projects
Each project is written as a research note: result, problem, approach, what failed, finding, and artifact.
resultOut of sample Sharpe near 0.7 after fees.
the hard part was reporting the modest number after fixing the inflated one.
Estimating hidden state in noisy market data, then testing whether that estimate survives fees and a clean split.
read projectresult122 tests, 85% coverage, documented API limits.
the hard part was proving which historical data did not exist.
A prediction market signal framework for sports, built around validation, fees, and data limits rather than a clean story.
read projectresultPaper entry, canonical writeup lives in Writing.
the hard part was accepting that a well trained Bayesian agent can still stay wrong too long.
A short Projects stub for my Barrett honors thesis. The full paper writeup lives in Writing.
read writing entryresultResearch prototype with repo; no public performance number claimed.
the hard part was making path structure matter without overclaiming the metric.
A rough volatility hedging prototype comparing smooth GBM intuition with jagged volatility paths and learned path dependent hedges.
read projectresultBayesian covariance prototype with repo; no universal return claim.
the hard part was watching the optimizer exploit covariance noise.
A Bayesian covariance estimation project built after sample covariance made mean variance portfolios unstable.
read projectresult#194 algo, top 1.4% of 18,800 teams.
the hard part was adapting strategy across rounds under sparse feedback.
A multi round algorithmic trading competition where simple models, inventory discipline, and careful iteration mattered more than fancy machinery.
read project