Research
How we think about strategy development, validation, and the difference between a backtest and a real edge.
Philosophy
Every strategy is validated through walk-forward backtesting before consideration. We model realistic transaction costs, account for survivorship bias, and apply deflated Sharpe ratio corrections to guard against overfitting. Strategies that cannot survive out-of-sample testing with costs are discarded regardless of in-sample performance.
Academic foundations: Fama and French on factor models, Jegadeesh and Titman on momentum, Avellaneda and Lee on statistical arbitrage, Bailey and Lopez de Prado on backtest overfitting.