extraordinarily smart (such that even a highly-motivated person with less intelligence couldn't compete)? Don't be afraid to hurt feelings.

How old are you? How long did it take to get as good at what you do as you are right now?

I assume that your job is stressful sometimes. Is there more "good" stress than "bad" stress? Are there some days where you hate your job?

How many hours per week do you work? Do you spend a lot of time thinking about your job? Is there time for socializing on the weekend?

Other than your current job, what types of jobs have you worked?

Are there any books on C++ that you can recommend? I'm looking for something that teaches good object-oriented programming style. Just to give you an idea of where I'm at, I have a background in Lisp, Matlab, and Java. My Java background is pretty weak and I learned most of my programming skills in my spare time.

What languages besides C++ do you recommend learning?

Are there any books on numerical methods that you can recommend? My math background: differential equations, vector calculus, linear algebra, complex analysis, a couple discrete math classes, and I'm currently taking a Scientific Computing class.

Again, I truly appreciate you taking the time to read this and answer questions.

Thanks,
doinok

2. Nov 13, 2009

### twofish-quant

First of all, don't think in terms of ideals. There are a dozen different types of jobs. I didn't move to Wall Street until I was 37, and I spent about seven years of my life doing general C++ computer programming elsewhere. You really do need to be flexible. No one knows what will happen next year, and if the bottom falls out of quant jobs, but people are looking for medical physicists or pastry chefs, you really have to be prepared.

A *lot* of computer programming is just bug fixing, and for that you just need warm bodies of moderate intelligence. If you have one super-genius and 10 million lines of code, that's just not going to work. This is why banks need large numbers of physics Ph.D.'s and universities don't. One Albert Einstein can change the world, but once he has, there's nothing left for the second guy to do. Now if you have 10 million lines of code to babysit, you are going to need thousands of people to deal with it.

A masters degree in mechanical engineering + a lot of experience programming somewhere else would work. Once you have work experience what your degree is in doesn't matter much.

Intelligence is overrated. I'm not particularly smart. Persistent as heck, yes, but I'm not a math genius. If you are brilliant, but you can't take orders and work with other people, then you are going to be have serious, serious problems on Wall Street. Being pleasant and cheerful under stress, is probably a much, much more important job qualification than intelligence. If you are average intelligence but you are cool when everything falls apart, you are going to do better than someone that is a super math genius that people can't stand to talk to or who completely freezes when everything goes bad.

The reason Wall Street likes Ph.D.'s has little to do with intelligence. If you have a Ph.D. then this is proof that once in your life you sat down and did a project on your own, and did whatever you need to do to get the job done even if you were bored and depressed.

One problem that Ph.D's have is that a lot of us get bored easily. I don't mind setting in front of a computer all day fixing bugs, but some people do. I do know people that do pretty much exactly what I do and they *HATE* it. Also, a lot work involves human relations, politics, and general working with other people. Personally, I like that part of the job. If you are a shy person that wants to sit in a room alone and program all day, it's going to be uncomfortable since you are always on the phone, on e-mail, chatting.

There are people here that are super math geniuses. I'm not one of them, but it's fun to learn from people that are. As I said, raw intelligence is not that important. Also if you are a super-genius, but you don't know something you need, it's not going to help you. If you are a super-genius, but you are *unwilling* to learn something you need, it's going to be really really bad.

The reason Wall Street likes people that are good with math is that if you are good with math, you are probably very flexible. Suppose you are really, really good with algebraic topology. Now it turns out that algebraic topology isn't that useful, but suppose you find out that stochastic statistical auto-coorelations models are import. Someone that doesn't have a math Ph.D. may just wait around looking for someone to teach them this stuff, but no one has time. If you are really good with math, then you go on google, download papers, buy some books from Amazon, and teach themselves what they need to know.

Same goes with computers. It's happened pretty often that someone just dumps in front of me, a computer program that I've never seen, written in a language that I've never encountered, and I need to fix it. Cool!!!!

