PROJECTS

Learning quantitative research. Over the past year I’ve built a small-cap trading algorithm end to end — from data pipeline to automated live execution. It has taught me about statistical biases and to distrust a good result.

workflow

Workflow

I research signal sets in-sample and verify them out-of-sample to avoid curve fitting.

second-res data in-sample out-of-sample live cross-check most die here

The first question is how a signal is measured, not whether it works. The same indicator reads differently on the same candle depending on prints, hours and smoothing, so I pin that down with a probe before any longer autoresearch run.