June 30, 2022 - In recent years, insurance companies have increased their investment in CLO liabilities, possibly because CLO notes have lower impairments (and often higher yields) than other, equivalently-rated instruments. However, the NAIC recently proposed increasing risk-based capital (“RBC”) on insurance companies’ investments in some CLO liabilities, which could cool insurance company appetite. Importantly, this proposal is moving quickly, necessitating a recent LSTA request for sufficient time to comment thoughtfully. Below, we discuss these developments and next steps.
First, there may be good reasons to like CLO investments; as page 7 of the recent LSTA White Paper demonstrates, CLO notes have experienced far lower default rates than equivalently rated corporate bonds. BofA expanded this analysis to CMBS and global structured finance; again, CLO impairment rates tend to be well lower than their equivalently rated counterparts.
And insurance companies have increased their exposure to CLOs. According to the NAIC, insurance company investments in CLOs increased from $157 billion in 2019 to $193 billion in 2020. Insurance companies hold more than half of the CLO A and BBB rated notes. Thus, if these notes – particularly BBB notes – became less attractive to insurance companies, there could be ramifications for CLO originations generally. On a more individualized basis, insurance companies hold more than $10 billion of CLO equity, oftentimes related to investments in their CLO manager affiliates. If CLO equity investments attract materially more RBC, this could impact the universe of insurance-affiliated CLO managers.
A recent proposal by the NAIC might materially reduce insurance companies appetite for CLOs. On May 25th, the NAIC issued an exposure draft discussing the “Risk Assessment of Structured Securities – CLOs.” The NAIC says that by buying a vertical strip of CLO tranches, an insurance company could materially reduce the RBC it otherwise would be charged when investing directly in the single-B rated loans in the asset pool. As demonstrated on p. 2 of the NAIC proposal, a CLO strip would have a 2.9% risk based capital charge, while investing directly in single-B loans would impose a 9.5% capital charge. Because the credit risk of single-B loans and the credit risk of a vertical strip of a CLO invested in a single-B rated loan portfolio should theoretically be the same – more on this momentarily – the NAIC argues that investing in a vertical strip of CLO notes creates an RBC arbitrage benefit of 6.6%.
To address this, the NAIC’s Structured Securities Group “can model CLO investments and evaluate all tranche level losses across all debt and equity tranches under a series of calibrated and weighted collateral stress scenarios to assign NAIC Designations that eliminate the RBC arbitrage.” Staff also recommended increasing the highest risk capital charge from 30% to 100%.
The NAIC initially provided a comment period of 30 days – to July 9th – to comment on the proposal. That comment period is far too short for such a complex and impactful methodological shift, and in response, the LSTA requested a 75-day comment period. (The NAIC subsequently moved the deadline to July 15th.)
As the market considers its response, several observers have pondered the assumptions underlying the NAIC’s proposed methodological changes. BofA research suggests that, conceptually, investments in a CLO holding single-B rated loans might have less risk than investing in single-B rated loans directly. First, CLOs invest in a diversified pool of loans; loan investments in uncorrelated industries should have less credit risk than a single loan. Second, CLOs are actively managed; managers typically sell loans well before default and often before there is a material price deterioration. Exhibit 7 in a recent BofA report illustrates this by looking at the cumulative four-year par build for 2016 vintage CLOs. After four years, the cumulative par build of the CLO was 99% – at a time when the average annual default rate in the S&P/LSTA Leveraged Loan Index was 2.07%. Exhibit 7 demonstrates that, assuming a 2% annual default rate (matching the LLI’s experience) and a 50% recovery rate, a passive portfolio would have seen a 96% par build. Crudely flipping the par build over into loss experience suggests that the actively managed CLO portfolios saw a cumulative 1% loss over 4 years, while a diversified passive portfolio (the LLI) saw a cumulative 4% credit loss over 4 years. In addition to the benefits of diversification and active management – which provide value across the capital stack – rated notes in a CLO benefit from structural safeguards. Finally, the higher rated notes also benefit from subordinated notes beneath them.
To respond substantively, analytically and constructively to the NAIC, the LSTA is constituting an NAIC RBC Comment Letter Working Group. Interested members should contact mcoffey@lsta.org.