At this point social skills are as important as technical skills, since the first things that you need to do are to find out who wrote this program. When you find out its someone that left the company two years ago, then you need to track down anyone that knows anything about the system. And then you have someone panicking on the other end of the phone, because he says that he needs it fixed tomorrow, and you have to calmly explain that yes you might get something done in about three or four days, but it's not happening tomorrow, so what can we do to deal with the immediate problem.

Again, this is one reason why Wall Street likes Ph.D.'s since this is the type of stuff that you have to do for your dissertation. If you have a Ph.D. and you don't know something, your first reaction is to go to the library to read about it, or to call someone up and talk to them about it. If you don't have a Ph.D., then you might just be sitting around waiting for someone to tell you what to do, which is just deadly, since everyone is too busy to tell you want you need to do, and no one really knows.

39

I've been programming continuously for about thirty-three years, and I'm still not that good. Part of the reason that I really like my job is that I get to work with and learn from people that are much better computer programmers than I am.

I haven't ever hated my job. Sometimes I dislike parts of it.

50-60

Yes. When I'm not at work I try not to think directly about work issues, but I'm always thinking about something financial.

Yes, but I'm not really social.

Programming continuously since I got my Ph.D.

For scientific programming, people are starting to use Python. Also learn something really, really bizarre. Prolog. Scheme.

Numerical Recipes. Also a lot of numerical coding isn't very well formalized so there is a lot of trial and error.

Last edited: Nov 13, 2009
3. Nov 13, 2009

### twofish-quant

One thing that you probably should to is to get involved in some research project. Pretty much any important research in astrophysics relies on massive amounts of computing power. One problem with the standard physics undergraduate curriculum is that it was developed in the 1950's long, long before you had cheap computers. There have been complaints that the standard physics curriculum is outdated, but no one is sure what to do about it. One problem is that curriculums are very inflexible. Any time you even whisper that you are doing to add or drop a required class, you are looking at a *MASSIVE* screaming and painful shouting match among faculty.

This is one reason why graduate programs like undergraduates with research experience. It's difficult to teach real world scientific coding in a classroom, because no one knows quite what to teach.

The other thing is that I personally thing that the whole framework of an education consisting of classes, grades, electives, or even universities, might be wrong. The current framework of university education with classes, electives, courses, etc. was invented by Charles W. Eliot at Harvard in the late 19th century. It's important to realize that someone *invented* the system, because if you don't realize that human beings created the system, then it becomes impossible for you to think that there might be some other way of doing it. This is why it's important to talk to people in education and social science departments because a lot of them have researched this sort of history. (And why do universities have departments? Who invented the admissions committee? Who decided how the admissions committee should work? Why do physicists look down on people in science education? Who decided that you get more prestige if you are good at math? Who invented the SAT? Who invented the GPA? Why do I care? Why should I care?)

Personally, I think that the purpose of an undergraduate education is to get people to *THINK* and *ASK QUESTIONS* (and who came up with that idea?) Personally, I think that bad things happen when people stop asking questions. I like asking questions, and I found it to me useful. You look at the COMMITTEE, and you think to yourself "so who put you guys (and why are you all guys) in charge?" "why am I scared to ask them and challenge their authority?"

Sometimes it's just too dangerous to ask a question out loud, but if you are in a situation in which you can't even *think* of the question, then this is really bad. Personally, I'm very curious, and it's impossible to get me to just ask questions about the big bang, and not ask other questions (like why are you in charge? why are you making more money and more powerful than me? what are you trying to do to me?)

It may be too early to make definite decisions in careers. Also, you should really look at getting a Ph.D. in something heavily numerical. People just have to stop thinking about a Ph.D. as solely as training for professors. There are a lot of careers in physics that have nothing to do with being a professor. Also, there are lots of physics professors that are pretty good programmers.

Last edited: Nov 13, 2009
4. Nov 13, 2009

### twofish-quant

The cool thing about google is how much you can learn if you have the right keywords. Some of the useful keywords are:

cultural hegemony
cultural reproduction
Antonio Gramsci
Pierre Bourdieu
Jurgen Habermas

The cool thing about these concepts is that they are *useful*. Once you read up on the idea of "cultural hegemony" and "cultural reproduction" you start thinking to yourself, so *that's* what's going on.

