Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Healthy skepticism is warranted and welcome of course! My account is currently at 1.4M. Taxes and living costs are taken out regularly. I'm not giving anyone access to the source of the models, only the output signals and trade history. I do give brief descriptions of some of the indicators that each take into account, but none of the details of how the complex system works which is the tricky part. You can find huge list of potential indicators for algotrading for free lots of places online.

Lastly, the 75% is a bit of an anomaly. As you can see, the top model is only expected to return about 60% pre-tax. It's unusual that in live trading we have exceded the back test--2020 was a bizarre year and we also started after a big bear market so we benefited for a huge bull that year + nailed the June and September crashes (check the trade history). 2021 was a below average year by contrast, and over the long term, any algotrader should expect the live results to be a bit under the backtest as despite our best efforts (and I consider my anti-overfitting engine to be the biggest accomplishment here), all systems are to some degree overfit to past data.

Lastly, as I mentioned in another comment, I make no guarantees of always publishing this site and signals. If I decide in the future (perhaps once my account is to that 50-100M range) that the site is more of a liability for me than an asset, I may close it, with or without grandfathering. However, obtaining 1% of Seeking Alpha's userbase would be 5M+ per year in revenue from the site which is a few multiples of my current net worth. Even with a hypothetical 50M net worth 10 years from now from 40% post-tax, post-living expense returns, 5M is still meaningful. If we fail to gain that much traction, or the bot does so well that I hit that kind of figure much earlier than expected, I may halt the publishing.

As of now, and for the forseeable future, the site makes financial sense to me, and it's been a lot of fun, so it also makes pleasure sense for me. Life is about more than money of course.



> If I decide in the future (perhaps once my account is to that 50-100M range) that the site is more of a liability for me than an asset, I may close it, with or without grandfathering.

I'd like to offer to take care of it and manage all aspects of it if you ever get annoyed by it. These algorithms are life-changing and the more (and especially lower income) people have access to them, the better.


What's the trickiest statistical method you employ that isn't embedded in this model?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: