Skip to main content

Blog

Backtesting research and guides — page 2

Petrol and aqua artwork of a staircase descending into darkness beside a pale haze of light

Aug 11, 2026

Four levels of backtest honesty

An EMA test moved from +34% frictionless to -59% with its fullest cost and fill model. Review the assumptions, trade-count changes, and limits.

Petrol and aqua artwork of a single sharp peak rising from a grainy plain against a luminous horizon

Aug 11, 2026

We overfit a strategy on purpose

The best of 72 Bitcoin EMA configurations gained 17.96% in sample, then lost 14.32% out of sample. See the search, split, and limits.

Teal and black glitch artwork of digital price data dissolving into a horizon

Aug 6, 2026

We tested ChatGPT's trading strategies

Four ChatGPT-labelled Bitcoin strategies lost money in a dated backtest. Review the recorded results, missing settings, and limits of the experiment.