A Few Questions for Twofish-Quant.

[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

 

[twofish-quant]

What does Wall Street's ideal Quant candidate look like?

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.

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?

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.

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?

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.

Are you and your colleagues extraordinarily smart (such that even a highly-motivated person with less intelligence couldn't compete)?

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.

How old are you?

39

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

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

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

How many hours per week do you work?

50-60

Do you spend a lot of time thinking about your job?

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

Is there time for socializing on the weekend?

Yes, but I'm not really social.

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

Programming continuously since I got my Ph.D.

Are there any books on C++ that you can recommend? I'm looking for something that teaches good object-oriented programming style.

Brian Eckel. Also start downloading code and start programming.

What languages besides C++ do you recommend learning?

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

Are there any books on numerical methods that you can recommend?

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

 

[twofish-quant]

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.

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?)

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.

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.

 

[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.)

 

[Vanadium 50]

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.

 

[maverick_starstrider]

There are grad programs in Computational Finance btw.

 

[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.

 

[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.

 

[twofish-quant]

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

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

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".

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?

 

[twofish-quant]

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.

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.

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.

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.

Opportunities may change by the time you graduate as well.

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.

 

[George Jones]

But if you show up at the door with money that you want to deposit, no one is going to say no.

Are you saying that one can buy one's way into grad school at MIT? If you so, do you have documented eveidence?

Also why *does* there need to be a throttle?

So, you agree with Alfred Schild? In the August 1989 issue of Physics Today, Engelbert L. Shucking wrote

... he felt one should get rid of students who attend it (GJ: university) solely for the purpose of obtaining a diploma. Hs solution? "Grant every American a Ph.D. at birth."

about Schild.

 

[twofish-quant]

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

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.

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".

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.

 

[twofish-quant]

Are you saying that one can buy one's way into grad school at MIT?

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.

If you so, do you have documented eveidence?

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

Now you do have to be an "executive" to get into these programs. But if you have like US$25 million and you say you are an executive, who is going to say that you aren't.

The hyper-rich can do whatever the hell they want. What I want to see is that the middle class have some of the perks that the hyper-rich already have.

So, you agree with Alfred Schild? In the August 1989 issue of Physics Today, Engelbert L. Shucking wrote

No. I said that I think that MIT should *admit* anyone that wants in. I didn't say anything about granting degrees. In you want a driver's license, you have to pass the test. There needs to be a "test" that you need to pass for MIT to certify that you have a bachelor/masters/Ph.D. level of knowledge. For Ph.D.'s it's pretty easy since we just use the defense/committee system we have now.

The problem is that it's not clear what an "MIT admissions" really means. Is it a reward for being a good student in high school, or is it a chance to learn new things? I think it's a bad idea to combine the two.

Maybe a better analogy is the Marines. If you have some very basic qualifications, then if you show up and want to join the Marines, they'll let you in. Why won't want to get into MIT? Well for the same reason not everyone wants to join the Marines. MIT is a cool place, but it can be gut wrenching, painful, and agonizing. You will be pounded with a ton of facts and information, it is painful, brutal, and a ton of work.

In an "open admissions" MIT, you'd have *anti-admissions* counselors. People that go to high schools telling people how tough the place is to scare people into not going. Now after telling you how painful, stressful, scary, and tough the place can be, you still want to go, then you're in.

 

[twofish-quant]

Or you can have MIT stop issuing diplomas at all. Personally, it might be better if we don't have grades at all. If you want to take freshman physics, great! You'll get no grade, no record of your attending the class, and absolutely nothing on your transcript.

One idea is to make MIT a pure research center. If you want to learn freshman physics at MIT because it's a great place to learn physics, just show up to class. The catch is that there are no grades and no transcript, so if your only goal is to get a piece of paper, that will get you another piece of paper that will get you some money, there's no point in going.

Personally in most intro astronomy classes at state schools that I've taught, about 90% of the people aren't really interested in studying the material. They are just going through the motions in order to get a credit. You end up with a much better educational experience if you get rid of the 90 students that just want the grade, and just teach the 10 that are really interested in learning something.

 

[twofish-quant]

Also MIT and Harvard are for-profit institutions. MIT has $10 billion and Harvard has $35 billion. The business model of MIT and Harvard is to get smart people admitted, make their alumni rich and powerful, and they collect the cash each year when some freshman calls you up for donations. You *can* buy your way into either MIT or Harvard, but it requires more money than anyone reading this will ever see. (I'd say $5 million or so in cash will do it).

