Investing in Litigation Finance

On March 27, 2026 the US Court of Appeals for the Second Circuit reversed, by two votes to one, the $16.1B judgment that Petersen Energía and Eton Park had won in the district court against the Argentine Republic over the 2012 renationalisation of the oil company YPF. Burford Capital’s shares fell more than 45% that day and closed at about $4.10, halted repeatedly on the way down.

The majority, in an opinion by Judge Denny Chin, gave two independent grounds, and either one disposed of the case on its own. The first was that YPF’s by-laws never created the shareholder-to-shareholder obligations the claimants’ contract theory required: the promise they needed was one shareholder owing another, and as a matter of Argentine law the by-laws did not contain it. The second was that Argentina’s General Expropriation Law supplied the exclusive procedure for compensating a taking, so an expropriation authorised by a state’s own law could not simultaneously found a New York contract claim for damages. Judge José Cabranes dissented.

The arithmetic behind that morning is worth holding on to, because it is the arithmetic of the whole asset class. Petersen’s claims had been sold to a Burford subsidiary for €15M, on terms giving Burford 70% of anything the claims produced. In the years that followed, the claim acquired a $16.1B judgment, and Burford carried the position at a fair value that made it the single largest asset it owned. In the quarter that followed the reversal, Burford recognised a $2.4B capital provision loss; its capital provision assets fell from $5,609,949 thousand at December 31, 2025 to $3,120,499 thousand at March 31, 2026, and it reported a net loss of $1,632,069 thousand, or $7.46 per share, against net income of $0.14 a share in the same quarter a year earlier. On June 2, 2026 the Second Circuit refused to rehear the case.

Nothing about the underlying claim changed on March 27. No fact was discovered, no witness recanted, no debtor went bankrupt. Three judges read a set of by-laws and a statute, two of them read those documents differently from the trial court, and an asset carried in the billions became an asset worth arguing about. That is what you are buying when you fund a lawsuit, and it is why the marketing multiple and the realised return in this market are further apart than in any other corner of private credit.

A litigation funding agreement is a non-recourse advance secured on the proceeds of a legal claim. The funder pays some or all of a claimant’s legal costs; if the claim produces money, the funder is paid first out of the proceeds, on a formula agreed in advance; if the claim produces nothing, the funder is paid nothing and has no right to chase the claimant for the money. That single feature — no recourse to the borrower, only to the outcome — is what separates litigation finance from a loan and what makes the return look like equity while the paperwork looks like credit.

Four shapes cover almost everything in the commercial market.

Single-case funding advances the cost of one claim, usually commercial, arbitral or patent, in exchange for a contracted return out of that claim’s proceeds. It is the purest expression of the asset and the most binary: one judge, one panel, one outcome.

Portfolio funding advances against a bundle of claims held by one law firm or one corporate claimant, cross-collateralised so that proceeds from any case in the bundle repay the facility. It is the form the market has moved toward, because diversification inside a single contract does what a fund’s diversification does, at lower cost, and because a law firm with twenty contingent-fee cases is a better credit than any one of them.

Monetisation advances cash against a judgment or award the claimant has already won but cannot yet collect. The legal risk is gone; the enforcement risk is not. This is the corner of the market where a court’s reversal on appeal, or a sovereign’s refusal to pay, does the damage.

Claim purchase buys the claim outright, so the funder becomes the claimant. Petersen’s claims against Argentina were sold this way, to a Burford subsidiary, for €15M and 70% of whatever the claims produced. Where it is lawful, this is the cleanest structure and the one that raises the sharpest questions about who is really running the lawsuit.

Consumer legal funding — the pre-settlement advance offered to an individual personal-injury plaintiff — is a different business with different economics and its own state consumer-credit statutes, and it is treated separately below.

There is no index of litigation finance returns. There is no Cliffwater, no Lipper, no S&P; there is no supervisor collecting returns from funders the way the SEC collects them from business development companies. What exists is a set of numbers each funder publishes about itself, computed on a basis each funder chooses, and one listed company’s audited accounts. Any honest account of this asset class has to start by saying that the record is self-reported, and that the three structural biases in it all point the same way.

The first is concluded-case bias. Almost every published return in this market is calculated on matters that have already resolved. Burford’s 2025 annual report gives a cumulative 83% return on invested capital and a 26% IRR on concluded matters over a roughly 15-year history and about $3.8B of lifetime realisations. Those are the company’s own figures, on the company’s own definition, and they exclude by construction every case still running — which, in a business where the bad cases are the ones that drag, is the population most likely to be going badly. A funder that concludes its winners quickly and litigates its losers for eight more years will report a rising ROIC the whole time.

The second is multiple-versus-rate confusion. Return on invested capital is a multiple. An 83% ROIC means $1.83 came back for every $1.00 that went out, in total, over whatever period it took. It says nothing about the years. The IRR is the number that says something about the years, which is why the two figures sit side by side in the disclosure and why the gap between them — 83% cumulative against 26% annualised — is the honest measure of how long these cases take.

The third is the distance between cash and marks. A funder’s reported return in any year is dominated by fair-value movements on cases that have not ended; the one figure in its accounts that no court can later revise is cash actually received. At Burford those two lines moved in opposite directions in 2025. Cash receipts fell to about $530M from nearly $700M the year before and reported revenues fell 24% to $413M, while new definitive commitments rose 39% to $872M. Writing more business while collecting less proves nothing on its own. It is the pattern under which a concluded-case record gets managed, and it is the reason to read the cash line before the return line.

