Warren A Sherman wandrewsherman@kamakuraco.com
Mark Slattery mslattery@kamakuraco.com
Eric Penanhoat epenanhoat@kamakuraco.comNEW YORK, November 18, 2020: “Investing in CLO Combo Notes allows the investor to invest in lower-rated CLO debt with high-yield returns, while achieving a nominally investment grade rating.”Through creative structuring and the legerdemain of agency rating models, Combo Notes are consistently awarded a higher rating than their underlying CLOs. The underlying loan collateral is identical, but Combo Notes restructure the tranches so that monthly interest payments are applied to pay down principal rather than disbursed as coupon payments to Combo Note holders.
This permits insurance investors to buy and hold lower-rated CLO collateral at ratings grades that avoid higher corresponding capital charges. With a nod to the great Stanley Kubrick, it seems there’s a certain amount of willful visual impairment going on here.At the least, ratings agency attitudes toward CLOs and Combo Notes have been quixotic and are sometimes inexplicable.
While the idea of pooling and securitizing loan debt has considerable appeal from an investor perspective, the true risk of correlated default must be accurately evaluated. During the 2007-8 financial crisis, more than 13,000 AAA rated CDOs defaulted. To avoid a repetition, the risk analysis must proceed from the individual loan level and loan obligor default probabilities, rather than from reliance on the estimation of Conditional Default Rates (CDRs), or credit ratings assigned to the senior tranches of CLO structures. It is common for the junior tranches not to be rated at all, and ratings for individual loans are infrequent, given that many of the loan obligors are unrated.
In our approach, the functional relationships between relevant macro variables and loan defaults are first determined, then applied in large-scale Monte Carlo simulations by which the impact of changes in selected factors on loan default probabilities can be assessed. The best practice in loss simulation is that in which selected macro factors are the drivers, rather than ratings transitions (which is the NAIC standard), or historical sector default rates, as is often done.
Determining risk factor coefficients for macro factors provides a mathematical link between the factors and loan obligor default probabilities. While independent, the factors exhibit correlated movement and, by means of a modern simulation engine, the correlations themselves can be user manipulated and stress-tested. Bad correlation assumptions were at the heart of the financial crisis and its securitized mortgage debt. (Perhaps it’s an opportune time to watch “Margin Call” once again!)
The argument by which the reduction of principal constitutes a credit strategy is elusive. Certainly, there’s a calculable impact on LGD and recovery rates. But the loan collateral is the same, except for the addition of Treasuries for Combo Notes when that’s the case. Obligor and loan defaults are subject to the same macro factor impacts, so there’s no reason to assume that loan collateral will be more resilient.
Of course, this is an example of regulatory arbitrage, since the NAIC assesses capital reserves based on ratings assignments. Note that most Combo Notes comprise mainly unrated CLO tranches, so investors aren’t even getting the benefit of ratings agency scrutiny down the stacked tranches.
Interestingly, CLO issuance has increased over the year, despite sharply dropping sales and revenues for many industries typically favored in securitized structures (Figure 1). According to Bloomberg,October’s sales of $13.8 billion made it the busiest CLO sales month since early 2019.
