From attending three consecutive NAIC national meetings and tracking every catastrophe risk subgroup exposure, we have watched how peril-specific politics delayed wildfire integration for nearly two years before model vendor consistency finally overcame regulators' objections. On March 23 and 24, 2026, at the Spring National Meeting in San Diego, three things happened in rapid succession: the NAIC formally consolidated its catastrophe oversight into a single Natural Catastrophe Risk and Resilience (EX) Task Force, the P&C RBC Working Group adopted a binding wildfire Rcat charge (Proposal 2025-20-CR), and the Catastrophe Risk Subgroup finalized the separation of earthquake and hurricane loss experience in PR100 filings (Proposal 2025-19-CR). Together, these actions represent the most consequential restructuring of P&C catastrophe risk-based capital oversight in over a decade.

This article maps the full arc of Rcat modernization: why three governance bodies became one, what the updated formula means for capital calculations, how the wildfire vendor evaluation process finally cleared ASOP No. 38 scrutiny, and what P&C actuaries filing 2027 annual statements need to prepare for now.

Why the NAIC Merged Three Bodies Into One Task Force

The consolidation addresses a coordination problem that had grown acute as catastrophe losses climbed. Before Spring 2026, catastrophe risk oversight was split across three bodies with overlapping but distinct mandates.

The Climate and Resiliency (EX) Task Force focused on broad climate policy, disclosure frameworks, and scenario analysis. It held its final meeting on February 24, 2026, and its revised charges were formally absorbed into the new task force at the March 24 inaugural session.

The Catastrophe Insurance (C) Working Group addressed insurance availability and affordability in catastrophe-prone markets. It has been formally disbanded according to the NAIC committee directory.

The FEMA coordination functions handled emergency management interactions, federal program alignment, and disaster preparedness protocols across multiple NAIC bodies.

Each body reported through a different committee chain, creating delays when proposals required cross-body coordination. The wildfire Rcat charge is the clearest example: the vendor model evaluation involved the Catastrophe Risk (E) Subgroup under the Capital Adequacy Task Force, but the policy implications of requiring carriers to hold wildfire capital touched the Climate and Resiliency Task Force's mandate on climate-related financial regulation. This dual reporting structure added months to what should have been a straightforward model validation exercise.

The new task force reports directly to the Executive Committee, the NAIC's highest governance body. Commissioner Ricardo Lara of California chairs the task force; Commissioner Mark Fowler of Alabama serves as vice chair. The elevation to Executive Committee level signals that the NAIC views catastrophe risk oversight as a top institutional priority rather than a subcommittee specialty. The task force's scope explicitly covers atmospheric rivers, wind, water, wildfires, severe convective storms, hail, hurricanes, landslides, and earthquakes.

The New Organizational Architecture

Two new working groups operate under the task force, each with focused mandates that reflect the NAIC's split between proactive mitigation and reactive peril management.

The Pre-Disaster Mitigation and Risk Modeling (EX) Working Group, chaired by Commissioner Timothy Temple of Louisiana with Commissioner Glen Mulready of Oklahoma as vice chair, carries four charges: developing mitigation model laws (including the Strengthen Homes Act), analyzing how catastrophe models assess risks to identify community mitigation priorities, coordinating with the Catastrophe Risk Management Center of Excellence on research, and building formal coordination protocols between state departments of insurance and their respective state emergency management agencies.

The Severe Peril (EX) Working Group, chaired by Director Angela Nelson of Missouri with Director Heather Carpenter of Alaska as vice chair, evaluates peril-specific protection gaps and market conditions across severe perils and leads the NAIC's national flood risk awareness initiative. This group absorbed the availability and affordability mandate from the disbanded Catastrophe Insurance Working Group, pairing it with the analytical resources of the new task force structure.

The Catastrophe Risk (E) Subgroup remains active under the Capital Adequacy Task Force and P&C RBC Working Group for technical RBC formula work. Wanchin Chou of the Connecticut Department of Insurance continues to chair the subgroup, with Jane Nelson of Florida as vice chair. Tom Botsko of Ohio chairs the parent P&C RBC Working Group. This means the task force sets strategic direction while the subgroup handles the actuarial mechanics of Rcat formulas, vendor evaluations, and annual statement instructions.

