NAIC Flags an RBC Formula That Undercharges $150B in Loans

Life insurers now hold roughly $150 billion in residential mortgage loans against a 68 basis-point risk-based capital charge that its own regulator calls not empirical-based, unchanged since holdings were $18 billion in 2018 (NAIC RBC Model Governance Task Force, August 13, 2026). It is one line on a longer list of formula gaps the task force just put a number on.

The document is called, plainly, the List of Potential Gaps. The NAIC's Risk-Based Capital Model Governance (EX) Task Force exposed it for comment on August 13, 2026, with responses due by close of business September 29. Eighteen months into a mandate the NAIC Executive Committee handed it in February 2025, the task force has stopped describing the RBC formula's shortcomings in the abstract and started attaching dollar figures, growth rates, and worked capital examples to each one. Fifteen distinct components made the list, spanning the Life, P&C, and Health formulas, each tagged with a 2025 materiality benchmark and an internal effort rating of moderate or significant.

A Gap List With Its Own Materiality Column

Every prior RBC governance effort in the NAIC's history proceeded case by case: a working group noticed a factor looked stale, commissioned a study, and adjusted it years later, one asset class at a time. The task force that Ohio Insurance Director Judith French and Wisconsin Insurance Commissioner Nathan Houdek co-chair was built to replace that pattern with something closer to a standing audit function. Announcing the task force's creation, French described the goal as building "a more coherent framework for governing changes and modifications to the RBC formulas" (NAIC, February 2025). Houdek's framing was more pointed: a decade and a half of low interest rates had already pushed insurers toward complex, yield-seeking investments the formula was never built to price, and RBC needed to keep pace or cede ground on international competitiveness.

That mandate produced a sequencing most actuaries will recognize: principles first, gaps second, calibration third. The task force spent 2025 building the rulebook for how future adjustments get made, adopting nine guiding principles, among them Materiality, Equal Capital for Equal Risk, Objectivity, and a data-driven Process standard, on September 23, 2025, after a July exposure draft and an August stakeholder meeting (Dechert, October 2025). Only with that governance layer in place, detailed further in actuary.info's coverage of the framework's spring 2026 adoption, did the task force turn to naming the specific formula components those principles would apply to first. The August 13 exposure is that naming exercise, produced with outside consultant Bridgeway Analytics after a multi-month comment process the American Academy of Actuaries' cross-practice RBC group weighed in on with a March 16, 2026 letter of its own.

The Mortgage Loan Line That Grew Eightfold

No item on the list pairs a starker growth curve with a flatter capital charge than residential mortgage loans. Life insurers reported approximately $150 billion of residential mortgage holdings in 2025, growth the task force's own materiality table calls "significant, over 30%" in that year alone, up from $18 billion in 2018 (NAIC RBC Model Governance Task Force, August 13, 2026). The C-1 charge covering that exposure, 68 basis points, sits near the cheapest end of the entire Life RBC investment grid, close to the 0.82% charge for A-rated corporate bonds and well below the 1.52% charge for BBB-rated bonds (SS&C Technologies, 2026). That charge has not moved through eight years of an eightfold expansion in the exposure it prices.

actuary.info tracked how that 68 basis-point figure got as cheap as it currently is: a 2024 Schedule BA line item first extended the factor to affiliated mortgage funds, and NAIC's Life RBC Working Group unanimously voted this spring to extend the same treatment to unaffiliated joint ventures, partnerships, and LLCs, cutting the applicable factor from 1.75% to 0.68% for that structure. A December 2025 amendment to SSAP No. 37 added a look-through reporting option for qualifying statutory trusts, effective 2027 (SS&C Technologies, 2026). Each of those moves made the mortgage-loan charge cheaper and the reporting paths into it wider. None touched the question the gap list now poses directly: whether 68 basis points, calibrated years ago against a much smaller and more homogeneous pool of loans, still reflects the risk in a $150 billion book the document itself says ranges "from a few hundred thousand to a several hundred million dollar townhouse development," with materially different risk characteristics bundled under one factor. A capital charge built for a starter book of $18 billion and left unchanged as that book grew eightfold is, on its face, the kind of drift the task force's own Materiality principle was written to catch.

