AI in Fundamental L/S Equity

Curious what the use of AI/ML is for these types of funds. Is it most useful as a form of "research assistant", helping synthesize large amounts of data? Is it used as a more effective screener, analyzing information/trends to determine companies/industries that could be worth a look?

Comments (13)

Mar 6, 2019

Would love some background on this as well. I think there's definitely ways for AI to be implemented, but I believe the human element is always going to be necessary in terms of stock selection.

Mar 7, 2019

There is lots of applications of NLP for processing news, SEC filings, earnings calls and synthesizing that data or creating alerts. so the ML would be built into the NLP part, as opposed to using ML to value the companies themselves

    • 1
Mar 6, 2019

Are there any firms that specialize in doing this analysis? I.e. data vendors that aggregate the info. Or is much of the work brought in house to the fund itself?

Mar 7, 2019

There are a few firms that are vendors of data/visualization/signals to hedge funds and AMs. A lot of these are on the smaller side or startups. For example: https://www.marketwatch.com/story/ge-earnings-can-...

Mar 7, 2019

Anonymized credit card transactions are sold to funds who use those to predict earnings reports. Right now some funds use analysts to go through the credit card data to predict earnings. More sophisticated investors use algorithms to automatically trade on this data.

Learn More

814 questions across 165 hedge funds. 10+ Sample Pitches (Short and Long) with Template Files. The WSO Hedge Fund Interview Prep Course has everything you'll ever need to land the most coveted jobs on the buyside. Learn more.

Mar 14, 2019

no one does this in the way you are talking about at traditional l/s funds. the "data scientists" that work at traditional l/s funds scrape data in very basic ways and output it. that's it. definitely no ML.

if you are a junior person that knows how to do this good luck trying to convince anyone that's been doing this for 10 years to listen to you or to give you the resources / time to work on this. it just won't happen. generally speaking you will look like a complete idiot if you bring this up in a mixed setting.

i am not saying i think it's not a good idea. it might be if done very well. i am just saying there is a very real institutional bias against doing this in my experience.

    • 1
    • 1
Mar 6, 2019

What about in event-driven scenarios? For example, using ML to predict likely M&A candidates, or to derive likelihood of a positive/negative event in the short-term.

Mar 16, 2019

Yeah that's a retarded question. As someone who does ML/AI, there is simply no way to usefully apply it here. When you say ML....you really mean to say statistics?

    • 1
Mar 7, 2019

This is wrong. Dont speak on things you dont understand

Mar 14, 2019

mike: feel free to mention at least 2 marquee / well-known TRADITIONAL LONG SHORT funds that people have heard of and that do this in a meaningful way. To be specific, point us to 2 articles that verify your claim. If you can't then i will rest my case.

Mar 7, 2019

Traditional long short funds that arent adapting to advanced data analytics are getting their faces ripped off, its a dying hedge fund model. Take your pic of a high profile failure. Like Balyasny who fired their whole equities division. Once one of the best in the industry.

Also TBH, ive got better things to do than look up articles. I browse this website sometimes, the amount of misinformation about the industry is appalling. Just because funds arent talking about what techniques theyre using, doesnt mean it isnt happening. Hottest job on wall street is a data scientist, not a yale history degree pouring over 10ks.

Mar 14, 2019
Comment