At the task force's inaugural meeting on March 24, 2026, five priority charges were adopted: implementing the NAIC National Climate Resilience Strategy for Insurance (published March 2024); serving as the coordinating body for all catastrophe risk discussions across NAIC committees; assessing financial regulatory strategies addressing catastrophe risk solvency, availability, and affordability; coordinating communications on solvency strategies and mitigation programs; and acting as catalyst for the Center of Excellence on Catastrophe Modeling and Risk Management.

An Aon presentation at the inaugural meeting reinforced the urgency: in 2025, severe convective storm losses surpassed hurricane losses for the first time, insurers face the combined pressure of rising climate-related losses and higher rebuilding costs, and the three available insurer responses (rate increases, non-renewals, and mitigation investment) require coordinated regulatory oversight to avoid further market contraction.

Wildfire Becomes a Binding Rcat Charge: Proposal 2025-20-CR

The most consequential technical action from Spring 2026 was the adoption of Proposal 2025-20-CR, which moves wildfire from an informational-only disclosure on PR027C to a binding RBC charge component. Filed by Wanchin Chou on November 12, 2025, exposed for a 60-day comment period ending January 11, 2026, and adopted on March 23, 2026, this proposal completes a vendor evaluation cycle that began with an ad hoc review group re-established in March 2025.

Before this change, carriers with material wildfire exposure reported modeled wildfire losses on PR027C for informational purposes only. Those figures appeared in annual statements but carried zero weight in the RBC calculation. The new framework makes PR027C a binding charge page, with the same structural format as PR027A (earthquake) and PR027B (hurricane). Modeled losses are required at the 1-in-50, 1-in-100, 1-in-250, 1-in-500, and 1-in-1000 return periods, with the 1-in-1000 figure shown for informational purposes only.

The binding charge structure on PR027C includes: Net Wildfire Risk (with a factor of 1.000), Contingent Credit Risk for Wildfire Risk (factor 0.018), and Total Wildfire Catastrophe Risk on either an Annual Exceedance Probability (AEP) or Occurrence Exceedance Probability (OEP) basis. Companies qualifying for an exemption may file a disclosure in lieu of model-based reporting.

The Vendor Evaluation Process Under ASOP No. 38

The evaluation followed ASOP No. 38 (Catastrophe Modeling for All Practice Areas), the Actuarial Standards Board's guidance on how actuaries should use and validate catastrophe models. The process included high-level analysis, confidential assessments, and detailed impact studies across four participating vendors:

  • Moody's Risk Management Solutions (RMS)
  • Verisk Extreme Event Solutions / AIR
  • KCC (Karen Clark & Company), joining as a new participant in this evaluation cycle
  • CoreLogic

Starting in June and July 2025, all four vendors collaborated on a second round of impact analysis using consistent exposure inputs. The group reconvened on September 25 to address feedback from the impact analysis presentations. The central finding was decisive: a comparative review of the initial 2022 assessment against the 2025 evaluation revealed that model outputs have become increasingly consistent across vendors. This convergence gave the subgroup "greater confidence in the reliability of these models and their suitability for risk management applications."

The vendor convergence matters for a practical reason. When the initial wildfire model review occurred in 2021-2022, model output divergence across vendors was wide enough that regulators hesitated to make wildfire a binding charge. If one vendor's model produced a 1-in-100-year wildfire loss three times higher than another's for the same portfolio, the capital charge would depend more on vendor selection than on actual risk exposure. The 2025 re-evaluation showed that gap had narrowed substantially, driven by vendors incorporating updated wildfire behavior data, improved defensible space and vegetation models, and refined building-level vulnerability functions after the 2017-2018 and 2020-2021 California wildfire seasons.

The approved third-party commercial vendor catastrophe models for the updated instructions now include AIR, CoreLogic, RMS, and KCC for earthquake, hurricane, and wildfire. ARA HurLoss and the Florida Public Model remain approved for hurricane only.

