Quantitative analyst internships: Do they exist?

[Metagame]

I've searched through the internet and only a few firms on Wall Street come up with these types of internships. Do they even exist in Toronto and how would I go about applying for one? What skills and knowledge do they typically look for in students? I also don't care if they are unpaid, as I'm looking more for the experience more than anything else.

[ramazorg]

They do exist but are typically not posted and generally found through connections. I have never heard of unpaid positions and a respected shop would typically never offer such a thing. As far as the required skill set, typically the knowledge of math/statistics as well as the ability to program is required.

[adehbone]

Had friends at UW and other schools. Typically that is the target audience. All the major banks, some funds and asset managers.

The main skills looked for is, SQL/Matlab/C#/top notch math skills. But the main thing is just incredible programming skills.

[ukilledanoob]

they exist .. i had a interview for a quant job at a big 5 for co-op

[Phoenix3434]

Interesting that you guys are saying "super programming" is important. Can anyone tell me what a typical day would look like of such a person? Mathematical simulations? Programming? Talking with ppl?

[ramazorg]

Probably doing some dirty work such as fixing Excel spreadsheets.

[Phoenix3434]

Probably doing some dirty work such as fixing Excel spreadsheets.

Then why did abehbone say this?

The main skills looked for is, SQL/Matlab/C#/top notch math skills. But the main thing is just incredible programming skills.

Or is it one of those professions where they get the best people, and relegate them to doing **** work?

[Metagame]

Had friends at UW and other schools. Typically that is the target audience. All the major banks, some funds and asset managers.

So, if you don't come from a target school, then you essentially have no chance? Also, does your major of study have to be "mathematical finance" or would math/physics/engineering/CS suffice?

The main skills looked for is, SQL/Matlab/C#/top notch math skills. But the main thing is just incredible programming skills.

What do you mean by "top notch math skills"? Solving PDE's or proving abstract theorems? Because both require very different skill sets and I know many who are skilled at the former (computations) but can't do proofs and vice-versa.

Also, what are the GPA requirements and could a very high GPA make up for coming from a non-target? How competitive are these internships and can you apply over the summer independently or do you need to apply through your school?

[resu]

^Why restrict yourself to Toronto? All the big banks will operate their east coast quant teams out of NYC. I think MS has some teams in Montreal though.

I've seen some Big 5 'quant' job postings and they don't seem like real quant positions IMHO. More like a glorified coder. If you want to be a quant, look outside of Canada!

You will need at least a masters degree (PhD much more likely) and the mathematical foundation to understand and test/implement topics from research papers. Most quant teams will also work with C++ and maybe some functional language.

If you're looking for a quant position with just an undergrad degree, well you won't find a real quant positon at a half-decent company.

[adehbone]

Probably doing some dirty work such as fixing Excel spreadsheets.

Hmm, lots of questions I will try my best to give some colour. Starting with this.

No sir, that is what I did back in the day as a short-term bonds desk intern. I worked in VBA/Excel and so on. The quant intern (for the rates/FX derivatives desk) was expected to build algorithms in C/C++/C# (choose one) and SQL/Matlab. Basically the main requirement from day 1 was to be a master of matlab. A good background in "R" also helps. The main focus is you will either be on a prop desk where you are looking at things like gamma trades or yield curves etc...Or on a flow desk where then you are building algorithms to deal with customer flow and so on. You will need much more than just excel.

Typical day...Just like any other programming intensive internship. You are expected to sit on your own, focus on doing coding for about 80% of the day, the other 20% will do with talking to the traders/team members about what they need and the results of your algorithms.

GPA wise, like any trading desk best to be in top 15 or 25% of your class. Top "notch math skills", at a minimum if you took the GRE you should be able to kill it and qualify to apply to Columbia. Solving PDE's would be very helpful, mastering Statistics courses as well. Operations research is always a popular minor too.

Very hard to come from a non-target. You will need to network hard and get someone to basically take a chance at you.

Most shops prefer Software/Computer/Electrical engineers first. CS majors as well. Stats and Act Sci good too. Mathemical finance is fine, but you need to have killed your math classes.

PS; resu is correct. Don't expect to be adding a ton of value beyond coding as undergrad intern. Most interns (friends I know) were told to go do their MMF after. Most desks will give very good recommendations to Uoft/Stanford/etc...

[resu]

^That is much better info than what I provided

It seems like the quantitative analyst role in Toronto is a bit different than the ones in NYC and elsewhere?

[Metagame]

^Why restrict yourself to Toronto? All the big banks will operate their east coast quant teams out of NYC. I think MS has some teams in Montreal though.

I've seen some Big 5 'quant' job postings and they don't seem like real quant positions IMHO. More like a glorified coder. If you want to be a quant, look outside of Canada!

I was under the impression that it would be impossible to get a job in the U.S, seeing how competitive it is and that preference will always be given to U.S citizens.

You will need at least a masters degree (PhD much more likely) and the mathematical foundation to understand and test/implement topics from research papers. Most quant teams will also work with C++ and maybe some functional language.

If you're looking for a quant position with just an undergrad degree, well you won't find a real quant positon at a half-decent company.

I was planning to do graduate studies in applied math/engineering (still deciding between SP or optimization). Out of curiosity, why is this so? I have read from mathematicians and physicists that work on quants from other websites (mainly quantnet and WSJ), and their experience is that the job surely did not require their advanced mathematical training and that it could be done efficiently by strong undergraduates. I just find it strange that in other areas of finance, PhD's are looked down upon and have worse prospects when compared to undergrads, but the same doesn't apply in quantitative finance?