What a Coarse Covariance Assumption Costs in RBC Points

The gap list's most consequential item may not be a single asset-class factor at all, but the covariance structure that combines every risk category into one number. Life RBC covariance assumptions are "coarse, generally set at 100% or 0%," the task force's own description reads, a binary that cannot recognize partial diversification or partial concentration across asset classes carrying genuinely different risks, corporate bonds, municipal bonds, and consumer-backed securitizations among them (NAIC RBC Model Governance Task Force, August 13, 2026). The task force did not leave that observation abstract. Its worked example: a hypothetical insurer with 3% of its portfolio in common stock would see its RBC ratio drop from 446% to 384% if a 50% stock-bond correlation were applied instead of the current treatment, the same correlation assumption Solvency II already uses.

A 62-point RBC ratio movement from a single assumption, on a modest 3% equity allocation, is the kind of number that reorders capital-adequacy conversations at the board level once it appears in a regulatory exposure document rather than an internal stress test. Insurers holding meaningfully more than 3% in correlated risk assets, common stock alongside high-yield credit and real estate, would see a larger effect if the structure moved toward a continuous correlation matrix rather than the current all-or-nothing treatment. The gap list rates this item significant effort, the same tier assigned to reinsurance and structured securities, which signals the task force expects the fix to require real model-building rather than a single-factor adjustment.

Structured Securities, Grid Mismatches, and a Marked-Down Mortgage Book

Structured securities carry their own cluster of flagged inconsistencies, and here the dollar figures are large enough to matter on their own terms. Life insurers held roughly $133 billion of RMBS and $148 billion of CMBS in 2025, the ninth- and tenth-largest life asset classes respectively (NAIC RBC Model Governance Task Force, August 13, 2026). The Intrinsic Price designations governing both securities deviate meaningfully from agency credit ratings, including a maturity effect the task force sized directly: an A2-rated five-year tranche can require more than 40 times the capital of an A2-rated one-year tranche carrying the identical letter designation. Whether corporate C-1 bond factors, built for single-name credit risk, are even the right template for securitizations with that kind of structural sensitivity to tranche thickness and maturity is the second question the gap list poses for the same asset class.

A separate mismatch runs through how the grid treats identical exposure reported on different schedules. Fixed-income assets on Schedule BA are charged against a coarse pre-2021 six-class grid layered with the 2021 portfolio adjustment factor, while the same credit quality reported on Schedule D-1 draws from a granular 20-class C-1 grid built specifically for bonds. Most Schedule BA Designation 1 holdings, $27 billion, and Designation 2 holdings, $11 billion, receive a materially lower RBC charge as a result, one the task force says "can be less than 50%" of what the identical exposure would draw on Schedule D-1. A related valuation-basis mismatch compounds the effect on commercial mortgages specifically: Schedule BA commercial mortgage funds carry a $32 billion balance, 32% of it valued below cost with an average markdown of negative 20%, yet the RBC formula applies the same bond and mortgage factors built for assets carried at amortized cost under Schedule B, ignoring that Schedule BA funds are measured under the U.S. GAAP equity method. Three separate items, one root cause: a formula built around a single accounting and reporting convention applied unchanged to assets that now flow through several.

Catastrophe Correlation and Two Charges Never Tested Empirically

The P&C and Health formulas get shorter treatment on the list, but the figures attached to them are no less pointed. Severe convective storm exposure, the peril actuary.info has covered as one of the industry's fastest-growing secondary-peril cost drivers, grew from $19 billion in 2019 to $47 billion in 2024, a nearly two-and-a-half-fold increase the task force says the Rcat correlation calculation has not been updated to reflect (NAIC RBC Model Governance Task Force, August 13, 2026). Correlations among hurricane, wildfire, and severe convective storm risk sit inside that same calculation, and a formula calibrated before SCS losses reached their current scale is, by the task force's own logic on the mortgage loan item, exactly the kind of stale assumption due for reassessment.

Two items get flagged for a reason distinct from growth or grid mismatch: they were never empirically grounded in the first place. The operational risk charge, a flat 3% applied against RBC before the operational add-on, "is not based on an empirical study," the task force states directly, even as it now totals roughly $0.6 billion for Life RBC net of the C-4a component, $14 billion for P&C RBC, and $2.7 billion for Health RBC in 2025. The task force weighed conducting a literature review to check for recent studies but ultimately judged the risk "not obviously material," a materiality call the P&C figure alone, fourteen times the Life figure, makes worth revisiting. The Health formula's H-2 underwriting risk charge draws similar scrutiny, covering $1.278 trillion of direct incurred claims and contract reserve change, with $34.2 billion in stop-loss exposure, $23.3 billion in long-term care, and a managed care credit proxy sized at $1.03 trillion, none of which the task force says has been reassessed alongside current diversification credit and managed care practice.

