Systematic trading model
Machine learning · FinanceAn ongoing experiment in whether I can find any real edge in markets. Emphasis on experiment - I'm nowhere near trusting it with real money.
Paper-traded for now, with costs included. A private research repo, so there's no public instance, but ask me how it's going and I'll be honest about it.
What it does
I'm trying a few systematic trading systems across stocks, commodities and crypto: turning signals into positions, and testing them carefully enough to spot when I'm fooling myself.
It's research, not a fund. I care less about the returns than about learning where these methods actually help and where they just overfit.
How it works
Each system is tested walk-forward: fit on the past, test on the next unseen window, then roll the window forward. That ordering matters: it's the difference between a real test and accidentally peeking at the future.
Everything is backtested with trading costs included, and runs in paper mode for now, to see if anything actually works before any real money is involved.