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.

 

[doinok]

Thanks for the replies. I appreciate it.

 

[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...

 

[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.

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.

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.

 

[vanesch]

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.

 

[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.

 

[atyy]

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.

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

 

[AsianSensationK]

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

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.

 

[atyy]

What makes these hot shot schools worthy is that they only admit a small number ("the best")

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.

 

[doinok]

A couple of essays that seem apropos to the discussion:

http://www.paulgraham.com/colleges.html

http://philip.greenspun.com/teaching/universities-and-economic-growth

 

[twofish-quant]

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

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.

 

[twofish-quant]

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).

I don't think we have a choice. The thing about branding is that brand perception can change in an instant. MIT like any other bureaucratic institution will refuse to change if it can survive without changing. But I don't really think it has a choice. The latest financial crisis are creating *huge* financial problems for the prestige universities.

In 1950, if you wanted to learn freshman physics, you basically *had* to go to MIT. Today, you can download the entire class on the internet. That radically changes the world, and MIT has to figure out how to change with it.

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.

University of Phoenix.

 

[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?

I worked for about five years as a software developer, my Ph.D. involved huge amounts of programming, then there is programming I did at MIT, then you can go back to when my father brought home the TRS-80 Model I that I used when I was six. Also, it's just not jobs, but I spend a lot of my free time programming.

I do assume the types of projects you do are going to be dramatically different, but I'm just a little curious.

It's not that different.

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.

It's really not. No employer that I know of will hire someone because of their resume or cover letter. The problem is that the resume and cover letter is just something that you rapidly look through to eliminate people that have no chance of getting the job. You take the stack, arrange them in an order, and then you call up the top ten candidates. Then you send in the top five for several rounds of face to face interviews.

You cannot find out how someone thinks from a piece of paper, and it's easy to create a good looking resume even if you know nothing. The resume is used to screen out candidates that obvious won't get the job, and then if you want to see how someone thinks, you give them math problems.

 

[twofish-quant]

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.

There's really very little information conveyed in a brand. A lot of emotion, but very little information. Coca-Cola is a powerful, powerful brand. It's just fizzy water in bottles. But Coca-Cola tastes *great!* Hmmmm... Could it be that it tastes great because you see so many commercials convincing you that it tastes great.

Also this is really why you are in a better shape if you take some marketing courses. Graduate schools are trying to sell you something. *PROFESSOR* is a brand.

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.

And the money that you get from donations go into glossy brochures that convince everybody else that the brand is cool. However, the trouble is that to maintain a brand, you need large amounts of cash, and most of the major elite universities got hit really, really hard by the financial crisis. Harvard lost $5 *billlion* and MIT lost something like $2 *billion*. It's much worse, because they were counting on 15% growth for money. If you end up with 3-5% growth, everything falls apart.

My argument is that in the next three or four years, it will become obvious that MIT has to do something really radical to survive.

One problem is that if you think of education as an industry, there is only room for about three or four prestige brands. The *BIG* money is where the University of Phoenix is at. Think Walmart.

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.

The last part is where they are going to have problems... I went to MIT. I'm smart. I'm successful. Why should I throw money at MIT when in fact I'll do much more social good if I throw money at a community college or raise money to improve inner city high schools? If MIT is going to focus on executive MBA programs and buying up Cambridge real estate, then that's great, but why should my money go to that.

Now if MIT were spending my money improving science teaching at in poor neighborhoods in Boston, I can go for that, but then the system falls apart when you have large numbers of people that now have decent skills, but can't get into MIT because their backgrounds aren't perfect.

 

[twofish-quant]

So if in fact "the best" is not true

The problem here is "best" is subjective. OK you are a brilliant research mathematician but you stink at teaching algebra. Who made up the rule that you are "better"? (Any my answer, is that in the Cold War both the US and the Russians needed relatively small numbers of physicists and mathematicians to work on weapons systems, at so research got defined as "better.")

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.

That's fine but then you run into a problem in that it's weird that people that keep on saying that there is no "best" usually end up with money, power, prestige. If you aren't "better" than me, then why are you telling me what to do?

You are in a great position of power if you can define what "best" is, and people that get to define what the "best" is usually define it so that the "best" people are people that look, think, and act like they do. Actions speak louder than words, and you can usually figure out what someone really believes not by what they say, but how they act.

 

[twofish-quant]

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

It's easy to be an elitist, if you think you are going to be a member of the elite. I'm not an elitist because it turns out that its very likely that I'm *not* going to be in the elite. I'm a loser and a failure.