Reserve Credit, Modco, and the Offshore Slice

Reinsurance drew the largest dollar figures on the entire list and a significant effort rating in both the Life and P&C formulas. Life reserve credit and modified coinsurance arrangements reached approximately $2.85 trillion in 2025, with roughly $135 billion of that tied to non-reciprocal jurisdictions, offshore reinsurance relationships that fall outside the NAIC's reciprocal-jurisdiction framework for automatic credit recognition (NAIC RBC Model Governance Task Force, August 13, 2026). The task force's description is blunt about the current treatment: RBC "includes a reserve credit and recoverables charge, but they are coarse," and the growth in non-reciprocal arrangements has outpaced the framework's ability to distinguish counterparty and structural risk within that $2.85 trillion pool. The P&C formula has a parallel problem at smaller scale: ceded reserves to non-reciprocal jurisdictions reached $42 billion in 2025, with the task force calling for the same counterparty-quality and collateral-based reassessment applied to Life.

That offshore slice of the reinsurance figure connects directly to a supervisory track already underway. actuary.info's coverage of Actuarial Guideline 55 detailed the asset-adequacy and collateral controls regulators built specifically for offshore life reinsurance, largely Bermuda-domiciled sidecar and affiliated arrangements of the kind driving much of the private-credit-linked annuity growth NAIC and Treasury officials discussed directly in a May 2026 meeting. AG 55 addresses reserve adequacy and collateral quality policy by policy; the gap list's reserve-credit item is the RBC-formula companion question, whether the capital charge assigned against counterparty and jurisdictional risk in that $135 billion non-reciprocal slice is calibrated to match. Recalibrating both in the same direction, tighter reserve-adequacy testing under AG 55 and a steeper reserve-credit charge under RBC, would compound rather than offset for carriers with concentrated offshore modco books.

What the Task Force Left Off the List

Deciding what to exclude carries as much information as what made the cut, since the effort ratings and materiality benchmarks are themselves the prioritization mechanism the nine governing principles called for. Health reinsurance did not make the priority list despite growth the task force calls worth monitoring: incurred claims plus contract reserve change totaled $1.3 trillion in 2025 against reinsurance recoverables under $30 billion, a recoverables-to-exposure ratio the task force judged too small relative to the rest of the Health formula to warrant immediate attention. Life policyholder behavior risk, an item industry commenters pushed for, was acknowledged as legitimate but set aside as too resource-intensive relative to the higher-priority items already queued. Business practices tied to concentrated investment strategies at specific carrier segments were judged better handled by individual state insurance departments than by RBC formula changes. And several P&C limitations, tail-risk lines like personal injury protection and mass tort, elevated underwriting risk at smaller companies, and catastrophe perils the Rcat adjustment does not yet cover beyond hurricane, wildfire, and severe convective storm, were flagged as plausible candidates that still need a materiality assessment before they can compete for the task force's limited capacity.

The Recalibration Sequence From Here

Comments on the List of Potential Gaps are due September 29, 2026, after which the task force will prioritize items for assignment to the working groups that actually own each formula, Life RBC, Capital Adequacy, P&C RBC, and Health RBC among them. That handoff is where the nine adopted principles get tested against real dollar figures for the first time: a Materiality threshold that screens out the $30 billion health reinsurance recoverables pool while advancing the $2.85 trillion life reserve-credit item, and an Equal Capital for Equal Risk standard applied to close a Schedule BA-versus-Schedule D-1 gap that currently prices identical credit quality at less than half the charge depending on which schedule it lands on.

For appointed actuaries and capital-planning teams, the practical takeaway is sequencing, not urgency. None of the fifteen items becomes a filed RBC change this cycle; each first needs a working group assignment, a calibration study, and typically a multi-year phase-in of the kind the NAIC used for the CLO factor overhaul and the collateral loan look-through framework already moving through Life RBC. But the gap list functions as an early-warning roadmap for where that phase-in work is headed. Carriers with residential mortgage concentrations approaching the growth curve the task force flagged, structured-securities books skewed toward longer-maturity RMBS and CMBS tranches, or reinsurance programs weighted toward non-reciprocal counterparties now have a two-year-plus runway to model what a steeper charge would do to their RBC ratio, using the task force's own worked examples, a 62-point covariance swing, a sub-50% Schedule BA discount, a factor unchanged through an eightfold growth in the book it prices, as the starting assumptions rather than waiting for a proposed factor to arrive without warning.