The steady drumbeat of corporate rating actions in U.S. broadly syndicated (BSL) collateralized loan obligation (CLO) collateral pools slowed down during October. While downgrades continue to accumulate, fewer widely-held companies have seen ratings lowered into the 'CCC' category and below in recent weeks. Across the notable downgrades in September through October 20th, we note only two widely held obligors (top 500) were downgraded to 'CCC+' from 'B-' and one was downgraded to 'SD' from 'CCC+', while six other issuers were downgraded to 'B-' from 'B' (table 2).

The rally in loan prices since June has abated somewhat, and average portfolio prices declined slightly in October. Across our index of 508 rated reinvesting BSL CLO transactions this month, 'CCC' category exposures ticked up modestly (15 basis points [bps]) since the end of August, increasing to 7.11% from 6.96%, while non-performing exposures declined to 0.58% from 0.60%. About 41.00% of the CLOs in our index are exceeding their 7.50% 'CCC' exposure limit, 4.30% of the CLOs have less than a 1.00% cushion in their junior overcollateralization (O/C) test, and 1.00% of the CLOs are failing, similar to last month.

Table 1

Table 2

Variance In CLO Performance And Metrics By Manager Size

We divided our index of 508 U.S. BSL CLOs this month into three cohorts based on the dollar amount of the U.S. BSL CLOs that the manager has closed since the start of the COVID-19 pandemic, as noted in the CLO Global Databank maintained by Pitchbook. Managers that have not issued a new CLO since the COVID-19 pandemic were aggregated into the third cohort. Using this framework, we found that CLO performance metrics varied notably between the cohorts.

BSL CLO manager groups:

• Group 1: More than approximately $2.3 billion of CLOs issued;
• Group 2: Between approximately $1.0 and $2.3 billion of CLOs issued; and,
• Group 3: Less than approximately $1.0 billion of CLOs issued.

Chart 1

Chart 2

We note that the Group 1 cohort started with a higher average par and junior O/C cushions. This stands to reason because the managers in our framework had issued the most deals, and thus this cohort included more newer transactions with cleaner metrics.

Additionally, we find that Group 1 managers were more likely to have purchased fixed-rate bonds in 2022 at a discount resulting in par increasing (hence, many transactions would have above target par balances a year ago). All three cohorts experienced a loss in par starting around March and April of this year, resulting in junior O/C cushion declines, though we find the Group 1 managers experienced the least amount of par loss and junior O/C cushion declines across the three cohorts.

Chart 3

Chart 4

Transactions from Group 1 managers were more likely to have had lower 'CCC' asset exposures a year ago relative to Group 3 managers, though that gap has narrowed slightly of late. Group 1 transactions experienced a slightly larger increase in 'CCC' exposures over the past year. We find some of the transactions from Group 3 managers had higher 'CCC' exposures a year ago; and therefore, were more likely to be exceeding their 7.50% 'CCC' thresholds as corporate rating downgrades into the 'CCC' category increased in momentum. It appears that this led some Group 3 managers to react by selling 'CCC' assets at a discount, resulting in greater par loss (see chart 1); consequently, they experienced a smaller increase in 'CCC' exposures relative to Group 1 managers but at a cost in par. Meanwhile, the Group 1 managers had more cushion to absorb the downgrades into the 'CCC' category before exceeding the 7.50% threshold and facing O/C test haircuts, thus facing less par loss over the one-year period.

Looking to recovery rating distributions, Group 1 transactions had lower recovery rating distributions relative to Group 2 and 3 transactions. We find that Group 1 transactions have greater exposure to the larger widely held CLO obligors (exposure to the top 250 obligors held across the average Group 1 manger transaction was 54.00% compared to 48.60% across Group 2 transactions and 50.70% for Group 3). We find these larger widely held obligors have a higher credit rating distribution with higher rating stability, more liquidity, and higher average loan prices. Meanwhile, despite the higher credit ratings, the loans from these widely held obligors tend to have lower recovery ratings, which accounts for the difference in the credit distributions and recovery distributions across the three groups of transactions (chart 3 and 4).

Appendix

The U.S. CLO Insights Index averages CLO portfolio metrics across a large sample of S&P Global Ratings' rated U.S. BSL CLO portfolios that have been reinvesting for the prior one-year period (13 months from start to end), where:

• We have fully processed trustee report information for the first 11 months of the one-year period;
• The credit ratings (including outlook and CreditWatch data), recovery ratings, and loan price values of the portfolio exposures are set to end-of-month values for each of the first 11 months;
• Data for in-progress months (i.e., the last two months within the one-year period through 13 months from start to end) are based off the latest processed trustee report information, where the ratings (including outlook and CreditWatch), recovery ratings, and loan prices values of the portfolio exposures are set to the date indicated within the in-progress months; and
• Transactions that exit the reinvestment period during the one-year period, or transactions that become effective during the one-year start-to-end period, or transactions where we do not have complete portfolio information processed across the first 11 months of the one-year period are excluded from this index.