Even before the public health effects of the new coronavirus outbreak went global, with policy responses throwing economies and financial markets into turmoil, the European collateralized loan obligation (CLO) market was facing a challenging environment. A scarcity of collateral was putting transaction economics under pressure, while underlying corporate credit quality was deteriorating. The CLO market's role in supplying debt funding to leveraged corporates was raising questions over how the transactions might behave if the credit cycle were to turn.

Now, the twin shocks of the coronavirus pandemic and the collapse in oil prices have triggered a tightening in financing conditions and a potential wave of credit distress. While a raft of central bank and government counter-measures may provide short-term liquidity support for businesses, impaired operating performance could still pressure credit quality over time, with the duration and severity of the health emergency unknown. This could raise particular challenges for the low rated speculative-grade companies backing CLOs.

Senior Note Ratings Are Only Affected In Severe Scenarios

To provide an indication of how rising pressures among speculative-grade corporates could affect our European CLO ratings, we analyzed a typical transaction in 10 hypothetical stress scenarios of varying severity, grouped under the following themes (see table 1):

• Increase in 'CCC' category rated assets;
• Asset defaults;
• Widespread asset downgrades;
• Lower asset recovery ratings; and
• Higher correlation assumptions.

Our analysis shows that the CLO rating migration would generally be greater further down the capital structure (see figure 1). The 'BB-' rated tranche would see the greatest rating change, suffering a downgrade of one to three rating notches in all but one of the 10 scenarios. By contrast, the 'AAA' tranche rating would be resilient, and the 'BBB' rated tranche would remain investment-grade in most cases. However, in the most severe scenarios--where all of the underlying obligors are downgraded or 10% of them default--CLO ratings throughout the capital structure could fall by one to three notches.

Figure 1

Table 1

Downturn Scenarios Could Change Various Aspects Of Our CLO Ratings Analysis

A deteriorating corporate credit environment could manifest in various possible changes to the inputs and assumptions used in our CLO rating analysis. We therefore considered a variety of scenarios.

'CCC' category rated assets

CLO transaction documentation typically specifies a threshold of 7.5% for the proportion of portfolio assets in the 'CCC' rating category. In the calculation of overcollateralization ratios, CLO documentation usually assigns a lower-than-par value to 'CCC' rated holdings in excess of this level, so a significant build-up of 'CCC' rated assets could lead to failure of the junior overcollateralization test, for example (see our video "CLOs Simplified: How CCCs Impact Coverage Tests" at https://www.spglobal.com/ratings/en/research-insights/videos/20200324-clo-simplified-how-cccs-impact-coverage-tests).

The proportion of CLO portfolio assets that are rated 'B-' or in the 'CCC' category has recently been an area of market focus. Although 'CCC' exposures in most transactions have remained well below the 7.5% threshold, exposures to assets rated 'B-' are much higher and have been rising (see chart 1). At the end of February 2020, the average portfolio exposure to 'CCC' category rated assets among European CLOs that we rate was less than 3%. However, the average exposure to 'B-' rated assets was closer to 20%, and up to almost 30% in some transactions.

A greater concentration of 'B-' rated credits in the portfolio puts CLOs at greater risk of eventually breaching their 'CCC' concentration limits in the event of a corporate downturn that leads to 'B-' rated credits being downgraded into the 'CCC' category.

Chart 1

In our analysis, we investigated five scenarios in which the 'CCC' category asset exposure increased to 5%, 10%, 15%, 20%, and 25%. In general, the resulting effect on our CLO ratings was moderate: even a 'CCC' category concentration as high as 20% led to only one-notch downgrades for tranches lower in the capital structure, with no effect on the 'AAA' and 'AA' rated tranches. Of these five scenarios, only the most severe led to a two-notch downgrade and still only for the 'BB-' rated tranche.

In these scenarios we did not model the effects of a potential breach in the junior overcollateralization test. In reality, such a breach could over time have a counteracting credit positive effect for some noteholders, as the CLO transaction would pay down its liabilities and de-lever earlier in this scenario.

Defaults

The default rate among European speculative-grade corporates that we rate has been low and stable for almost a decade, at close to 2% (see chart 2). However, given the economic stresses surrounding the COVID-19 pandemic, we believe that funding channels available to more vulnerable companies may now be limited, and that many--facing a few months of very severe weakness in operating performance--may become insolvent. We therefore expect that the European speculative-grade corporate default rate could rise to 8% over the next year.

