For four years, American insurance regulators worked to tighten the rules on one particular flavor of risky debt sitting inside life insurers' portfolios. This month, they finally finished the job. By the time they did, the industry had already moved on.
The debt in question is the collateralized loan obligation: a security stitched together from bundles of corporate loans, most of it lent to companies with junk credit ratings. Regulators had grown uneasy about how much of it insurers were holding, and how thinly the reserves behind it were stretched. So the National Association of Insurance Commissioners spent years designing new capital rules meant to make CLOs more expensive to hold in bulk. When those rules finally cleared this month, covering an estimated $314 billion of insurer-held structured debt, they were, by most accounts, milder than the industry had originally feared. And insurers, in the meantime, had already redirected fresh money toward other kinds of structured debt that resembled CLOs in almost every way that mattered, except one: none of it was subject to the new rules.
It is a pattern showing up on both sides of the Atlantic right now, in different products and under different regulators, but with a common shape. Insurers holding long-term liabilities to pensioners and policyholders are reaching further into complex, hard-to-price debt in search of yield, and the agencies meant to police that behavior are consistently a step behind it.
Insurer holdings of CLOs roughly doubled between 2018 and 2022, according to NAIC data, before interest rates on the instruments grew less attractive and regulators began signaling a crackdown. What happened next is the part that should concern anyone underwriting or rating these companies: CLO growth slowed to single digits a year, but total structured securities held by insurers - including debt backed by student loans, auto payments, and even music royalties - kept climbing at a steady clip of roughly 10% annually. The industry hadn't reduced its appetite for this kind of debt. It had simply found new wrappers for it.
"They created other forms of oftentimes similar structured securities to invest in," said Aaron Sarfatti, the former chief risk officer at insurer Equitable and now a member of the Federal Reserve's Insurance Policy Advisory Committee, describing the shift away from CLOs specifically.
The incentive to do this traces back to a memo few outside the industry ever saw. In May 2022, an NAIC staffer named Eric Kolchinsky, then the agency's director of structured securities, circulated an internal analysis showing that an insurer buying every tranche of a CLO could satisfy its capital requirement with roughly one-third of what it would need to hold the underlying loans directly. Separately, Pratik Gupta, Bank of America's head of CLO research, ran his own numbers on Kolchinsky's proposed model and found it would sharply raise capital requirements for CLO tranches rated single-A, a common rating for insurer holdings. The American Council of Life Insurers pushed instead for a rival methodology developed by the American Academy of Actuaries, arguing it rested on cleaner data and a more defensible read of actual risk. Regulators ultimately adopted that version. It lowered capital requirements for single-A holdings rather than raising them. Kolchinsky left the NAIC for an insurance company at the end of last year.
Athene, the world's largest annuity seller and a unit of Apollo, offers a case study in where the money has gone instead. The company trimmed its $25 billion CLO portfolio last year, citing falling yields relative to Treasurys, while adding other structured credit - including a $676 million triple-B-rated position called Fox Hedge LP C, purchased from its own sister company, Apollo Asset Management. Had that instrument been structured as a CLO, Athene would have had to hold twice as much capital against it.
The ACLI, for its part, disputes that regulation is driving the asset allocation at all. What draws large money managers into the life insurance business, the trade group has said, is closer to "the natural fit between long-term liabilities and long-term investment grade assets" than anything about the capital rules. The NAIC's own public comment on the pattern stayed procedural, noting only that regulators "adjust their focus as market conditions and risk profiles evolve."
Total structured credit still represents only about 13% of insurer holdings industry-wide. But that average obscures a wide spread: some carriers hold almost none of it, while others - typically the private-equity-affiliated insurers that have become dominant sellers of annuities - carry concentrations heavy enough that a sharp repricing in structured credit would land unevenly and, in places, hard.
Three thousand miles away, UK insurers that specialize in pension risk transfer - buying out the liabilities of corporate pension schemes, a market expected to absorb up to £500 billion in British pension obligations over the next decade - are running a similar experiment with a similar blind spot.
An S&P Global Ratings analysis found that the most opaque category of private assets, the kind that rarely trade and lack observable pricing inputs, now makes up more than 10% of the investment portfolios at Legal & General, Standard Life, and Brookfield-owned Just Group. That figure is almost certainly an undercount of the industry's real exposure. Pension Insurance Corporation, recently acquired by Apollo-backed Athora, likely holds much of its own private credit exposure in an entirely different accounting category that S&P's methodology didn't capture.
"It's difficult for market participants to fully understand" how much private credit sits on those balance sheets, said Charles-Marie Delpuech, an S&P analyst.
Part of the difficulty is structural: UK insurers aren't required to disclose which sectors their private loans are exposed to, where the underlying borrowers are based, or whether the debt is held directly or repackaged through tranches - the same layering technique at the center of the American CLO story. Much of this capital is now funding the same borrowers reshaping credit markets everywhere: AI software companies and data centers, increasingly financed by asset managers rather than banks, with insurers' balance sheets absorbing much of the resulting paper.
Insurers defend the shift as a natural response to their own liabilities, which can stretch across decades and require assets to match. Roman Hederer, head of portfolio management at L&G, said the firm turns to private credit partly "to diversify our exposure in the longer-duration segment." L&G struck a deal last year with Blackstone specifically to expand its access to US private credit. Standard Life's Nuwan Goonetilleke, head of asset management and retirement at the firm, called private assets "a really good match" for insurers' long-dated liabilities, provided a company can source the right ones.
S&P's own stress test offers a partial reassurance: modeling a hypothetical UK insurer with roughly 12% private credit exposure, and assigning the underlying loans a realistic range of investment-grade ratings, the firm found the portfolio would retain enough capital to survive a shock on the scale of 2008. The caveat is that the test depends entirely on those ratings holding up - the same assumption regulators are now questioning on the American side of the Atlantic.
Neither story, on its own, describes a crisis. Ratings have held up and stress tests have passed, and no insurer named in either report has missed a payment to a policyholder. But together, the two episodes describe something insurance professionals are increasingly being asked to underwrite and price around. Assets keep drifting toward whatever is hardest to observe and hardest to value in a downturn, and the rules meant to contain that drift keep arriving a step behind it, on both sides of the Atlantic.
Insurance Business has tracked this drift closely over the past year, from Moody's warning that life insurers' private credit push is creating liquidity and concentration risks, to a Morningstar analysis finding illiquid asset bets rising even as the NAIC tightens oversight, to a separate report on the NAIC's revamped capital treatment of complex assets that, notably, is still working through comment periods rather than final rules. The same dynamic shows up in reinsurance, where the US Treasury has now joined regulators scrutinizing the $1.52 trillion Bermuda reinsurance market, much of it built by the same private-equity-affiliated insurers now expanding into structured credit onshore. And it shows up in the market's own pricing: hedge funds have built more than $5 billion in short positions against US life insurance stocks, a bet that the gap between disclosed risk and actual risk in these portfolios is wider than carriers are letting on.
None of this means the industry is wrong to reach for private credit. Insurers carry genuine, decades-long liabilities, and private credit genuinely does offer the long-duration yield a pension buyout or a fixed annuity needs. The ACLI's underlying argument, that this is a business model story before it's a regulatory one, has real merit. But the CLO episode and the S&P findings point to the same practical lesson for anyone pricing, rating, or reinsuring this business: the asset class insurers describe when they talk about their books keeps turning out to be narrower than the one a regulator or analyst finds once they look closely. How large that gap really is remains an open question. Whether it's manageable is one insurers, rating agencies, and regulators on both sides of the Atlantic are still working out in real time.
