NAIC answers Warren with a $1.2 trillion private credit perimeter and the charge behind it

The National Association of Insurance Commissioners has put a number on the private credit US insurers carry—about $1.2 trillion at the end of 2025, roughly 13% of cash and invested assets and 21% of bond holdings—and, in the same document, described the machinery it is building to charge capital against it. The vehicle was a response to Senator Elizabeth Warren of Massachusetts, who had asked about insurers' exposure to private credit, private equity and affiliated investments; what came back addresses the credit and the reinsurance and reads less like a disclosure than like a work plan with dollar signs attached.

The narrower figure is the one risk committees should carry: $544 billion in privately rated bonds, business development companies and private credit funds, which the NAIC flagged for particular attention because of their complexity or lower transparency. Those are the holdings whose capital treatment runs through a private rating rather than a public price, which puts them first in line for a framework still being drafted around them.

Life insurers' bond portfolios, where the policyholder guarantees sit, show the shift most sharply: privately placed securities were 48.4% of life industry bonds at the end of 2025, against 37.4% five years earlier, according to S&P Global Market Intelligence. Eleven points in five years is a migration rather than a drift, and it happened while the balance sheet itself was growing—US life and health admitted assets rose 4% to $10.31 trillion in a six-month stretch.

The NAIC's leaders framed the work as maintenance, telling Warren that regulators have regularly updated capital requirements, reporting standards, supervisory tools and analytical capacity rather than relying on a static framework, with solvency and policyholder protection as the constant. The posture is familiar; the venue is not, since a senator's inquiry produced a public accounting of dollar amounts and named asset types that a supervisory memo would not have, and the political channel has carried this further, faster than the comment process would have managed alone.

Size explains why the classification decisions inside this response matter at portfolio scale: 21% of the industry's bond holdings now sits in a category whose definition changed in January 2025, with another 13% of cash and invested assets alongside it. Move even a modest share of that $1.2 trillion out of bond treatment and the reported capital picture would shift for the industry, not for a position.

The principles-based bond definition that took effect in January 2025 replaces a test built on an instrument's legal form with one that looks at its underlying economic characteristics, so what the wrapper says stops deciding the answer. That reclassification does repricing work because capital treatment follows classification—one filing at a time, with no new charge adopted and no headline attached to it.

Read against the portfolio data, the sequence is hardly subtle: private placements moving from 37.4% to 48.4% of life bonds prompted regulators to rethink how those holdings get sorted, making the bond definition the first move in that response rather than an isolated housekeeping item, with the ratings framework and the actuarial guidelines as the second and third.

Valuation gets parallel treatment: the NAIC is strengthening asset-adequacy testing through Actuarial Guideline 53, with more attention to structured assets, illiquidity and Level 3 valuations—the positions priced from models rather than from a market. A similar approach is being applied to reinsurance through Actuarial Guideline 55, which turns the same scrutiny on certain reinsurance structures.

Reinsurance is where the two halves of this meet: private credit held on a life balance sheet is a capital question, while the same assets held through an offshore cession carry a jurisdiction question on top of it. Actuarial Guideline 55 goes at the assets, and a separate NAIC track is going at the venues, so a structure that once answered a reserve problem by moving liabilities now has to answer for the assets that travel with them.

The widest-reaching instrument may also be the least visible: regulators now receive supporting information on how a private letter rating was reached, and the NAIC is developing a framework to judge whether rating providers' methodologies and their mappings remain appropriate for regulatory purposes. That moves the argument upstream of the individual asset, where a methodology finding could reach every rating a provider has mapped into the formula at once—a different order of event from a disputed designation on one holding.

It is also harder to answer: a rating can pass an internal credit committee and still fail a review of how it was mapped into the formula, and when that happens the cost shows up as capital rather than as a write-down—no impairment headline, no mark on the asset, just a charge applied to the holder.

The flagged $544 billion is where this bites hardest, because those are precisely the holdings whose treatment depends on a rating rather than a traded price. Breadth is the point: privately rated bonds, business development companies and private credit funds are not obviously the same kind of exposure, yet the NAIC groups them anyway, consistent with the logic of the bond definition, and that narrows the room to argue that a structure falls outside the perimeter because of what it is called.

Sizing resists precision, which is itself informative: the NAIC's $1.2 trillion measures private credit, while Moody's put private and illiquid holdings at $807 billion at the end of 2025. The labels are not interchangeable, and the distance between the two numbers shows how much depends on where the surveyor draws the line—a capital rule is nothing but a line. Moody's supplied the concentration detail that makes the line consequential: the ten largest holders accounted for 44% of that total.

That concentration is why a framework in development can change behavior before it becomes a rule: where a small number of holders carry close to half the exposure, the framework does not need to be finished to be felt; the largest holders have to answer the question whenever it is asked, and their answers become the industry's working standard. Their practical deadline arrives earlier than the comment clock suggests.

Two tracks are already moving, and both have been visible in our reporting: statutory reporting and capital treatment for illiquid assets and offshore reinsurance sit on a short comment clock, and the charge that eventually lands in the RBC formula will outlast every deal priced ahead of it. Running alongside, a memo from the NAIC's National Meeting instructed the Life Risk-Based Capital Working Group to develop a capital charge for cessions outside reciprocal jurisdictions, putting the Cayman gap into the formula and turning recognition that jurisdictions like Bermuda have earned into a capital advantage. The Warren response is the third track, and it is the only one that names assets.

This publication has argued that the NAIC is moving faster than its own text and that offshore arbitrage will be repriced by rating and capital committees before the final RBC language is adopted; the response to Warren extends that call rather than complicating it. The bond definition is in force, the ratings framework is in development, and the actuarial guidelines set testing standards rather than charges, so the first real cost of this regime is likely to surface inside a filing—a methodology finding, a Level 3 mark that does not hold—rather than as a headline RBC change.

The general account's binding constraint is capital text rather than yield, and the text is being written now: in a ratings framework, in an actuarial guideline, and in a bond definition that has already changed what counts as a bond. The $544 billion is the figure to carry into next quarter's risk committee meetings: it is the portion of the book the NAIC has already said the existing framework cannot see clearly enough, and it will be repriced by someone's committee long before it is repriced by any adopted formula.