Set against that, the record of the vehicles an outsider can actually buy is poor. Burford’s shareholders owned, at March 31, 2026, a company with $3.17 of tangible book value per share after a single case took $2.4B out of it in one quarter. Omni Bridgeway, the largest listed funder outside the United States, spent 2025 restructuring rather than compounding: in April 2025 it sold a 70% interest in a continuation vehicle holding more than 150 legal assets to funds managed by Ares Management for about A$320M in cash, a transaction its own announcement described as delivering a day-one cash multiple above 3× on the assets sold, and which was also, plainly, a liquidity event for a balance sheet that needed one.

Therium, one of the industry’s founding names, stopped raising capital and stopped backing new claims; in April 2025 it laid off staff across its offices, and on June 11, 2025 it handed day-to-day oversight of its portfolio to Fortress Investment Group.

US commercial litigation finance is smaller than its press coverage implies, and it has stopped growing in a straight line. The best annual count is the Westfleet Insider, published each spring by the broker Westfleet Advisors from data the funders give it. New capital committed to US commercial litigation finance deals ran at $3.2B in 2022, $2.7B in 2023 and $2.3B in 2024, a 16% fall in the last of those years, before rebounding about 23% in 2025 to roughly $2.8B. That is the whole American commercial market’s annual new business: less than a single mid-sized private credit fund raises in a quarter.

The count of firms doing the committing has not grown either, and it does not fall in a line; it oscillates inside a narrow band. Westfleet identified 39 active capital providers in its 2023 report, 42 in its 2024 report and 39 again in its 2025 report, with only one new entrant in 2025 and several participants entering wind-down. Its earlier counts were 41 in 2019, 46 in 2020 and 44 in 2022, so the population of this industry has sat between about 39 and 46 firms for seven years while the capital behind them turned over. The 2025 rebound, on Westfleet’s own reading, came from a small group of established funders deploying more, not from new money arriving.

The most useful thing in the 2025 report is a number Westfleet stopped publishing. For years the standard headline for this market was an industry assets-under-management figure — $16.1B in the 2024 report, up from $9.5B in the 2019 report and $15.2B in the 2023 one. In the 2025 report Westfleet declined to publish an industry AUM figure at all, on the grounds that AUM is not a reliable measure of annual financing activity and that the figures had frequently been misunderstood or mischaracterised. That is an unusual thing for a market’s own scorekeeper to say, and you should take it seriously: an AUM number in this business mixes committed-but-undrawn capital, capital already deployed into live cases, and in some counts the gross face of the claims themselves.

Deal sizes are large enough that a private investor cannot participate case by case without a fund or a broker. The 2024 report put the average commitment at $8M, with single-case deals averaging $6.6M and portfolio deals $16.5M, and noted that 19% of new commitments carried insurance of some kind. Co-investment between funders is rare: on Westfleet’s new 2025 measure, only 7% of deals involved two or more funders sharing a position.

Five groups meet in every funding deal, and what each of them wants explains most of the price.

The claimant is usually a company with a good claim and a bad reason to avoid paying for it: the legal budget is an expense that hits this year’s earnings while the recovery, if any, arrives in four years. Funding converts a certain cost into a contingent one. A claimant who could comfortably fund its own case and chooses not to is telling you it does not rate the case as highly as you do — the adverse-selection problem at the heart of the asset class.

The law firm is often the real counterparty. Portfolio facilities are written against firms, not cases, and a firm that puts twenty contingent matters into one facility is hedging its own cash flow. This is also where the ethics rules bite: in the United States, a funder cannot take a share of a lawyer’s fee directly, because the professional conduct rules of almost every state prohibit fee-sharing with non-lawyers — Arizona, which abolished its Rule 5.4 in August 2020 and licenses alternative business structures, is the conspicuous exception — so the economics are routed through the claim’s proceeds or through the firm’s revenue rather than through the fee itself.

The defendant and its insurer are the source of every dollar the asset class earns, and they are also the asset class’s most effective political opponents. The US Chamber’s Institute for Legal Reform and Lawyers for Civil Justice have been the drivers of the disclosure campaign described below; the funding industry’s trade bodies argue the opposite case. Every rule fight in this market is a fight between people who pay judgments and people who collect them.

The funder is the manager: it sources, underwrites, prices and monitors, and it is compensated as an asset manager, not as a lender. Underwriting is done by litigators, not credit analysts, and the diligence question is not “can this borrower pay” but “will this claim win, when, and against assets we can reach”.

The capital behind the funder changed character over 2024 and 2025, and that matters more than it sounds. Litigation finance used to be funded by dedicated closed-end funds raised from endowments, family offices and specialist allocators.

Increasingly it is funded by private credit: the Ares purchase of 70% of Omni Bridgeway’s Fund 9 in April 2025 was the industry’s first continuation fund, and Therium’s chief executive, explaining the firm’s retreat, pointed to exactly this — the flood of money into private credit and the appetite of those investors for legal assets. Our hub flagship, Investing in Private Credit, describes what that capital wants and what it does when it is asked for its money back, with the listed and semi-liquid vehicles covered in Investing in Listed BDCs and Investing in Interval Funds and Non-Traded BDCs; the point here is that a claim which used to be held to conclusion by a patient ten-year fund may now sit inside a vehicle carrying liquidity promises of its own.

Litigation funding contracts do not quote an interest rate. They quote a claim on proceeds, and the claim almost always takes one of two forms, or the greater of both.

