Subrogation recovery is one of the most predictable sources of net loss reduction available to P&C carriers, and one of the most systematically underperformed. Industry data consistently shows that carriers recover significantly less in subrogation than their potential -- not because subrogation rights do not exist, but because they are identified too late to be exercised effectively. Most subrogation potential is visible at FNOL. Most subrogation identification happens months later.
What subrogation potential looks like at FNOL
Subrogation potential at the FNOL stage appears in the narrative and structured fields in recognizable patterns. Third-party vehicle involvement is the most obvious: any auto or transportation claim that mentions another vehicle, a named driver who is not the insured, or an intersection event has third-party liability potential. Equipment or product failure is the second most common: a claim describing electrical fire that started from an appliance, roof damage that mentions recent contractor work, or water damage from a defective HVAC installation has potential against a manufacturer or contractor.
Commercial premises cases are a third category: a claim filed by a tenant against their carrier for damage that was caused by a landlord's failure to maintain the property, or a commercial claim where another business on the property bears responsibility. Landlord-tenant subrogation is particularly common and particularly frequently missed, because the FNOL describes the tenant's damage without explicitly naming the landlord as the liable party -- the handler has to recognize the liability implication from the description of how the damage occurred.
Why late identification kills recovery potential
The practical obstacles to subrogation recovery grow significantly with time. Physical evidence -- the appliance that started the fire, the failed component in the HVAC system, the vehicle that caused the accident -- is subject to repair, disposal, or degradation. Witnesses' recollections fade and become less reliable in deposition or at trial. The third party's insurer may have already made decisions about the claim or entered settlement negotiations based on their understanding of liability.
When subrogation potential is identified at FNOL, the carrier can act immediately: notify the third party's carrier that a subrogation claim may be forthcoming, preserve evidence through a timely demand letter, and coordinate the investigation of the loss cause with the pursuit of recovery rights. This coordination is impossible when the subrogation potential is first identified three months after settlement, by which time the evidence is gone and the third party's insurer has closed their file.
The coverage-reserve interaction in subrogation cases
Subrogation potential affects both how a file is reserved and how it should be documented from the start. When subrogation potential is identified at FNOL, the file should be noted with a subrogation flag that prompts specific documentation practices: preserve the claimant's account of how the loss occurred, identify all potentially responsible parties, document the condition of any relevant equipment or property at the time of loss, and request the insured's cooperation in the recovery process.
Reserve setting in subrogation cases should account for anticipated recovery. A file reserved at gross expected loss without a subrogation credit will develop favorably when the recovery comes in -- which is a form of reserve inaccuracy even when it goes the right direction. Adjusters who flag subrogation potential early and set a reserve net of expected recovery produce more accurate initial reserves and more stable development triangles for those claim types.
Automated subrogation indicator extraction
The most reliable way to ensure consistent subrogation identification at FNOL is to extract subrogation indicators automatically from the narrative and structured data at intake. The extraction targets are: named third parties other than the insured, equipment or product failure language, contractor or service provider references in property claims, multi-vehicle language in auto claims, and premises liability indicators in commercial claims.
False positives are manageable because the subrogation flag triggers a review step, not an automatic action. A flag on a file that turns out to have no recovery potential costs a few minutes of handler review time. A missed flag on a file with genuine recovery potential may cost tens of thousands of dollars in lost recovery. The asymmetry of costs justifies a sensitivity setting that errs toward flagging. Handlers who review flagged files and find no subrogation potential can clear the flag in one click; no harm done.
Building a subrogation metrics baseline
Carriers that want to improve subrogation recovery rates need a measurement baseline. The key metrics are: subrogation identification rate (percentage of closed claims where subrogation was evaluated), identification-to-filing lag (average time between claim open and subrogation demand letter), and recovery rate on evaluated files (gross recovery as a percentage of gross paid losses on files with documented subrogation evaluation).
Most carriers can calculate these metrics from their CMS data, but many have never done so. The identification rate is often surprisingly low -- in manual intake environments, 30 to 40% of files with genuine subrogation potential may never be flagged. The identification-to-filing lag is often measured in months rather than days. Establishing these baselines before implementing automated subrogation flagging creates the measurement framework for demonstrating the impact of earlier identification on recovery rates.