Systematic trading strategies, backtested properly.
Each article takes one well-known strategy and gives it the treatment it rarely gets: the exact rules, a full metric sheet against a benchmark, and the robustness tests that decide whether the result means anything — parameter sensitivity, in-sample versus out-of-sample, and walk-forward. Every backtest runs on the same open-source engine and free data, and every result file is committed to GitHub, so you can reproduce or refute any number here.
New to this? The glossary defines every term used across the series in plain language.
Written by Abhinav Taduka. Code at github.com/AlgoStrategyLab.
Faber's moving-average filter over 1987–2026, head to head with the momentum filter from article #1. They make the same call 88% of the time, it switches 48 times against momentum's 27, and every moving-average length longer than the famous 10 months beat it while trading less.
Gary Antonacci's Global Equities Momentum over 1997–2026. Before the book was published it beat the S&P 500 by 5.1 points a year; after, it trailed by 5.4. The trend filter does all the work, optimising the lookback makes it worse, and the risk reduction survives what the returns do not.
Donchian channel breakout · RSI(2) mean reversion · cointegration pairs trading · volatility targeting · cross-sectional momentum · the variance risk premium