The Verdict That Vanished: Inside the “Take Care of Maya” Funding Fight

The Verdict That Vanished: Inside the “Take Care of Maya” Funding Fight

A $213M verdict became a $42M loan and a $60M insurance policy. Then an appeals court erased the judgment and sent the case back for a smaller retrial — the one outcome that breaks the economics while triggering none of the protection. Florida’s highest court has now refused to disturb that, so the number everything was built on is gone for good.

I write about litigation finance as an investor, not a lawyer, and I no longer pick cases. For years I bought slices of individual funded claims through retail platforms — 42 LexShares cases, for a whole-portfolio return of about 1.25% a year. New money now goes into funds run by people who do this professionally. So I have no position in the Kowalski matter, and no special access: everything below comes from public filings and reporting.

What I do have is a standing question. If I’m paying specialists to underwrite legal assets, I should know what the hard part of that job actually is. The Kowalski file is the clearest public answer I’ve seen — and the answer is not what the headlines suggest.

The story everyone tells about this case is that a huge verdict got overturned, which is a story about being wrong. The more useful story is that every party in the deal was protected against losing, and none of them was protected against waiting. That distinction is the whole post.

The headline everyone remembers

Maya Kowalski’s story became a 2023 Netflix documentary, Take Care of Maya. In November 2023 a Florida jury found that Johns Hopkins All Children’s Hospital had falsely imprisoned Maya as a 10-year-old and that its conduct contributed to her mother Beata’s suicide. The award was roughly $261 million, trimmed by the trial judge to about $213 million. It was the kind of number that makes headlines, and — it turns out — the kind that attracts financing.

The money built on the headline

A $213M judgment under appeal is not $213M in the bank. It is a claim: contingent, years from final, and reversible. But it can be borrowed against. After the verdict the Kowalskis took roughly $42.1 million as an advance from HPS Investment Partners, a large private-credit manager acquired by BlackRock in 2025. To support the deal they bought a $60 million judgment preservation insurance (JPI) policy from Ambridge Group, brokered by Willis Towers Watson.

Run the ratios, because they tell you how the professionals sized this risk. The advance was 19.8% of the judgment — a lender taking about one dollar for every five on paper, which is a sober number and roughly what you’d expect from someone who assumed the verdict might not survive. The insurance was 1.43x the loan. That is not a rounding error, and the likeliest explanation for the extra cover is accruing interest and costs — which implies the parties expected this to run for years.

Step Amount Note
Jury award, Nov 2023 $261M The number that made the news
Reduced by the trial judge $213M Judgment as entered
HPS advance against it $42.1M 19.8% of the judgment
JPI policy (Ambridge) $60M 1.43x the advance
Reportedly paid out of the advance ~$30M ~71% of gross; the rest to costs, brokerage, premium
Former counsel’s fee claim ~$9.9M 23.5% of the advance; disputed
The judgment today $0 Reversed; reversal now unappealable; retrial set for March 2027

The redacted credit agreement, surfaced in filings and reported by Bloomberg Law, shows the covenants funders actually write. The family kept “sole, final, and unconditional” control of the appeal, retrial and settlement — but not really. They cannot settle in a way that “adversely affects the interests of the lenders,” must keep HPS informed of every development, share draft filings, and give notice of settlement offers. They can be put in default for settling against the lender’s interest or changing attorneys without HPS’s consent. Control on the cover page; a lender’s hand on the wheel in the covenants.

Then the verdict evaporated

On October 29, 2025, Florida’s Second District Court of Appeal reversed the judgment entirely. The court held that the trial court had erred in refusing directed verdicts and had allowed testimony and argument that blurred the hospital’s statutory good-faith immunity for reporting suspected child abuse with conduct that might be tortious. It remanded for a new trial on a narrowed set of claims — Maya’s intentional infliction of emotional distress, plus the remaining false imprisonment, battery and medical negligence counts. The wrongful-death claim, the fraudulent-billing claim and all punitive damages are gone. The hospital then tried to move the retrial out of Sarasota; that was refused in April 2026, the judge citing the cost of moving an eight-year-old case to a new circuit.

Then the appellate route closed. On August 4, 2026 the Florida Supreme Court declined to review the reversal, which settles it: the $213M judgment is not coming back. The claim goes to a jury again, smaller, with the calendar reset.

Note precisely what has happened and what hasn’t. Nobody has decided that Maya loses — the merits are still live. What has been decided, finally and unappealably, is that the number the financing was built on no longer exists.

The trap: insurance that cannot fire

Here is the part the coverage skipped, and it is the reason this case is worth an investor’s attention rather than a documentary viewer’s.