RBC Impact Analysis: Minimal Disruption by Design

The impact analysis included in Proposal 2025-20-CR showed that adding wildfire to the binding Rcat charge produces minimal disruption to existing RBC action levels:

RBC Action Level2024 Current Formula2024 With Wildfire2025 Current Formula2025 With Wildfire
No Action868868863863
Trend Test9744
Mandatory Control Level4411
Total Companies881881868868

No additional companies move to Mandatory Control Level or Authorized Control Level when wildfire is included in the binding Rcat charge. Two companies moved from Trend Test to No Action in 2024, a counterintuitive result likely driven by the diversification benefit of the square-root-of-sum-of-squares formula. The deliberate calibration ensured a smooth implementation path, consistent with how the NAIC has historically introduced new RBC components.

This result reflects a pragmatic regulatory choice. The wildfire charge does not introduce new data; carriers were already reporting wildfire modeled losses on PR027C. Making the charge binding simply moves existing numbers into the formula. The companies most affected are those with concentrated wildfire exposure in California, Colorado, and the Pacific Northwest, where wildfire modeled losses are material relative to surplus. For diversified multi-state writers, the square-root aggregation dilutes the wildfire charge against existing earthquake and hurricane components.

Separating Earthquake and Hurricane Loss Experience: Proposal 2025-19-CR

Adopted alongside the wildfire proposal on March 23, 2026, Proposal 2025-19-CR addresses a data quality problem in catastrophe experience reporting. Previously, hurricane and earthquake experience data were reported combined in the PR100 pages (Schedule P Part 1XX catastrophe experience). Wildfire and severe convective storm losses were already reported separately, creating an inconsistency that limited the analytical utility of PR100 filings for both regulators and the subgroup itself.

As subgroup chair Wanchin Chou explained during the November 12, 2025, meeting: the proposed change seeks to separate hurricane and earthquake losses, aligning their reporting with that of other perils. This adjustment enables both the Subgroup and the Working Group to more effectively address each risk, taking into account their unique characteristics and impacts.

The revised PR100 structure now includes separate column groups for four perils:

Column GroupPerilData Points
Columns 24AI through 28BIEarthquakeNet losses unpaid, losses incurred net (U.S. and non-U.S.)
Columns 24AII through 28BIIHurricaneNet losses unpaid, losses incurred net (U.S. and non-U.S.)
Columns 24AIII through 28BIIIWildfireNet losses unpaid, losses incurred net (U.S. and non-U.S.)
Columns 24BII through 28VConvective StormsNet losses unpaid, losses incurred net (U.S. and non-U.S.)

Each section captures accident year data from 2017 through 2026, with a totals row. Column 28C now reports total losses and expenses incurred net excluding all four catastrophe perils. This granularity enables the subgroup to track development patterns by peril, validate vendor model calibrations against actual loss emergence, and identify where specific peril reserves are developing favorably or adversely relative to initial picks.

For P&C actuaries, the separation creates a richer dataset for reserve analysis. Hurricane and earthquake development patterns are fundamentally different: hurricane claims typically develop over 18 to 36 months with litigation tails in states like Florida and Louisiana, while earthquake claims can develop over 5 to 10 years depending on the severity and complexity of structural damage assessments. Combining them in a single column obscured these distinct development patterns and made it harder to calibrate peril-specific IBNR factors.

The Updated Rcat Formula: Square Root of Sum of Squares

The binding Rcat formula on line 5 of PR027 is now:

Total Rcat = SQRT(L(1)2 + L(2)2 + L(3)2)

Where:

  • L(1) = Total Earthquake Catastrophe Risk (from PR027A)
  • L(2) = Total Hurricane Catastrophe Risk (from PR027B)
  • L(3) = Total Wildfire Catastrophe Risk (from PR027C), newly added by 2025-20-CR

Line 5a on PR027 shows an informational-only version of the same formula with a fourth component:

Informational Rcat = SQRT(L(1)2 + L(2)2 + L(3)2 + L(4)2)

Where L(4) = Total Convective Storms Catastrophe Risk (from PR027D). Lines 3, 4, and 5a remain informational only, pending the convective storm vendor model review.

