A book in run-off has a specific cost structure problem: the expense ratio rises as the premium base shrinks, but claims continue to develop and require professional handling. The traditional response — maintain a full claims and actuarial staff until the book winds down — becomes economically untenable within two to three years of the last policy expiration. The claims that remain open at that point are typically the most complex: long-tail bodily injury, occupational disease, and environmental liability claims that require specialized expertise but do not justify full-time staffing at the volume that remains.
The run-off management problem is not one of finding enough claims to justify the staff — it is one of allocating specialist attention to the claims that genuinely require it, and handling the remaining volume efficiently with more junior or generalist adjusters. That is a triage problem, and it has a triage solution.
The Cost Structure of a Book in Run-Off
When a carrier stops writing new business on a line or exits a market entirely, the expense ratio typically worsens immediately. Premium earned continues declining as policies expire, but the fixed costs of claims administration — adjuster salaries, system maintenance, actuarial reserve review — decline more slowly. The run-off manager's job is to accelerate claim resolution on the bulk of the remaining open claims while preserving the specialist capacity needed for the complex cases that resist easy settlement.
The claims remaining on a book two years into run-off are not a random sample of the original book. They are disproportionately the high-severity, long-development cases that resisted resolution during the active period: large bodily injury claims with disputed causation, occupational disease claims with extended latency periods, commercial property claims with ongoing coverage disputes. The easy cases have mostly closed. The hard cases are the ones remaining.
This selection effect means that the average cost per open claim in run-off is higher than the average cost at the time the book was active. Staffing decisions made based on average claim counts from the active period will underestimate the specialist capacity needed, because the residual open claim population is more severe than the average that informed the original staffing model.
Triage on a Legacy Book: The Data Challenge
Applying triage to a run-off book is more data-intensive than applying it to new FNOL intake. Legacy open claims may have been in the system for three to seven years, with complex reserve histories, multiple adjuster assignments, and disputed coverage positions documented in unstructured notes rather than structured fields. Building a current severity triage classification for these claims requires extracting usable signals from a claims file that was not designed for systematic analysis.
The most reliable severity signal for legacy open claims is the reserve development history. A claim that has been strengthened three times in 24 months, from $50K to $150K to $300K, is almost certainly more complex than the current reserve suggests — reserve development patterns are persistent, and claims with multiple prior strengthenings are more likely than stationary reserves to require future strengthening. The reserve velocity metric we discussed earlier applies directly to legacy open claims as a triage signal.
The second signal is the presence of litigation. Claims with attorney involvement, active discovery, or scheduled trial dates are unambiguously Tier 3 regardless of current reserve level. The litigation status field in most claims management systems is a reliable binary indicator for this purpose. Generating a litigation-flagged subset of the open claim population is usually a straightforward query, even on legacy systems with limited analytical capabilities.
Allocating Specialist Attention Efficiently
The run-off triage output should produce three categories: claims that can be resolved through fast-track settlement (low reserve, no litigation, straightforward coverage), claims that require active management but follow a standard resolution pathway (mid-complexity, unrepresented), and claims that require specialist handling (litigated, high reserve development, contested coverage).
For a book with 400 remaining open claims, a reasonable expectation from triage is that 40 to 60 percent fall into the fast-track category, 25 to 35 percent require active but standard management, and 10 to 20 percent require specialist attention. The exact split depends on the composition of the book, but the point is that it is not 100 percent specialist attention for all claims — which is what an unstratified approach implicitly assumes.
The cost saving from the triage allocation comes from the difference between fast-track handling cost and specialist handling cost, applied to the claims that genuinely qualify for fast-track treatment. If the average fast-track settlement cost is $12K per claim and the average specialist handling cost is $45K per claim, routing 200 claims to fast-track instead of specialist handling produces a direct expense reduction of $6.6M — before any improvement in settlement values from faster resolution of uncomplicated claims.
Reserve Adequacy Review Under Run-Off Conditions
Actuarial reserve review in run-off operates differently from active-book reserving. The LDF method, which we discussed in our article on LDF vs. Bornhuetter-Ferguson, is less reliable on run-off books because the development patterns derived from prior accident years may not apply to the specific claim population remaining. A run-off book retaining mostly litigated claims will show faster reserve development from current maturity to ultimate than the historical patterns from the full book suggest.
The appropriate adjustment is to segment the reserve analysis by claim type — litigated versus non-litigated, large-deductible versus standard — and apply separate development factors to each segment. Industry run-off development benchmarks, available from actuarial consulting firms specializing in run-off operations, provide external reference points for claim populations that differ from the historical book average.
Actuarial reserves on run-off books are also subject to a specific timing pressure: if the run-off operation is being managed with the goal of portfolio sale or reinsurance commutation, the reserve adequacy at the transaction date determines the economics of the deal. Buyers and commutation counterparties will apply their own reserve assessment, and the gap between the carrier's carried reserves and the buyer's best estimate reserve is a negotiated item. Run-off managers who can demonstrate rigorous reserve methodology — segmented analysis, development factors supported by claim-level data, documented rationale for deviations from industry benchmarks — are better positioned in those negotiations.
Technology Fit for Run-Off Operations
Run-off operations often retain legacy claims systems that the carrier's active business has already migrated away from. Claims data may be stored in multiple systems from different vintage periods, with varying data structures and limited connectivity. Building triage analytics on top of these legacy systems requires a different architecture than integrating with a modern PAS — often a data extraction and normalization layer that pulls claim data from multiple sources into a common format before analysis.
The investment threshold for this infrastructure work needs to be evaluated against the remaining claim volume and the expected duration of the run-off. For a book with $50M in outstanding reserves and five to seven years of expected development, a data infrastructure investment of $200K to $400K that materially improves triage accuracy and claim resolution efficiency has a clear ROI. For a book with $5M in outstanding reserves and two years of expected development, the infrastructure investment may not be justified.
The minimum viable triage capability for most run-off operations is a combination of the litigation flag query (which most legacy systems support) and a reserve development velocity report (which requires claim-level reserve change data, also typically available but not surfaced in standard reports). These two data elements alone support a credible first-pass triage allocation that can meaningfully improve claim handling efficiency without requiring full analytics infrastructure.
Conclusion
Run-off books are not inherently difficult to manage efficiently. They become expensive when treated as a uniform population requiring equal attention regardless of claim complexity. Triage-based allocation — segregating the 15% of claims that require specialist attention from the 55% that can be resolved efficiently through standard channels — is the operational change that makes run-off economically manageable without staffing at levels that the declining premium base cannot support. The data infrastructure for that allocation exists in most claims systems; what is often missing is the analytical process to extract and use it systematically.
Managing a run-off book and evaluating triage automation?
Contact us at support@riskvertx.com or speak with our team.