Few parts of the global economy have remained unscathed by the coronavirus pandemic. From a corporate credit standpoint, the greatest deterioration has been most noticeable in those sectors and industries vulnerable to falling demand, supply chains disruptions, and tighter financing conditions.

As S&P Global Ratings' corporate analysts continue to reassess the creditworthiness of speculative-grade companies, an increasing volume of the collateral backing European broadly syndicated loan (BSL) collateralized loan obligations (CLOs) have seen pressure build on tranche ratings. As of April 21, 2020, almost 88 obligors (non-financial corporations) included in Europe, Middle East, and Africa (EMEA) CLO portfolios that we rate have had their ratings lowered and/or placed on CreditWatch with negative implications (see "COVID-19: Coronavirus-Related Public Rating Actions On Nonfinancial Corporations And Affected European CLOs," published April 21, 2020).

With market sentiment fluctuating as the COVID-19 effects evolve, CLO structures and their associated documents are being reassessed to adapt to the current market landscape. Whether it may be perceived as CLOs being reactive or just simply prepared, in this publication we highlight some of the most commonly observed developments in CLO documentation, structures, and the underlying loans that they are exposed to. At the same time, we also offer insights into how certain loan or CLO features appear to be developing in the market place.

S&P Global Ratings acknowledges a high degree of uncertainty about the rate of spread and peak of the coronavirus outbreak. Some government authorities estimate the pandemic will peak about midyear, and we are using this assumption in assessing the economic and credit implications. We believe the measures adopted to contain COVID-19 have pushed the global economy into recession (see our macroeconomic and credit updates here: www.spglobal.com/ratings). As the situation evolves, we will update our assumptions and estimates accordingly.

Lower Leveraged CLO With Structural Protections To Accommodate Downside Risk

Higher credit enhancement levels and shorter reinvestment periods will likely set the CLO landscape

It is now more likely that any immediate new issue CLO structure in the current environment will look vastly different from the traditional arbitrage-driven structures that have dominated the CLO landscape for the past several years.

To address the increased downside risks on corporate downgrades, defaults, and generally wider CLO liability spreads, we are observing new issue CLOs with higher credit enhancement levels (in line with what we would expect) and shorter reinvestments. For example, CLOs attaching 'AAA' credit enhancement at levels above 40% with a one-year reinvestment and non-call period, versus a typical CLO 2.0 that would attach 'AAA' credit enhancement levels around 38% and include a four- to five-year reinvestment period. Static CLOs (those with no reinvestment period) have also been considered and appear to be gaining traction in the market place. Some of these are not fully static though and may allow limited reinvestment of unscheduled principal proceeds and/or the sale of credit risk assets, which will still allow CLO managers to passively manage the credit risk in the portfolio.

Lower rated tranches less likely to be issued for now, reducing leverage

Perhaps not surprisingly, we are seeing a reduced prevalence of lower rated issuances ('BB' and below) in CLO structures. The reasons are mainly two-fold: first, the significant widening in CLO liability spreads has substantially increased the overall cost of debt for structures where such tranches are included; second, these tranches may be under-collateralized on a mark-to-market basis.

Consequently, we also note in some instances that larger equity checks are being written into CLO structures in part to help support credit enhancement levels; of course, this also largely depends on the price of such issuances.

Raising par value test thresholds and offering shorter non-call periods may become more common

CLO structures like those described above may come with added buffers and optionality, such as par value tests being set at higher levels (and potentially offering larger cushions) and shorter non-call periods.

Par value tests set at higher levels generally would imply that the tests are likely to breach sooner, resulting in the deleveraging of the rated notes in sequential order in order to cure the test.

With regard to non-call periods, typically all CLOs are structured with debt investors shorting call options to the subordinated notes, which generally may only be exercised following the end of a two-year non-call period. Since 2015, it has been common practice for CLOs to refinance the original liabilities by issuing new ones that typically pay lower spreads after the end of the non-call period.

