LTV drift alerter

The threshold breach happened in March. You found it in June.

Why LTV alerting should follow the evidence, not the reporting cycle.

Propalt Team · For secured lenders

A loan crosses your LTV policy boundary the day the market moves, not the day you next run the report. On a quarterly cycle the gap between those two dates can stretch to weeks. During that window the loan is misclassified in every internal view that depends on the mark, and the first time anyone acts is after the breach has aged. The calendar decided when you looked, and the market did not consult the calendar.

To be fair to the quarterly cycle, it exists for a reason. Revaluing an entire book was expensive, so lenders batched the work into scheduled runs and lived with the lag. But the trigger for review does not have to be the reporting date. It can be the event itself: the asset moving across a threshold, or a comparable transaction nearby that changes what the asset is worth. Watch the evidence, and the alert arrives when the risk does.

Two triggers, one alert

There are two distinct events worth alerting on, and they are easy to conflate. The first is the asset crossing a threshold on its own revaluation. The second is a nearby transaction, a comparable sale that moves the local evidence and therefore the mark on your loan, even though nothing about the loan itself changed. A house price index at fine geography lets you catch the second kind, where a single sale down the road quietly repriced your collateral.

The table is illustrative. What matters is that the second and third rows are driven by evidence outside the loan, which a scheduled revaluation on that single asset would have missed until the next cycle.

From quarterly to daily, without a bigger team

Moving to event-driven alerting sounds like more work, and done manually it would be. Automated, it is less. Instead of a quarterly exercise that reprices everything and produces a long exceptions report to triage, you receive a short daily list of the specific loans where something actually changed. The volume of alerts tracks the volume of real movement, not the size of the book.

For the risk function this changes what you can evidence. A breach caught the day the evidence appeared, with the triggering transaction attached, is a cleaner story for internal audit and model validation than a breach discovered at quarter-end with no record of when it truly began. Under PRA expectations that collateral values stay current, being able to show the date a loan crossed the line, and why, is worth more than a tidy quarterly report that was already out of date when it printed.

The market does not breach your thresholds on reporting dates. It breaches them on transaction dates.

Follow the evidence, not the calendar.

Try the LTV drift alerter → · propalt.ai

Valuation and price data is drawn from HM Land Registry and the Propalt intelligence layer. Figures shown are illustrative. This article is general information for lending risk professionals.