By Yair Knijn · March 3, 2026
Counted twice: how replacement vehicles overstate the fleet and inflate every premium calculation
The fleet manager who runs the schedule is conscientious. A lease comes off, a replacement comes on, and the new registration gets added the day the keys arrive. Adding the new vehicle is easy to remember because someone is driving it. Retiring the old one is the step that slips, because by then it is parked at the dealer and nobody is chasing the paperwork.
The wrong assumption is that a replacement is one event. It is two: an addition and a retirement, separated by a window where both vehicles legitimately sit on the books. Treat it as one and you keep adding without subtracting, and the schedule grows past the fleet you run.
Why replacement, not addition or removal, is where fleet counts go wrong
A pure addition is self-correcting: you add a vehicle because it exists. A pure removal is forced on you, because the lease ends and someone wants the premium credit back. Replacement is the one event that can be half-done and stay half-done, because the half you did, the addition, feels like the whole job.
So the errors do not scatter across the fleet. They concentrate on replacements, and each one that overlaps without a clean retirement adds a phantom vehicle nobody drives and nobody questions until naverrekening.
The overlap window: when both vehicles really are on the books
For a few days or a few weeks, the overlap is honest. The outgoing van is still registered and still insurable; the incoming one is on the road. Pro-rata premium is charged on time on risk, so a vehicle on the schedule for nine days of a month is a real, chargeable nine days. The problem starts when the outgoing vehicle never leaves: its overlap window has no closing date, and it sits at full risk for the rest of the policy year.
One replacement done this way is a rounding error. Twenty a year, each leaving a permanent ghost, is a fleet that reads materially larger than the metal in the yard. The schedule still looks orderly, every line has a plausible registration, and nothing flags, because nothing in a flat vehicle list knows that two of those lines describe the same swap.
How a systematic double-count inflates premium and distorts loss ratio
The premium hit is the obvious one and the smaller one: you pay time on risk for vehicles that carry no risk because they are gone. The expensive damage is to the denominator. Loss ratio is claims over premium, and the insurer reads that ratio at renewal to set your rate. A fleet that books a better loss ratio in one year carries that position into the next, so an inflated schedule corrupts the exposure base your loss experience gets normalised against.
- Claims per vehicle looks artificially low, because the base is padded with ghosts that can never have a claim.
- Premium per real vehicle is too high, because you funded the ghosts.
- Frequency benchmarks rank you against a fleet size you reported wrong, so you land in the wrong peer band.
None of this is in your favour. A loss ratio built on an overstated fleet costs you twice: once on the premium you overpaid, again on a renewal priced off numbers you handed over yourself.
A replacement workflow that retires the outgoing vehicle as cleanly as it adds the new one
Make replacement a single transaction with two required legs. You cannot add the incoming vehicle without naming the outgoing one and stamping its end date. Reconcile the schedule against RDW registration status continuously, so a kenteken that has been deregistered or transferred cannot keep accruing time on risk in silence. The bordereaux you send should already reflect closed overlaps, not a count you reconcile in a panic the week before naverrekening.
That is the discipline a fleet program should enforce by construction: every addition tied to a retirement, every overlap window with a closing date. FleetLedger treats replacement as the two-sided event it is, so the fleet you insure stays the fleet you run, and the loss ratio you renew on is one you can defend.