• 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!

From Academy to Quant research

The "elephant in the room" is the issue of how to write maintainable Python programs. No one talks about it... Sapir-Whorf strikes again.
I'm talking about programming as a discipline and not ad-hoc "coding" as it is sometimes called.

Maybe people just accept undocumented unmaintainable code??

BTW you mentioned Optiver a few times They are active in Meetup C++ here in Amsterdam and the meetings are all about C++>

Just asking. Now falls a wall of silence :giggle:
 
Last edited:
Thanks!

I particularly like to find patterns and backtesting strategies. I don't think that this kind of job requires more speed than academic research. Which role can I have in the industry?

Of course, I can also code a strategy after a succesful backtest.
 
While on the 'bathwater' analogy ...
In ML PDE/FDM with C++ is often used in the training for several reasons and researchers are working on new methods such as RKHS and nd PDE for ML algos.
Here is a recent thesis with interesting conclusions.



This whole area is in its infancy.
 
Last edited:
The "elephant in the room" is the issue of how to write maintainable Python programs. No one talks about it... Sapir-Whorf strikes again.
I'm talking about programming as a discipline and not ad-hoc "coding" as it is sometimes called.

Maybe people just accept undocumented unmaintainable code??

BTW you mentioned Optiver a few times They are active in Meetup C++ here in Amsterdam and the meetings are all about C++>

Just asking. Now falls a wall of silence :giggle:
C++ is mainly used on the execution side as well as for pricing libraries - everything extremely latency sensitive. So naturally you'll find high-frequency trading firms active in this space, hosting events, ... Python is also used in large scale in production at these companies, often not for trading systems but in the analysis stack which is what quants mostly interact with. These are developer-built and maintained production systems. Python has developed a really good support for type hints and you typically find lots of unit and integration tests. Optiver also hosted Python meetups in Amsterdam in the past by the way. Matter of fact is that in order to get a quant job at a prop. trading firm these days, C++ is often not a key aspect that is assessed in the interviews and PDE pricing even less. General programming aptitude is.
 
Thanks!

I particularly like to find patterns and backtesting strategies. I don't think that this kind of job requires more speed than academic research. Which role can I have in the industry?

Of course, I can also code a strategy after a succesful backtest.
Alpha strategy development and backtesting is typically done by quantitative researchers. The question is a bit what you bring to the table compared to candidates with a stats, econometrics or machine learning background when it comes to this role. Graduates from these disciplines will be your main competition.
 
My main problem is that not every hedge fund has a "quant research" position. I assume that this job is done by traders instead.
Also, I am not good enough to pass the quant tests of companies like Jane Street.

That's why I am asking for smaller (or less selective) hedge funds.
 
Today I received the rejection from Dare - VCMT. I guess the reason is my score in their game-based test. Again, if the aim of that test is to select people who are happy to spend 6 hours at day monitoring market prices I can fully understand. This will not be my business. But... is there anyone who hires graduates based on problem solving skills and academic background? Problem solving and useful skills are totally different from the tests I have done so far, Imao.

Is it possible that all trading firms use exactly the same criteria to hire graduates?
 
Today I received the rejection from Dare - VCMT. I guess the reason is my score in their game-based test. Again, if the aim of that test is to select people who are happy to spend 6 hours at day monitoring market prices I can fully understand. This will not be my business. But... is there anyone who hires graduates based on problem solving skills and academic background? Problem solving and useful skills are totally different from the tests I have done so far, Imao.

Is it possible that all trading firms use exactly the same criteria to hire graduates?

I explained you in my previous reply what the purpose of these types of assessments are - and its not “to spend 6 hours a day monitoring market prices”. I also mentioned how this is just the first round filter and how you’d face more puzzle-like questions in the later rounds.

Again - if you apply to trader positions, then you are more likely to come across these tests than when you apply for quant research roles. For trader positions, your PhD provides you very little to no edge over your competition which is mainly STEM Bachelors and Masters graduates. Age, having worked in a slower-paced research environment for a while and lack of relevant internships might even disadvantage you.

Trading companies typically aim to hire very sharp and technically strong candidates into these positions but don’t care much about specific academic skills (probably and statistics are probably most useful). Given the extremely high number of applications they typically receive (100 per hire is not uncommon), these gamified first rounds are a good way for them to filter down the field. These tests are a necessity to pass unfortunately and your competition in the more problem solving-focused rounds later will not be lighter. Maybe these more operational roles are just not for you?
 
