The Premium That Bought Nothing: When ATE Insurance Doesn’t Pay

The Premium That Bought Nothing: When ATE Insurance Doesn’t Pay

£19.5M of after-the-event premiums bought policies that never paid out on a single claim, and the funder whose money paid for them is now suing the insurer to get them back. In the same month, seven claimants learned that their £16.2M of the same cover was facing a £34.5M costs bill. In a funded claims model, the insurance policy isn’t a safety feature bolted onto the collateral. It is the collateral.

---

I don’t own an after-the-event policy and never will — there is no loser-pays costs rule in the US courts where most of my exposure sits. My interest in this is narrower, and I think more useful. I’m a limited partner in private litigation-finance funds, and what arrives each quarter is a case name, a procedural status and a cost basis. I have never once been shown an insurance schedule. So whenever a structure’s downside is described to me as covered, I’m being asked to take the most consequential term in the deal on trust, and this month produced two public demonstrations of what that trust is actually worth.

---

The Policy Is the Collateral

England and Wales run a loser-pays costs rule: lose, and you owe the other side’s legal bill, which in heavy litigation can dwarf the amount in dispute. After-the-event insurance is the product sold against that exposure — a policy bought once a dispute already exists, covering the claimant’s liability for adverse costs, with the premium itself usually deferred and contingent so that it only becomes payable on a win.

For a claimant, that policy is what makes a no-win-no-fee arrangement genuinely no-risk rather than merely no-fee. For a funder, it is something else entirely, and this is the part that gets glossed over: the policy is the security.

Follow the cash through a high-volume consumer-claims model. A lender advances disbursement funding to a claimant firm, and that funding pays the case expenses — including the ATE premiums. If the claims win, the defendant pays the costs and the loan is repaid. If the claims lose, the policy pays the expenses and its own premium, and the loan is repaid anyway. Both branches end at repayment, which is precisely what makes the model financeable. It also means the lender is not really underwriting a portfolio of lawsuits. It is underwriting a portfolio of insurance policies, bought with its own money, on which its entire downside recovery depends.

That layer is being built out rather than wound down. Ignite Specialty Risk, one of the specialist litigation insurers, has written more than US$2 billion of litigation capital since launching in 2022 — including US$360 million of US policies in 2024 sitting over litigation assets valued above US$5 billion — and opened Australian operations this month to sell ATE, litigation risk and contingent risk cover into a market that previously had one serious provider. More claims are being wrapped, in more jurisdictions, every year.

---

Failure Mode One: The Policy That Never Responded

SSB Law ran cavity-wall insulation claims at volume — homeowners suing over defective insulation, taken on under conditional fee agreements, disbursements financed by lenders, adverse-costs risk placed with an ATE insurer. The firm’s client-care letters, as described in the court filings, told each client that the arrangement involved a financial institution to lend the money for expenses and a legal expense insurer, and that if the claim was unsuccessful the policy would pay all the expenses and its own premium. Clients were told they would not be out of pocket personally.

None of that held. The claims did not succeed, ATE insurers in cavity-wall cases repudiated their policies, and losing clients who had been told they would pay nothing found successful defendants and their insurers enforcing costs orders against them. SSB collapsed, and its fallout is one of the reference points for the Solicitors Regulation Authority’s current crackdown on high-volume consumer claims firms.

What is new this month is who is suing whom. Katch Fund Solutions — part of the same Katch operation whose open-ended fund I wrote up in July as the one vehicle of three that survived a long delay in UK consumer claims, because it owed nobody money on a fixed date — was owed £63M by SSB when the firm went down. Its exposure here is a different book from the motor-finance one in that story: disbursement lending to SSB against cavity-wall claims. Katch has now taken an assignment of SSB’s own claim from the administrators and issued proceedings in the Commercial Court against Stonefort Insurance S.A. (formerly Builders Direct S.A.), a Luxembourg company, seeking restitution of £19.5M plus interest.

£19.5M is what Stonefort was paid in premiums from 2020 onward. Against those premiums, according to the claim form, not one of the claims succeeded or qualified as a win under the terms of the clients’ conditional fee agreements. Katch’s case is unjust enrichment: the insurer retained premiums for policies that never delivered the protection they were intended to deliver.

Set that against the structure and the position is worse than a bad recovery. The lender’s capital bought the policies; the lender’s repayment on the losing branch ran through those same policies. When the policies didn’t respond there was no winning case, no insurance recovery, and an insolvent borrower in between. A loan book that looked fully covered turned out to be an unsecured bet on one insurer’s willingness to pay — and the claim had to be purchased out of an administration, because by the time anyone could sue the insurer, the firm holding the contractual relationship no longer existed.

