Quant Research Lab
Reproducible strategy experiments — explicit hypotheses, sound methods, and honestly reported results.
Self-contained research experiments on market data. The purpose is not to publish profitable strategies — it is to ask well-posed questions and answer them rigorously. A negative result, analysed carefully, is as valuable here as a positive one.
Every experiment follows the same structure:
- Research question
- Hypothesis
- Dataset
- Methodology
- Code
- Results
- Limitations
- What I learned
- Next experiment
The tone is deliberately that of a researcher, not a signal seller. Not “this model predicts Tesla with 80% accuracy,” but: this experiment tests whether lagged return, volatility, and volume features contain statistically useful information for short-horizon directional classification, evaluated with walk-forward validation to reduce look-ahead bias.
Experiments
Published experiments appear below. More are in progress.
Planned experiments
- Can short-horizon Tesla returns be predicted from lagged return, volatility, and volume features?
- Does the volatility regime materially change a strategy’s performance?
- Can a Hidden Markov Model identify distinct market states out-of-sample?
- Does volume confirm momentum, or add nothing once price is known?
- Does regime-filtering improve a simple RSI rule, or just overfit it?
- Does news sentiment improve entry timing beyond price-based features?
- Can reinforcement learning improve position sizing under a fixed signal?