The multiple waterfall

The funder is entitled to a stated multiple of the capital it actually deployed, stepping up with the time the case takes. A schedule of that kind might read 2× if the case resolves in the first year, 3× in years two and three and 4× thereafter, with the steps written into the contract; the levels vary by funder and by case, and the shape is what matters. Because the entitlement is capped by the multiple, a case that produces an enormous recovery pays the funder no more than the schedule allows, and the claimant keeps the upside. Because the multiple rises with time, the structure partly protects the funder against the one risk it cannot control.

The proceeds share

The funder takes a stated percentage of gross or net recovery — commonly in the twenties or thirties for commercial matters — regardless of how much capital was deployed. Where the deployed capital ends up small relative to the recovery, this is by far the more valuable entitlement, which is why most contracts are written as the greater of a multiple and a percentage, and why the two clauses are the first thing to read in any funding agreement.

Why the two are not comparable

The step-up schedule is the funder’s answer to duration risk, but it is a partial answer. Look at what a fixed multiple is worth as an annual rate at different case lives, and the entire economics of the asset class fall out of one table.

A 3× contracted multiple sounds like a different asset class from a 2×. Over eight years it is worth 14.7% a year against 9.1%, and both of those are gross of every fee the investor pays to reach them. The headline number in this market is a multiple; the number that pays for your retirement is a rate; and the bridge between them is the one variable nobody in the transaction controls.

Every other risk in litigation finance can be diversified. Duration cannot, because the thing that lengthens one case — a crowded docket, an interlocutory appeal, a bankruptcy stay, a change of counsel — tends to lengthen the others at the same time, and because the cases that go long are disproportionately the cases that go badly. Duration is the asset class’s systematic risk, and it is the reason a portfolio of twenty uncorrelated lawsuits is less diversified than it looks.

The mechanism runs through three doors. Appeals are the first: a first-instance win is not a resolution, and the appellate stage adds years to the cases most worth appealing, which is to say the largest ones. Enforcement is the second: a judgment against a defendant who will not pay begins a second lawsuit in a second jurisdiction, and the clock restarts. Settlement timing is the third and the most under-appreciated: a defendant with a strong balance sheet and a weak case has every incentive to settle late, because the funder’s cost of carry is not the defendant’s problem, and because the claimant’s willingness to accept less rises with every year the case runs.

The practical consequence for an outside investor is that the fund’s stated term is not the relevant number. A ten-year closed-end fund that deploys over four years into cases with a three-and-a-half-year expected life will, on any realistic tail, be seeking extensions in year nine. That is normal in private equity and it is priced there; it is not usually priced here, because the marketing arithmetic tends to present a case-level multiple as though it were a fund-level annual return.

The last risk in the chain is the one that ended the largest position in the industry’s history, and it is the one an outsider is least equipped to underwrite. A claim has to survive three tests to pay: it has to win, the win has to survive appeal, and the defendant has to have attachable assets inside a jurisdiction whose courts will hand them over. Each test is independent of the others and the asset is worth nothing if any one fails.

What happened in Petersen

The chronology is worth setting out in full, because it is the only case in this market with a complete public record from purchase to reversal. Argentina renationalised YPF in 2012 without making the tender offer that YPF’s own by-laws required of a controlling shareholder. Petersen Energía, a minority holder driven into insolvency by the expropriation, had its claims sold in 2015 by its Spanish bankruptcy administrators to Prospect Investments LLC, a wholly owned Burford Capital subsidiary, for €15M, with Burford entitled to 70% of anything the claims produced and Petersen’s estate to the remaining 30%.

Burford later funded a parallel claim by Eton Park. In September 2023 the US District Court for the Southern District of New York entered judgment for the claimants in the amount of $16.1B, the largest damages award ever entered against a foreign state by a US court.

Then it went the other way. On March 27, 2026 the Second Circuit reversed, 2–1, on two independent grounds, either of which disposed of the case. The first, and the one the majority took first, was that YPF’s by-laws did not create the shareholder-to-shareholder obligations the contract claim required under Argentine law: the claimants needed a promise running between shareholders, and the by-laws did not give them one. The second was that, even if the by-laws had, Argentina’s General Expropriation Law provided the exclusive route to compensation, so an act of state authorised by domestic law could not simultaneously found a New York contract claim. Judge Cabranes dissented. On June 2, 2026 the same court denied rehearing en banc.

Burford has said the claimants expect to seek certiorari from the Supreme Court, with a filing deadline of September 28, 2026 and the petition eligible to reach the Justices’ conference by November 12, 2026, while noting that the Court declines most such applications; as of September 10, 2026 no petition had been reported filed. Burford has also said that investment treaty arbitration against Argentina — under the Spain and United States bilateral investment treaties that cover Petersen and Eton Park respectively — remains a viable prospect.

What it cost, and what it did not

The accounting damage is precise. In the first quarter of 2026 Burford booked a $2.4B capital provision loss on the YPF position; capital provision assets fell from $5,609,949 thousand at December 31, 2025 to $3,120,499 thousand at March 31, 2026; the quarter’s net loss attributable to shareholders was $1,632,069 thousand, or $7.46 per share, against $0.14 a share of net income a year earlier. Tangible book value per share stood at $3.17 at the end of that quarter, and management guided to a Burford-only figure excluding YPF of about $3.40.

The cash damage was smaller than the accounting damage, and that distinction matters. Burford had sold 38.75% of its Petersen entitlement to third-party investors over the preceding years, taking real money off the table against an asset that had not yet paid: $136M of cumulative cash proceeds by June 2018 on the first 28.75%, a further tranche priced at $30M alongside the Eton Park transaction, and reported cumulative proceeds from those secondary sales of about $236M by April 2026. That is the single best argument for the discipline of realising part of a position that has become too large to be a position. What the reversal destroyed was the carrying value, and the carrying value is what an equity investor had been buying.

