A verifiable track record is a sequence of recorded results that can be tied to a specific researcher and a specific time, then checked by someone else.

A verifiable track record is a sequence of recorded results that can be tied to a specific researcher and a specific time, then checked by someone else. It differs from a screenshot because a screenshot is only a static display of output, while a verifiable record preserves the information needed for independent audit.

For quantitative research, that distinction matters because strategy results are produced through many iterations of testing, selection, and revision. Sonar describes a workflow in which research moves from development to publication with structured records around the process rather than isolated images of results. In that framework, a track record is stronger when each result is attached to researcher identity, time, and an auditable history of how it was generated and reviewed.[1]

The mechanism is auditability. To verify a track record, an external reviewer needs more than the final metric. They need a record of when a result was created, who created it, and what analytical path led to it. Sonar’s backtest overfitting audit tool is designed to produce this kind of review context by auditing backtests for selection bias and overfitting risk rather than accepting headline performance at face value.[2]

One element in that audit trail is the deflated Sharpe ratio. Sonar defines the deflated Sharpe ratio as a Sharpe ratio adjusted to account for non normal returns, sample length, and multiple testing. Its purpose is to estimate the probability that an observed Sharpe ratio exceeds what could arise by luck after many trials.[3] In practice, that means a track record is more informative when the reported result is accompanied by an audit record showing whether the result remains statistically credible after accounting for repeated testing.[2][3]

This is why a verifiable track record is not just a list of outcomes. It is a chain of evidence. Identity links the result to a researcher. Time links the result to a particular moment in the research process. Audit records link the result to the methods used to evaluate robustness. Together, those elements allow another party to inspect the record instead of relying on an image or summary.

Claims about cryptographically signed logs, tamper evident ledgers, immutable timestamps, or third party verification reports are not supported by the cited sources.

Covered in depth in the From research to publishing pillar hub.