By Yair Knijn · May 14, 2025

Premium leakage hides in the endorsements nobody priced: the mid-term adjustments that quietly cost a fleet

The controller signs off the fleet renewal once a year and treats everything in between as the broker's job. That is the wrong assumption. The renewal is a snapshot; the cost lives in the twelve months of changes after it. Every vehicle sold, added, or re-rated mid-term is a priced event, and if those events only exist as a reply-all thread, the schedule the insurer bills against stops matching the fleet you run.

I watched one mid-sized fleet pay premium on two vans written off eight months earlier while under-declaring three new high-value tractor units. Net leakage in both directions, and nobody could say when it started because there was no ledger to point at, only inboxes.

What a mid-term adjustment actually changes in the priced schedule

A mid-term adjustment is not an admin note. It is a re-rating. Insurers handle it as either an AP (additional premium) when exposure goes up or an RP (returned premium) when it goes down, calculated pro-rata against the remaining term. Add a vehicle in month four and you owe roughly eight-twelfths of its annual rate; remove one and you are owed the unexpired balance back. That math only runs if the change reaches the schedule.

Most motor fleet programs also carry a quarterly declaration where you confirm the vehicles actually on cover and the premium is trued up. Between those checkpoints, the policy runs on whatever the insurer last knew. The declaration is only as good as the events you fed it.

Leakage in both directions: paying for ghosts, under-declaring additions

Leakage is rarely fraud and almost always omission, and it cuts two ways. A vehicle leaves the fleet and the RP never gets requested: you keep paying for a ghost. A vehicle joins and the AP never gets declared: you are under-insured on an asset that is on the road right now, and if it has a loss, the insurer can average the claim or query cover entirely.

The dangerous part is that the two errors can net to nearly zero on the invoice. Total premium looks roughly normal, so nothing trips an alarm, while underneath you are simultaneously overpaying on ghosts and carrying uninsured exposure on new units. A flat top-line number hides both.

Why email-based endorsements are the leak

Email has no state. A request to add a vehicle and a confirmation it was rated are two messages that may never get connected, and there is no field that says this endorsement is still open. Three failure modes show up every time:

• The change is sent but never acknowledged, so it stays unpriced and you do not know it.
• The change is acknowledged in a thread the finance team never sees, so it never hits the books.
• The effective date in the email and the date the insurer applied differ, and the pro-rata is quietly wrong.

None of these are exotic. They are the default outcome of running a priced process through an unstructured channel. The fix is giving every endorsement an identity and a status, not a tidier inbox.

An endorsement ledger that reconciles to the insurer's billing every month

Treat each adjustment as a tracked record: vehicle, change type, effective date, expected AP or RP, and a status that stays open until the insurer's document confirms it. Then reconcile that ledger against the insurer's billing every month, not at naverrekening when the year is already closed and the dispute is twelve months stale. The question a controller should be able to answer on any given day is simple: which endorsements are open, and what should the next invoice be.

That is the discipline a fleet program needs to make leakage visible instead of structural. FleetLedger keeps the endorsement ledger as the system of record, every add, removal, and re-rating carrying its priced impact and open-or-closed status, and reconciles it to the insurer's billing each cycle, so the premium you pay matches the fleet you actually run.