Katch’s solicitor, Erich Kurtz of Hugh James, stated the diligence lesson more plainly than funder marketing ever does:

“An ATE policy alone is not sufficient protection. Underwriting quality, satisfaction of conditions precedent, and the financial standing of the insurer all require independent scrutiny before capital is deployed.”

Three separate checks in one sentence, none of them the same as reading the claim file, and none of them visible in the word “insured.”

Two caveats, because they matter. The proceedings are at an early stage, Stonefort has not yet answered the claim, and nothing has been established against it. And there is a reading of these facts in which the insurer is entirely in the right — if the conditions precedent in those policies were never satisfied, an insurer declining to pay is the policy operating exactly as drafted. I’ll come back to that, because it is the strongest argument against the way I’m framing this.

---

Failure Mode Two: The Policy That Was Too Small

The second failure requires no allegation at all, because the arithmetic is already public. Seven claimants — including the Duke of Sussex, Baroness Doreen Lawrence and Sir Elton John — brought a privacy claim against Associated Newspapers Limited, publisher of the Daily Mail, alleging unlawful information gathering. Mr Justice Nicklin dismissed the claim in its entirety on 7 July 2026. Both sides accept the claimants must pay ANL’s costs.

The claimants held a combined £16.2M of ATE cover against adverse costs. ANL has reported legal spend across a four-year case and an eleven-week trial of £34.5M — more than £18.6M above its own approved budget. The gap between the cover bought and the bill presented is roughly £18.3M, and whatever survives assessment above the policy limit falls on the claimants personally.

The live question is standard basis versus indemnity basis, and it decides the size of that bill. On the standard basis, proportionality constrains what a winner can recover, so a defendant that overshoots its own budget by £18.6M does not simply hand the overshoot to the losing side. On the indemnity basis, that constraint comes off. So the honest way to state the gap is that £18.3M is the shortfall measured against the costs being claimed, and the final figure depends on a ruling nobody has made yet. It could land well below that.

The structural point survives whatever the assessment produces, and it is more uncomfortable than the first story. Nobody repudiated anything here. The insurer will presumably pay its £16.2M in full. The policy did exactly what it said it would do. It was simply sized, at the outset, against an estimate of what the opponent would spend over the following four years — and the opponent spent more than half again its own approved budget. There is no policy you can buy today against a number the other side chooses later.

---

Four Ways the Wrapper Fails

Set this month beside the UK funder collapses of the last year and the failure modes of adverse-costs cover sort into four. They are genuinely independent of one another, which is the whole problem:

The four don’t share a fix, and that is what makes this structural rather than two unlucky files. Reading the policy wording catches the first two and tells you nothing about the third. Sizing the limit properly means forecasting your opponent’s legal budget, which is not a document anyone gets to read. Underwriting the insurer is a credit assessment, not an insurance one. A diligence process that checks one of these and moves on has covered a quarter of the ground.

---

What a Limited Partner Can Actually See

Here is why this travels beyond the English costs rule. Insurance doesn’t eliminate case risk; it converts it. What used to be a question about a docket becomes a question about a contract and a counterparty — and those are the two things an outside investor is least equipped to inspect, because neither the policy nor the insurer appears anywhere in the reporting. I can tell you the procedural posture of cases I’m exposed to. I cannot tell you which of them carry cover, for how much, subject to what conditions, or written by whom.

The American version of this problem is already on my own site. When I took apart the $213M verdict that a $60M judgment preservation policy couldn’t rescue, the lesson was that the trigger matters more than the limit: a policy payable only on a final, non-appealable judgment does nothing at all when an appeals court orders a retrial. This month is the other half of that sentence. Sometimes the trigger is fine and the limit is the problem. Either way the number in the marketing is the least informative thing about the protection.

---

The Honest Handicap

Several things cut against the way I’ve framed this.

ATE mostly works, and I’m reading a selected sample. This is a decades-old product sitting over an enormous number of ordinary cases that resolve without anyone writing about them. Two visible failures inside a fortnight is not a base rate — it is what my sources exist to surface. Had ATE routinely failed, the funded consumer-claims model would never have been financeable in the first place, and it was, for years.

The insurer may simply be right. Katch’s case is that the cover didn’t deliver what it was intended to deliver. The mirror image is that these policies contained conditions the firm was required to satisfy case by case and didn’t. On that reading Stonefort declining is not a wrapper failing; it is a wrapper performing, and the money was lost by whoever was supposed to be monitoring compliance. Note that this version is worse for funders, not better: it moves the defect from “the insurer behaved badly” to “nobody was checking,” and a lender with £63M outstanding against a high-volume firm has to answer for the monitoring either way. Nobody knows which version is true until Stonefort responds.

The two specimens aren’t the same market. One is high-volume consumer claims financed by disbursement lending; the other is heavyweight private-client litigation where the claimants bought their own cover. Filing them together under “ATE fails” is a rhetorical convenience. What they genuinely share is only the mechanism — a fixed sum of adverse-costs protection purchased at the start, against an uncertain liability at the end.