Last edited:
I explained you in my previous reply what the purpose of these types of assessments are - and its not “to spend 6 hours a day monitoring market prices”. I also mentioned how this is just the first round filter and how you’d face more puzzle-like questions in the later rounds.

Again - if you apply to trader positions, then you are more likely to come across these tests than when you apply for quant research roles. For trader positions, your PhD provides you very little to no edge over your competition which is mainly STEM Bachelors and Masters graduates. Age, having worked in a slower-paced research environment for a while and lack of relevant internships might even disadvantage you.

Trading companies typically aim to hire very sharp and technically strong candidates into these positions but don’t care much about specific academic skills (probably and statistics are probably most useful). Given the extremely high number of applications they typically receive (100 per hire is not uncommon), these gamified first rounds are a good way for them to filter down the field. These tests are a necessity to pass unfortunately and your competition in the more problem solving-focused rounds later will not be lighter. Maybe these more operational roles are just not for you?
Thanks. I was missing a part of your previous message. Now it's clear.

I think I can get better results in problem-solving because it's closer to my skillset and I am much more trained. I can solve most of the brainteasers I found online for such interviews.
 
, C++ is often not a key aspect that is assessed in the interviews and PDE pricing even less. General programming aptitude is.

What is "general programming aptitude"? Sounds fuzzy. Is it just to write prototype proof-of-concept code or at the other extreme multithreaded event-driven frameworks in C++ or C#?
 
I have always had a choice so far. Id est C++ :)
The best way to get a good grounding in programming is C++. All the other languages might be OK in places, but thy don't cut the mustard.
And that's the way it is.
 
Last edited:
, C++ is often not a key aspect that is assessed in the interviews and PDE pricing even less. General programming aptitude is.

What is "general programming aptitude"? Sounds fuzzy. Is it just to write prototype proof-of-concept code or at the other extreme multithreaded event-driven frameworks in C++ or C#?
In earlier stages this is typically assessed through LeetCode-like online tests. In later stages, the candidates often receive a take-home assignment. A focus there is is on clarity of code, choosing good abstractions, extensibility, using relevant data structures and algorithms. Both rounds can typically be completed in a language of the candidate's choosing (within a predefined set).
 
Update: I will get an interview with the first company. I cannot state his name right now, but its initial aptitude test was totally different from the others and not based on the speed (even though the limited available time requires some speed too).
 
In earlier stages this is typically assessed through LeetCode-like online tests. In later stages, the candidates often receive a take-home assignment. A focus there is is on clarity of code, choosing good abstractions, extensibility, using relevant data structures and algorithms. Both rounds can typically be completed in a language of the candidate's choosing (within a predefined set).
In my book this does not prove programming ability. It is however better than multiple choice questions.
Do traders need to be able to write their own (production) C++ code?
 
Do traders need to be able to write their own (production) C++ code?
In a company where traders are "quant traders" it is expected that they can code a strategy. If the same company also hires "quant developers" or "quant researchers", then I guess trader is just a boring job :)
 
Some random remarks
I suppose the world is changing. A while back traders would have quant developers helping them with Excel addins, coffee etc.
I wrote a book on C# in 2013 with Andrea Germani (Lodi, Milano) who at the time was a trader and wrote his own code.

I know several traders who have done both QN courses.

What does a trader typically do these days? Depends on the kind of trader I suppose.
 
Last edited:
I wrote a book on C# in 2013 with Andrea Germani (Lodi, Milano) who at the time was a trader and wrote his own code.
Wow, that's interesting. It's possible that I will need that book at some point.

What do you think of companies such as "Into City Prep"? They offer a paid internship (in the sense that I need to pay them) and guarantee I will find a job after that. It is worth mentioning that they were among the fastest in replying to my application :D
 
Wow, that's interesting. It's possible that I will need that book at some point.

What do you think of companies such as "Into City Prep"? They offer a paid internship (in the sense that I need to pay them) and guarantee I will find a job after that. It is worth mentioning that they were among the fastest in replying to my application :D
I can imagine they are fast :)
 
Back
Top