Chart 2

We investigated two scenarios in which either 5% or 10% of assets in our hypothetical CLOs portfolio defaulted. In these scenarios, the collateral par balance was correspondingly reduced, along with the CLO tranches' credit enhancement. As a result, the breakeven default rates from our cash flow analysis decreased, meaning that some of the tranches could no longer support their initial ratings.

The 5% default rate scenario indicated no rating change for the 'AAA' and 'AA' rated tranches, one-notch downgrades for the 'A' and 'BBB' rated tranches, and a two-notch downgrade for the 'BB-' rated tranche. However, the 10% default rate scenario led to an implied one-notch downgrade at the 'AAA' and 'AA' levels, with up to a three-notch downgrade for the 'BB-' rated tranche.

For these scenarios, we assumed a recovery rate for the defaulted assets of 35%, consistent with the weighted-average recovery rate we would apply in this CLO for a 'AAA' rating stress. Higher recovery rate assumptions would likely lead to a less severe implied ratings change. In addition, we made the simplifying assumption that portfolio defaults only affected our cash flow analysis but not our credit analysis. In reality, if defaults occurred on lower-rated assets in the portfolio, the scenario default rates produced by our credit analysis may decrease as these weaker assets leave the portfolio, again leading to a less severe ratings change.

Downgrades

An economic downturn or distress among corporate credits would likely lead to changes in outlook, CreditWatch placements, and downgrades among the corporate issuers we rate. Outlook changes do not directly affect our CLO analysis. By contrast, CreditWatch negative placements and downgrades for corporate obligors included in CLO portfolios directly change our assumptions about those issuers' future default propensity in our CLO credit analysis. All else being equal, they would increase the scenario default rates we assume for different CLO rating stresses.

In our scenario analysis, we considered a severe situation in which all the underlying corporates in our hypothetical CLO portfolio had their ratings lowered by one notch or placed on CreditWatch negative. The implied effect on the CLO tranches was the same as in the 10% default rate scenario, with one-notch downgrades to the 'AAA' and 'AA' rated tranches, rising to a three-notch downgrade for the 'BB-' rated tranche.

Recoveries and correlation

Our final two hypothetical scenarios explored the effect of other possible changes to our CLO ratings analysis in a corporate downturn.

The post-default recovery rate assumptions in our CLO analysis are linked to the recovery ratings of assets in the CLO portfolio. These recovery rate assumptions are also a function of the rating stress under consideration. Our lower recovery rate scenario estimated the effect of a decline in asset recovery ratings, to the extent that our weighted-average recovery rate assumptions in the CLO cash flow analysis would fall by about 10%. For example, in the 'AAA' rating stress, the weighted-average recovery rate assumption would be 32%, rather than 35%. In this scenario, the 'AAA' and 'AA' tranche ratings remained stable, with one- or two-notch implied downgrades further down the capital structure.

Finally, we calculated the potential CLO ratings change that would result from an increase in our correlation assumptions. We increased the correlation between obligors in the same industry sector to 25% from 20%, and for those in different sectors to 10% from 7.5%. Although a corporate downturn would not generally cause us to change our correlation assumptions in this way, this analysis acts as a proxy for a scenario in which sector concentrations in CLO portfolios increase over time. This could happen if CLO managers begin to shun certain distressed sectors and adjust their portfolio accordingly. In addition, we may in some cases modify our modeling assumptions. Given the geographic scope of the current health emergency, correlations that are usually observed at a domestic or regional level may be more global in nature while these extraordinary circumstances persist, so we believe it is appropriate to assess these sensitivities. This scenario led to implied downgrades of one notch throughout the CLO capital structure (except at the 'AA' level).

In Reality, The CLO Ratings Migration Would Differ Between Transactions

We intend this analysis to be broadly representative of how our ratings on European CLO structures may move in a variety of downturn scenarios. However, the estimation approach we have used includes some simplifying assumptions and limitations.

We applied the downturn scenarios to only a single hypothetical CLO transaction. While most outstanding European CLOs that we rate are similar in structure and portfolio composition, they are not identical. In reality, credit deterioration among speculative-grade corporates would affect various CLO transactions differently, depending on their underlying portfolio composition by sector and obligor, for example. In terms of the capital structure, higher or lower initial levels of credit enhancement cushion in different transactions could make ratings less or more sensitive to stress, respectively.