Litigation finance poses one of the hardest fair-value problems in finance, and the industry’s largest company has spent its listed life arguing about it. A funding position is a financial asset carried at fair value, which requires an estimate of what a willing buyer would pay for a claim on the proceeds of a lawsuit that has not been decided. There is no observable market, no comparable transaction and no discount rate that means anything; there is an internal judgement about the probability of winning, the probable quantum, the probable date, and the probability of collecting. In accounting language these are Level 3 inputs, and in plain language they are the manager marking its own homework.

The dispute became public on August 7, 2019, when Muddy Waters Research published a short thesis accusing Burford of manipulating its returns metrics and misleading investors about how it accounted for gains, calling the marks subjective and the governance arrangements inadequate, and comparing the company to Enron. The shares closed about 46% lower that day, having already fallen some 19% the day before on speculation about the report, and traded as much as 66% down intraday. The accounting sequel is the more interesting half.

Following engagement with the SEC, Burford revised its fair-value approach for capital provision assets under ASC 820, and on May 2, 2023 its management and audit committee concluded that the financial statements for 2019, 2020, 2021 and the six months to June 30, 2022 should be restated. The restatement corrected a material understatement of capital provision assets and capital provision income — that is, the correction ran in the company’s favour, not against it.

Both facts should stay in your head at once. A short seller alleged the marks were too aggressive; a restatement four years later concluded they had been too conservative; and the same asset that was allegedly mis-marked in one direction or the other lost $2.4B of its carrying value on a 2–1 appellate vote in 2026. The honest conclusion is not that the marks were fraudulent or that they were sound. It is that in this asset class a fair value is a forecast of a court’s behaviour, and the range of defensible forecasts is enormous.

For an investor, three practical rules follow. Read the concentration disclosure before the return disclosure, because a portfolio’s fair value is dominated by its largest position. Watch for realisations, not marks: cash received on concluded cases is the only number in a funder’s accounts that a court cannot revise. And treat any period in which unrealised fair-value gains exceed realisations as a period in which the manager, not the market, produced the return. The same distance between a mark and a realisation is the subject of Investing in Pre-IPO Shares and Investing in Music Royalties on this hub, in markets where the mark is a funding round and a catalogue multiple rather than a court’s behaviour.

Funding a stranger’s lawsuit for a share of the proceeds was a criminal offence in England until sections 13 and 14 of the Criminal Law Act 1967 abolished the crimes and torts of maintenance and champerty, and the doctrines behind them have not entirely gone away. They govern whether your contract is enforceable, which is a more fundamental risk than any of the credit risks in this guide.

Champerty and maintenance

Maintenance is meddling in someone else’s litigation without a bona fide interest in it; champerty is maintenance for a share of the proceeds. American states inherited both, and most have narrowed or abandoned them, but they have done so one state at a time and by different routes.

Minnesota is the clearest worked example. A funder advanced $6,000 to a personal-injury plaintiff against her eventual settlement, on terms adding 30% every six months until the case settled, subject to a cap; the district court and the court of appeals both held the contract unenforceable as champertous. In Maslowski v. Prospect Funding Partners, 944 N.W.2d 235 (Minn. 2020), the Minnesota Supreme Court reversed and abolished the common-law prohibition on champerty outright, holding that the evolution of social needs had made the doctrine unnecessary and that litigation financing may increase access to justice, and remanding the contract for examination on other grounds. The court could do that because Minnesota’s prohibition rested on common law rather than statute; where a state’s rule is statutory, only the legislature can move it.

The practical upshot for an investor is that champerty risk is a choice-of-law question in every deal, and a fund that cannot tell you which states’ law governs its portfolio is not managing the risk.

Disclosure in the federal courts

There is still no single national rule requiring a party to disclose that its case is funded. The federal Advisory Committee on Civil Rules has had the question in front of it since 2014 and has repeatedly declined to act, leaving a patchwork: some judges require disclosure by standing order, some districts by local rule, and multidistrict litigation judges increasingly ask as a matter of case management.

The pressure to standardise is real and organised. On March 10, 2026 the US Chamber’s Institute for Legal Reform and Lawyers for Civil Justice filed a joint proposal with the Advisory Committee for specific rule language amending Rule 26(a)(1)(A) to require disclosure, in the initial disclosures, of the identity of any non-party funding the action and holding a financial interest in it, together with the agreements that define that interest. In Congress, Senator Chuck Grassley introduced the Litigation Funding Transparency Act of 2026 (S.3826) on February 11, 2026, with Senators Tillis, Kennedy and Cornyn; it would require disclosure of the funder and the funding agreement in class actions, multidistrict litigation and coordinated federal proceedings involving 100 or more cases, and would bar funders from directing litigation or settlement. A separate Grassley bill of February 2026 is directed at foreign third-party funding.

The states have moved faster

Where the federal rulemakers have hesitated, legislatures have not. Three states moved in 2024: Indiana’s HB 1160, signed on March 13, 2024, requires disclosure, bars funders from obtaining proprietary data and prohibits them from controlling the litigation; Louisiana’s SB 355, effective August 1, 2024, limits foreign funding, requires disclosure within thirty days and bans funder control of settlement decisions; and West Virginia extended its consumer statute to commercial funding through SB 850, signed on March 27, 2024.

