Actuarial update shows higher near-term claims; subcommittee weighs captive-insurance options and statutory changes

Consulting actuaries told the Arkansas Legislative Council Executive Subcommittee that updated claims data through Sept. 30, 2024, show materially higher projected losses for the coming policy year and prompted renewed discussion about whether a state captive should retain some catastrophe risk.

Charlie Lehi (Per? transcribed as "per night") and Kyle Hales, principal and consulting actuary, presented the updated actuarial report. The presenters said several large claims incurred late in the 2023–24 year were not reflected in the June 30, 2024, dataset and that five claims increased in measured value by more than $27 million in one quarter. Combined with an increase in insured values of about 10.3%, the updated data produced a substantial upward revision to the projected losses for the 2024–25 renewal period.

The consultants described results from catastrophe (CAT) modeling used by the three pooled entities cited in the presentation (transcript: “M8, ASBA, and APSIT”) and a third-party CAT modeler identified as Cadence. For severe storms, modeled average annual loss (AAL) from the CAT modeling was roughly $31 million, while the historical average annual loss in the claims record was about $27 million. For earthquake and flood combined, the CAT model AAL exceeded $35 million while the historical average over the 16–23 year lookback was less than $1 million (with flood accounting for most of the modeled difference). The consultants described that divergence as an “opportunity” to consider retaining some flood and earthquake risk in a captive over time rather than paying market reinsurance prices.

When asked about the total expected annual variability of losses, the consultants said a “current ground up, unlimited loss” estimate was about $55 million and emphasized wide year-to-year variability.

Members asked operational questions about how a captive would be structured and how premiums might be allocated across school districts with different exposure profiles. The consultants said underwriting and premium allocation can be adjusted by district-level exposures (for example, flood-prone versus lower-risk areas). Representative Brooks asked whether Bentonville’s tornado losses were included; presenters said Bentonville is not included in the historical claims dataset used for the pooled analysis because that district is not on the program, and they offered to follow up.

On implementation timing, consultants and staff described two simultaneous procurement work streams: an RFP for captive management and a separate RFP for brokerage services. The subcommittee was given a planning timeline that included vendor shortlisting in February, vendor pitches in March, earliest contract awards in April, and a realistic expectation that full placement and operation could be six months or longer — with an outside-case expectation of achieving a fourth-quarter policy placement rather than a July 1 start. The presenters emphasized the need to avoid coverage gaps and said aggressive acceleration could leave the state “bare,” which they said must be avoided.

Jill Thayer of the Bureau of Legislative Research told members statutory changes will be required to allow a state-owned captive because current law reserves captive ownership to private businesses. She said a package of bills will need to be prepared to authorize state ownership and to determine which state entity would run the captive.

Members asked consultants to probe bidders on whether they could meet an accelerated July 1 implementation date and to explain reasons if they cannot. The subcommittee agreed to continue the discussion in subsequent meetings and to coordinate follow-up with staff and potential vendors.

Ending: Staff and consultants will continue due diligence, follow up on Bentonville data, refine RFP timelines, and draft statutory language options for any future captive-authorizing bills.