Time to outcome
Time to outcome measures how long a distressed property takes to reach resolution — from a defined starting point to the event that ends the window.
For Minnesota foreclosure the starting point is the sheriff's sale, because that is what every statutory deadline counts from.
What the data shows#
Across resolved windows with a dated outcome:
| Outcome | Observations | 25th pct | Median | 75th pct |
|---|---|---|---|---|
| Owner exit | 109 | 50 days | 106 days | 165 days |
| Resale by buyer | 109 | 187 days | 258 days | 320 days |
Both measured in days from the sheriff's sale.
Owners act early, not at the deadline#
The finding worth sitting with.
A six-month redemption period runs 181 to 184 days depending on the calendar. The median owner exit occurs at 106 days — well inside it, and the seventy-fifth percentile at 165 days is still short of expiry.
That is not what the deadline structure would predict. If redemption were mainly about assembling money against a fixed date, outcomes would cluster near the end.
They cluster in the middle, which suggests most resolutions are sales completed during the period rather than last-minute redemptions. An owner with equity who sells has to allow weeks for a transaction to close, so they start early and finish before the deadline rather than at it.
The practical implication for a homeowner is the same either way: the useful window is months one to four, not month six.
The resale timing follows the statute#
The median resale at 258 days sits about 76 days after a six-month period expires around day 182.
That is the certificate holder taking title, preparing the property, and listing — roughly two and a half months of holding cost between the redemption period closing and the property being sold on.
For anyone underwriting a sheriff's sale purchase, that is the real holding period: six months waiting, then a further two to three months to exit. Not the six months the redemption period suggests.
What is excluded#
Windows still running, and windows where the outcome could not be dated.
Only uncensored observations — those with a known event and a known date — are in the medians above. Censored windows, where the period ended without an observable event, are handled by the survival curves rather than by median arithmetic, because including them would bias the medians downward.
Reading it honestly#
These are medians on 109 observations each, in the counties currently tracked, over the period tracked.
They will move as more windows resolve. They are published with their n so that they can be checked and so that a later figure can be compared against this one rather than replacing it silently.