Six more moved in 2025 — Arizona, Colorado, Georgia, Kansas, Montana and Oklahoma. Georgia’s SB 69, the Courts Access and Consumer Protection Act, is the most complete of them: signed on April 21, 2025, it made the existence and terms of funding agreements of $25,000 or more discoverable at once, and required funders to register with the Georgia Department of Banking and Finance through the Nationwide Multistate Licensing System from January 1, 2026. Kansas’s SB 54 requires a funding agreement to be disclosed within thirty days, along with every contracting party and any foreign funder. Arizona, Montana and Colorado legislated mainly against foreign funding, and Oklahoma requires agreements to be produced on request. Three states in one year, six in the next: the count is rising, and so is the reach, from disclosing that a funder exists to producing the agreement that says what it is owed.

North Carolina bans it outright

House Bill 315, the Prohibit Litigation Investments Act, was signed on June 22, 2026 and took effect the same day. It prohibits litigation investment — defined as providing money for the costs of a civil proceeding in exchange for a return contingent on its outcome — in civil proceedings commenced on or after that date and under contracts entered into, renewed or amended on or after it. Contingency fees, an insurer’s duty to defend or indemnify, non-profit legal services and money from immediate family are carved out.

The enforcement is what gives it teeth. The Attorney General may sue to enjoin violations, civil penalties run to $50,000 per violation, and a person harmed by a prohibited investment has a private right of action for common-law damages or treble statutory damages. It is the first outright state ban in the country, and it is the most consequential fact in this section: in North Carolina a commercial funding position is not a regulated asset, it is a prohibited one. The question to put to any manager with a national book is not whether other states will follow but what happens to the North Carolina matters already in the portfolio.

The United Kingdom, and why it matters here

The most consequential single decision for funders anywhere was handed down in London on July 26, 2023. The UK Supreme Court held in the PACCAR case that a litigation funding agreement entitling the funder to a share of damages is a damages-based agreement, and is therefore unenforceable unless it complies with regulations almost no funding agreement complied with — and cannot be used at all to fund opt-out collective proceedings in the Competition Appeal Tribunal.

The Civil Justice Council’s final review, published in June 2025, made 58 recommendations, among them that PACCAR be reversed by legislation as soon as possible and that funding be put on a light-touch statutory footing. On December 17, 2025 the government told Parliament that it accepted the two principal recommendations — legislation clarifying that funding agreements are not damages-based agreements, with prospective effect only rather than the retrospective effect the Council had urged, and proportionate regulation of those agreements — to be introduced “when parliamentary time allows.” The 2026 King’s Speech did not include it. As of September 10, 2026 no bill has been introduced and PACCAR remains the law. The reason it matters to an American investor is portfolio composition: several large funders run global books, and a decision that makes a class of contracts unenforceable in one jurisdiction reprices the assets everywhere.

Of all the drags on this asset class, tax is the largest and the least discussed, and it is the one that decides which wrapper you should hold. Nothing here is advice; the rules below are the general federal position for a US individual and the treatment of any particular agreement turns on its own terms.

Character: ordinary in the ordinary case

The default expectation should be that litigation funding profits are ordinary income, taxed at rates up to 37% federal in 2026, plus the 3.8% net investment income tax under IRC §1411 above $200,000 of modified adjusted gross income single or $250,000 joint, plus state tax. Two doctrines produce that result. The first is the substitute-for-ordinary-income doctrine: an amount received in place of what would have been ordinary income is itself ordinary, the principle the Supreme Court applied in United States v. Midland-Ross Corp., 381 U.S. 54 (1965). Where a law firm sells a share of its future contingent fees, the property sold is the right to fees, and the gain is ordinary for that reason.

The second is more basic: long-term capital gain requires the sale or exchange of a capital asset held more than a year, and a funder that simply collects its contracted share of a settlement has no sale or exchange at all. The IRS has taken exactly that position in a private letter ruling denying capital treatment to a litigation fund on the ground that there was no sale or exchange.

At the top of the schedule the spread is worth roughly 17 percentage points — about 45.8% on ordinary income against 28.8% on long-term capital gain, taking 37% and 20% federal, the 3.8% NIIT and a 5% state rate. On a $110,000 profit that is close to $19,000. Character is not a technicality here; it is the largest single line in the cost stack.

The prepaid forward, and what it defers

The dominant structure in the market is the prepaid forward purchase agreement: the funder pays cash now for a contractual right to a share of future proceeds, and the tax event is deferred until the contract settles. The underlying authority is Revenue Ruling 2003-7, in which the IRS treated a properly structured variable prepaid forward contract as an open transaction — neither a current sale under IRC §1001 nor a constructive sale under IRC §1259 — so that the cash received up front is not taxed until the contract settles.

For a funded law firm, the effect is that the firm reports its fee as ordinary income when earned and separately reports ordinary gain or loss on settling the forward, so that it is taxed on the net amount it keeps. For funders, and particularly for non-US funders, prepaid forwards have also been structured in an attempt to reach capital treatment — the practice that produced the 2025 legislative push described below. Deferral is real and valuable; conversion of character is contested, and you should assume your fund’s profits are ordinary unless its tax counsel has given a reasoned opinion otherwise and you have read it.

The loss side is where the asymmetry bites

A litigation portfolio produces total losses routinely, so the treatment of losses matters as much as the treatment of gains. The worst outcome is asymmetric: profits ordinary, losses capital. That result is available in principle, because IRC §1234A treats gain or loss from the termination of a right with respect to property that is a capital asset in the taxpayer’s hands as gain or loss from a sale or exchange, and because an individual’s net capital loss is deductible against ordinary income only to the extent of $3,000 a year under IRC §1211(b), with the remainder carried forward indefinitely. A $250,000 loss that is ordinary shelters $250,000 of other income this year; the same loss treated as capital, in a year with no capital gains, shelters $3,000. Ask the question in diligence and get the answer in writing.

