Not about funding, about analysis. Useful whether or not you ever seek capital. Seven core lessons, a real case study that ties them all together, and a provision-by-provision walkthrough of the funding agreement itself for when you're ready to actually sign one.
These seven lessons work as a hypothetical, or you can apply them to something real. Tell the AI Concierge about your matter and we'll assess it step by step, in lockstep with these same lessons, applied to your actual facts instead of an example.
These are two different questions, and conflating them is the single most common mistake claimants make when sizing up their own case. Liability asks: are you legally right? Did the other side breach the contract, infringe the patent, act negligently? Damages asks a completely separate question: assuming you're right, what is that actually worth in dollars? A claimant can be nearly certain to win on liability and still have a weak investment case, because the damages are small, speculative, or hard to prove, and a claimant can have a genuinely messy liability picture and still attract serious funder interest, because the damages at stake are enormous and well-documented.
Funders weigh both, but they often weigh damages more heavily than claimants expect, for a simple reason: damages set the ceiling on the entire pie. A funder's return comes out of the eventual recovery, so a claim that's certain to win but worth very little is frequently a worse investment than a claim that's 60% likely to win but worth ten times as much. This is why the first real question in any serious case assessment isn't "will I win", it's "if I win, how much is that actually worth, and can I prove it."
A damages figure a claimant cites off the top of their head, "this cost us $5 million", is not the same thing as a damages figure a funder can underwrite. Between those two things sits a substantial amount of work: a damages expert building a model (lost profits, unjust enrichment, a reasonable royalty in a patent case, diminished value in a commercial dispute), testing that model against the specific facts of the case, and producing a number that can survive cross-examination, not just sound plausible in conversation.
Sophisticated investors don't take a single damages number at face value, they stress-test it across a range: a best case, an expected case, and a conservative case, then apply a probability-weighted view across that range rather than anchoring on the headline figure. They also scrutinize who built the model. A damages theory built by a credible, experienced expert who has survived Daubert challenges before carries real underwriting weight; the identical number produced without independent expert analysis is treated as, at best, a starting hypothesis.
Go deeper: How Investors Think: Underwriting a Legal Claim →
Here's a scenario that trips up almost everyone outside the industry: a claimant wins a $10 million judgment against a defendant who is, by the time judgment is entered, functionally insolvent. Legally, that's a complete victory. Financially, it may be worth close to zero, a judgment is only as valuable as the ability to actually collect on it. This is collectability, and funders treat it as a screening factor every bit as important as the strength of the underlying legal claim.
Collectability analysis asks: is the defendant solvent, and will they still be solvent by the time a judgment is enforceable? Is there insurance coverage that will actually respond? Are there identifiable, reachable assets, and in what jurisdiction? A funder will often pass on a near-certain win against a judgment-proof defendant and take real interest in a messier liability picture against a defendant with clear, liquid assets, not because the merits don't matter, but because merits without collectability produce a moral victory, not a financial one.
A funder cannot independently re-litigate every case before deciding whether to fund it, so they rely heavily on proxies for case quality, and two of the strongest proxies are who is trying the case and who is explaining it. Experienced litigators self-select into matters they genuinely believe in, know how to build a clean record, negotiate more effectively, and are far less likely to make the procedural missteps that quietly sink otherwise strong claims. Track record and resourcing become a real, if imperfect, signal of case quality itself.
On technical and damages questions, which in patent, IP, and complex commercial litigation frequently decide the case outright, the expert witness matters just as much as counsel, arguably more. The December 2023 amendments to Federal Rule of Evidence 702 strengthened trial courts' obligation to police expert testimony before it ever reaches a jury. An expert who cannot survive that gatekeeping doesn't just weaken a case; they can end it before trial, regardless of how strong the underlying facts are. Sophisticated funders increasingly ask not just "has an expert been retained," but "is this expert credible enough to survive a Daubert challenge, and persuasive enough to actually move a jury."
A dollar recovered next year is worth more than the same dollar recovered in five years, a basic finance principle that gets underweighted in most conversations about case merit, because claimants naturally focus on whether they'll win, not on how long winning will take. Litigation duration is genuinely uncertain and routinely extends through appeals, and every additional year a matter is outstanding is a year of capital tied up, unable to be redeployed, with the underlying risk of the case still live.
This is a large part of why funders target internal rates of return in the 30% range rather than something closer to a typical lending rate: the target isn't just compensating for the risk of losing outright, it's compensating for years of illiquidity on capital that can't be touched until the case resolves. It's also why funding agreements are frequently structured with tranched capital releases tied to procedural milestones, rather than a single upfront payment, a way of managing duration risk by only committing more capital as the case actually progresses.
Underwriting a single matter has real fixed costs, diligence, expert review, ongoing monitoring, and those costs don't scale down proportionally for smaller claims. That's why most institutional funders have effective minimum claim sizes, often in the $1-5 million range: below that threshold, even a genuinely strong case frequently can't generate enough expected return to justify the underwriting cost, no matter how favorable the liability and damages picture looks. A smaller, purpose-built tier of the market exists specifically to serve claims below that floor, but pricing there tends to be less favorable, reflecting the same fixed-cost economics at a smaller scale.
Claim size isn't the only reason a strong case gets declined. A funder may already be overexposed to a particular sector, defendant, or geography and be managing concentration risk at the portfolio level, entirely independent of any individual case's merits. Reputational or publicity concerns, potential conflicts with other matters in the funder's book, or simply a mismatch with a fund's stated investment thesis can all produce a pass on a case that, evaluated purely on the facts, would score well.
