← Frozen TRS dashboard · View experimental results

Experimental configuration · paper-only execution

Backtest controls

Configure a reproducible experiment. Sign in to submit a paper-only historical backtest. The frozen benchmark is not modified.

Strategy and stock universe

Selections follow the frozen benchmark purchase-priority order. Per-symbol historical coverage still requires server validation.

Historical period

Capital and costs

Configured costs and position limits apply only to experimental simulations. Frozen benchmark defaults remain unchanged.

Run your experiment

A date or stock-universe change requires a matched-period, matched-universe baseline for fair comparison. Experiments must record code and dataset versions.

Review settings to validate.

Submission saves a nonsecret experiment receipt, then follows the exact GitHub run automatically. Results appear below when complete. After a browser interruption, re-enter the access token to resume; no new experiment is required.

Experiment progress

No experiment selected.