The wrapper decides the outcome

Because the income is ordinary and lumpy, the wrapper is worth more than the manager. A tax-deferred account removes the largest single drag in the guide at a stroke, for the same reason it does in private credit generally. Tax-exempt investors and retirement accounts should nonetheless ask about unrelated business taxable income: a fund’s income from funding is not obviously income from an operating trade or business, but leverage at the fund level creates debt-financed income under IRC §514, and the answer is fund-specific. A domestic fund will issue a Schedule K-1, typically late, and may create filing obligations in several states.

A listed funder, by contrast, is simply a share: your gain on the shares is a capital gain if you have held them more than a year, which is the single largest tax advantage of the listed route and one reason to compare it seriously against a fund. Note that a funder incorporated outside the United States — Burford was incorporated in Guernsey in 2009 — raises passive foreign investment company questions for US holders. Burford’s own filings carry the risk factor, and state that its current view is that it is not a PFIC; PFIC status is tested annually and can change, and the consequences if it did are outside the scope of this guide.

The bill that has not passed

In May 2025 Senator Thom Tillis introduced the Tackling Predatory Litigation Funding Act (S.1821), with Senator John Husted as lead cosponsor and a House companion from Representative Kevin Hern, proposing a new tax on “qualified litigation proceeds” at the highest individual rate plus 3.8% — 40.8% under the rates then in force — while barring funders from netting litigation-funding losses against those proceeds, and applying to taxable years beginning after December 31, 2025.

It was cut from the 2025 reconciliation package before enactment; a revised version at 31.8% was subsequently discussed, and Senator Tillis has said he will bring the measure back. As of September 10, 2026 the bill sits in the Senate Finance Committee, has passed neither chamber, and no such tax has been enacted. Treat it as a live tail risk rather than a forecast: the bill’s design would tax gross proceeds without loss offset, which in a business where total write-offs are a routine outcome is not a rate change but a change in the shape of the asset.

The cost stack in litigation finance is a private-equity stack applied to an asset with private-equity duration and none of private equity’s exit routes. There are four layers, and the reason they matter so much here is that the gross return arrives as one lumpy multiple rather than as a coupon, so every layer takes its cut from the same place.

Management fees. The market standard for a dedicated fund is 2% a year, charged on committed capital during the investment period and on invested capital or net asset value thereafter. Over an eight- or nine-year fund life that is roughly 16% to 18% of your commitment, paid out of the commitment, which means the capital that actually reaches cases is materially less than the cheque you wrote. On a $250,000 commitment at 2% for eight years, $40,000 goes to fees and $210,000 goes to lawsuits.

Carried interest. Typically 20% of profits above a preferred return of around 8%, with a catch-up. In a portfolio whose gross multiple is under 2×, the preferred absorbs most of the profit and the carry looks modest in dollars; above that level the carry accelerates sharply, because the catch-up hands the manager 100% of the next dollars until it has its full 20%.

The broker and the origination layer. Deals reach funders through brokers and specialist advisers; those fees are generally paid by the claimant out of its side of the waterfall rather than by the fund’s investors, which is a good thing for you and a reason the claimant’s economics are worse than the headline split implies.

The wrapper. A listed funder charges you nothing directly and everything indirectly: its operating expenses, financing costs and management compensation come out of the same case proceeds, and you can read them in the accounts. This is the only route in the asset class where the whole cost stack is published.

Every risk in this guide so far is a risk to the return. These are the risks to the capital.

Adverse selection at the point of sale

A claimant who can fund its own case and chooses not to has made a judgement about the case, and it is not a flattering one. This is the market’s original problem and no amount of diligence fully solves it, because the claimant knows things about its own conduct that no data room contains. The mitigants are structural: fund portfolios rather than cases, insist on the claimant retaining meaningful skin in the outcome, and pay attention to whether the law firm is taking contingency risk alongside you. A firm charging full hourly rates on a funded case has no exposure to the result.

Adverse costs and security for costs

Outside the United States, and in some US proceedings, the loser pays the winner’s costs. A funder backing a claim in England, Australia or an international arbitration can therefore lose more than it advanced, which is why after-the-event insurance and capital adequacy rules matter in those jurisdictions and why courts there can order a funded claimant to post security for the defendant’s costs. Ask any fund with a non-US book how adverse-costs exposure is capped and who bears it.

Funder insolvency and the run-off

Cases outlive funders. When a funder stops writing new business, its existing commitments still need to be honoured, and a claimant left without the money to finish a case is a claimant who settles cheaply. Therium’s retreat is the live example: after April 2025 layoffs across its offices, on June 11, 2025 it transferred day-to-day oversight of its portfolio to Fortress Investment Group. That is an orderly outcome. The disorderly version — a funder unable to meet drawdowns on live matters — destroys value in every case in the book at once, and it is why capital adequacy, not just track record, belongs in your diligence.

Control, ethics and the discovery of your own terms

US professional conduct rules bar a lawyer from sharing fees with a non-lawyer and require that the client, not the funder, controls the litigation and any settlement. A funding agreement that gives the funder settlement approval, veto rights or day-to-day direction risks unenforceability, disqualification of counsel, or waiver of privilege over the funder’s diligence materials — and several 2024 and 2025 state statutes now make funder control an express prohibition and the agreement itself discoverable. The commercial consequence is that your economics can end up in the hands of the party you are suing.