One problem with getting into the "inner circle" is that once you get in, you find that there is an "inner inner circle". Once you get into the "inner inner circle" you find that there is an "inner, inner, inner circle"

At some point you just throw your hands up and just want to blow the system up.

 

[AsianSensationK]

There's really very little information conveyed in a brand. A lot of emotion, but very little information. Coca-Cola is a powerful, powerful brand. It's just fizzy water in bottles. But Coca-Cola tastes *great!* Hmmmm... Could it be that it tastes great because you see so many commercials convincing you that it tastes great

Actually, yes, this is the best way to phrase it. A brand is most like an emotional experience, or an attitude. There isn't a lot of relevant stuff conveyed in a brand name. But they're powerful things because of how people think.

I have a quote in my marketing book from the CEO of McDonald's basically saying, "If we lost every single asset on our books today through some freak disaster, we would still make tons of money because of our brand. We could borrow what we need, and continue business with relatively little trouble. Heck, we could give you every single asset of ours, keep our brand, and still make more money than you this next quarter."

And the money that you get from donations go into glossy brochures that convince everybody else that the brand is cool. However, the trouble is that to maintain a brand, you need large amounts of cash, and most of the major elite universities got hit really, really hard by the financial crisis. Harvard lost $5 *billlion* and MIT lost something like $2 *billion*. It's much worse, because they were counting on 15% growth for money. If you end up with 3-5% growth, everything falls apart.

My argument is that in the next three or four years, it will become obvious that MIT has to do something really radical to survive.

I believe it. They have to make more money somewhere in order to meet their budgets.

 

[Edgardo]

Hello twofish-quant,

some of my fellow students work at banks which got me interested. Does your job involve "financial mathematics" as described on this website http://guava.physics.uiuc.edu/~nigel/finance.html" (for example option pricing)?

Do you use any of the books cited on the website?
And have you read Derman's "My life as a quant"?

You mentioned your C++ skills. Is there anything else that can make me interesting for banks?

Thanks in advance.

 

[twofish-quant]

some of my fellow students work at banks which got me interested. Does your job involve "financial mathematics" as described on this website http://guava.physics.uiuc.edu/~nigel/finance.html" (for example option pricing)?

More or less. The trouble with that website is that it's a bit outdated. The good/bad news is that there are about a dozen different types of jobs for physicists on Wall Street. Something interesting about mathematical finance is that it's a tiny, tiny part of finance, but finance is so large that it keeps physicists quite strongly employed.

Do you use any of the books cited on the website?

Sort of, but again the website is out of date. Curiously it doesn't contain one of the books which I think is essential reading which is Kuznetsov's "The Complete Guide to Capital Markets" and Fusai and Roncoroni's Implementing Models in Quantitative Finance: Methods and Cases. Also you should look at Springer-Verlag and Wiley Finance since they will get you an idea of the type of problems that you need.

And have you read Derman's "My life as a quant"?

Yes. Interesting book, but again, somewhat out of date. It's a very, very useful book if you look at it was history, but don't take it as a accurate representation of what Wall Street is like right now. One of the larger differences is that these are much. much more heavily computational than they were in the late-1990's. This is bad if you are a string theorist. Good if you are doing computational fluid dynamics.

Just to give one example of how Derman's book is a bit old. No one right now is that interested in modeling complex interest rate derivatives. People are interested right now in modeling negative and zero interest rates and counterparty default. You won't find a book that gives the standard method for doing that, because if there were a standard method for doing it, they wouldn't need you to figure out how to do it. Next year, they'll be interested in something very different.

You mentioned your C++ skills. Is there anything else that can make me interesting for banks?

One big thing is to be a physicist and focus on writing a quality dissertation. There's nothing wrong with being a finance or economics Ph.D. except that if your Ph.D. is in physics, you are going find it much easier to to be a physicist with basic finance knowledge rather than an expert in finance. If you try to turn into an MBA or finance expert, then you'll probably lose the skills that make you interesting to banks and hedge funds.

One other curious thing is that you will be seriously, seriously miserable if you go into finance in order to make lots of money. It's really weird, because once you are in finance, you will make money that most people outside of finance consider to be huge, but because you are constantly meeting people that make even more money, so you end up feeling quite poor. Finance is also a horrible place to be if you like to be on top of the heap, for the same reasons. Also you'll probably end up living in the NYC area, which is both good and bad.

The one thing that would be useful is to read some standard finance texts. There is a huge amount of jargon. It's not that difficult to pick up, and it helps if you know it.

