A litigation investment that produces a loss is not necessarily evidence of poor underwriting.
That proposition is easy to accept in theory and considerably harder to maintain once an investment has consumed several years of management attention, substantial legal expenditure and significant capital.
The difficulty is structural. The quality of an investment decision must be judged against the information available before the outcome is known. Performance becomes visible afterwards. The two are related, but they are not interchangeable.
A matter may have been properly diligenced, rationally priced and attractive when capital was committed, yet still realise an adverse outcome that fell within the original range of possibilities. Conversely, an inadequately investigated or poorly priced investment may generate an excellent return because of an unexpectedly favourable settlement, a procedural development or a change in the defendant’s commercial position.
For litigation funders, this matters because post-investment review can either improve future capital allocation or distort it.
The useful question is not simply whether the investment won or lost. It is why realised performance differed from the investment case.
What the outcome actually reveals
Not all losses contain the same information.
A loss may expose an underwriting failure. A factual weakness may have been available at inception but inadequately tested. The legal analysis may have been too optimistic. Enforcement risk may have been understated. The budget may never have been credible.
A different loss may reveal something more systemic. Duration assumptions may consistently prove too short. Quantum may be more volatile than the model allows. Capital requirements may repeatedly exceed initial expectations. Where the same variance appears across several matters, the problem may sit less in the individual case than in the assumptions through which the portfolio is being assessed.
There is also a third possibility: a risk that was identified, understood and appropriately priced simply materialised.
Those explanations can produce the same financial loss. They should not produce the same institutional response.
A missed issue may require better diligence. A recurring error in duration or recovery assumptions may require recalibration. But where the adverse outcome sat within the original range of reasonably contemplated results, the loss itself does not establish that the investment decision was defective.
This becomes harder to recognise once the ending is known.
Litigation outcomes tend to acquire an artificial coherence in retrospect. A witness problem that developed gradually appears inevitable. A procedural risk that once sat among many possibilities becomes the event everyone remembers. A genuinely contestable legal issue begins to look as though only one answer was ever available.
The temptation is to conclude that the problem should have been obvious.
Sometimes that criticism is justified.
But an investment committee that repeatedly converts adverse outcomes into proof that the original investment should never have been made risks becoming more conservative without becoming better. It begins to underwrite against the last loss rather than improving the next decision.
The inverse is equally dangerous. A profitable investment can conceal weak analysis. An early settlement can rescue an aggressive budget. Unexpected commercial pressure on a defendant can make poor pricing look astute. A favourable procedural development can overwhelm weaknesses that should still matter when the next opportunity is assessed.
P&L is therefore evidence about underwriting quality. It is not, by itself, proof of underwriting quality.
The deeper review is one of attribution. It requires separating what was reasonably knowable at inception from what became knowable later, assumptions that were wrong from risks that simply went against the investor, and isolated variance from patterns that indicate a structural problem.
A portfolio does not present those lessons neatly. The same economic result can arise from very different underlying causes. The value lies in identifying which cause, if any, should alter the next decision.
Re-underwriting after capital is deployed
The separation between outcome and decision quality becomes more important once funding has commenced.
As litigation develops, evidence changes, experts revise their views, quantum moves, budgets expand, procedural decisions alter leverage, defendants become more or less able to pay, and time reshapes the economics.
An investment thesis can nevertheless acquire institutional weight simply because capital has already been committed to it.
That is understandable. The committee has approved the matter, lawyers have advocated for it, internal teams have diligenced it and external experts may have supported it. Once significant capital has been deployed, revisiting the thesis can feel uncomfortably close to conceding that the original assessment was wrong.
But a rational decision at inception can be followed by an equally rational decision to change course later.
The relevant issue is not whether the original investment thesis can still be defended. It is whether the investment remains attractive on the information now available.
For lawyers, this requires a willingness to report material deterioration even where the original case was presented strongly and the revised assessment is commercially uncomfortable.
For funders, it means separating confidence in the professionals running the matter from confidence in the economics that justified the investment.
For insurers, it may require recognising that the exposure being carried today is not identical to the exposure originally priced, particularly where duration, procedural posture or recoverability has changed.
None of this requires constant second-guessing. It requires the investment case itself to remain open to challenge.
Capital already spent is relevant to the economics of the matter, but it cannot become the principal argument for capital still to be spent.
That is especially important where further funding is discretionary, reserves are being reconsidered, settlement is available, or the next procedural stage materially changes the risk-reward profile.
The original case theory deserves continuing weight only to the extent that the evidence and economics continue to support it.
Precision, resilience and risk allocation
Litigation finance has become progressively more sophisticated in its use of models, data and portfolio analysis. That is a positive development.
The danger lies in allowing precision in the model to imply equivalent precision in the underlying asset.
A spreadsheet can calculate an expected return to two decimal places. The litigation producing that return will not behave with equivalent accuracy.
Probability of success, duration, damages, costs and settlement timing remain estimates. Some are increasingly informed by useful data. Others continue to depend heavily on legal and commercial judgment.
The important question is therefore not only whether the base case works. It is how sensitive the investment is to the base case being wrong.
A material extension in duration can change the economics while leaving the legal merits untouched. Lower recoverable damages can alter the effective funding ratio. Additional capital may change both expected return and the relative position of stakeholders. Earlier settlement may improve duration while reducing absolute recovery.
These are not merely modelling questions. They concern where risk ultimately sits.
Some uncertainty can be reduced through better evidence or process; some can be allocated contractually; some can be transferred to an insurer prepared to price it; some belongs in the return requirement or capital structure; and some is better managed across a portfolio than within a single matter. Other risks remain inherent in contested litigation, however extensive the diligence.
The sophistication lies less in eliminating uncertainty than in deciding how each material uncertainty should be carried.
This is where prediction and resilience begin to separate.
A case may appear compelling in the base model yet prove economically fragile if a relatively modest change in duration, cost or quantum destroys the return. A portfolio may contain many claims and remain highly concentrated if several depend on the same legal proposition, defendant behaviour or procedural assumption. Insurance may remove a defined exposure from one participant while leaving the underlying litigation uncertainty unchanged.
Those vulnerabilities rarely reveal themselves in a headline probability of success.
They emerge from the relationship between legal risk, duration, capital, structure and the parties ultimately carrying each exposure.
The strongest underwriting processes are therefore unlikely to be those that appear most certain about a single future. They are more likely to be those that understand which assumptions matter most, recognise when those assumptions have changed, and allocate capital in a way that remains rational across more than one plausible outcome.
Litigation funding does not exist for a single reason. It responds to uncertainty, liquidity constraints, differences in risk appetite and the capital required to convert legal rights into recoveries.
The objective is not to remove uncertainty from the asset class. It is to decide which uncertainties warrant further investigation, which should affect pricing, which can be transferred or diversified, and which the investor is ultimately willing to carry.
The same discipline applies when the result eventually arrives.
A loss should not automatically be treated as proof that underwriting failed. A profit should not automatically be treated as proof that it succeeded.
Over time, consistently good decisions should produce good portfolio economics. If they do not, the assumptions, pricing or investment process deserve scrutiny.
But individual outcomes will always contain noise.
The more difficult task is distinguishing that noise from a genuine signal.
The objective is not to construct an investment that works only when the future arrives exactly as forecast, but to understand whether it remains rational when it does not.