Data Science in Hedge Fund
Hi,
I was wondering if anyone has any experience working as a data scientist in a hedge fund? I have a data science background, but currently work in equity research. My aim is to combine these two skillsets to become a hedge fund analyst.
I want to do some side projects similar to what data scientists do in hedge funds in order to hone my skills. What are the key languages/technologies you use? What does a typical project look like? What would you focus on learning? What would be some good side project ideas be?
Any input at all is extremely helpful, however irrelevant you think it may be.
Thanks in advance, I would really appreciate any input!
bump
BUMP
Almost nobody wants this combo of skills.
it’s a good product but there’s no market for it sadly
best bet is being more quanty — they have more tolerance for weirdos
a fundamental pm won’t understand what you are doing or care
There might be exceptions here and there but good luck finding them - they are rare. I’m sure people will tell me I’m wrong but feel free to mention even 1 prominent exception to what I’m saying.
Two different routes. The closest you can get to is a fund that believes in a quant overlay to single stock picking. But then that fund will have separate quant PM/strategist and fundamental PM. Fundamental PMs will not care about what statistical model the quant strategist used, as long as they tell the fundamental PM what (s)he should do with the single stock positions or overweight / underweight certain sector / factor / asset class. Quant PM will not have looked at a single financial statement in his/her life.
When I was interviewing out of grad school I did a couple interviews for data science roles at asset managers (not hedge funds), and the vibe I got was that they were there mostly to support the discretionary decision makers, ie here's a bunch of data, here's a hypothesis I have, figure out whether or not the data supports that or if not, what is the data saying.
I agree with some posters above who say that it's a better career path to either go more or less quant. What I mean is either try and specialize more and go into a fully systematic fund (I assume this is doable from a data science role but not sure - if it isn't you could consider an MFE type degree and then transitioning to the buyside afterwards). Personally, I chose to go the other way, and now trade fixed income instruments at a hedge fund. While I still use some programming / math in my day-to-day, it's much less than a data science role.
When I say a "better" career path, I mean that in those cases you would have a route to a role tied to P&L - this isn't necessarily better for everybody. The data science type roles at asset managers are still great careers, and I would bet that they are not too difficult to transition to from your position. Feel free to DM if you have more questions--for reference, my background is 2 years of trading at a bank -> MFE type program -> hedge fund.
Would love to chat further - PM me if you have a minute!
currently a quant, can you DM me?
DM'ed both of you guys.
interesting...i was a desk analyst at a bank for a few years (lots of excel and VBA back in the day) and then traded rates for a few years and did "ok" (made 10mm trading US rates my last year on a tight leash, then got pushed out because of politics...my former boss didn't respect me ans when they needed to cut headcout, i was out...to be fair, i was young and immature back then, but feels like i've been blackballed). I've taken a number of years away from markets and want to get back into it, but can't seem to even get an interview when the rare seat opens up that i feel i'm qualified for (my old boss is still a desk head at a bank and has been around for a long time).
i have 2 strategies
1) rates RV in the long end 10-30yr curve and fly stuff (not as much opportunity these days vs pre-2008)
2) semi-systematic outright direction of long rates based on pattern recognition of 10yr note futures (mostly intraday stuff)
i haven't traded in a number of years, but am looking to get back into it, but can't even get an interview for an institutional seat.
any advice?
Depending on your age, you could always try going back to grad school for an MFE type degree if you have the background, and trying to make the switch to the buyside from there. When I was in grad school, the oldest member of my class was 34, and he successfully made the jump from a risk management role at a bank to a hedge fund (he did have a PhD in applied math), so there are some older candidates.
Other than that, again depending on your age, you might consider looking into junior level roles (entry level or just above) at a small fund or trading firm, if you'd be willing to do that. I say small fund or trading firm because larger, established funds tend to have recruiting pipelines either directly from undergrad or grad school or from banks/other funds, while smaller firms tend to be more open to unorthodox candidates.
To be honest, I think it would be difficult to transition directly to a risk-taking role having taken several years off from trading without taking some intermediate step in the middle. From my (pretty limited) number of years in the industry, the market tends to evolve quickly, and those who are able to stick around are constantly adapting with it.
I have a similar background to yours, graduated from my MFE, but had an awful time recruiting through the pandemic. Settled for a data science analyst role at a good systematic hedge fund. The work feels more like a data engineer to me. I have been looking to make the shift towards a QR role, the market seems good for recruiting now. Is it possible to have a chat about this?
Major financial companies are using machine learning and advanced statistical data techniques for trading and research. In this tech-driven environment, data science in financial services has changed an entire trading scenario.
Developing a data analytics system assists traders to uncover trade patterns, trends, and correlations that can outperform the market with a large dataset.
In hedge funds, traders use quantitative techniques to predict future returns and risks. According to a research report of a financial firm, around 13,060 hedge funds are using artificial technology and machine learning tools to trade. Several new predictive techniques have emerged in the last 2-3 years such as artificial neural networks, random forecasting tools, decision tree techniques, etc.
Data science in the financial industry has an application in quant hedge funds. Quant hedge funds are hedge funds that employ automatic trading strategies despite traditional stock selection based on technical/fundamental analysis. Funds buy large electronic datasets and apply them in algorithms to derive profit-generating financial strategies. Experts at Acuity Knowledge Partners state that dealing with trillion megabytes of data to derive actionable insights is the biggest issue.
Advanced predictive machine learning techniques and data analytics techniques are coping up with this huge data. These analytics apply statistical concepts to predict future trends and performance. A report of Barley Hedge's Hedge Fund survey has stated that over 56% of total survey respondents used artificial technology and data science tools to inform financial information. With this trend, data science and quantitative trading will become a necessity of the financial industry.