In addition, our analysis here is based solely on changes in the results of our CLO credit and cash flow analysis, assuming a static underlying portfolio of corporate credits. In reality, CLO managers may be able to take action to navigate a stress scenario, helping to mitigate the effect on CLO tranche ratings. CLO portfolios have historically seen less credit migration than the wider universe of speculative-grade corporate issuers that we rate, likely due to CLO managers' ability to steer away from problematic credits and potentially build par through opportunistic trades of loans trading in the 70%-80% price range.

Appendix: Hypothetical CLO Transaction And Scenario Analysis Approach

For the analyses in this article, we applied various downturn scenarios to an illustrative, hypothetical CLO transaction with a capital structure typical of European CLO transactions issued since 2013, i.e. so-called "CLO 2.0" structures (see table 2).

Table 2

The underlying collateral portfolio's weighted-average spread was 3.9% and the weighted-average coupon spread on the tranches was 2.0%. We assumed that this CLO has a fully ramped portfolio and is in its reinvestment period.

We made the simplifying assumption that each downturn scenario would affect one element of our CLO ratings analysis: either the scenario default rates (SDRs) or the breakeven default rates (BDRs). SDRs are a result of our CLO credit analysis. The SDR is an array of gross default rates in the CLO's underlying collateral portfolio, each associated with a different rating level. For example, the SDR at the 'AAA' rating level may be 65%, while the 'BBB' SDR may be 45%. Among other factors, the SDRs for a given CLO are a function of the credit quality (i.e., credit ratings), asset weighted-average life, and industry and geographic diversity of obligors in the underlying portfolio. BDRs are a result of our CLO cash flow analysis. Each CLO tranche has an array of BDRs, which represent the maximum default rate in the underlying portfolio that the tranche can withstand and still pay timely interest and ultimate principal. Like SDRs, BDRs are also associated with different rating levels. For example, we assume lower recovery rates on portfolio defaults in a 'AAA' rating stress than in a 'BBB' stress. Among other factors, BDRs could change due to a change in credit enhancement, asset recovery rate assumptions, and weighted-average spread.

In general, we determine the appropriate CLO tranche rating by finding the highest rating for which the BDR exceeds the SDR. Our various downturn scenarios either raise the SDRs or lower the BDRs, which in some cases is sufficient to lead to an implied change in the tranche rating.

Our ratings analysis makes additional considerations before assigning ratings in the 'CCC' category (see "Criteria For Assigning 'CCC+', 'CCC', 'CCC-', And 'CC' Ratings," published on Oct. 1, 2012). This effectively floors ratings at 'B-' if the criteria for assigning a 'CCC' category rating are not met, and typically affects tranches at the bottom of CLO capital structures. For this reason, the class F tranche in the hypothetical CLO described above initially has a 'B-' rating, despite having a negative cushion. We therefore also do not present scenario analysis results for this tranche.

Related Criteria

• Global Methodology And Assumptions For CLOs And Corporate CDOs, June 21, 2019
• Criteria For Assigning 'CCC+', 'CCC', 'CCC-', And 'CC' Ratings, Oct. 1, 2012

Related Research

• Recession Likely Will Spur The European Speculative-Grade Default Rate To Rise Toward 8%, March 31 2020
• European ABS And RMBS: Assessing The Credit Effects Of COVID-19, March 30, 2020
• Coronavirus Impact: Key Takeaways From Our Articles, March 27, 2020
• COVID-19: The Steepening Cost To The Eurozone And U.K. Economies, March 26, 2020
• European Corporate Securitizations: Assessing The Credit Effects Of COVID-19, March 26, 2020
• European CLOs: Assessing The Credit Effects Of COVID-19, March 25, 2020
• Global Covered Bonds: Assessing The Credit Effects Of COVID-19, March 25, 2020
• European CMBS: Assessing The Credit Effects Of COVID-19, March 24, 2020
• European CLO Performance Index Report Q4 2019, March 24, 2020
• European CLO Performance: An Interactive Look At 'CCC' Exposure And The SPWARF, March 24, 2020
• COVID-19 Macroeconomic Update: The Global Recession Is Here And Now, March 17, 2020
• COVID-19 Credit Update: The Sudden Economic Stop Will Bring Intense Credit Pressure, March 17, 2020
• Credit FAQ: Will Mortgage Payment Suspensions Related To COVID-19 Affect European RMBS?, March 13, 2020