The Shadow of Private Credit Over U.S. Insurers

If one had to pick Warren Buffett’s single smartest investment decision as the world’s most renowned investor, it would likely be his 1967 acquisition, in Berkshire Hathaway’s early days, of National Indemnity for $8.6 million. National Indemnity is a commercial insurer that provides passenger liability and motor insurance to transportation companies that carry passengers or freight.

Insurers are a type of financial institution and, in some respects, share a similar profit model with banks. Typically, insurers sell policies to a large number of customers and pay out claims when events that fall within the agreed coverage (accidents, disasters, etc.) occur. Because they price premiums based on the probability of such events, they usually have a low likelihood of loss, and they often reinsure a portion or a large share of the policies they sell to reduce exposure and risk. Put simply, if they collect 10 billion won in premiums, they might pay out 9 billion won in claims. The details vary by product, but because insurers also have high selling, general, and administrative expenses, the profitability of the insurance business itself is not dramatically different from that of banks.

Compared with banks, the real advantage of insurers lies in their access to cash. Banks generally see a large outflow of cash at the moment a loan is made and then recover it over the agreed term (strictly speaking, under the modern banking system this is not quite accurate, as the loan proceeds are credited to a deposit account). Insurers, by contrast, collect premiums from customers up front and pay claims later. The more policies they sell and the more customers they insure, the larger the pool of cash they hold.

Cash that is not technically the insurer’s own money (it is classified as a liability in accounting terms) but in practice functions much like its own capital is called “float.” Insurers (in theory) earn returns by investing this float well. Premiums serve as a steady stream of cash inflows, while claims are paid out probabilistically over time. This allows insurers to keep only a portion of collected premiums in claim-paying accounts and allocate the rest to investment accounts. Alongside major pension funds, this is why insurers have emerged as some of the biggest players in financial markets. Berkshire Hathaway acquired multiple insurers that generate this float and used their cash flows as a form of permanent capital to make high-quality, long-term investments.

Around the turn of the 21st century, the main challenge facing insurers was a persistent downtrend in interest rates. The era of free trade that blossomed under U.S. President Bill Clinton in the 1990s dramatically boosted the manufacturing capabilities of Korea and China. Thanks to cheap labor in developing countries and pro-business trade infrastructure, low value-added factories flooded into the Asia-Pacific region. As a result, long-term inflationary pressures in the United States entered a clear downward trajectory, and the neutral rate gradually declined as well.

This posed a serious problem for the insurance industry, which primarily invested in fixed income products such as bonds (Treasuries, corporate bonds, etc.) and real estate that pay fixed interest or rent. As the neutral rate fell, bond yields declined in tandem, and rental yields on assets like commercial real estate—now classified as alternative investments—also shrank. To meet their target returns, insurers had to increase exposure to new types of assets, and the star of this shift was private credit.

Insurance regulation differs by country, but in the United States it is based on insurers’ Risk-Based Capital (RBC). RBC aggregates, in financial terms, an insurer’s (1) risk from affiliates, (2) investment asset risk, (3) insurance underwriting risk, (4) interest rate and other market-related risks, and (4) business risk to calculate the minimum required capital.

As the formula above shows, while there are slight differences depending on the type of insurer (life, property and casualty, health), investment assets are included in the covariance term. This applies the logic that, on the assumption that “bad things are unlikely to happen all at once,” total risk is less than the simple sum of individual risks. Intuitively, the probability that the Soon family’s house will be flooded plus the probability that the Jung family’s house will catch fire is much higher than the probability that both disasters will occur simultaneously.

The United States is still using the asset-class correlations defined in 2001 for RBC calculations, even though they no longer reflect reality at all. In the formula, the correlation between equities and interest rates is set to zero, which any investor knows is far from how markets actually behave. Since last year there have been efforts to adjust these assumptions to better match reality, but nothing has been finalized yet.

Insurance RBC regulation also incentivizes diversification within each asset class. For fixed income portfolios, this is done by applying a risk-adjustment factor based on the number of bond issuers held in the portfolio.

Let’s look at the fixed income portfolio of a hypothetical insurer A:

Without any risk adjustment for diversification, insurer A would need $1.36 billion of RBC capital to run a $30 billion fixed income portfolio. But because there are only three issuers in the portfolio, a diversification adjustment factor of 2.5 (for fewer than 50 issuers) is applied, and the actual RBC requirement jumps to $3.4 billion. In this way, it appears that insurers operate within a framework that controls risk not only through diversification across asset classes but also through regulatory incentives for diversification within each asset class.

According to the article above, the IMF recently warned in its Global Financial Stability Report that private credit funds now account for as much as 35% of U.S. insurers’ portfolios, and that “potential conflicts of interest and the lack of transparency” between private equity firms and the insurers they own require special attention.

This is highly puzzling. Under the current RBC framework, raising the share of private credit fund assets to 35% would incur such a steep risk-weight penalty that it should be uneconomical.

To meet the RBC capital requirements of the insurers they acquire, private equity firms transfer a portion of the insurers’ assets to reinsurers in Bermuda or the Cayman Islands. In many cases, however, these reinsurers are in fact subsidiaries of the very insurers owned by the private equity funds. The underlying risk remains the same, but by exploiting differences in capital regulation across jurisdictions, the policyholder assets of the U.S. parent do not show up as liabilities (i.e., risk-bearing exposures) on its balance sheet. The notional amount of insurance assets shifted off the books in this way is said to reach as much as $2 trillion.

In addition, instead of participating directly in private credit funds as LPs (where commitments are classified as equity investments), insurers have been able to invest in structured loan notes backed by the cash flows of private credit funds. Because these are classified as bonds with lower risk weights, insurers can satisfy RBC rules while still serving as a powerful funding source for private equity firms.

The NAIC is well aware of these practices. Starting this year, it has formalized the PBBD (principles-based bond guidance) policy and mandated its gradual implementation. As the name suggests, PBBD looks past the legal form of an investment product and classifies it based on its economic substance: if the economic outcome resembles equity, it is treated as equity; if it resembles debt, it is treated as a bond. As a result, a significant portion of insurers’ private credit fund exposures that had been treated as stable bonds will be reclassified as equity investments going forward.