5 ms·

I think it's pretty safe to say finance will be a big one. Finance has a large amount of individuals and firms researching the applications of ML methodologies

by pmulv 8y ago

I think it's pretty safe to say finance will be a big one. Finance has a large amount of individuals and firms researching the applications of ML methodologies to financial indicators. With the semi-recent rise of quant firms, I think this research is only going to get more aggressive, and HFT will become more lucrative and more automated as long as regulation does not get in the way.

a008t 8y ago
HFT - yes. But longer-term investment (i.e. Buffett - or even with a horizon of a couple of years) is unlikely to be transformed soon - ML needs vast historical data, which is very slow to generate. Waiting 10 years only gives you 10 years of history, which is 5 non-overlapping 2-year forward returns, and maybe 1 or 2 economic/financial regimes. This is also a problem with new datasets being generated - there is not nearly enough history available to test them or feed them to a ML system. Furthermore, arguably, longer-term investment requires forward-looking modelling of scenarios, based on the kinds of inputs that were not seen in history. ML is not very applicable when you get big covariate shifts. So I would say human financial analysts are not going anywhere, and any improvements would be relatively small and incremental.
mikert5671 8y ago
HFT is not profitable. Its completely commoditized.
- dasil003 8y ago
  I would think commoditization of HFT is impossible by definition due to the fundamentally competitive zero-sum nature of securities markets.
  - mikert5671 8y ago
    great! do you work in finance? Are you aware of how costly it is to advance speed in HFT and how that eats away the profits you get from the advance in speed? Have you looked at virtu financials stock? Do you know anything at all about the field?
    - dasil003 8y ago
      You already know the answer to that based on your cocky response. It still would help those of us less enlightened than you to understand how something as specialized as HFT can be commoditized. The simplest CRUD apps that any Bootcamp graduate can produce are not commoditized even though hundreds of thousands or even millions of people can build them.

by pmulv 8y ago

a008t 8y ago

HFT - yes. But longer-term investment (i.e. Buffett - or even with a horizon of a couple of years) is unlikely to be transformed soon - ML needs vast historical data, which is very slow to generate. Waiting 10 years only gives you 10 years of history, which is 5 non-overlapping 2-year forward returns, and maybe 1 or 2 economic/financial regimes. This is also a problem with new datasets being generated - there is not nearly enough history available to test them or feed them to a ML system. Furthermore, arguably, longer-term investment requires forward-looking modelling of scenarios, based on the kinds of inputs that were not seen in history. ML is not very applicable when you get big covariate shifts. So I would say human financial analysts are not going anywhere, and any improvements would be relatively small and incremental.

mikert5671 8y ago

HFT is not profitable. Its completely commoditized.

dasil003 8y ago
I would think commoditization of HFT is impossible by definition due to the fundamentally competitive zero-sum nature of securities markets.
- mikert5671 8y ago
  great! do you work in finance? Are you aware of how costly it is to advance speed in HFT and how that eats away the profits you get from the advance in speed? Have you looked at virtu financials stock? Do you know anything at all about the field?
  - dasil003 8y ago
    You already know the answer to that based on your cocky response. It still would help those of us less enlightened than you to understand how something as specialized as HFT can be commoditized. The simplest CRUD apps that any Bootcamp graduate can produce are not commoditized even though hundreds of thousands or even millions of people can build them.