Fraud and the claim that was never real

The asset is a legal claim, and legal claims can be manufactured. The cautionary case is the Ecuadorian environmental litigation against Chevron. On March 4, 2014 the US District Court for the Southern District of New York held that the $9.5B Ecuadorian judgment had been procured by fraud and racketeering and was unenforceable in the United States, a ruling the Second Circuit affirmed unanimously on August 8, 2016. Burford had been one of the plaintiffs’ largest financial backers, and it was Burford that gave sworn testimony documenting the misconduct by which its own funding had been obtained. The lesson does not depend on the details: in a market where the collateral is a story told to a court, a funder’s fraud controls are part of its investment process, not its compliance function.

Concentration, which is the one that actually happened

Every risk above is survivable in a diversified book. Concentration is not, and it is the failure the industry’s largest company actually experienced. One position, marked into the billions, took $2.4B out of Burford’s capital provision assets in a single quarter of 2026 on an appellate ruling. If you cannot see a funder’s largest position as a percentage of its book, you cannot underwrite the funder.

One cheque, three routes, so that the fee stack, the duration, the tax character and the exit can be compared in dollars. The assumptions are stated in each route and in the chart caption, and the point is the shape of the result, not the decimals.

Route A: a private litigation finance fund

$250,000 committed to a dedicated fund with a nine-year life, 2% a year on commitments for eight years, and 20% of profits over an 8% preferred return with a full catch-up. Fees take $40,000, so $210,000 reaches cases. The portfolio returns 1.85× gross on deployed capital — a respectable result in this market — producing $388,500. Profit over contributed capital is $138,500; the preferred, at 8% a year on capital outstanding for a weighted-average 5.5 years, is $110,000, so the manager’s carry is $27,860 and you keep $110,640 of profit. Pre-tax you have $360,640, or 1.44× over nine years, 4.2% a year. After tax at 45.8% on ordinary income you have $309,967, or 2.4% a year. In a Roth or an IRA, 4.2%. A 1.85× gross multiple has become a low-single-digit annual return, and nothing went wrong.

Route B: shares in a listed funder

$250,000 of a listed funder bought at 1.0× tangible book, held nine years, with the company compounding tangible book per share at 8% a year — roughly 2.0× over the period. Sold at the same 1.0× multiple of book, you have $499,751 pre-tax, or 8.0% a year, and $427,823 after capital gains tax at 28.8%, or 6.2% a year. The tax rate is 17 points lower than Route A’s and the position is saleable on any trading day. The catch is the multiple: sell at 0.6× book and the same nine years of compounding returns $285,494 after tax, 1.5% a year; sell at 1.4× and it returns $570,152, 9.6% a year. Between those two outcomes there is no difference in the cases. On March 27, 2026 the market repriced Burford’s shares by more than 45% in a single morning.

Route C: one case, funded directly

$250,000 into a single commercial claim on a contract paying the greater of 3× deployed capital or 30% of proceeds, with an expected 3.5 years to resolution. Three outcomes cover most of the distribution. The claim fails, or wins and cannot be collected: you receive nothing, and the $250,000 loss is worth $114,500 in tax relief if it is ordinary and offsets other income, but shelters only $3,000 a year if it is capital and you have no gains. The claim settles early at 1.3×: $325,000 gross, $290,650 after tax, 4.4% a year. The claim wins at the contracted 3×: $750,000 gross, $521,000 after tax, 23.3% a year. Weight those at 30%, 30% and 40% and the expected gross multiple is 1.59×, about 14% a year before tax — a fine expected value that you get exactly one draw at.

Three things fall out. The fund route, on a gross portfolio result most managers would be pleased with, delivers 2.4% a year after tax for nine years of complete illiquidity — the fee stack and the ordinary-income rate together take more than half the profit. The listed route pays a lower tax rate and offers daily liquidity, and its entire outcome range is set by a valuation multiple that has nothing to do with the cases. And the single-case route is the only one that can pay you 23% a year, which is precisely why it is the one you should never fund with money you cannot lose in full.

There are four doors into this asset class and only one of them is open to everybody. They are set out below in order of how visible the price is, which is the order in which a newcomer should consider them.

The listed funders

A share in a publicly quoted funder is the only route with a daily price, an audited balance sheet, no minimum and no lock-up. Burford Capital is the largest, listed in New York and London; Omni Bridgeway is the largest outside the United States, listed in Sydney; a handful of smaller specialists trade in London. You are buying a manager, not a portfolio: the operating costs, the leverage and the fair-value judgements are all inside the price, and so is a valuation multiple that can move violently for reasons unconnected to any case. The compensating advantages are real — full public disclosure, capital-gain tax treatment on your shares, and the ability to leave.

Private funds

The institutional route is a closed-end fund, typically Delaware or Cayman, with an eight- to ten-year life, 2-and-20 economics, an accredited-investor or qualified-purchaser test and minimums that generally start at $250,000 and are frequently higher. This is where most of the industry’s capital sits and where the diversification is. It is also where the fee stack, the ordinary-income character and the illiquidity compound against you at the same time, which is what Route A above measures. Specific managers’ terms and minimums change and should be taken from the current offering documents rather than from any guide.

Platforms, and their history

Between 2014 and the early 2020s several platforms sold fractional participations in individual cases to accredited investors at minimums in the low thousands. LexShares was the best known of them; in August 2024 it cancelled the third fund it had been raising, cut most of its staff and moved into what it called harvest mode, leaving its investors in a run-off portfolio while it managed the existing book. That is the reason to treat any new single-case platform with care: the offering is the easy part of this business and the ten-year workout is the hard part. If you use one, assume the platform will not be there at the end and ask who administers the positions if it is not.

