• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

So how much does a master's GPA matter?

Joined
6/6/08
Messages
1,194
Points
58
Okay, I'll make it quick...

My grades for my stats masters just came in and I got a 4.0. My question is...how much does this really matter, considering master's programs are one year? Will they just skip the 4.0 and punish me again for the undergrad GPA?
 
I would have to say the answer is ( almost surely ): it depends.

In one scenario, you could be competing with someone with a similar skill set for a job, and the hiring manager could use that GPA as the deciding factor. This was the case in one interview I recently had.

On the other hand, the hiring manager might give a damn. But since it is a 4.0, you lose nothing by putting it on your resume. Can't change the past, but you can demonstrate potential.
 
My 2 cents:

For the vast majority of jobs, nobody asked me what my GPA was. I also know a number of people who found jobs and didn't include their GPAs on their resume. That having been said, listing a 4.0 on your resume can probably only help. I'd stick it right next to your degree title, so it sticks out.
 
Did so. And Morgan Stanley rejected me a day after for a quant position.

This is ridiculous.
 
Just to make sure my sarcasm isn't completely busted, you don't actually mean that, do you?
 
Did you forget to bold and use red color big font for the GPA? That may help draw their attention away from your school name.
Rutgers is a respectable school, IMHO. Maybe state schoolers would do better on the street if we stopped sniping at each other's schools quite as much. There are fewer state schoolers on our trading floor than there are UPenn graduates, and all of our experiences were more similar to each others' than they were with those at UPenn or the private schools.

Rutger's basketball team is going to get totally crushed by the Illini this year, but it's a decent school, and I've met more competent programmers from Rutgers than I have from many Ivy Leagues. I can't imagine their stats program doesn't produce people qualified to work on the street.
 
Well, see the one thing about their stats program is this...

We can probably blow anyone out of the water when it comes to data analysis and R coding.

However, what we're not exposed to is C++.

So there's the trade-off there.

I've coded 4000 lines of Java in two weeks before and am sure I could do similar things with C++, but I've forgotten much of the syntax of the java that I coded. I have no issue understanding algorithms, and am familiar with big-O notation and all of that. It's just the exact syntax that gives me nightmares.

That said, I'm pretty confident I can out-stat a lot of people without PhDs.

The question is does Wall Street really value people who have more of a variables-and-data background rather than code-and-data-structures background.
 
The question is does Wall Street really value people who have more of a variables-and-data background rather than code-and-data-structures background.

This is a flawed question, but in a very interesting way. The good news (for you, at least) is that correcting the flaw leads to good news (for you).

You see, Wall Street is not a monoculture. I'm sure there are some IBs who value variables-and-data over code-and-data-structures, and I'm sure there are some IBs who value the latter over the former. Even within individual IBs there are wide variations between what certain teams are looking for. When my own team was hiring last year, we looked for slightly different skill sets for two separate hires, even though the two positions were technically the same. We wanted as broad a mix of skills across the team as possible.

Long story short: You can't take it personally. If you didn't get this job, it's not necessarily a reflection on you. It just means this job wasn't the right fit. I'm sure, somewhere out there, is the right position for you. As the late Henry Gibson sang, keep a-goin'.
 
Well see, it isn't so much that which I'm taking personally, but the fact that I haven't had a paid offer in around two years, and that paid offer was for an actuarial internship.

I realize that we just got bombed with the 80 year flood and all, and that I'm not out of MIT, but at the same time...I feel that if this doesn't do the trick, that what exactly will?
 
Maybe switching program/transferring to Baruch MFE will do the trick ;-) the admission is in progress... I believe you would learn a quant-must-have-skills including design and a hardcore C++ programming of financial models and derivatives... so writing a multithreaded pricing model for a Barrier option on multiple assets will be like a breeze... and passing the interview will be like a joke
 
Actually I'm applying to Stony Brook (and to Lehigh) for PhDs--if I can't get a job through the recruiters I work with.