In the current environment where CLO liabilities have widened, we have seen proposals for shorter non-call periods. For example, this would allow CLOs issued this year to refinance as soon as the dust settles following the COVID-19 dislocation, potentially reducing their weighted average cost of debt. As we note in the next section below, a shorter non-call period may add more value to debt investors in the current market to the extent that their notes are issued at discounted levels. In this regard, market participants may consider whether these newly formulated CLOs may actually be considered as "pull-to-par" products, versus the typical arbitrage transactions that the CLO market is familiar with.

Discount on issuances cannot be ruled out

While loan prices are likely to trade at discounted rates under a current market dislocation, so too will all CLO liabilities. In our view, any CLO liabilities issued in today's environment will be issued at discounted levels, rather than at par, which typically has always been the case for investment-grade CLO notes.

The downside to this scenario is that the CLO has fewer sources of proceeds to deploy for their uses, such as allocating note proceeds towards transaction costs and expenses, crediting reserve accounts, and purchasing assets for the CLO portfolio.

The upside, in our view, is the benefit a discount on issuance provides to CLO investors in case bond prices fall because of rising liability spreads. Combined with shorter non-call periods, where debt investors are repaid at full par plus accrued interest at the time of a refinancing, these structural features are likely to help incentivize and spur new issue demand.

Conservative Covenants To Tackle Corporate Downgrades

As can be seen from chart 1 below, the S&P Global Ratings' weighted average rating factor (SPWARF) metric (for the corporates most commonly backing CLOs in EMEA) shows how credit quality has been deteriorating. As such, existing CLOs are bracing for a period of corporate credit distress, managing their portfolios by avoiding names that may be downgraded, and at the same time taking the opportunities afforded by lower asset prices to build par.

For new transactions, it appears that CLOs are adding more headroom to accommodate future credit deterioration when proposing a capital structure.

For example, we have noticed some CLOs build new structures by factoring covenants into scenario default rates (SDRs) at higher levels than those indicated by the portfolio's current credit quality. For a particular rating level, the SDR represents the minimum gross level of asset defaults that we expect a CLO tranche with that rating to be able to withstand, according to our rating criteria (see "Global Methodology And Assumptions For CLOs And Corporate CDOs," published June 21, 2019). For example, if our model generated a 62% SDR under a 'AAA' rating scenario, we are receiving requests to assess ratings assuming an SDR of 64% under a 'AAA' rating scenario. In our view, this is to accommodate the possibility of rating downgrades in the CLO's underlying portfolio between the pricing, settlement and effective date.

Chart 1

Similar protective covenants (e.g., up to 1%) have been proposed when formulating the S&P Global Ratings recovery rate, which in turn has an impact on break-even default rates (BDRs) in our analysis. The BDR measures the maximum level of defaults that a transaction may sustain while still repaying noteholders in full and on time.

More Principal Available For Protection Now:

As CLO asset prices have plunged, the rated CLO investor community understands the possibility of building par with discounted assets, which may lead to note proceeds being used to flush excess amounts to equity noteholders. To avoid this scenario and maintain sufficient principal protection in the vehicle, recent CLOs have required stricter hurdles to manage any principal leakage.

Whether it's through higher overcollateralization test thresholds or a strong percentage limitation on the amount of par that can be flushed on the effective date, it is clear that secured note investors are conscious of the need for any par creation to stay in the transaction. We would expect this trend to continue in the uncertain economic environment.

CLO Par Value Tests: Testing The Waters With Higher 'CCC' And Lower Discount Thresholds

As the market prepares for an economic downturn after years of growth, new issue CLOs, where possible and to the extent that investors are amenable, are increasingly adapting their deal documents to include some breathing space to address pressures on par value tests.

A par value test compares the ratio of aggregate collateral (numerator in the test) to the balance of respective liabilities to be repaid (the denominator) and acts as a protection feature for senior noteholders. If this test fails, CLOs are required to divert interest payments away from more junior notes and therefore, equity noteholders as well, in order to repay principal on the most senior class(es) of notes.

