MS data science for Quant trading
I'm currently in undergrad studying economics at a target school. I'm in my earlier years and wondering if getting an MS in data science (1 year) would be worthwhile for quant trading. My two questions are how do you build up your resume for QT without being a math/comp sci/DS major and what kind of internship path would make sense to follow
bump, in the same position except doing discretionary trading. would a masters in data science be useful for rates / FX trading?
IMHO I think an MS in Statistics is more valuable and useful than an MS in Data Science. Statistics provides the fundamental building blocks for Data Science and understanding these fundamentals well, will greatly assist your trading career more so than Data Science. Data Science is still the flashy [relatively] new buzzword much like ML or AI - it sounds cool, but it's usefulness for trading is limited. From my 15+ years experience in the industry, I've always found the best trading models to be based on traditional statistical techniques but with some creative tweak. I've never seen a complex ML model ever be successful in trading (yet they maybe applicable for trading adjacent projects) because of overfitting and lack of understanding. This is even coming from experience working in teams entirely filled with PhDs (but I don't have one myself!). If you thoroughly understand the foundations then you should be able to express your idea/view/edge using standard approaches. Only jump to advanced techniques to improve on what you have otherwise you'll get lost in complexity and lose sight of what you're trying to model. Statistics gives you the core knowledge and tools you need to robustly evaluate trading ideas. Data Science will give you the ability to data mine but at the risk of finding the illusion of alpha.
Thanks for the POV. My main concern is my lack of experience with CS because everyone makes it seem like CS is essential to QT. Is it possible to do QT with no CS experience/knowledge?
Yes, I agree that CS is highly useful for QT (I did a CS major at college myself), however, from my experience, the most useful aspects of CS are the practical components i.e. programming. If you can pick up a couple of practical programming courses in say Python, and get to know it reasonably well, then you can get away with not having to understand all the CS theory.
You're aware that alot of Maths, Stats or Physics majors go into QT because firms only really care about their ability to problem solve and think logically. These guys don't necessarily have any CS knowledge, but firms know that they could probably learn how to code pretty easily. In the end, they just want traders to be able to test and implement their ideas and programming facilitates this.
what exactly is Quantitative Trading - that also seems like a buzz word. The basic fundamentals of finance are there, but you are just trying to find patterns in data, which requires programming, to get an idea on intrinsic values?
I guess the difference is that Quantitative Trading is basically it's own distinct area within broader trading. Data Science is just a combination of several disciplines packaged up and relabelled. So I'd be hesitant to say that QT is a buzzword.
Quantitative Trading can be loosely defined as using any type of mathematical or statistical model (with a vast range of possible input variables) to generate a signal to trade on. Clearly that presents a large spectrum e.g. quantmental, semi-systematic, fully systematic, HFT etc. This style of trading is relatively modern and only became popular in the early 2000s. Contrast this with very traditional forms of trading where traders make discretionary calls based on fundamentals (equity research) to decide to buy/sell a stock e.g. stock-pickers.
When you mentioned "intrinsic value", this would refer to the actual pricing of a given asset. Pricing is a field/concept in itself. Within the context of QT this might be the instantaneous mathematical pricing of a future or option, which trading firms then quote a bid-ask spread around (derives its roots from pit traders/market makers). In the context of traditional trading, this might be the fundamental valuation of a stock/company, which then drives a decision as to whether to buy/sell.
Then when you say "patterns in data", this is another concept. Which means you're following the movement of prices through time and trying to predict the next step in a time series.
These are all separate but heavily linked concepts. It can be hard to tell you exactly where one ends and another starts..
.
So I am a Quant, but not a QT - so don't take my word on everything.
Candidates who break into QT have a little different backgrounds than those who go into QD/QR. QT guys are often coming straight from a target undergrad, with degrees in math, statistics, and some other tangentially related areas (physics, maybe engineering, CS). I saw another comment talking about CS, yes CS can be a path to QT. CS focusses a lot of systematic logic, building discreet solutions to problems. With that being said, programming is not a skillset most QT's need. I'm sure there are plenty of QT's that know how to code, but they do not need to be as familiar with this skill set as a QD. QT's mostly use quantitative models, whereas QD will build, QRs will research. In smaller shops, the line that separates these categories may be a little blurry.
I am giving you this background mostly to say, what you will learn in a MS DS degree doesn't translate very well to a QT career. Generally speaking, DS is not quantitative enough for ANY of the quant career paths (you could probably land yourself in some back office sell side risk team). If you are to do a MS, go pure/applied math, CS, or maybe an MFE if you feel like spending 150k in a year and a half.
What I would actually do if I were you:
- change your major to math or statistics (at the very least pick up a minor)
- start looking at employer sponsored competitions and do as many as possible
- try to work with one of your professors on research that is a little more quanty
- look for ANY internships in the QF, tech, engineering space.
Your focus in economics is definitely hurting you here, you need to prove your skills in quantitative areas with internships, competitions, personal projects, research, and a major change if you can work it in (seriously this is your best bet). Then let the target school on your resume do the rest and network.
Quisquam occaecati cupiditate dolores corrupti cum officiis. Consequatur voluptas error porro modi. Facere et dolor ut ut saepe ratione totam.
Iure recusandae et quia harum quaerat ducimus omnis consequuntur. Est impedit laboriosam ut sit delectus. Est praesentium sit sint. Vitae voluptatem nesciunt ipsum hic odio.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...