Transitioning into Quant Roles from Enterprise Risk

[lax_milgram]

Hi All.

To give a bit of background - I have a bachelor's degree in engineering and I recently completed a master's degree in computational fluid dynamics where I wrote a thesis on designing and optimizing new numerical non-linear PDE discretizations (energy/entropy-stable methods). During my master's I coded mostly in MATLAB, FORTRAN, bash, and some C++.

After I graduated I obtained a position in enterprise risk for a large bank in Toronto, Canada doing model implementation for regulatory stress-testing credit risk models i.e. estimating PD, LGD, EAD (been here for about 10 months). This doesn't intersect with the front-office at all and is primarily about the bank's loan/credit portfolios - I spend most of my day doing model testing, UAT, coding in SAS and bash, ad-hoc stress-testing runs with our platform, investigating bugs, looking through the models and communicating with model development. Most of the models are statistical models: linear/nonlinear regression, Markov chains, Gaussian copulas, hazard models etc.

I'd like to transition to a front-office quant role in capital markets or market risk at one of the large Canadian banks (my current bank for example) doing model development. In my current role I found I'm more interested in something which is a bit more mathematical and related to my Master's research. I've already found some groups in my bank that are doing exactly what I want to do i.e. developing models for pricing exotics, IR derivatives, XVA etc. or working with these groups to develop risk models. However, I realize that I'm lacking some skills (mostly in stochastic calculus) so have the following tentative plan for the next year:

1. Refresh linear algebra (since I've forgotten a lot since undergrad and only used basic numerical linear algebra during my master's research)
2. Refresh probability: Introduction to Probability by Blitzstein and Hwang
3. Options, Futures, and Other Derivatives by Hull
4. Financial Calculus by Baxter and Rennie
5. Stochastic Calculus for Finance II by Shreve
6. Code up some derivative pricing models in C++/Python

I was wondering if anyone any suggestions about how to make myself competitive for these types of roles or any other things I should be working on? I don't really have any interest in getting an MFE degree since they're extremely expensive and I feel confident that I can learn most of these things on my own. If anyone currently works in Toronto in a quant role I'd be happy to hear from you.

Thanks.

 

[KillingField]

Instead of asking random, anonymous people identifying or aspiring to be quants (whatever their definition of the term may be), I would encourage you to network internally. You're going to get much better colour on your locale, and a better understanding of what the teams you wish to join expect of an entry level candidate. The sooner you express your interest and start making moves, the better: It is easy to get pigeonholed.

While having the basics nailed down with something like Shreve is beneficial, knowing and understanding models actually used by practitioners is even more so. You may already have or be able to gain access to the internal model documentation or if not the full thing in detail, it is likely some presentations have been circulated to a wider group. They may be written well enough and be detailed enough that you can learn out of them, and maybe even implement some of the models yourself. Network, and find out.

 

[lax_milgram]

Instead of asking random, anonymous people identifying or aspiring to be quants (whatever their definition of the term may be), I would encourage you to network internally. You're going to get much better colour on your locale, and a better understanding of what the teams you wish to join expect of an entry level candidate. The sooner you express your interest and start making moves, the better: It is easy to get pigeonholed.

While having the basics nailed down with something like Shreve is beneficial, knowing and understanding models actually used by practitioners is even more so. You may already have or be able to gain access to the internal model documentation or if not the full thing in detail, it is likely some presentations have been circulated to a wider group. They may be written well enough and be detailed enough that you can learn out of them, and maybe even implement some of the models yourself. Network, and find out.

Good advice, thanks. I think I will plan to network internally after I build a bit of a foundational knowledge in the subject (as given by my study plan). Unfortunately, I don't think I have access to our internal financial engineering libraries - that would be a great resource to have. I know that they are primarily developed in C# server applications. I'd imagine it's probably something quite similar to quantlib. The desk quants happen to be on the capital markets side of the bank which is a different branch from where I sit, which is on the enterprise risk side.

 

[Squasher]

I am an undergrad in math and stats
I am going to start my MS applied math degree in fall 2021
which quant roles use concepts like finite difference method, ode, numerical methods ? what other skills are needed for such roles?

