Alternative Data Vendor to HF?
Exactly what the title says - can it be done?
Has anyone done this or seen anyone make this leap???
Trending Content - Hedge Fund Forum
| +23 | % of pods making 9 figures consistently? | 29 | 11h |
| +12 | Fastest Way to MMHF? | 5 | 1h |
| +6 | Average single manager slope | 0 | 5h |
| +6 | Bad PM | 1 | 1d |
| +6 | Credit HF Guys: How much do you think abt "value" | 7 | 1d |
| +4 | Amakor Capital - Who are these guys? | 2 | 3d |
| +3 | culture at citadel macro | 1 | 6d |
| +3 | 4 YOE in ER & 30 years old | 5 | 1d |
| +2 | Walleye SA 2027 | 3 | 2d |
Career Resources
Have seen alt data vendor employees join data science groups at hedge funds. We look at candidates from that background, but few have strong enough programming skills to meet our bar.
Do they actually get to take on risk or just crunch numbers all day?
Number crunching. I’ve seen people go from vendor to data scientist and data scientist to investment professional, but never vendor directly to investment professional.
Is this regarding quant funds? We don't really have a lot of need for data scientists on the discretionary side.
Many large discretionary L/S equity funds have data scientists now, including both single PM shops and the multi-manager platforms. This has been a trend for the last 5-ish years.
It's definitely the trend, but I don't know a single PM who takes input from a data scientist too seriosuly. It's mostly used as confirmation of a story you already believe in. Not saying it is not used, I'm sure it is, just not what I have seen.
When you think about it, that’s odd, right? It’s well-known that alt data is more accurate than street for many names, and there are clear insights to be drawn from the data outside of calling quarters. (And yes, even with Yipit being a large contributor to buy-side bogey, the data are still relatively accurate, and thereby need to be paid attention to.) So why wouldn’t someone pay attention to their data scientists versus using it as confirmation bias?
Personally, I think it comes down to a few things. First, the communication skills of the group. Not many data scientists can do the work while also distilling the results and takeaways for the investment professionals. In other words, they can’t think like an investor. Second, overly complex models. Too many data scientists want to use fancy techniques, even if it causes skepticism. And finally, a lack of trust within the firm. A data scientist shows up and says NFLX is going to miss on net adds. You have no idea if this tracking is good or not. How do you believe it? This is cultural, and it’s the hardest one to solve.
There are funds that do legitimately use data science as more than just marketing, but I completely get why many see it as such.
And sentiment like this is why there’s still tons of alpha in the data. Collecting, tagging, ingesting, and delivering the data is a massive undertaking which accrues large benefits to the platforms across all sectors
Out of curiosity, what are the hard programming requirements for a fund such as yours?
The range of technical chops among DS at the MMs and SMs is so wide based on my experience with some individuals basically just building scripts in Jupyter Notebooks that spits outs results into CSVs and others essentially being full fledged ML engineers with strong programming/data engineering skills + statistical knowledge.
Culpa voluptatem voluptas molestias adipisci. Labore ullam quos modi error rerum enim consequatur.
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...