Why a Reserve Moves by Seven Figures "Overnight," and Never Actually Does

Nothing changed overnight. The reserve had been wrong for a while. The actuarial review just measured the gap for the first time.

The actuarial review lands, and a reserve everyone assumed was stable moves by seven figures overnight. Someone on the finance side asks what changed. The honest answer is almost always the same: nothing changed overnight. The reserve had been wrong for a while. The actuarial review just measured the gap for the first time.

That distinction changes what you're actually diagnosing. A reserve that moves overnight because of real new information, a verdict, a settlement demand, a diagnosis that changes the prognosis, is doing its job. A reserve that moves seven figures because eighteen months of small, uncorrected drift finally got measured is a different problem, and a far more common one than most self-insured programs want to admit.

How the drift happens

A claim comes in, and someone sets an initial reserve on early information: the incident report, a rough sense of exposure, maybe a first look at medical records. That number is reasonable at the time.

Then the file develops. Treatment runs longer than expected. A specialist gets involved. Litigation gets filed, and discovery turns up facts the initial reserve never accounted for. Each development should nudge the reserve. In a well-run file, it does, in small increments, tracked and documented as they happen.

In a file without that discipline, none of those developments move the reserve at all. Your adjuster manages the file day to day and has no particular reason to revisit a number set months ago that hasn't caused a problem yet. Nobody flags the growing gap between what the file looks like and what the reserve says it's worth, because flagging it isn't anyone's specific job. So the reserve sits there, technically wrong and practically invisible, until an actuarial review measures every file at once.

That's the moment the seven-figure move happens. It looks sudden because measurement and drift got separated. The drift ran gradually. The measurement did not.

Why this hits long-tail programs hardest

For self-insured healthcare providers and public entities, this is an expensive way to learn about a problem. Your reserves usually attach to long-tail claims, professional liability, general liability, exposures that take years to resolve and give small drift plenty of time to compound into something large.

A claim that sits open for three years with a reserve nobody revisits accumulates error at whatever rate the file actually develops, not at some steady, predictable pace. Nobody's watching closely enough to know that rate until the actuary reports it.

The actuary isn't your problem

Your actuary is doing the job correctly, measuring what's actually there. Your fix is closing the gap between when a file develops and when someone revisits the reserve to reflect it, so that by the time the annual review lands, it confirms numbers that were already tracking reality instead of correcting numbers that drifted away from it.

That takes a claims operation built to flag reserve-relevant developments as they happen, not one that waits for a scheduled review to notice them. It takes someone with the authority and the habit to ask, on a live file, whether the number on record still matches the file in front of them, well before your actuary asks the same question with less patience.

Want a second set of eyes on your reserving discipline? That's a cheaper conversation in month six than in year three.