A JPI policy pays only once the judgment becomes final and non-appealable. Not when the verdict wobbles — when every avenue is exhausted: intermediate appeal, rehearing, petitions for discretionary review. Policies typically carry a self-insured retention of 10–25% so the plaintiff keeps skin in the game, and they conventionally exclude settlements, defence costs and the defendant’s inability to pay.

Now apply that to a reversal-and-remand. There is no final judgment. There is no judgment at all — the number the policy was written around has been vacated, and the case is back to being an unresolved claim. So the policy has not been triggered, cannot yet be triggered, and may not be triggered for years. Meanwhile the loan it was bought to support presumably keeps accruing.

That is the structural point: vacatur-and-retrial is the outcome that maximises damage while activating none of the protection. An affirmance pays the lender. A clean, final reversal pays the insurer’s money to the lender. A remand does neither — it converts a defined instrument back into an open-ended one. The policy is simultaneously intact and useless.

The Florida Supreme Court’s refusal to take the case makes that permanent rather than temporary, and it is worth being exact about why, because the intuition runs the wrong way. “The appeals are over” sounds like the condition a policy waiting on finality has been waiting for. It isn’t. What became final is the vacatur — the ruling that there is no judgment. A policy that pays on a final, non-appealable judgment needs a judgment to attach to, and the one it was underwritten against has been erased with no possibility of reinstatement. The trigger hasn’t been met and now never can be, on this judgment. Any payout would have to wait on a second verdict, at a second trial, on narrower claims, with punitive damages off the table — a different and smaller number, if it arrives at all, well over three years after the policy was bought.

And the exclusions bite from the other side. If the family settles at or before retrial, a conventional JPI policy may not respond at all — while the credit agreement restricts settling in ways that hurt the lender. Read those two documents together and the family is boxed: the fast resolution is the one the insurance was never written to cover, and the slow one is the one the insurance cannot pay on until it ends.

What makes this worth an investor’s attention is that nobody involved was careless. The advance rate was conservative, the cover exceeded the loan, the covenants were tight, and the collateral was a verdict a jury had already returned — better collateral than most litigation finance ever sees. Every one of those protections addressed the question “what if we lose?” The outcome that arrived was “nobody decides for another several years,” and there was no instrument in the stack pointed at it.

The market had already voted on this trade

This is not an isolated misfire, which is what makes it worth generalising from. The JPI market has contracted sharply, for exactly this reason.

Aon’s claims data tells the arc plainly: while litigation-risk insurance was mostly M&A-adjacent between 2015 and 2021, carrier loss ratios sat in the low single digits. As capital migrated into single-case judgment preservation, the industry took negative development on large policies — and appetite fell not only in JPI but across the broader contingent-risk market, including lines with no relationship to it at all. After the BMC Software reversal — a $1.6B judgment Liberty Mutual had insured for a reported $500–750M, overturned on appeal — the carrier withdrew from pending JPI deals and paused quoting new active-litigation liability. Capacity above $100 million contracted, pricing hardened, underwriting tightened. In some venues the base rates are brutal: in Texas, reversal rates on granted petitions hit 88% in 2024.

The consequence is that appellate monetisation — funders lending against verdicts directly — has returned as the primary source of post-judgment liquidity, because insurance got repriced out of reach for the cases that most need it. Aon’s observation about which carriers weathered it is the familiar one: the ones deploying across diversified portfolios, where each risk is non-systemic. Concentrated single-case exposure is what hurt. That is the same lesson my own 42-case book taught me from the other direction, and it is why the funders I read now are moving away from single big swings toward portfolios.

The sleeper ruling: a loan became a “recovery”

The sharpest live dispute isn’t the family against the hospital. It’s the family against their former lawyers — and it produced a holding with implications well beyond this case.

In May 2026, Judge Hunter Carroll ruled that the loan constitutes proceeds of a judgment, and ordered an evidentiary hearing on whether the contingency fee was excessive and whether the firm violated ethics rules by providing improper financial assistance. The Kowalskis had argued the loan proceeds weren’t a “recovery” at all, so no fee was owed.

Sit with what that means. If borrowing against a verdict creates a recovery, then monetising a judgment crystallises fee obligations on money that is a liability, not a gain. The loan has to be repaid. The fee does not get refunded if the judgment later evaporates — which here it did. Should the March 2027 retrial produce little or nothing, the family will have paid a multi-million-dollar contingency fee on borrowed money while still owing HPS. The waterfall inverts: counsel converted a paper verdict into cash at the top of the stack, and the client absorbed the reversal at the bottom. Anyone underwriting appellate monetisation should treat that as a structural feature of the product, not a quirk of one Florida courtroom.