 

[seamusfinnegan]

Instead of asking random, anonymous people identifying or aspiring to be quants (whatever their definition of the term may be), I would encourage you to network internally. You're going to get much better colour on your locale, and a better understanding of what the teams you wish to join expect of an entry level candidate. The sooner you express your interest and start making moves, the better: It is easy to get pigeonholed.

While having the basics nailed down with something like Shreve is beneficial, knowing and understanding models actually used by practitioners is even more so. You may already have or be able to gain access to the internal model documentation or if not the full thing in detail, it is likely some presentations have been circulated to a wider group. They may be written well enough and be detailed enough that you can learn out of them, and maybe even implement some of the models yourself. Network, and find out.

What causes the pigeonholing exactly?

 

[KillingField]

I am an undergrad in math and stats
I am going to start my MS applied math degree in fall 2021
which quant roles use concepts like finite difference method, ode, numerical methods ? what other skills are needed for such roles?

You're putting the cart before the horse, and rather than choosing a job based on some tools you like, I suggest you learn the tools for a job you find interesting. Now that said, PDE etc methods are mostly used in derivatives pricing at banks. You should also know the basics of financial mathematics for these roles (changes of measure etc).

What causes the pigeonholing exactly?

This is not quant specific, nor is it specific to finance. Recruiting processes are standardized at the entry level, and often these will leverage university pipelines or internships - something not really available for experienced hires. It's tough to move to a new type of position at a more senior level if you have to be trained up almost like a fresh graduate yet be paid like someone who knows what they're doing (it's not very common to take a "demotion" in corporate title, and sometimes this is prohibited by internal mobility policies): Not only will you not be immediately making a positive contribution in your new role, but the previous team will have lost someone experienced and proficient (the latter concern is only relevant when moving internally and can cause some drama in the form of office politics). This is not to say that these moves don't occur, but they usually don't happen by accident and require a fair bit of proactivity.

 

[seamusfinnegan]

You're putting the cart before the horse, and rather than choosing a job based on some tools you like, I suggest you learn the tools for a job you find interesting. Now that said, PDE etc methods are mostly used in derivatives pricing at banks. You should also know the basics of financial mathematics for these roles (changes of measure etc).


This is not quant specific, nor is it specific to finance. Recruiting processes are standardized at the entry level, and often these will leverage university pipelines or internships - something not really available for experienced hires. It's tough to move to a new type of position at a more senior level if you have to be trained up almost like a fresh graduate yet be paid like someone who knows what they're doing (it's not very common to take a "demotion" in corporate title, and sometimes this is prohibited by internal mobility policies): Not only will you not be immediately making a positive contribution in your new role, but the previous team will have lost someone experienced and proficient (the latter concern is only relevant when moving internally and can cause some drama in the form of office politics). This is not to say that these moves don't occur, but they usually don't happen by accident and require a fair bit of proactivity.

What sort of proactivity should you do? As far as I can see most companies want to provide essentially no training beyond the first few weeks of a grad programme maybe.

Also what are the internal mobility policies at the Analyst/Associate levels? (FO--> FO, MO --> FO) Which banks or other financial institutions are easier to move around than others?

 

[KillingField]

What sort of proactivity should you do? As far as I can see most companies want to provide essentially no training beyond the first few weeks of a grad programme maybe.

Also what are the internal mobility policies at the Analyst/Associate levels? (FO--> FO, MO --> FO) Which banks or other financial institutions are easier to move around than others?

Most companies definitely provide training on a continuous basis. It's not formal, course based education, but more akin to an apprenticeship. For example, the models used by practitioners are usually not published for general consumption, and in the rare cases they are, many of the accompanying key insights are not. The internal documentation may be shoddy, and there are no classes, yet as a quant you will learn the models through osmosis.

As for being proactive, you go and have a look around the landscape, see who to talk with, how to impress the relevant people and have the network so you'll hear about any potential openings through the grapevine. All circumstances are unique (and there's little point in general speculation of different firms etc), and so there's nothing more concrete to say, really, and that is the point: be proactive, figure it out. Go and hustle.