Everything up to this point has been about evaluating one case in isolation, the vantage point of a claimant looking at their own matter. Zoom out to how a fund actually operates, and the picture changes. Litigation outcomes are close to binary at the individual-case level, you tend to win big or recover little, with less middle ground than in most asset classes, and any single case genuinely is closer to a coin flip than most claimants are comfortable admitting.
Funders manage that the way venture investors manage individual startup risk: not by trying to make every single case a winner, but by diversifying across many matters and accepting that a meaningful share, commonly cited in the 20-40% range, will produce little or no recovery, relying on a smaller number of outsized wins to carry the portfolio's overall return. That's also, notably, a genuine institutional selling point: litigation outcomes are largely uncorrelated with broader financial markets, which is part of why endowments, pensions, and hedge funds have been drawn to the asset class as a diversifier.
In 2012, Argentina expropriated a controlling stake in YPF, the country's largest oil company, without paying the tender offer the law required minority shareholders. Burford Capital eventually acquired the resulting legal claims, held by Petersen Energía, a bankrupt former YPF shareholder, out of a Spanish insolvency proceeding, and funded the case in U.S. federal court for roughly a decade. In September 2023, the Southern District of New York entered judgment for the plaintiffs at $16.1 billion, the largest funder-backed outcome on record. Argentina has resisted paying ever since, and enforcement is still being fought out across multiple jurisdictions today. It's a single real case that happens to touch nearly every lesson above.
Liability and damages: the underlying claim was unusually clean on liability, Argentina's own bylaws required the tender offer, and it simply never made one, which is part of why the case attracted funding despite its scale. But the damages side is where the real work happened: three decades of foregone value in a nationalized oil company is not a number you can eyeball, and both sides fought for years over the valuation methodology before the court adopted a figure closer to the plaintiffs' model.
Collectability, the central risk: this is the lesson this case illustrates best. Winning a $16.1 billion judgment against a sovereign is not the same as collecting it. Argentina is fighting enforcement on multiple fronts, and Burford itself has been candid that collection could take years more and may never reach the full judgment amount. A case can have the strongest liability and damages picture imaginable and still carry serious collectability risk simply because of who the defendant is.
Duration and counsel/expert quality: from claim acquisition to judgment took roughly a decade, with enforcement still ongoing years later, a real-world illustration of just how long "long-duration" litigation finance can mean. And a valuation this size and this contested does not survive a decade of litigation without a damages case built by top-tier experts able to withstand exactly the kind of scrutiny Module 2 describes.
Financeability economics and portfolio context: almost no funder in the world has the balance sheet to take a position this size, which is exactly the point of Module 6: financeability is bounded by a funder's own capital and portfolio construction, not just a case's merits. And the size of Burford's eventual return (a large majority of the proceeds, by public reporting) became its own controversy, a reminder from Module 7 that a single outsized position, even a winning one, carries concentration and reputational risk at the portfolio level, distinct from whether the underlying case was a good bet.
Go deeper: the Dispute Library entry on Petersen Energia v. Argentine Republic →
Everything above is about deciding whether a claim is a good investment. This module is about the moment that decision becomes a contract. Litigation finance agreements are not standardized the way a mortgage or a car loan is, terms vary widely across funders, claim types, and jurisdictions, and most claimants see their first one with no real benchmark for what is customary versus unusual. The closest thing this field has to a benchmark is the annotated model agreement Maya Steinitz and Abigail C. Field published in the Iowa Law Review in 2014, developed through an open research process where each provision was posted publicly for comment before the final version went to print. This module walks through the provision categories it identifies, in plain language.
Funding mechanics: the agreement sets out the amount committed and how it is released, as a single payment or, far more commonly, in tranches tied to case milestones such as surviving a motion to dismiss or completing key depositions. Read this section for what triggers each tranche and what happens if the funder declines to advance a later one; that is often a contractually anticipated outcome, not a breach.
The return waterfall: typically a return of the capital advanced, plus the greater of a fixed multiple (commonly in the 2 to 4 times range) or a percentage of recovery, sometimes capped, sometimes not. Ask specifically whether the multiple and the percentage stack or whether only the larger of the two applies, and what happens to the waterfall if the case settles for less than expected or resolves faster than expected.
Control and consent rights: this is usually the most heavily negotiated section, and the one worth the most careful reading. Well-drafted agreements reserve litigation strategy, settlement authority, and privileged communications to the claimant and counsel, while giving the funder information rights and case updates, often with a consultation right on major decisions short of an actual veto. A funder consent right over settlement is not standard market practice; when one appears, as it did in the Burford-Sysco dispute covered in our Dispute Library, it is worth understanding exactly how it is worded and what "unreasonably withheld" would actually mean if it were ever tested.
Termination and default: specifies what happens if the claimant settles without required consent, voluntarily dismisses the case, or the funder itself stops advancing capital partway through. These provisions rarely get read carefully at signing and matter enormously later.
Confidentiality and assignment: governs who else may see the agreement, and whether the funder's interest can be sold or syndicated to another party without the claimant's consent. Also worth asking: whether anything in the agreement could become discoverable by the opposing party in your specific jurisdiction, since courts remain split on this question nationally.
Governing law and dispute resolution: which state's law governs, and whether disputes between claimant and funder go to arbitration or court. This clause is easy to skip past and can end up mattering more than almost any other provision if the relationship with the funder ever sours.
Go deeper: What's Actually In a Funding Agreement → · Foundational Scholarship: the full annotated bibliography →
Answer honestly about a hypothetical, or your own matter. This is a rough, educational self-check, not an assessment; for a real one, talk to the AI Concierge.
A one-page PDF version of these seven questions, for whenever you're reviewing a matter without your laptop open.