[Metagame]

Typical day...Just like any other programming intensive internship. You are expected to sit on your own, focus on doing coding for about 80% of the day, the other 20% will do with talking to the traders/team members about what they need and the results of your algorithms.

So, it's essentially a glorified programming job like the above poster said?

Very hard to come from a non-target. You will need to network hard and get someone to basically take a chance at you.

So, would it be better if I just wait until graduate school and do the internship as a graduate student at a target? Also, what exactly are the target schools for quantitative finance? For example, in the U.S, I know Cambridge is a strong target for quantitative finance, but not sure for other areas of finance.

Most shops prefer Software/Computer/Electrical engineers first. CS majors as well. Stats and Act Sci good too. Mathemical finance is fine, but you need to have killed your math classes.

Interesting that you list computer engineering and not physics, even though the former isn't even close in terms of mathematical theory to the latter. Why do shops prefer computer engineers and are physics majors completely ignored?

PS; resu is correct. Don't expect to be adding a ton of value beyond coding as undergrad intern. Most interns (friends I know) were told to go do their MMF after. Most desks will give very good recommendations to Uoft/Stanford/etc...

I thought so, which is why I have no problem doing an unpaid internship. I know I would be next to useless working alongside PhD's that were well established in their field before switching. Anyways, thanks for that very helpful and informative post.

[GreedIsGood]

For equities, get your degree then definitely go south for hardcore quant. The diversity, size and liquidity of that market facilitates quantitative finance.

Here's how it is in Canada for equities.

I've worked both sell side (before) and buy side (now) quant in Canada (equities). In Canada and for equities specifically, quant is light weight compared to the US. You need to understand investment principles and what works in Canada (which is pretty basic) more so than complicated math. This comes from experience, not school. Coming out of school everybody wants technically smart people, hence the need for top 15%, technical degrees etc. But it doesn't actually mean you'll be using any of that stuff. Coding skills are a must though. In fact, it's critical to be good and fast with this. You'll be constantly backtesting ideas many of which are adhoc where your "engine" won't accomodate. So you need to be able customize, fast.

Quant is probably THE WORST field when it comes to academics flexing their muscle with extremely interesting yet totally useless trading/investment ideas (for Canada that is). This stuff is good for impressing the innocent but that's about it. If you are on the sell side you probably make a living spamming peoples inboxes with these types of studies. Problem is, most of these studies are S&P 1500/Russell based and not representative of what would happen in Canada where there are essentially 3 sectors and only about 75 names are liquid enough to trade without moving the market.

Basically what I'm trying to say is that the nature of the stocks traded on the TSX don't allow for (need?) that level of sophistication. Hence why it is harder to find internships here. There are just so few firms that specialize in quant investing. Heck, even full time gigs are hard. If there was a market for this, I think the big US quant shops would already have teams working on it. As it stands, not too many US firms think it's worth their while to "quant" Canada. Probably a good thing for me cuz I'd hate to be on the other end of their trades given our horsepower here.

Education wise, don't split hairs between Math/CS/Engineering/Physics, just kill whatever non artsy degree you choose. But masters at most. Don't PhD it, it's really not necessary. With PhD wou'll feel compelled to employ 30 dynamic factors and Brownian Motion on a neural network when in fact my Excel 3 factor model using grade 10 math gets the same results. My previous bosses had degrees in math, physics, and even economics.

[victorwooten]

For equities, get your degree then definitely go south for hardcore quant. The diversity, size and liquidity of that market facilitates quantitative finance.

Here's how it is in Canada for equities.

I've worked both sell side (before) and buy side (now) quant in Canada (equities). In Canada and for equities specifically, quant is light weight compared to the US. You need to understand investment principles and what works in Canada (which is pretty basic) more so than complicated math. This comes from experience, not school. Coming out of school everybody wants technically smart people, hence the need for top 15%, technical degrees etc. But it doesn't actually mean you'll be using any of that stuff. Coding skills are a must though. In fact, it's critical to be good and fast with this. You'll be constantly backtesting ideas many of which are adhoc where your "engine" won't accomodate. So you need to be able customize, fast.

Quant is probably THE WORST field when it comes to academics flexing their muscle with extremely interesting yet totally useless trading/investment ideas (for Canada that is). This stuff is good for impressing the innocent but that's about it. If you are on the sell side you probably make a living spamming peoples inboxes with these types of studies. Problem is, most of these studies are S&P 1500/Russell based and not representative of what would happen in Canada where there are essentially 3 sectors and only about 75 names are liquid enough to trade without moving the market.

Basically what I'm trying to say is that the nature of the stocks traded on the TSX don't allow for (need?) that level of sophistication. Hence why it is harder to find internships here. There are just so few firms that specialize in quant investing. Heck, even full time gigs are hard. If there was a market for this, I think the big US quant shops would already have teams working on it. As it stands, not too many US firms think it's worth their while to "quant" Canada. Probably a good thing for me cuz I'd hate to be on the other end of their trades given our horsepower here.

Education wise, don't split hairs between Math/CS/Engineering/Physics, just kill whatever non artsy degree you choose. But masters at most. Don't PhD it, it's really not necessary. With PhD wou'll feel compelled to employ 30 dynamic factors and Brownian Motion on a neural network when in fact my Excel 3 factor model using grade 10 math gets the same results. My previous bosses had degrees in math, physics, and even economics.

Or you can get a PhD, become a god, and then work at Renaissance Technologies. Also, in that order.

Helpful career advice on rfd wut? Keep the advice coming guys! It's a field I'm considering but I have to do far more research