The square-root-of-sum-of-squares (SRSS) aggregation assumes the perils are independent, meaning a major earthquake event and a major hurricane event in the same calendar year are treated as uncorrelated. This is a reasonable assumption for earthquake and hurricane, which are driven by entirely different physical mechanisms. The independence assumption is somewhat more debatable for wildfire and hurricane, since both can be influenced by drought conditions and atmospheric patterns, but the NAIC's actuarial analysis indicates the correlation is low enough that SRSS remains appropriate.

The practical effect of SRSS aggregation is a diversification benefit: the combined Rcat charge for a carrier exposed to all three perils is less than the sum of the individual peril charges. A carrier with $100 million each in earthquake, hurricane, and wildfire Rcat components would face a total Rcat of $173 million (the square root of $30 billion) rather than $300 million under simple addition. This structural benefit partly explains why the impact analysis showed no additional companies moving to action levels when wildfire was added.

Convective Storms: The Next Rcat Candidate

Convective storms remain the fourth peril in the Rcat framework, reported on PR027D for informational purposes only. At the Spring 2026 meeting, the joint group heard an update on the severe convective storm impact analysis, with a vendor model review expected to begin in April 2026.

This timeline follows the same multi-year path that wildfire traveled: informational reporting first, then vendor evaluation under ASOP No. 38, then binding adoption once model consistency is demonstrated. Based on the wildfire precedent (initial review in 2021, re-evaluation in 2025, binding adoption in March 2026), a binding convective storm Rcat charge could potentially be adopted in the 2028-2029 timeframe if vendor model outputs converge on acceptable bounds.

The stakes for convective storms are substantial. Severe convective storms overtook hurricanes as the costliest insured peril in 2025, and global insured losses from convective events have exceeded $45 billion for three consecutive years. A binding convective storm Rcat charge would affect a broader set of carriers than wildfire, since convective storm exposure extends across the entire central and eastern United States rather than concentrating in California and the Pacific Northwest.

One complicating factor: Proposal 2026-08-CR, exposed for public comment at the Spring 2026 meeting with a 30-day comment period ending April 22, 2026, would eliminate two convective storm exemption questions (D13 and D14) from the PR027INT exemption interrogatory. The rationale is that no clearly defined geographic areas are currently considered "prone to convective storms" in the way that earthquake fault zones or hurricane coastal corridors are defined. Removing these questions reduces ambiguity in exemption determinations while the subgroup works toward a binding charge framework.

PR027 and PR027C Instruction Updates

The adopted proposals include several instruction changes that affect how carriers complete their annual statement catastrophe risk filings:

Approved vendor models. The instructions now list AIR, CoreLogic, RMS, and KCC as approved third-party commercial vendors for earthquake, hurricane, and wildfire. Previously, wildfire was excluded from the vendor approval list. ARA HurLoss and the Florida Public Model remain approved for hurricane modeling only.

Own-model provisions. Carriers that develop proprietary catastrophe models may use them for Rcat calculations, subject to the disclosure and documentation requirements in PR027INT. The own-model option has historically been used by a handful of large carriers with in-house catastrophe modeling teams, primarily for earthquake and hurricane. The extension to wildfire opens this path for carriers that have invested in proprietary wildfire models, though most carriers rely on third-party vendors.

AEP versus OEP reporting. PR027C requires carriers to report wildfire catastrophe risk on either an AEP (Annual Exceedance Probability) or OEP (Occurrence Exceedance Probability) basis. The choice between AEP and OEP affects the capital charge because AEP captures the aggregate risk of multiple wildfire events in a single year, while OEP captures the risk from the single largest event. For wildfire, where multiple concurrent events are common (as California experienced in both 2017 and 2025), the AEP basis typically produces a higher charge than OEP at the same return period.

Exemption disclosure. Companies qualifying for an exemption from model-based wildfire reporting must complete a disclosure on line 12 of PR027C in lieu of the modeled loss data. The exemption criteria parallel those for earthquake and hurricane, based on geographic exposure concentration, reinsurance arrangements with non-affiliates, inter-company pooling structures, and insured-value-to-surplus ratios.

What P&C Actuaries Should Prepare for 2027 Annual Statements

The adopted proposals take effect for the next annual statement filing cycle, meaning carriers will need to incorporate these changes into their year-end 2026 filings due in early 2027. Several preparation steps warrant attention now.

