Ask HN: What’s it like being a Quant or Quantitative Programmer

18 points by mraza007 ↗ HN
What can you expect in a quantitative role. Do you have to be extra smart or you can learn and work hard to become a quantitative programmer. Lastly what’s the most popular programming language among quants and what kind of workloads you can expect.

I’m coming from a software background who did internships as I’m getting into field soon I’m trying explore different roles where I can do meaningful work

Would love to hear from the community

15 comments

[ 3.3 ms ] story [ 46.7 ms ] thread
My company has something like a quantitative programmer but doesn't give it that exact title. I worked in an adjacent role and looked at moving into that sort of role and even a trader role, so I know a bit about it.

At my company the main tech they use is AWS, Python, R, and various homegrown and vendor financial software (like BlackRock Aladdin). They also use some excel, Tableau, and stuff like that for reports.

You have to be a quick learner, but I wouldn't say you need to be super smart. If you are super smart, then you will be more likely to be successful. You basically have to be competent in two domains - technology and finance. At least in my company, you would work closely with the actual quants (we call them analysts) and they can give you feedback and direction on your work. This type of work is usually for larger projects and permanent systems. This is more software dev and less finance.

We also have a role called something like trade desk programmer. These are typically business people with a background in trading/analysis that learn enough Python, scripting, R, and excel macros to be able to write small scripts or programs to do some sort of analysis or filtering as part of their trading activity. Each trading team would probably have 1 out of 5-10 people in that role. This type of work is smaller projects that are usually temporary or change often. This is more finance and less software dev. Hours are usually something like 7am to 5pm.

Just curious do you work for a HFT and do you think quantitative programmer is equivalent to Trade Desk Operations roles where helping the traders
I work for a large financial company. We don't build any HFT systems, but we send our orders through various vendor systems and homegrown ones. The typical strategy is to consolidate orders so we aren't placing a bunch of small orders. This is usually done for funds. I believe the retail orders are placed instantly.

Most of the quantitative systems we have do not actually place trades. Instead they provide data to the fund managers or traders and they make the final decision on placing a trade.

It's hard to say if the roles are equivalent. Role responsibilities and structure vary from company to company, even for something that's named the same. The typical trade team structure is something like 1 desk manager, 2-3 traders, and 3-5 analysts. Usually the desk programmer is not from IT, but rather one of the analysts that knows how to code scripts, Python etc. Their scripts are usually some part of the quantitative analysis process, mostly on current or targeted data.

They also use systems developed by IT. The IT systems are bigger and span multiple trade teams. These systems rely more on historical data or market patterns. An example would be a system used by all equity trade desks to identify potential sources of alpha using big data.

So the desk programmer is really more quantitative since they are an analyst whereas the engineers working on the alpha system rely more on input from analysts on what algorithms to implement.

One other thing to mention is that desk operations is a separate type of role that is more like a better version of help desk support.

This is just how my company is structured. I'm sure others are different.

I see I think that makes a lot of sense I get an impression you probably work for a bank or IB firm but I could be wrong anyways thanks for replying
It's similar to Fidelity, Vanguard, BlackRock, etc.
Nice !!
Actually, I hate my job.
Haha i can relate to you i actually worked for two big banks as SE
I will reply for the quantitative developer role, given your description I suppose that it's what you are looking for.

Succinctly, I would say that it's a strong computer scientist/developer highly interested in finance to be able to understand/discuss/implement business requests.

It's important to be able to work under pressure (when there is a problem on an automaton for instance, etc.). It's also important to be able to frequently switch between tasks.

I would not say that you have to be extra smart.

I started as a computer scientist in a bank, turned quantitative developer and then trader a couple of years later.

How can I get started into this I did worked for a bank but didn’t find it interesting I’m currently looking for a path where i can do automation and maybe write software that can be helpful or even develop machine learning models
Would recommend the book The Complete Guide to Capital Markets for Quantitative Professionals, it's more an intro to how finance works rather than quant dev but will give you a sense of what it involves at a high-level.
You work as what we call a 'data monkey' for some economist or trader who can't work computers, thinks a linear fit is the state of the art, would jack off over an R² of >0.05, wants you to spend all your time automating p-hacking, and is almost certainly a sociopath or sex offender.
As far as programming language I'd recommend KDB/Q as it's one of the few you can just self learn, prove it in an interview call for 15-20 min, and then start Monday. Pay is good and KDB devs are rare.
Just curious what’s KDB
A database optimized for queries on time series.