Criterica Group — The institutional data science platform for regulated outcomes. A Splitifi company.
Underwriting

Concentration Risk in Legal Portfolios

A legal portfolio can look diversified by case type and jurisdiction while concentrating in a single judge, a single legal theory, or a single defendant. The four axes a standard credit framework misses.

September 2026

Concentration risk in legal-asset portfolios is often assessed with a framework borrowed directly from conventional credit portfolios: diversify by counterparty, diversify by sector, check correlation across those two axes, and call the concentration review complete. That framework is not wrong so much as it is incomplete, because legal assets carry correlation structures specific to how litigation actually works, and a portfolio that looks well diversified on counterparty and sector can still be dangerously concentrated along axes the borrowed framework never checks.

The first legal-specific axis is judge and venue concentration. A portfolio holding many positions pending before the same judge, or clustered in the same venue, shares exposure to that judge's specific tendencies on dispositive motions, that judge's typical pace to resolution, and that court's general procedural posture, in a way that a summary table organized by case type or defendant industry will never reveal. Two positions in different case types, held against different defendants, can still move together because they sit in front of the same decision-maker, and a concentration review that stops at case type has not measured that shared exposure at all.

The second axis is legal-theory concentration, and it is easy to miss because it is not visible in a portfolio summary the way a defendant name or a case type is. Multiple positions built on the same underlying legal theory, the same statutory reading, the same causation standard, share exposure to a single appellate ruling that can move the value of every position resting on that theory in the same direction on the same day, regardless of how different the positions look on every other dimension a standard portfolio review checks.

The third axis is counterparty concentration in the specific legal-asset sense: multiple positions held against the same defendant, or against defendants who share counsel, insurance coverage, or a common corporate parent, share exposure to that party's settlement posture and litigation strategy. A defendant that decides to litigate aggressively rather than settle, across every matter it faces in a given period, moves every position against it in the same direction at the same time, and a portfolio that diversified by case type while concentrating unknowingly in defendant relationships has diversified along the wrong axis.

The fourth axis, duration correlation, is distinct from outcome correlation and just as easy to miss. A portfolio can hold positions with genuinely uncorrelated outcome probabilities while sharing a correlated duration exposure, because every position was underwritten against the same court system's typical pace, and a systemic slowdown, a docket backlog, a procedural rule change, extends every position's holding period simultaneously even though each position's probability of a favorable resolution remains independent of the others.

Standard portfolio tools miss all four of these axes because they were built for asset classes where counterparty and sector genuinely are the dominant correlation structures. A bond portfolio's concentration risk really is mostly about issuer and industry. A legal-asset portfolio's concentration risk runs through decision-makers, legal theories, and procedural systems that a sector or issuer classification simply does not capture, and applying the borrowed framework produces a concentration report that looks rigorous while measuring the wrong thing.

Proper concentration modeling for this asset class requires mapping every position along all four axes at once, judge and venue, legal theory, defendant relationship, and duration exposure, and checking for clustering along each independently rather than assuming that diversification on the conventional axes implies diversification on the legal-specific ones. This is more operationally demanding than a standard credit concentration review, because it requires the portfolio system to track metadata a conventional system was never built to hold, but the additional demand is exactly proportional to the risk it is meant to catch.

A practical illustration makes the gap concrete. Consider a portfolio of one hundred positions spread across fifteen case types and twelve states, a profile that would clear almost any conventional concentration limit built around sector and geography. If forty of those positions happen to sit before three judges in adjoining counties of the same state, and thirty more rest on a single statutory interpretation currently pending before a circuit court of appeals, the portfolio carries two live concentration events that its own summary reporting would never surface, because neither judge assignment nor legal theory appears as a field in a reporting structure built around case type and state.

Building the legal-specific axes into a portfolio system also changes what a capital partner should expect to see in a periodic reporting package. A report that states diversification by case type and jurisdiction alone should be read as incomplete rather than reassuring, because it answers a narrower question than the one that actually determines whether the portfolio is exposed to a single correlated event. A capital partner reviewing a fund's concentration disclosure should ask specifically whether judge and venue clustering, legal-theory exposure, and defendant-relationship mapping are tracked, and should treat the absence of an answer as informative in itself.

None of these four axes is static once established at underwriting, and a concentration review conducted only at the point of origination will miss clustering that develops afterward, as new positions are added to a portfolio that already carries meaningful exposure along one of these dimensions. Concentration monitoring, like duration and outcome monitoring, has to be a continuous process checked against the current portfolio composition, not a snapshot taken once and treated as durable, with each new position screened against the existing book's exposure before it is committed rather than only reviewed retrospectively at the next scheduled portfolio review.

An institution that reviews concentration only through counterparty and sector has produced a concentration report, not a concentration finding. Legal-asset concentration risk is genuinely multidimensional, and a portfolio that looks diversified on the two axes a borrowed framework checks can be sitting in front of a single judge, resting on a single legal theory pending appellate review, or exposed to a single defendant's litigation posture across what appears, on paper, to be a wide and uncorrelated book, invisible until someone finally builds the concentration report along the axes that actually govern this asset class.

← All InsightsRequest Access →