Feel like a lot of high profile funds/tiger Cubs have in house data science ppl. D1, Viking, eminence etc all have them. Seems like a solid role
Hey, I come from a data science background to and am starting work in equity research.
I'm wondering what you ended up doing ?
Cheers!
Ended up as an 'investment' data scientist in a fundamental hedge fund. Would avoid at all costs lol (based on my experience and majority of others I have spoken to anyway).
Oh wow, that sounds like high paying and fun, what don't you like about it?
If you don't mind me asking, would you have taken otherwise and where are you looking to move going forward?
I could write an essay explaining it, but it would take me a while to explain what I have seen and heard within the industry.
In summary, fundamental shops/PMs have no real understanding of how to use data science to provide significant value to their process. And a lot of the more senior/more established PMs aren't too interested in experimenting too much. Data scientists within these funds are nothing more than business analysts (at a push), there is no 'science' - It is very loose calling these individuals 'data scientists', but I think this is primarily done for marketing reasons. They typically use it for simple things such as trend following (e.g. tracking changes in a company's job listings, credit card spend changes, product pricing data, advertising data, etc.), as well as process improvement, such as simplifying monotonous tasks that they previously had to do manually, web-scraping, as well as small uninventive ad-hoc tasks. Due to this, it isn't typically fun or high paying (in the majority of shops that I have heard of), there is of course a number of exceptions. The roles will typically be sold as 'investment' data science roles, but you will most definitely be seen as a support function. In my opinion a lot of the value of having these underpaid teams is being able to sell it to investors claiming it enhances the investment process, and differentiates from their peers.
Given that you are starting in ER, if you want to work in a fundamental HF, I would highly recommend going down the investment route while continuing to develop and utilise your data science skillset. This will hopefully allow you to develop a very desirable skillset going forward. A huge problem I find is that hedge funds typically silo their data science teams to some extent (as much as they deny it). In order for the team to add value it needs to be embedded in the investment process from start to finish. Alternatively, a 'softer' quant researcher role might be a good fit either if you'd prefer to stick to the data science side of things.
This is a very brief and rushed summary, but if I get more time I will write out my thoughts properly. Again, this is based on my own experience and from talking to participants across the industry over the last number of months.
Not sure what you mean by 'would you have taken otherwise'?
Where are you looking to move going forward?: Not really sure what to do moving forward tbh, I work for a very large and well respected fund, but just in the wrong position. I think I need to leave unfortunately, but not sure what options are available for someone with my skillset. I want to have more of an impact on the investment side while utilising my strong data science skillset, but I am not sure a lot of funds are looking for this atm. Any advice from anyone would be appreciated here.
Good advice. There is a gap to bridge between the investment side and the data science side. Most fundamental analysts don’t know programming enough to get the most out of applying the methods to the research process beyond simple forecasting of drivers. There is much potential to use data science to understand the business and unit economics and a deeper level and to use it for idea generation. Ultimately investment research is about identifying the key questions and then trying your best to answer those using any/all available info. Available info is proliferating creating new opportunities for creative edge. But you have to understand the business and the market to know what you should be “asking” of the data and you have to know the data to be efficient and know what questions are answerable and have a sufficient return on time to do the data work. Like some shops have separate roles just to build models or just to do expert calls, I don’t believe these components can be compartmentalized. The data scientist often ends up underutilized, handling vague/poorly defined projects or just making tableau visualizations all day. The investor gets credit for the pnl. So seek to be an investor with data science skills vs. a data scientist with investment skills and generally I think you’ll do better yourself.
Echo this. If you want to chat, DM me.
.
I know it's not a hedge fund but for anyone reading this that has a data science skill set and is interested in trading/financial markets - look into commodity trading shops. Shops like Cargill, Vitol, Trafi, Bunge, etc are all starting to build out there data science groups more and more. Just saw a posting on LinkedIn from Cargill looking to add to their data science group at pretty much all levels. Commodity trading is all about data so the people there actually listen to the data guys and they are generally well respected.
This is blatantly false at most physical commodity shops. Those firms are incredibly boomer led and if you're not a trader, get ready to be treated like garbage. Most likely they'll put you under the tech team to save money and if you're tech, why not head to a company that will actually pay you more for programming?
Guess we have had different experiences. I do a lot of data science and also trade my own book. Met a couple other people that have similar stories.
Go join Citadel
Can confirm Citadel has roles like this.
Removed for anonymity. Removed for anonymity.
Dolorem voluptatem consequuntur sunt in doloribus et illum facilis. Sed sit et dolor dolore. Quod repellendus eaque ducimus sit. Tenetur quaerat et rem.
Voluptatem doloremque culpa quisquam consectetur perferendis aut hic. Corrupti non ipsa praesentium tempora adipisci. In distinctio eligendi deleniti nemo est aut. Aliquam dicta pariatur porro dignissimos.
Aperiam quia qui quo cumque. Cupiditate animi at mollitia. Ipsam porro aut quia dolorem sit. Aut recusandae adipisci porro omnis soluta. Odio qui officiis rerum reiciendis.
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...
Maxime voluptatem vel sit a qui. Aut aut sed cumque autem dolor.
Sint est et quis culpa nam possimus maxime. Quidem cum facilis sapiente repellat corrupti voluptatem reprehenderit. Dolore quia fugiat aut fugit alias. Eligendi omnis ut rerum nesciunt.
Sed molestias et maxime nam vero. Tenetur incidunt eum numquam voluptas. Culpa quibusdam eius veritatis neque eaque modi totam.