News for hackers. Stuff that matters.
Re:Ask HN: What are you working on? (August 2026) by qurren on Monday August 10, 2026 @10:48AM (#49244443)
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.

Re:Ask HN: What are you working on? (August 2026) by DLA on Monday August 10, 2026 @12:18PM (#49245723)
I call BS on 30,000% return in 3 months. If true, OMG!
Re:Ask HN: What are you working on? (August 2026) by qurren on Monday August 10, 2026 @01:12PM (#49246683)
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.

Re:Ask HN: What are you working on? (August 2026) by senordevnyc on Monday August 10, 2026 @01:46PM (#49247139)
How much was your initial capital? How niche are we talking about here?
Re:Ask HN: What are you working on? (August 2026) by qurren on Monday August 10, 2026 @03:01PM (#49248129)
$150 -> $48K. Gradually and consistently over 100K trades over a well-understood (by me) market mechanism.
Re:Ask HN: What are you working on? (August 2026) by bkovacev on Tuesday August 11, 2026 @03:53AM (#49254722)
Would you be willing to share more? I don't mind if I have to pay for it!
Re:Ask HN: What are you working on? (August 2026) by qurren on Tuesday August 11, 2026 @11:29AM (#49259885)
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.

Re:Ask HN: What are you working on? (August 2026) by patatino on Monday August 10, 2026 @04:03PM (#49248951)
Did you use any libraries?
Re:Ask HN: What are you working on? (August 2026) by qurren on Monday August 10, 2026 @06:34PM (#49250808)
No.
Re:Ask HN: What are you working on? (August 2026) by danhudlow on Tuesday August 11, 2026 @12:07AM (#49253270)
What service / API are you using to execute the trades? Has latency been an issue?
Re:Ask HN: What are you working on? (August 2026) by c-fe on Tuesday August 11, 2026 @05:46AM (#49255574)
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?
Re:Ask HN: What are you working on? (August 2026) by qurren on Tuesday August 11, 2026 @10:14AM (#49258756)
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.