One reason that I got my Ph.D. was that once I understood the rules of the game, it was obvious that if I quite grad school, and said the "the system stinks let's overthrow it" I get marked as a loser. Now that I have a Ph.D. and I say "the system stinks let's overthrow it" then the system can't play that card.

Personally, I think that for MIT to have any admission's, committee at all is really, really stupid, and that MIT should admit anyone to physics grad school that wants to attend. Now if I wasn't an MIT alumni with a Ph.D., you can totally ignore that idea. But since I understand the rules of cultural hegemony, you can't play that card against me. The really cool thing is that since I'm an MIT alumni with a Ph.D. and I can say that I think it's stupid for MIT to have any admissions committee at all (I mean does google have an admissions committee? does Microsoft have an admissions committee?), people that *aren't* MIT alumni now have permission to *think* that graduate admissions committees are stupid and that MIT should have open enrollment like the University of Phoenix.

Now of course, people will argue that MIT must have admissions committees, but the thing is that now they have to argue the point rather than just assume it and have people accept it without thinking, and you look into the eyes of people on the COMMITTEE, and suddenly they have very worried looks on their faces.

Cool......

(And MIT is just basically a community college who ended up having some very powerful political friends.)

5. Nov 13, 2009

Staff Emeritus
What happens when MIT gets 1000 graduate students? Where do they put them? How do they fund them?

There needs to be some throttle on the system somewhere - at your bank, it's HR, but presumably they are taking their orders from management, who tosses around words like "headcount". For graduate school, it's an admissions committee.

6. Nov 13, 2009

### maverick_starstrider

There are grad programs in Computational Finance btw.

7. Nov 13, 2009

### kote

Just to add, there may be programs that are be a better fit for your actual interests. My undergrad lab partner is at a top 10-15ish applied math PhD program, and banks were pitching jobs to her department her first year there.

Correct me if I'm wrong, but although physics PhDs historically make up a good portion of "quants," physics isn't necessarily preferred over other math-heavy subjects. If you aren't seriously considering even doing actual physics in industry, you might want to take a look at math, applied math, or some other field.

This would broaden your choices when it comes to jobs you may actually be interested in, and you may discover more opportunities along the way. Opportunities may change by the time you graduate as well.

8. Nov 13, 2009

### kote

Oh, if it helps she was Phi Beta Kappa with a double major in mechanical engineering and math and thesis honors in both majors. I don't know about GREs. YMMV, and I don't have any idea what range of schools banks recruit from.

Last edited: Nov 13, 2009
9. Nov 13, 2009

### twofish-quant

Don't know. But smart people can figure something out.

But if you show up at the door with money that you want to deposit, no one is going to say no. Also why *does* there need to be a throttle?

10. Nov 13, 2009

### twofish-quant

Correct. However, in recent years what people have been looking for are people with lots of computational skills. If you have C++ experience that puts you ahead of someone that just does "pencil/paper" math and physics.

Yes. Statistics is a hot field. Also do what you think is fun. One reason I end up in good shape is that I love coding, and so when I go home, I relax by doing C++ programming. If you *hate* C++ programming, then a job that requires it is going to be a major drag.

And they'll change after you graduate. One thing that you have to ask yourself is whether or not it's good for society that there are so many people in finance. It's possible that finance is a bubble that will blow up. Personally I don't think so, but you have to think about this for yourself.

11. Nov 13, 2009

### George Jones

Staff Emeritus
Are you saying that one can buy one's way into grad school at MIT? If you so, do you have documented eveidence?
So, you agree with Alfred Schild? In the August 1989 issue of Physics Today, Engelbert L. Shucking wrote