The other thing is to try to *think* about the "big questions." Are there too many investment bankers? Is what you are doing really economically useful? Personally, I think finance is economically useful, and there aren't too many investment bankers, but you have to think about this for yourself. The basic issue is that if you aren't really generating economic value, then you are just part of a bubble, and if you are part of a bubble, you are likely to get smashed.

The reason that you need to think about this for yourself, is that I really don't want to take the moral responsibility for encouraging you to go into a career that is going to blow up in three years. On the other hand, I don't want to the talk the moral responsibility for discouraging you from going into something that is going to boom. So I'll just say that its something that you have to think about.

 

[peppies]

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.

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.

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.

Twofish-Quant, I just happened to read your responses and wanted to ask for some advice. I was actually looking into these kinds of Wall Street jobs and got shot down since essentially all of them require MS and Ph.D and I currently have a BS in Physics and minor in Mathematics.

The only thing that is preventing me from getting an MS or Ph.D in physics is the money. If it wasn't for that and $100k already in student debt, I would be comfortably studying for an MS and Ph.D.

To me, it sounds like you are saying that Ph.D is mostly used as an indicator of proof that someone can work independently for a long long time without giving up easily and having the ability to adapt and learn new things (much like my time studying the physics major).

I have much of these qualities and have the determination to self-teach myself new things. In fact, I am moving towards a different career path as an Actuary (while still looking for physics related jobs.) So I had to go out, buy Actuary texts, study them and take special exams for this field. I've never taken any finance or Actuary classes in college, yet I am able to improvise. The benefit here is I can actually afford these books rather than a Ph.D. On top of that, I've self-taught myself in computer skills and programming to a point good enough to build and run an online business over the past 5 years (more proof of teaching myself new skills to make money.)

Now what am I suppose to do in order to get one of these Wall Street jobs by getting around the MS and Ph.D wall that is pretty much impossible for me to do because of a lack of money? To me it sounds like I can fit right in with that Ph.D mindset and research determination to do tough long jobs, plus I have some intelligence, but the only thing missing is the actual Ph.D or MS title. Any advice?

Thanks

 

[twofish-quant]

Twofish-Quant, I just happened to read your responses and wanted to ask for some advice. I was actually looking into these kinds of Wall Street jobs and got shot down since essentially all of them require MS and Ph.D and I currently have a BS in Physics and minor in Mathematics.

Most jobs on Wall Street require some sort of masters degree. There's very little available for people with only a bachelors. Most people in Wall Street don't have Ph.D.'s and in quantitative finance, it tends to be 50/50 with half the people having Ph.D.'s and the other half having a Masters in computer science or something similar + work experience.

There is a market for physics bachelors in management consulting, but that's a different market.

The only thing that is preventing me from getting an MS or Ph.D in physics is the money. If it wasn't for that and $100k already in student debt, I would be comfortably studying for an MS and Ph.D.

In the case of physics Ph.D. programs, money isn't a huge factor since typically the school pays you, and you can defer payment (and depending on the loan possibly get subsidized interest) while you get getting the degree.

To me, it sounds like you are saying that Ph.D is mostly used as an indicator of proof that someone can work independently for a long long time without giving up easily and having the ability to adapt and learn new things (much like my time studying the physics major).

It's proof that you can work semi-independently on a specific type of problem.

Now what am I suppose to do in order to get one of these Wall Street jobs by getting around the MS and Ph.D wall that is pretty much impossible for me to do because of a lack of money? To me it sounds like I can fit right in with that Ph.D mindset and research determination to do tough long jobs, plus I have some intelligence, but the only thing missing is the actual Ph.D or MS title. Any advice?

First of all, the title is important, because anyone can say that they can do X, but the employer is looking for proof that they can do X. Now there are some alternate ways of getting that proof (enlisting in the military), but there's a chicken/egg quality to them (i.e. you need a job to get experience, but you need experience to get a job.) The other thing to do is to get a job (any job) pay down the loans, get a masters, and then go off to Wall Street (assuming of course that it's still there in some recognizable form in about five years.) Now is a lousy time to move into finance. Things may be very different in two years.

One other thing to remember is that especially when an employer first looks at you. You are a cog in the machine. You are not a human being, but rather one of ten people looking for a position. Yes, you may very will be able to do the same quality work as someone else, but why should the employer take the risk?

The other thing is that I really don't think that money should deter you from getting a Ph.D. because the stipends are paid for by the school.