The Great British Write-Off

Three times now this series has watched Derek pay too much: for a No Claims Discount that discounted nothing, to a comparison website that showed him a fraction of the market, and for the sin of admitting he pays monthly. Each was a toll levied at the door of entry. This instalment concerns the exit, through which Derek has just been escorted by a bin lorry on the Kingston bypass. His car is a total loss. He is about to learn what nineteen years of premiums actually bought him.

An insurance policy is, when every clause has been read, a single promise: that money will change hands at the worst moment of a stranger’s year, and that the sum will be fair. Derek honoured his half for nineteen years. He is about to discover how the other half is kept.

That an insurer pays too little is no thesis to detain a serious person. Of course it does; a claims handler who paid too much would be seeking other work by Friday. The story is two facts stranger than that. First: the central manoeuvre of the written-off car is one the regulator has already met, already understood, and already, in a neighbouring aisle of the same shop, banned outright. Second: the market has been arranged, with no conspirator anywhere in it, so that the insurers who would have treated Derek honestly are precisely the ones he was prevented from finding.

Consider what he bought. An insurance contract has two halves, unequally administered. The premium is calculated to the penny, whole actuarial departments existing to make that one number exact. The payout is not. The policy promises the car’s “market value,” and the FCA, asked what that means, says it expects “a fair estimate” of it, and then stops. No rule says which valuation guide a firm must use, how many, or how far a settlement may fall below one. So there is a contractual promise to pay a number and no rule governing how the number is found; the number is found by the party who must pay it. Having paid into the arrangement faithfully for nineteen years, the first offer that reached Derek was an insult to the intelligence of a child.

The Thumb on the Scale

How is the number found? Three trade guides exist: Glass’s, CAP HPI and Percayso. They disagree, producing a spread on any given car, and that spread is the game. The insurer picks the guide; where the guides diverge, it picks which figure to call the truth; then, from the lowest plausible start, it deducts.

The reader of this series will feel a flicker of recognition: the deduction is the trick of the first instalment, where the No Claims surcharge was loaded onto a premium already priced for youth and postcode, charging one risk twice. The guide price already reflects the condition of a car of that age and mileage; that is what a guide is. Yet the FCA found firms deducting again for “wear and tear,” and lopping a flat twenty per cent off merely because a car was once repaired. The catalogue of deductions an a loss adjuster may reach for are not short.

Worse is the timing. The FCA found firms making first offers below market value, raised only when the customer pushes back. A deliberately light first offer; a fuller second one, released only to those who complain. It converts a contractual entitlement into a prize for assertiveness. And the errors run one way: they cluster below the guide, never above. An error that points one way every time is not an error. It is a policy.

The Twin That Was Already Outlawed

Now the first of the two strange facts.

In January 2022 the FCA’s pricing reforms came into force, and the industry lost a practice it had cherished for a generation: price walking, the loyalty penalty. An insurer quoted a keen price to win you, then “walked” the renewal upward year on year, relying on nothing but your inertia. The customer who did nothing paid more for doing nothing. The customer who telephoned and threatened to leave was discovered to qualify for a “retention” price, which is to say the price he should have had all along.

Observe what the regulator did. It did not issue guidance. It did not publish a roadmap. It banned the practice. It made it unlawful to charge a renewing customer more than an equivalent new one. The FCA had looked at a market where the price you paid depended on your willingness to complain, and ruled that such a market could not be permitted to exist.

Now hold that against the thin first offer. At renewal: the customer who does not complain overpays; the one who complains is made whole. At the claim: the customer who does not complain is underpaid; the one who complains is made whole. There is no daylight between those two sentences. They are not analogous practices. They are one practice: the same cross-subsidy, from the trusting to the assertive, running on the same engine of inertia. The only difference is which side of the contract the discretion sits on, the price paid in or the money paid out. A pound walked off a renewal and a pound shaved off a claim are the same pound. Derek’s bank balance cannot tell them apart.

If anything, the tolerated version is the worse. The walked customer is free; he can shop next year. The lowballed customer is captive, his car on a recovery truck, needing the money this week to get to work, very often in exactly the “vulnerable circumstances” the FCA’s own warning named. Price walking exploited laziness. The lowball exploits an emergency. One was banned outright. The other received three documents and a request that firms consider the findings.

So the question is not why insurers underpay. It is this: the regulator has already proven, by law it wrote itself, that it regards “your outcome must not depend on whether you complain” as a principle worth abolishing practices to defend. It has the template. It is the author of the precedent. Why has the lowball not been served the same notice? Banning price walking cost the FCA a headline and the industry some margin. Banning the lowball would mean telling several of the largest insurers in Britain that a revenue line they have run for a decade is closed. The distance between a principle and a policy is sometimes only the size of the firm you would have to enforce it against.