Consumer legal funding

Pre-settlement advances to individual personal-injury plaintiffs are a separate industry with separate economics: small advances, short expected lives, non-recourse, and rates that in disputed cases run into the tens of percent a year — the Minnesota advance that produced that state’s champerty decision was $6,000 repayable with 30% added every six months until the case settled, subject to a cap. It is regulated as consumer credit in a growing number of states, several of which cap rates or require registration. It is not a sensible destination for an individual investor’s capital and it is included here so you can recognise it when a broker describes it as litigation finance.

The sequence

1. Decide the wrapper before the manager. If the money can go into a Roth or an IRA, the ordinary-income drag disappears and Route A’s arithmetic improves by nearly two percentage points a year; if it cannot, the listed route’s capital-gain treatment is worth 17 points of rate on every dollar of profit.
2. Read one listed funder’s annual report end to end before you look at any private fund. It is the only complete, audited description of how this business earns money, and it will teach you what questions the private documents do not answer.
3. Ask every manager for deployed-capital IRR by vintage year, with unconcluded cases in the denominator, and for the largest position as a percentage of the book. Both answers should arrive in writing. If either does not arrive, stop.
4. Size the position as venture capital, not credit. A total loss on any single case is a normal outcome, and the fund’s own diversification is smaller than the case count implies because duration is common to all of them.
5. Set the term expectation at the stated fund life plus two years, and confirm what happens at the end: whether the manager can extend, on whose consent, and whether unconcluded positions are sold or distributed in kind.
6. Once invested, track realisations rather than marks. Each year, compare cash received on concluded cases with the change in fair value; a year in which unrealised gains exceed realisations is a year the manager produced the return.

These are the readings that would change our view, each with a threshold and a date, so a reader in 2027 can check them against the record of the day rather than against this guide’s assumptions.

The market’s own numbers

• Annual new commitments, from the Westfleet Insider each spring. $3.2B in 2022, $2.7B in 2023, $2.3B in 2024, and a rebound of about 23% in 2025 to roughly $2.8B. A second consecutive rising year would say capital is returning; a fall back below $2.3B would say the 2025 rebound was one group of funders deploying reserves.
• The active funder count. 39 in the 2023 report, 42 in the 2024 report, 39 again in the 2025 report, with one new entrant in 2025. Below 35, with wind-downs continuing, the market is consolidating into a handful of balance sheets and single-case pricing will move in the funders’ favour.
• Whether an industry AUM figure comes back. Westfleet stopped publishing one in its 2025 report, saying AUM is not a reliable measure of activity. Any AUM number quoted for this industry after that date deserves a source and a definition before it is believed.

Burford, because it is the only full public record

• The YPF endgame. Rehearing en banc was denied on June 2, 2026; the certiorari deadline Burford has stated is September 28, 2026, with conference eligibility by November 12, 2026, and the Court declines most petitions; no petition had been reported filed as of September 10, 2026. A grant would be a genuine surprise; a denial closes the US court route and leaves the arbitration path Burford has said it is considering.
• Tangible book value per share and the price paid for it. $3.17 at March 31, 2026, with management guiding to about $3.40 excluding YPF, against a share price of about $4.10 on March 27, 2026. Sustained trading above 1.5× tangible book means the market is again paying for unrealised marks; below 0.7× it is refusing to.
• Realisations against fair-value gains. Cash receipts fell to about $530M in FY2025 from nearly $700M in FY2024, while new definitive commitments rose 39% to $872M and FY2025 revenue fell 24% to $413M. Two consecutive years in which cash coming in falls while new business and reported returns rise is the signal that the concluded-case sample is being managed.

The rules

• The federal disclosure rule. The Advisory Committee on Civil Rules has declined to act since 2014; the Institute for Legal Reform and Lawyers for Civil Justice filed proposed Rule 26(a)(1)(A) language on March 10, 2026, and S.3826 would legislate disclosure for class actions, MDLs and 100-case proceedings. Publication of a proposed rule for comment would be the moment the federal position changes; until then the patchwork holds.
• The tax bill. S.1821 proposed 40.8% on qualified litigation proceeds with no loss offset and was cut from the 2025 package; a 31.8% version has been discussed and its sponsor has said he will return to it. Enactment in any form would change the shape of the asset, not merely its rate, because the denial of loss netting falls hardest on a business with a routine total-loss rate.
• The states. Indiana, Louisiana and West Virginia legislated in 2024; Arizona, Colorado, Georgia, Kansas, Montana and Oklahoma in 2025; Georgia’s registration regime took effect January 1, 2026; and North Carolina banned litigation investment outright with House Bill 315, signed and effective June 22, 2026. Watch two numbers: how many states make the funding agreement itself discoverable, because that is the provision that changes settlement behaviour, and whether any state follows North Carolina into prohibition, because that is the provision that voids the contract.

Our tape

Invest Alternative does not yet carry a litigation-finance series: there is no public price for this asset class to collect, which is itself the most useful thing our data can tell you about it. The nearest thing we hold is the Private Credit sub-index on the same hub, a BDC-proxied series built from the VanEck BDC Income ETF’s daily close, based at 100 on September 2, 2025 and standing at 83.157 on September 8, 2026, down 16.79% over twelve months. It matters here because private credit has become a principal supplier of capital to legal assets — the Ares purchase of 70% of Omni Bridgeway’s Fund 9 in April 2025 is the clearest example — and a private credit market that is repricing its own liquidity is not a market that will fund ten-year lawsuits generously.