- The equity tranche of investment-grade corporate synthetic CDOs has historically demonstrated higher risks of potential default losses. But in our stress-test scenarios, increasing the default correlations reduced mark-to-market losses.
- In the past decade, the more senior tranches didn’t incur any direct losses and may even have been viewed as stable by some market participants.
- However, our analysis based on a valuation model and three crisis scenarios illustrated that mark-to-market losses can potentially be devastating, even for the senior tranches.
As credit spreads have generally declined over the past five years, investors have increasingly been turning their attention to more complex instruments like synthetic collateralized debt obligations (CDOs) once again in the search for yield.1 However, market participants may also wish to analyze the potential for actual defaults, changes in default correlations and credit spreads. These issues have previously impacted the market value of all tranches. Even for senior synthetic-CDO tranches, it may be worth scrutinizing these potential risks, especially if the investor might need to sell these securities during times of market distress.
CDOs: The basics
Synthetic CDOs are structured credit products with streams of income that do not originate from loans or bonds, but rather are synthetically generated by corporate credit-default swaps.2 The cash flows are distributed into various tranches with very different risk profiles, which are then sold to investors. The equity tranche incurs the first default losses and in return offers the highest premium. The mezzanine and senior tranches start incurring losses only after a certain threshold is hit. Consequently, these tranches are viewed as riskier when the corporate defaults of various issuers within the same synthetic CDO are expected to be correlated. In addition, CDO tranches are considered highly leveraged, as losses of the original index are magnified in the tranche.
More defaults, lower returns
We used MSCI’s synthetic-CDO model to perform a number of stress tests to see how three scenarios might affect the performance of an investment-grade index tranche during a hypothetical five-year period beginning on Oct. 1, 2019, as shown in the exhibit below. We assumed the investments were kept until their five-year maturity. The upfront price initially received for selling the protection and the losses incurred by the tranche until maturity were the main drivers of investment performance.3
If no default occurred, we found that all the tranches outperformed the underlying index. However, if three defaults occurred, the equity tranche underperformed the underlying index. If 10 defaults occurred, both equity and mezzanine tranches performed worse than the underlying index.
Expected cumulative return of five-year investment in CDO tranches
Hypothetical performance of different tranches of the newly released investment-grade credit index, Markit CDX.NA.IG.33, for the five-year period starting Oct. 1, 2019, through Dec. 20, 2024. Source: IHS Markit, MSCI
To better understand how the senior tranche outperformed the underlying index for any given number of defaults (up to 15),4 we studied the underlying risks by evaluating tranche positions using MSCI’s synthetic-CDO valuation model in the hypothetical stress scenarios. We modeled the three scenarios based on the 2007-2008 crisis. The first consisted of a credit-spread shock, the second consisted of the same spread shock accompanied by an increase in default correlations and the third illustrated a spread shock with a few corporate defaults resulting in direct losses. We computed the instantaneous mark-to-market losses incurred by various tranches of the Markit CDX.NA.IG.33 credit index under these crisis scenarios (see exhibit below).
CDX tranche losses under three crisis scenarios
The spread shock moves the credit spread of the underlying index by +150 basis points (bps) instantaneously; the correlation shock increases the value of the first two base correlation nodes by +20% and +10%, respectively; and the default shock makes two of the 125 constituents default with a recovery rate of 40%. Data as of Oct. 1, 2019. Source: IHS Markit, MSCI.
Some effects of the various crisis scenarios were intuitive: The spread shock increased future default probabilities on all tranches, causing mark-to-market losses; an additional increase in default correlation resulted in default clusters,5 favoring equity-tranche investors at the expense of senior-tranche investors; a default shock negatively impacted all tranches by incurring losses directly or by removing their cushion. Other effects were harder to understand intuitively, but a reliable valuation model can help to assess them properly.
We found that all tranches suffered larger mark-to-market losses than the underlying index, under the three crisis scenarios outlined, reflecting their relatively large leverage. These losses would be realized immediately if the investor needed to close the position during times of market distress. This finding suggests that, even though the senior tranche, at first glance, may have seemed safe, mark-to-market losses can be far worse compared to those of the underlying index — posing potentially significant risks for investors with leverage constraints or those who may have to sell positions during market distress.
1Rennison, J. “Revival of crisis-era creation shows strain in credit markets.” Financial Times, May 3, 2019.
2Earlier versions of synthetic CDOs, on the other hand, typically referenced mortgage-backed securities. This made it more difficult to quantitatively assess and analyze their risk profile.
3The timing of the defaults is less important, as coupons are typically fairly low on investment-grade indexes. We assume an average recovery rate of 40%.
4This is true with the 40%-recovery-rate assumption. In case the average of the actual recovery rates is below 40%, fewer corporate defaults are enough to hit the senior tranche directly.
5Since the expected number of defaults remains constant, a correlation shock decreases the chance of few isolated defaults and increases the chance of higher number of defaults. Thus, higher tranches may be hit more often, and the equity tranche is expected to incur a somewhat smaller portion of the total losses.