A Free Option

The regulator’s defenders will say it has not been idle, and they are right. By late 2025 the FCA had announced that some 270,000 motorists would share £200 million in compensation for total-loss claims settled below what fair handling required.

The industry says the system worked. Look harder. £200 million across 270,000 motorists is an average underpayment near £740 a head: a decade-long, systemic practice the regulator now concedes in its own press release. Yet across four warnings since 2022, a review and a national petition, not one insurer has been fined. The money returned is redress, the repayment of sums always owed. Restitution is not deterrence.

Run an option trader’s eye over it: The lowball is an option the industry has written in its own favour. Its upside is the margin shaved from every uncomplaining claimant, banked yearly, compounding. Its worst case is that, years later, a regulator makes the firm hand back the money it always owed, without meaningful interest and without a penny of penalty. An instrument whose worst conceivable outcome is “return exactly what you took” has no downside at all — its expected value is simply its entire upside. No competent board would decline to write it. The lowball is not a scandal the industry failed to notice. It is a trade it has correctly priced.

Where the Good Insurers Went

Now the second strange fact, the one that should keep a regulator awake.

Some insurers pay fairly. This is not a consoling fiction: the same FCA review that caught the lowballers also found firms settling, on average, close to the guide prices.

Splendid. Now name one. Derek cannot. Many industry insiders cannot. The reason is the buried hinge of the whole affair: the quality of an insurer’s claims handling is invisible at the moment you buy. It is not a number, not a column on the comparison site, nowhere in the price. It is a fact about how a company will behave on the worst morning of a year that has not yet happened, and nothing on Derek’s screen can show it to him.

Here the four episode about Derek lock together. Derek does not choose an insurer; he chooses a comparison website, and the website ranks by price. Claims fairness has no rank, because it cannot be seen. The honest insurer and the lowballer sit side by side in the same typeface, separated by one figure: the premium. And the honest insurer’s premium is necessarily the higher. A firm that pays the full worth of every written-off car carries a heavier claims cost than one shaving £700 off each. Heavier cost compels a higher premium; a higher premium earns a lower rank; a lower rank falls below the fold of an eye trained over three renewals to drop to the cheapest line and stop.

So the comparison site, Derek’s “champion,” does not merely fail to reward the honest insurer. It punishes it. The fair-paying firm is undercut and out-sold by the rival that financed its keener price with money it has already resolved to withhold at claim time. Economists have a name for this: where buyers cannot tell quality apart before purchase, the good product cannot command its price, the bad undercuts it, and the good is driven from the shelf. They call it the market for lemons. Nor does the process stop. An honest insurer that learns its honesty costs customers and earns nothing has three roads: pay fairly and bleed its book to cheaper rivals; be bought by one of them and have its claims philosophy quietly retired; or study the lowballer’s combined ratio and instruct its own adjusters to start thinner. Two of those roads end with one fewer honest insurer in Britain. None ends with more.

That is where the good insurers are going: not hounded out by villainy, but competed out, by an engine that could not price the one thing that mattered and a clientele of Dereks never shown it. The cruelty is exact. The cheapest quote and the insurer that would have paid in full are very nearly opposite things, because the cheapness was manufactured out of the lowball. Derek optimised, faithfully, every year, for price. He was, the whole time, optimising against himself.

The Willing Victim

Derek’s car is gone. The meerkat, plush and button-eyed, sits on the mantelpiece where it has sat for three renewals, and watches the post. An envelope arrives. The cheque inside is plausible, not absurd, close enough to right that a tired man who has spent his week arranging lifts to work feels mostly relief that it is ending. Derek did not write the letter, he banked the cheque and said nothing. He supposed he had less grounds to argue than in fact he had. What passed between he and the loss adjuster was never a negotiation; it was a quiet, one-sided tax on the dignity of a man too shaken to haggle.

He was underpaid by something near £700, and will never know it, because the whole apparatus was built by people confident he would bank it. But the £700 is not the true measure of what was done to him. Somewhere on that comparison panel, three renewals ago, sat an insurer that would have paid this exact claim in full, on the first cheque, no letter required. Derek never saw it. It carried the modest price premium of a firm that meant to keep its promises, and his eye slid past it to the bargain. He did not weigh the honest insurer and reject it. He was never shown it. And the regulator, holding the very instrument that cured this disease on the pricing side, has chosen to leave the claims side untreated.

This series has been accused of being hard on Derek. It never was. Derek did everything asked of a good customer: he shopped around, paid on time, did not exaggerate, trusted the contract. Each of those virtues was, somewhere in the machine, repriced as a cost to him — and the hardness in these pages was only ever aimed at an industry that catalogued those virtues and read each one as an opportunity.

The meerkat says nothing. It never has. It was, of course, paid for by Derek. He thinks, even now, that he got rather a good deal.