Wildfire model selection and documentation. Carriers that have been reporting wildfire on PR027C for informational purposes already have model output. The transition to a binding charge requires verifying that the selected vendor model (AIR, CoreLogic, RMS, or KCC) produces results consistent with the carrier's actual wildfire exposure profile. If a carrier has been using a vendor for informational reporting without the same level of scrutiny applied to earthquake and hurricane models, the ASOP No. 38 documentation requirements now apply with the same rigor. This includes documenting model selection rationale, input data quality, and any adjustments or overrides to vendor defaults.

PR100 data separation. Finance and actuarial teams need to reclassify historical catastrophe experience data to populate the new peril-separated PR100 columns. Carriers with combined earthquake-hurricane experience in their systems will need to split that data retroactively back to accident year 2017. The data quality of this historical split depends on whether the carrier's loss coding system already distinguishes earthquake from hurricane at the claim level, or whether the split requires manual allocation based on event identification.

Rcat formula recalculation. Appointed actuaries should run the updated three-peril Rcat formula against current data to understand the magnitude of the wildfire charge and its interaction with existing earthquake and hurricane components. For carriers with minimal wildfire exposure, the impact may be negligible. For California-concentrated homeowners writers, the wildfire component could be the largest of the three peril charges, and the SRSS aggregation benefit will be the primary factor keeping total Rcat within manageable bounds.

Reinsurance program alignment. The wildfire Rcat charge is calculated net of reinsurance. Carriers should verify that their catastrophe reinsurance programs appropriately reduce the wildfire component of Rcat, including checking whether excess-of-loss treaties cover wildfire as a named peril or only as part of an all-peril catastrophe cover. The contingent credit risk factor of 0.018 on PR027C captures the counterparty risk associated with reinsurance recoveries on wildfire claims.

Why This Matters

The consolidation of three governance bodies into one task force is more than an organizational chart revision. It accelerates the path from catastrophe model approval to capital formula adoption by eliminating the coordination gaps that delayed wildfire integration for years.

For pricing actuaries, the binding wildfire Rcat charge creates a direct link between catastrophe model output and required capital. Carriers that underinvest in wildfire modeling accuracy will either hold insufficient capital (if the model understates risk) or hold excessive capital (if the model overstates risk). Both outcomes have competitive consequences: insufficient capital invites regulatory scrutiny, while excessive capital reduces return on equity. The four-vendor consistency finding gives pricing actuaries greater confidence in cross-vendor benchmarking, but it does not eliminate the need for careful model validation against actual loss experience. Swiss Re sigma data showing wildfire losses growing 12% annually should inform trend assumptions applied on top of vendor model outputs.

For reserving actuaries, the peril-separated PR100 data creates new opportunities for development pattern analysis. Tracking earthquake, hurricane, wildfire, and convective storm development separately enables more precise IBNR estimates for each peril class. This is particularly relevant for wildfire, where the claims development pattern differs meaningfully from hurricane: wildfire claims often involve total losses with relatively short development, but environmental remediation costs and utility litigation (as seen in the PG&E and Edison cases) can extend tails for years beyond the initial loss event.

For appointed actuaries signing statutory opinions, the binding wildfire Rcat charge adds a new dimension to the capital adequacy assessment. The NAIC's impact analysis showing no new companies at action levels is reassuring for the initial implementation, but appointed actuaries should consider stress scenarios where wildfire losses exceed vendor model expectations. The California regulatory shift toward forward-looking catastrophe models in rate filings means the same wildfire models feeding Rcat calculations will also drive filed rate levels, creating a consistency requirement between pricing and capital adequacy.

The broader signal is that the NAIC is moving toward a comprehensive multi-peril Rcat framework. Wildfire took roughly five years from initial informational reporting to binding charge. Convective storms are now entering the vendor review phase, following the same path. The task force structure, with its direct Executive Committee reporting line and dedicated working groups, is designed to compress future timelines. Actuaries at carriers with material catastrophe exposure should assume that every peril currently on PR027 in informational status will eventually become binding, and plan their modeling and capital management accordingly.

Further Reading