From a no name university, what should I do to get interviewed?

[OKCWBY]

After reading this forum for a while I figure out I should register and ask you guys for advice.

Right now I am in a no name public school (80-100 ranking), getting close to graduate with a Ph.D in physics. I feel really frustrated whenever I start thinking about my career. There is about zero chance for me to get a faculty position in this country with my current background. Even though I constantly out performed my peers, no one will care about that. So I figured out that I may have to change my career. But as much as I know about quant jobs now, I start feel the same. Wall Street only wants people from top universities. I may never get interviewed. But seriously I really think I can do as good as ivy league students in this field. I have solid understanding in derivative pricing. I can pass 3 actuarial exams with less than 4 months studying time (doing research at the same time). I have 5 years programming experience with Fortran 90. I know C++ and data structure. Basically I think I can handle most part of the interview questions, including the brain teaser. But again, who cares about a student from a no name school. Please don't ask me why didn't I go to another school for Ph.D or those kind of questions.

All these being said, here comes my question. What can I do in order to improve my stock and possibly get an interview?

 

[Andy Nguyen]

Step 1: identify a target
Step 2: Attack
Step 3: Profit!

1. What is your target? What kind of employer? What kind of position? What are the requirements?
2. Did you try anything? Did you talk to recruiter? Have you tried to network with people who may have a lead? How many resume have you sent out?

 

[Connor]

Also consider jobs not on Wall Street, other banks, risk management, etc

 

[DominiConnor]

It's not quite exactly true that Wall Street only want people from the Ivy Leagues, they want smart people, and that's a convenient place to look, they pay me to find the ones they miss, and having done AI, I apply rather smarter search algorithms, and your goal is to show up on the radar of people like me...

The demand for F90 is pretty small, but I'm assuming that this was used for PDEs and/or other numerical methods ?
That's pretty useful, and it's a lot harder to get that right than to learn the different syntax used in C++ as opposed to F90, some people's C++ looks remarkably like Fortran.

One option is to do the CQF or an MFE, lots to choose from, but not exactly cheap, one tip: if you do the CQF apply when you're still a student and you get a 35% reduction.

But I can't imagine you'd survive an interview at a top bank with what you currently know.
Data structure level C++ is old fashioned (I first did C before you were born, when I call it old fashioned that's really old.)
You really ought to be aiming at getting to the level of the Sutter C++ books, and also learning some more non-numeric algorithms. Can you write an exception safe constructor or a fast multi threaded singleton ?
What does
42;

Do in C++ ?
Why is it valid C++ ?
Why would you do that anyway ?
(I got asked that myself at Merrill Lynch)

I'm also at least a bit sceptical of your ability to do brainteasers to an excellent level, that's quite a call to make, very few people are there at your stage in life.

When you say "solid understanding" of pricing, what exactly do you mean, and how would someone in a physics department have a valid metric anyway ?
For instance, a bank asked one of my candidates with a background not unlike yours how he'd price a volatility swap. This candidate have zero formal education in quant finance, and had read himself to the right level, which is quite tough, but as he demonstrated entirely possible.
Could you do that ?

Why is there a N-1 term in standard deviation ?
In general what is your level in stats ?

Do you know how to determine if a time series exhibits mean reversion, how would you test for that and then for regime change ?

What intuition do you have for the fact that price movements often seem to follow a lognormal distribution ?
What other distributions might you use ? Why are they good or bad ?
Do you know how to write code to generate any of them ?

A cruel question is to hand you a standard calculator and ask you to calculate the volatility of a time series by hand. It's bloody sad how many people screw up that one.

I haven't even begun picking over your PhD yet, you'd probably enjoy that even less than what I've fired at you.

Your resume is also pretty shit as well.
I can tell that even though I've not read it, I'm that smart

Actually, I'm not that smart, I know that many people like you have shit CVs, and if you aren't getting any interviews at all, there's probably something wrong with it.

Andy makes a very valid point about the number of shots you've taken, I strongly suspect you've sent a mediocre CV to a few banks. Before you commit yourself to despair, you should at least send a good CV to a lot of banks.

None of the things I've mentioned are beyond a few months of intense self study, and since you are shortly to become unemployed, you will have time to do it in.

 

[TraderJoe]

Many of the Physics PhDs are able to get quant jobs after two months self-study of MFE course materials. A few of them actually do the MFE degree.
Even if you have a PhD degree there is surely no harm in being humble. After you have done a few quant interviews, maybe you may not be so confident about your knowledge and your intelligence level. The interviews are quite tough and you have to pass 7 - 8 interviews on various topics to get the quant job. Even if one person says you are not fit, that is the end of the job application for that company.
Preparation is the key and don't be overconfident.

 

[Andy D.]

What does
42;

Do in C++ ?

The meaning of life (sorry couldn't resist the Hitchhikers pun there).

Why is there a N-1 term in standard deviation ?

Having just studied variance in my undergrad Stats course, I presume this is a case of population versus a sample?

if you do the CQF apply when you're still a student and you get a 35% reduction.

Does this apply to undergrad students as well?

Thanks

 

[Steven]

The meaning of life (sorry couldn't resist the Hitchhikers pun there).

Having just studied variance in my undergrad Stats course, I presume this is a case of population versus a sample?



Does this apply to undergrad students as well?

Thanks

The N-1 term appears in variance instead of sd? N-1 is used instead of N because N-1 will make the sample variance unbiased.

 

[Andy D.]

The N-1 term appears in variance instead of sd? N-1 is used instead of N because N-1 will make the sample variance unbiased.

I figured that was what Dominic was probably eluding to in his post. Looks like I did remember something from the syllabus after all .

 

[Joy Pathak]

If by "wall street" you mean't finance in general then keep applying. IF you mean't only new york, BB, then apply to all banks and boutiques.

Apply to the top MFE programs where strong career services are provided. If you find schools where they can get you internships right when you start or a few months into you starting the program, target those schools. I know at my program, a couple students got part-time internships a month or two into the program through the career services provided. With an MFE program on your resume apply to several jobs in your first or second semester. This will get you interviews and possibly a job while you're in the program. Switch to part-time and finish the program while you work. When you finish you will probably get a raise too.

This would be the ideal scenario. Your backup is getting an internship in summer and then a job after graduation from the program.