Duration drift is the movement of the duration distribution for a matter or a book since underwriting or the last review. It is the honest record of what changed, preserved against the original estimate.
Drift is calculated by comparing the distribution at one cutoff date to the distribution at an earlier one. A rightward shift means expected delay has grown, and a narrowing means evidence has reduced uncertainty.
Preserving the original estimate is what makes drift measurable and what makes accuracy auditable later. An estimate that is overwritten cannot be compared with what happened.
Duration drift is not an error. A well-run estimate should move when evidence arrives, and the drift shows that it did.
Working a file that turns on this term?
Working through a diligence process?
Institutional partners evaluating a position against this platform's outcome and duration models are welcome to reach out directly.