The job is most preferred however.
 
Maybe switching program/transferring to Baruch MFE will do the trick ;-) the admission is in progress... I believe you would learn a quant-must-have-skills including design and a hardcore C++ programming of financial models and derivatives... so writing a multithreaded pricing model for a Barrier option on multiple assets will be like a breeze... and passing the interview will be like a joke

Lol is that sarcasm ? :-k
 
I agree with Adam. However, the other important factor is your presentation. From two years on this forum and meeting you personally, you may want to work on your presentation.
 
Maybe switching program/transferring to Baruch MFE will do the trick ;-) the admission is in progress... I believe you would learn a quant-must-have-skills including design and a hardcore C++ programming of financial models and derivatives... so writing a multithreaded pricing model for a Barrier option on multiple assets will be like a breeze... and passing the interview will be like a joke

First off, setting a whole bunch of requirements for a certain position may be painting with a bit of a broad brush. Quant development interviews usually don't get that complicated- at least the ones I've heard of- and I work with a number of quant developers at a major investment bank.

Second-off, though, the fact that you talk about multithreading as a major challenge indicates to me that you may be new to programming- more experienced programmers learn to abstract multithreading from the outset and don't spend much time thinking about it. Multithreading only starts to get tricky if you have to use the banker's algorithm to avoid deadlock rather than something simpler. Distributed pricing models? That's a bit tougher.

I realize that we just got bombed with the 80 year flood and all, and that I'm not out of MIT, but at the same time...I feel that if this doesn't do the trick, that what exactly will?
Good news is that it doesn't matter if you're out of MIT. Assuming you can land the interview, it all has to do with your competence and presentation.

You could benefit from toning it down a little. You can come off as a little self-promoting sometimes, which some interviewers may not like. Interviewing is sort of a balancing act- you have to demonstrate that you're a strong candidate by coming off as competent, but you have to demonstrate that you're a good fit for everyone by staying relatively modest during the process.
 
Heh...just had an interview with Analytic Partners in midtown the other day. Accidentally blew a question that you had a 4x4x4 cube and you removed one layer from each side. I said 4 brainlessly in the beginning, talked through it, and realized it was eight (palmfaced myself then and there) because I knew what I was doing and thinking in terms of 2s and 2s, but my mind just slammed on four by accident. Still got the right answer, but not before blurting out a rushed answer...honestly, I like the kinds of questions in the quantitative interview questions thread better that I don't feel like I'm supposed to nail it in 5 seconds, because the slightest tinge of nervousness can screw it up then and there.

Here's another question I got:

A company has a rental car for every 1000 people. How much mileage on one car can they expect in a year?

My answer: assuming that people take the car out for a work day's amount of time, and that the car is always in use, the car gets driven three round trips a day, and say, for ten miles one way to a destination. Since there's approximately 50 weeks in a year and the cars are driven five days a week (work days), each car will have approximately 60 x 5 x 50 = 15000 miles on it in a year.

Not sure how close I was. It was one of those "force of logic" questions ala how much would you charge to wash all the windows in Seattle sort of thing.

Interview other than that went well, I guess...got interviewed by a pair of Sr. Analysts and the software package there is SPSS.

Other than that, the world is still very, very quiet on my end.
 
Heh...just had an interview with Analytic Partners in midtown the other day. Accidentally blew a question that you had a 4x4x4 cube and you removed one layer from each side. I said 4 brainlessly in the beginning, talked through it, and realized it was eight (palmfaced myself then and there) because I knew what I was doing and thinking in terms of 2s and 2s, but my mind just slammed on four by accident. Still got the right answer, but not before blurting out a rushed answer...honestly, I like the kinds of questions in the quantitative interview questions thread better that I don't feel like I'm supposed to nail it in 5 seconds, because the slightest tinge of nervousness can screw it up then and there.

Here's another question I got:

A company has a rental car for every 1000 people. How much mileage on one car can they expect in a year?

