TradeLab research
Real tests, transparent assumptions, and practical lessons for turning trading ideas into systems you can inspect.
Research library
Review position sizing, leverage, concentration, and combined open risk without hiding assumptions inside the headline.
Separate development from evaluation, freeze the rules, and protect reserved history from repeated tuning.
Stress commissions, spreads, slippage, market impact, and turnover before interpreting the result.
Audit when data existed, when the signal became final, and which price could actually have been available for execution.
There is no universal minimum. Judge the count alongside market coverage, independence, concentration, and stability.
Validate the specification, implementation, and evidence, then classify the result as pass, investigate, or reject.
Compare what each stage tests, which should come first, and how to diagnose disagreement without confusing simulation for live execution.
Six required decisions and a reusable template turn a market opinion into a strategy specification another person can reproduce.
A result that rejects a weak idea, catches a configuration error, or prevents false confidence has done useful work.
A 20/50 SMA idea produced zero daily trades in a seven-day sample and 51 noisy trades on one-minute data. The failure was the lesson.