Writing
Notes on backtesting, risk, execution and prediction markets: what works, what overfits, and how to tell the difference.
Building prediction market models
Latest ·
Prediction-market models: reading prices as probabilities, starting from base rates, checking calibration and sizing an edge after fees.
Read · 6 min read
Kelly sizing for traders
The Kelly criterion for sizing trades: the formula, why half Kelly keeps 75% of the growth, and how to use it when your edge is only an estimate.
Risk Management · 6 min readMarket making in crypto
How crypto market makers earn the spread, why inventory and adverse selection eat into it, and how the Avellaneda–Stoikov model sets quotes.
Crypto · 6 min readTWAP, VWAP and beyond
How TWAP, VWAP, percentage-of-volume and implementation-shortfall algorithms work, and how to measure what your execution really costs.
Execution · 6 min readHow not to overfit a backtest
Why great backtests fail live, and a seven-step routine to catch overfitting: hold-out data, walk-forward tests, luck benchmarks and stress tests.
Backtesting · 8 min readThe Sharpe ratio, beyond the basics
What the Sharpe ratio measures, how uncertain a single backtest figure really is, where it misleads, and which numbers to read next to it.
Risk Management · 6 min read
New posts by email
One email when something is published. Unsubscribe any time.