4 ms·

AI quant trading (so that I have another income and can stop being threatened by managers that if I don't meet their unrealistic expecations that I won't have f

by qurren 2mo ago

AI quant trading (so that I have another income and can stop being threatened by managers that if I don't meet their unrealistic expecations that I won't have food on the table next year). So far it's going well, 30000% gains on my initial capital in 3 months very gradually over 100K trades. But I need to build some more sleeves because (a) the edge could die any time (b) I hit capacity of that edge and it can only grow linearly now.

patatino 2mo ago
Did you use any libraries?
- qurren 2mo ago
  No.
DLA 2mo ago
I call BS on 30,000% return in 3 months. If true, OMG!
- qurren 2mo ago
  It's severely capacity limited, that's why it's possible and Wall Street wouldn't be interested in it, yet it's too complex for most retail folks without a solid math and game theory background to execute. There are a handful of niche edges like this lying around that nobody at a quant firm would set up a trade desk around, it's not worth their time but it's meaningful money to someone in a personal capacity.
  - senordevnyc 2mo ago
    How much was your initial capital? How niche are we talking about here?
    - qurren 2mo ago
      $150 -> $48K. Gradually and consistently over 100K trades over a well-understood (by me) market mechanism.
      - bkovacev 2mo ago
        Would you be willing to share more? I don't mind if I have to pay for it!
        qurren 2mo ago
        No. It's a capacity-limited edge and a zero-sum game. If you traded it, it would eat into my edge. If you paid me less than I could make from it, it would be a bad deal for me. If you paid me more than you could make from it it would be a bad deal for you. If you paid the same it would be a waste of time and we'd both lose money to the IRS.
danhudlow 2mo ago
What service / API are you using to execute the trades? Has latency been an issue?
c-fe 2mo ago
What approach did you take with AI to make it discover an edge? Give the AI autonomy to search for edges itself and build e.g. a backtest it can use to confirm, or narrow it done with research (possibly AI) and then investigate specific approaches with AI? Or in other words, how autonomous was the AI in finding that edge?
- qurren 2mo ago
  AI couldn't really find any edges large enough to be worthwhile risk- and tax-adjusted. It's mostly me thinking of edge ideas that aren't on the internet and AI testing the hypotheses, building and battle-testing the infra, building risk models, implementing risk controls, monitoring it 24/7 for out-of-distribution behavior with several circuit breakers, etc.

by qurren 2mo ago

patatino 2mo ago

Did you use any libraries?

qurren 2mo ago
No.

DLA 2mo ago

I call BS on 30,000% return in 3 months. If true, OMG!

qurren 2mo ago
It's severely capacity limited, that's why it's possible and Wall Street wouldn't be interested in it, yet it's too complex for most retail folks without a solid math and game theory background to execute. There are a handful of niche edges like this lying around that nobody at a quant firm would set up a trade desk around, it's not worth their time but it's meaningful money to someone in a personal capacity.
- senordevnyc 2mo ago
  How much was your initial capital? How niche are we talking about here?
  - qurren 2mo ago
    $150 -> $48K. Gradually and consistently over 100K trades over a well-understood (by me) market mechanism.
    - bkovacev 2mo ago
      Would you be willing to share more? I don't mind if I have to pay for it!
      - qurren 2mo ago
        No. It's a capacity-limited edge and a zero-sum game. If you traded it, it would eat into my edge. If you paid me less than I could make from it, it would be a bad deal for me. If you paid me more than you could make from it it would be a bad deal for you. If you paid the same it would be a waste of time and we'd both lose money to the IRS.

danhudlow 2mo ago

What service / API are you using to execute the trades? Has latency been an issue?

c-fe 2mo ago

What approach did you take with AI to make it discover an edge? Give the AI autonomy to search for edges itself and build e.g. a backtest it can use to confirm, or narrow it done with research (possibly AI) and then investigate specific approaches with AI? Or in other words, how autonomous was the AI in finding that edge?

qurren 2mo ago
AI couldn't really find any edges large enough to be worthwhile risk- and tax-adjusted. It's mostly me thinking of edge ideas that aren't on the internet and AI testing the hypotheses, building and battle-testing the infra, building risk models, implementing risk controls, monitoring it 24/7 for out-of-distribution behavior with several circuit breakers, etc.