Importar datos de mercado

01 Use clean inputs

Backtests are only as credible as their input data. Confirm timestamps, missing bars, split adjustments, and symbol conventions before comparing results.

• 1Use consistent timezone handling across files.
• 2Avoid mixing adjusted and unadjusted data in one comparison.
• 3Check for large gaps before running optimization.

02 When results look wrong

Most surprising backtest results come from a mismatch between expected data shape and actual data shape. Inspect a sample of raw rows before changing strategy logic.