12. Nov 13, 2009

1000 graduate students at US$35,000/student works out$35 million a year in a $2.5 billion/year budget. If we start talking about 10,000 or 100,000 graduate students then it starts to get more difficult. MIT has about 4000 graduate students. However, people have been working on string theory for 20 years. What I hate are conversations that go "we can't think of a way of doing something in ten minutes so let's give up." It's amazing to me how people that presume to think grand thoughts about the big bang and want to be the next Albert Einstein suddenly get so unimaginative when you come to something like funding 1000 graduate students. If MIT has ten million students taking freshmen physics, there will be funding for 100,000 graduate students. The throttle is money. Management takes their orders from the market. The problem that MIT faces is that the budget structure right now just is not sustainable. If MIT were teaching freshman physics to ten million students, then it would need *TONS* of graduate students and tutors. Personally, I think that MIT *can* teach freshman physics to ten million students. If they won't. University of Phoenix might if they can see money in it. If UoP won't. *Someone* will, and I'd rather MIT lead this rather than follow. Last edited: Nov 13, 2009 13. Nov 13, 2009 ### twofish-quant Sure can. What do you think executive MBA programs are all about? And it's not that hard to buy your way into the MIT physics program. If you have money, you can study physics without worrying about a job, and pay for the best physics tutors and attend conferences to network with professors. Once you have all that, then you can get in without too much trouble. If all else fails, if you have like US$1 million to spend, you can pay MIT professors to give you one on one physics tutoring.

If you have enough money and average intelligence, you can always buy you way in. The question is how.

Personally, I don't have any problem with MIT teaching courses to anyone that can afford to pay for them. Granting degrees for money is something different.

Well there is a whole website on courses at MIT that pretty much anyone with money can take:

http://mitsloan.mit.edu/execed/

MIT will even create a custom graduate degree for you if you have the money.

http://mitsloan.mit.edu/execed/custom.php [Broken]

One reason elite universities will let people buy their way in is because much of the reason middle class families want their kids to go to Harvard/MIT is that your kids get the meet the kids of people that can drop $5 million to get their kids in. One thing about wealth it's just like knowledge. Now that I work on Wall Street, I feel a lot poorer than I did before I came up here. I'm making more money, but I'm exposed to people that make even more money. 16. Nov 13, 2009 ### doinok Thanks for the replies. I appreciate it. 17. Nov 14, 2009 ### twofish-quant There's another important reason I ended up in finance which you can see from the responses to my crazy ideas. The standard response to anyone that suggests any new and crazy idea is "that's nice, too bad we can't do this because we don't have the money." At which point you are supposed to stop thinking. "We can't do this because we don't have the money" is an answer I've gotten so many times that I figured that it might be a good idea to learn about money, which is why I went to Wall Street. It's part of my education. So now the conversation goes like.... Me: So I think it's a good idea if MIT has open admission for graduate students. Committee: That's crazy we don't have the money or the buildings to do that. I'm supposed at this point shut up, but that's not how I think the conversation should go.... Me: OK. So how much money do we need? Committee: Ummmmm.... Uhhhhh,,,, Me: OK. Since you haven't actually calculated this. Let me do this.$35,000/year for a grad student stipend. Let's say MIT has 100,000 grad students. That works out to a budget of $3.5 billion/year Committee: See that's crazy talk. You can't possibly raise$3.5 billion a year. The only people that have access to that kind of money are people on Wall.... Oh.......

Me: OK. So if we decide it's a good idea, and we can keep the budget under $5 billion/year, I'll figure out a way of raising the cash. So educationally spending, is it a good idea or not if MIT has open admissions. If it is we'll figure out a way of making the financing work...... MIT has$10B in its endowment. It's current budget is $2.5 billion. University of Phoenix has revenues of$2 billion/year and growing. $3.5 billion/year is not even a small company..... The annual revenue of the Coca-Cola corporation is$32 billion/year. The annual revenue for ATT is $110 billion/year. Apple makes$5 billion/quarter on the iPhone. Congress just spent $700 billion bailing out the banks.$5B is pocket change.

Having MIT admit 100,000 graduate students may be a good idea. It may be a bad idea. But it's not an impossible idea. So if we can do it, is it a good idea or not?

Committee: But the housing.....

Me: We'll run the numbers to see how much money we need. If we need to, we can open satellite campuses, distance learning centers. Walk down Main Street in Cambridge, you see lots of empty apartment buildings. I can find the names of the banks that hold the mortgages, and they might be glad to get rid of those properties really cheap.

Committee: But the political problems, marketing, legal issues.

Me: I can find lobbyists, salespeople, and lawyers. Look, the fact of the matter is that we *can* make MIT open admissions... The question is *should* we do it.