The unwrapped alternative has its own failure mode. Owning litigation risk with no insurance layer means owning it undiluted, and my own resolved case book — 1.04x gross over nearly nine years — is the evidence that unhedged selection can disappoint badly on its own terms. No insurance failed me there. I picked badly, and no wrapper would have fixed that.

---

The Tripwires I Use Now

For anyone looking at a funded claim, at a fund that reports its downside as covered, or at any structure where a policy is doing the work of a reserve:

1. Name the insured event, not the product. “ATE,” “contingent risk” and “judgment preservation” are categories. The only thing that pays is the defined event in the policy, and the useful question is always which realistic bad outcome sits outside it.
2. Find the conditions precedent, and ask who has to satisfy them. In a funded model the answer is usually the law firm — an intermediary whose competence, incentives and solvency are now part of your insurance recovery.
3. Size the limit against the opponent’s likely spend, not the claim value. A defendant with a reputational stake will outspend its own budget, and £34.5M against a £16.2M policy is what that looks like at the end.
4. Underwrite the insurer as a credit. Domicile, balance sheet, and how expensive it would be to sue if it says no. A small offshore carrier is a different asset from a rated one, whatever the certificate says.
5. Ask what happens to the protection if the intermediary fails. This claim had to be assigned out of an administration because the party holding the contractual rights had ceased to exist. Cover you can only enforce through a company that may not survive is worth less than the certificate suggests.
6. Treat “insured” in a report as a question. Insured against what, for how much, subject to what conditions, by whom. Four answers, or you have been handed a word rather than a protection.

---

Where I Land

I’m not against insurance in this asset class. It is one of the more rational things to have happened to litigation finance in a decade — real risk transfer, priced by people who do it for a living, in a business where individual outcomes are binary and a single reversal can take a fund’s year with it.

What I won’t do is let it net against the underwriting. If a deal only clears my bar because a policy sits behind it, then the policy is the investment, and I should be reading the policy — which I can’t, and wouldn’t be qualified to interpret if I could. That is an argument for pricing the risk unwrapped and declining it when it doesn’t pay enough, not for taking comfort from a document I have never seen and cannot ask for.

The thing to hold onto from Katch’s claim isn’t the £19.5M. It’s the sequence. A lender that believed it had bought protection at the outset is only now, well after its borrower’s collapse, at the start of contested litigation to establish whether it ever had any. That is the real cost of a conditional protection: not that it fails, but that its terms get settled at the end, by a judge, once the money is already gone.

The premium leaves on day one. The cover is an argument you have at the end.

---

Sources

• Legal Futures: SSB funder launches £20m claim against ATE insurer (Aug 7, 2026) — Katch Fund Solutions owed £63m by SSB Law; assignment of the claim from the administrators; £19.5m claim against Stonefort Insurance; Erich Kurtz of Hugh James on underwriting quality, conditions precedent and insurer financial standing; case at an early stage with no answer yet filed
• Law Society Gazette: SSB funders sue insurers for £19.5m over ATE policies — court papers showing Stonefort Insurance S.A. (formerly Builders Direct S.A.) paid £19.5m in premiums from 2020 onward; no claim succeeded or qualified as a win under the CFAs; the client-care letter description of the lender and of the ATE policy paying all expenses and its own premium; restitution plus interest sought
• Hugh James: ATE insurers, distressed law firms and the fight to recover funder capital — lessons from SSB Law — proceedings in the Commercial Court, King’s Bench Division; allegations undetermined and Stonefort yet to respond; counsel instructed
• Law360: SSB Law Seeks £19.5M From ATE Insurers In Cavity Wall Case (Jul 9, 2026) — the administrators’ original claim, before its assignment to the funder
• Legal Funding Journal — “Insurance Shortfall Leaves Prince Harry and Co-Claimants Facing £18 Million Costs Gap” (Jul 31, 2026), reporting Insurance Business: £16.2m combined ATE cover across seven claimants; ANL legal spend of £34.5m, more than £18.6m above its approved budget; claim dismissed in its entirety by Mr Justice Nicklin on 7 July 2026; two-day costs hearing on standard versus indemnity basis, with the shortfall beyond the ATE limit falling on the claimants personally. Also “Ignite Specialty Risk Enters Australian Market with Sydney Hire” (Aug 4, 2026): more than US$2bn of litigation capital written since 2022, including US$360m of US policies in 2024 over litigation assets valued above US$5bn
• Author’s own records, reconciled against the Portfolio page — resolved direct-case book

Commentary and personal experience — not investment, legal, or tax advice. Investing carries risk, including total loss of capital. Always do your own due diligence.