What the fee fight is actually about

The disputed fee is roughly $9.9 million — about 23.5% of the advance. The 2017 engagement entitled AndersonGlenn to 40% of any recovery up to $1 million plus 33% of everything above it. Florida Bar Rule 4-1.5(f)(4)(B) sets a presumptive ceiling for a case litigated through judgment: 40% of the first $1 million, 30% of the next, then 20% of anything above $2 million, with an additional 5% available once an appeal is instituted. Exceeding the schedule requires prior court approval, which the family says was never sought.

A flat 33% above the first million is well outside that 20% tier, and the gap is not academic. On my own arithmetic, applied to the $42.1M the court has now deemed a recovery, the contracted rate produces about $14.0 million against a Bar ceiling near $8.7 million — a gap of just over $5 million. That figure is worth holding next to two public statements. Greg Anderson wrote in an email quoted in the complaint: “To make sure we are following Bar Rules the best we can without destroying your case in the process, we escrowed the difference between the Bar recommendations and the 33.33%.” And he has said publicly: “We took 50 percent of $9.9 million and left over $5 million in a trust account for Judge Carroll to ultimately decide if we should get our full fee.” The escrow is, in other words, roughly the size of the overage — which suggests everyone involved understood the schedule was in play from the beginning.

The rest of the complaint, filed June 17, 2026 in Florida’s Twelfth Judicial Circuit, is heavier and entirely unproven. The Kowalskis allege constructive fraud, that the advance’s total costs, interest, repayment terms and case-control restrictions were never properly explained, and that a new fee agreement was presented to Maya during a hotel stay shortly after she turned 18 in December 2023, while she was emotionally vulnerable. They allege trust-account funds were misused, including roughly $4 million pledged as collateral so the Andersons could borrow to buy a house. These are allegations, not findings. The Andersons deny wrongdoing, say a judge has already resolved most of the issues, and have signalled counterclaims for fraud and defamation. Nothing here has been tested at trial.

But notice what the fight is about. Not whether the hospital wronged Maya. The mechanics of the money layered on top: the loan, the premium, the brokerage, the covenants, the fee schedule, and the order in which people get paid.

What this actually shows

  • The headline verdict is a marketing number, not a bank balance. $261M → $213M → $0-pending-retrial in about 24 months. Everything financial downstream inherited that fragility.
  • Outcome risk was hedged; duration risk wasn’t. The policy answers “what if we lose?” Nobody had an instrument for “what if nobody decides for another four years?” — which is the outcome that actually arrived.
  • Insurance triggers matter more than insurance limits. A $60M policy that pays only on a final, non-appealable judgment does nothing on a remand. The limit is the headline; the trigger is the product.
  • “Control” is a negotiated fiction. The claimant nominally ran the case but couldn’t settle freely or change counsel without the lender’s consent. Read the covenants, not the cover page.
  • The waterfall decides who actually bears the reversal. Fees came out of the advance at the top; the family carries the debt at the bottom. Position in line beat the size of the verdict.

Where I land

Two honest limits first. I’m reading public and partly-redacted filings, so I don’t know the loan’s interest rate, its maturity, or the policy’s actual trigger language — and bespoke policies can be drafted around the conventions I’ve described. It remains entirely possible the insurance ultimately does its job, the retrial goes well, and everyone lands roughly whole. I also can’t tell you which HPS vehicle holds this position, and I’d caution against assuming: a firm this size runs many funds, and the retail credit vehicles I wrote about separately are a different animal from a bespoke judgment advance.

What I take from it is narrower and more useful to me. I stopped picking cases because I’m not equipped to price reversal risk, read credit agreements, or stress-test an insurance trigger — and this file is a public demonstration that those three things, not the merits, decided the outcome. That’s an argument for paying a specialist, not for avoiding the asset class.

So it also gives me a sharper list for the managers I do pay, and I’d rather ask it now than after: not “what’s your case win rate,” but when does your protection actually pay. What triggers the policy — a reduction, a reversal, a remand? Does it survive a settlement, or does settling void it? Who controls the decision to settle, and what happens to the fee waterfall if the recovery is later vacated? How much of the book is single-case concentration versus portfolio? Those are the questions this file answers badly, and the ones I’d want answered before capital moves.

The verdict was the story everyone tuned in for. The financing is the story that outlived it — a $42.1M loan, a $60M policy that can never fire on the judgment it was written against, and a disputed $9.9M fee, all orbiting a number that no longer exists. In this asset class the figure in the headline is the one I trust least. The real money moves in the fine print underneath, and the fine print is mostly about timing.

Sources

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