Committee: Ummmm... Errrrrrr..... Uhhhhhh.....

18. Nov 14, 2009

### AsianSensationK

I know I didn't actually ask the question, but this thread has been really informative and helpful to me as well. I've appreciated reading your responses as well twofish-quant.

I do have something I'd like to ask though, if it's not too intrusive. What kind of programming jobs did you have before you made it to Wall Street? Does your current job make use of a lot of the skills you had to use/develop at earlier jobs? I do assume the types of projects you do are going to be dramatically different, but I'm just a little curious.

I love this piece of advice. As I get older and work a few more professional-type jobs, I'm beginning to realize it's not always about what you know but the qualities you've developed through the experiences you've had.

I think that's why the best employers typically want both a resume and a cover letter from candidates. Maybe just to get a small sample of how you think.

Last edited: Nov 14, 2009
19. Nov 14, 2009

### vanesch

Staff Emeritus
Just a small and maybe off-topic reply to the idea of open admissions at MIT. The same issues are often discussed about some famous schools in France: extremely difficult to get into it (but nothing special once you are there).

I think that what has not been integrated (strange for someone from Wall Street ) is that value = rareness.

If you distributed 50 diamonds to every girl in the world, they wouldn't love diamonds anymore. What makes these hot shot schools worthy is that they only admit a small number ("the best"), so having been there has value because it is rare.
The aim is not the quality of the teaching or the special skills you get there. The aim is to be able to say that you belong to the small and select club who went there. That you've been a winner of a particular kind of competition.

Make 10 million people follow freshman physics at MIT, and MIT doesn't mean anything anymore.

20. Nov 14, 2009

### Sankaku

I like your "out-of-the-box" thinking, twofish-quant. I doubt that MIT will ever change that drastically (or any of the other prestige universities as vanesch points out). However, it is the kind of thinking that gets some other university who is not in the "club" to change and suddenly eat everyone else's lunch.

21. Nov 14, 2009

### atyy

So if MIT admits everyone, then it will be valueless, then no one will want to go there, so it can admit everyone. QED

22. Nov 14, 2009

### AsianSensationK

Yes, this is exactly right. Make 10 million people freshman at MIT, and MIT loses it's prestige.

It's not just that value = rareness though. There are a lot of really small schools out there no one cares for. The trick is, MIT has an image and reputation that evokes ideas in a person's mind. It's a brand. There's a surprising amount of information conveyed in a brand name, and I can't say I blame people for wanting to be a part of the club. There's a lot of financial value in owning a brand/ being branded.

The school's position is probably a little different though. They have a brand that they need to protect and that is done by carefully selecting who they allow into "the club." They need people to go out in the world, attain respected positions and be successful. Then donate. That makes school look like a terrific place to attend and study.

It's kind of an organic marketing process. Get smart, successful people to attend. Make money, gain respect. Then respect draws more smart, successful people to the school. And the people who are willing to throw money at them.

Last edited: Nov 14, 2009
23. Nov 14, 2009

### atyy

So if in fact "the best" is not true - then the non-elitist advice would not be for MIT to admit everyone, but to find the best somewhere else - another school, another profession - maybe even one that doesn't involve science. I dare say many from MIT might even agree that there is no best - worthy human endeavours are more like the complex numbers than the reals.

24. Nov 14, 2009

### doinok

25. Nov 15, 2009

### twofish-quant

And it shouldn't mean anything.

The fact that you got your degree from MIT ***should*** be totally irrelevant. It's a *HORRIBLE* thing when people want to take a waves class at MIT when they could get a better educational experience at a community college. If a community college can teach freshman physics better than MIT, then people should go to the community college, and if they won't because of branding nonsense then we got to change this.

The fact of the matter is that classroom instruction at MIT is not particularly good. Your average community college has better teachers than MIT has. MIT's main focus is research, and it has motivated, brilliant students. If your students are good then you can have some professors that are totally incompetent at teaching. You can have a professor that just cannot teach, and then the students will figure out how to learn the material.

MIT has a wonderful culture, and one part of the culture that is wonderful is that MIT teaches you to hate brands, and hate MIT.