Research
Quantitative research
Reproducible studies, with their limits in plain sight.
Each study publishes the method, the derived data and the code needed to reproduce it. Weak or negative results stay on the page: they are part of the result too.
Studies
Calendar anomalies · Temporal stability
The Ghost of an Anomaly
How much of the turn-of-the-month premium still belongs to today's market, with independent replication and a clear split between exploratory and confirmatory work.
MethodHAC/Newey-West10-year rolling windowsMatched-sample replicationStata + PythonRead the studyReinforcement learning · Evaluation
Realistic evaluation of DRL portfolio claims
What survives of a DRL allocation claim once you add costs, explicit cash, matched benchmarks and statistical validation.
MethodWalk-forwardBootstrapTransaction costsMatched benchmarksRead the study
Educational and research content. It is not investment advice or a recommendation to trade.