My answer: assuming that people take the car out for a work day's amount of time, and that the car is always in use, the car gets driven three round trips a day, and say, for ten miles one way to a destination. Since there's approximately 50 weeks in a year and the cars are driven five days a week (work days), each car will have approximately 60 x 5 x 50 = 15000 miles on it in a year.
I don't think you did THAT badly. You corrected yourself before the interviewer told you you were wrong; in a researcher's, quant's, or financial programmer's book, that's still pretty acceptable. For a trader, it might not be. You don't have to be perfect, but you do have to be comfortable with numbers.

Just smile. One thing that I do in interviews is, "Just a sec, I need a minute to think about this." It takes some of the pressure off; you don't have to be lightning fast with the stuff you're not expected to know off the bat. I work on a fixed income analytics system and I can be expected to know the definitions (both mathematical summary and casual explanation) of the biggest risk measures without flinching, but if they ask me to price something or throw me a question from left field, I have every right to take my time and think- especially if I smile and comment that it's a really good (or tough) question.

One of the things that always happens in a technology- and often finance or engineering interview- are questions about "Have you seen this?" Two years ago, when I was interviewing for my first job, I would invariably say yes if I had remotely seen it. The problem is that those initial questions usually have something really tough lurking behind them. Now that I'm older and wiser, unless I know the subject like the back of my hand, my response is to smile and say, "'Seen' is a strong word if you're about to ask me a really tough question, but I'd like to take a crack at whatever you've got for me." If the interviewer's in a good mood, he'll laugh and give me the question. There won't be anywhere near as much pressure on me, and 75% of the time, I'll get the right answer, anyway.

Your resume is a marketing document, but the strategy in the interview is to downplay, downplay, downplay unless you really know something well. Manage the interviewer's expectations, and he'll come out of the interview thinking, "Wow. That kid didn't claim to know much about stochastic calculus, but he derived the black-scholes formula in five minutes." rather than, "This kid claimed to know everything, but he couldn't answer half my questions." This strategy tends to have a positive impact on most interviewers.

That said, the strategy doesn't work quite as well with interviews with traders (particularly the guys who focus on executing short-term trades- like brokers, market-makers, or scalpers). They cut through BS, ask pointed questions that they have already irrevocably deemed you incompetent of if you cannot answer before you even open your mouth, and want an answer fast. They are the epitome of the J-type personality on the Meyers-Briggs scale and show no mercy. If you can survive one of these kinds of interviews- and you'll know it when you're in one- without saying "I don't know" or leaving an awkward silence, you've probably done just fine.

Those interview books you bought for trading don't apply for analytics, banking, research, quant work, or technology (and the rules for the latter five are actually pretty similar). For the non-traders, just underpromise (at least in the interview), overdeliver, and if you have the opportunity, provide some tactful comic relief with a smile.
 
See the one thing about being a financial programmer that gets to me is that I don't stick on any one language for too long, since I use whatever language the class requires. Did some C++. Did some Java. Was able to code 4000 lines of Java in two weeks at one point last May. Now forgot most of the syntax as well.

I understand the whole idea behind algorithms and Big O notation and all of that stuff (I understand what pointers do, as well), and the only thing I haven't really covered is micro-managing the memory allocation.

The biggest issue I have is the syntax of the program. All the time, I think in terms of "what do I want to do" but not "how do I tell this to the computer"?

And the problem there is that either you know the syntax of that particular language, or you don't. For instance, Analytic Partners' use of SPSS really gave me a nagging feeling because I know that Google and the hedge funds use R as their statistical language (I've heard SAS as well).

Furthermore, another issue I have with this whole programming issue is this:

There are people out of MIT who have studied computer science and have 3.8 GPAs and the like. What the heck is someone who programmed on the side with their engineering majors going to do about holding a candle to that? At which point am I supposed to say "I didn't study computer science. I studied OR and statistics."?
 
Back
Top