In calculating the numerator of the par value test, haircuts are typically applied to the par value of certain risker assets, such as those rated in the 'CCC' category and discount obligations. Generally, a loosely defined numerator would imply that the par value test is less likely to fail and therefore, less likely to trigger earlier repayment of senior notes in times of stress.

Traditionally, CLOs have specified that 'CCC' rated assets (those rated 'CCC+', 'CCC' and 'CCC-') would have a threshold of 7.5% of the portfolio (for middle-market CLOs, the threshold stands at around 17.5%). If the exposure to 'CCC' rated assets rise above this limit, then the excess of such assets above the threshold are carried at their market value in the par value test, rather than at par.

Similarly, CLOs are also allowed to purchase discounted assets (typically, up to 25%), which are also carried at a haircut in par value tests; this time at their purchase price. This is to limit the CLO having an exposure to deeply discounted collateral, which market participants may view as higher risk assets. We are observing the following features evolve with respect to both types of credits:

'CCC' rated assets:  

Lately, there are proposals to increase 'CCC' excess thresholds (i.e., the amount permitted before any haircut kicks in) from the traditional 7.5% in BSL CLOs to between 15% to 35%. This perhaps doesn't come as a surprise in the current environment, where corporate issuers continue to face downward rating pressures. By raising the 'CCC' excess threshold, par value tests are less likely to fail, causing a subsequent diversion of proceeds to senior noteholders, than if the threshold was set at the traditional 7.5% level. Our guidance criteria, "Global Methodology And Assumptions For CLOs And Corporate CDOs," published June 21, 2019, outlines how we would typically address this in our analysis when rating a CLO. In our opinion, a CLO with a relatively larger 'CCC' bucket may require additional caution because it may affect the nature of the product, effectively changing it into a distressed type of instrument.

Discount obligations:  

Discounted obligations are obligations that are purchased at a price below a pre-defined limit that is set in the CLOs' documentation.

Generally, we found that both European and U.S. CLOs that purchase assets below 80% of par are considered to be discount obligations and are generally carried at their purchase price in the numerator for par value tests. In the current environment, where significantly more leveraged loan assets may be trading below the traditional 80% mark, we have seen proposals to amend this definition to a lower percentage, for example 70%. This would provide CLOs with more opportunity to carry assets purchased at significant discount at full par. The resulting impact on the par value test would be similar to what we observed in the proposals to increase the 'CCC' excess threshold above.

By way of extension, there have also been proposals allowing CLOs to consider benchmarking loan prices to certain indexes (e.g., the S&P European Loan Index) when classifying an asset as discounted. For example, as long as an asset is trading above a certain percentage of the index, it would not have to be categorized as a discounted obligation. Our CLO guidance outlines how we would typically address this in our analysis when rating a CLO.

Print-And-Sprint CLOs Likely To Prevail (For Now)

"Timing is everything," as the saying goes, and CLOs are no exception to the rule.

Broadly speaking, a print-and-sprint involves the process of CLO issuers assembling a model portfolio of loans, issuing CLO bonds, and then ramping-up the pool of loans in a brief time (e.g., three to five weeks). Therefore, print and sprint CLOs differ from typical CLOs, which spend months ramping up warehouse portfolios from both the primary and secondary loan markets.

In our view, accelerating the formation of a CLO by locking in CLO liabilities as soon as possible, and navigating volatility to buy cheaper but fundamentally attractive loans over a shorter period of time than usually observed, may become a common feature in the current environment. A shorter price-to-close CLO process, (i.e., securing investors at pricing and ensuring a relatively quick settlement), may help alleviate some of the pressures on existing warehouses, especially for senior lenders. In this regard, we believe that print-and-sprints may become more popular.

Print-and-sprints are not a new phenomenon, and have been common practice in the U.S. CLO market at times, such as during December 2018 when bank loan prices fell.

Historically, CLO managers would rely on a warehouse facility to acquire loans and gradually build their portfolios. Today, when most of the market is trading at discounted levels, it is more attractive to print-and-sprint, or issue CLO notes before acquiring any (or very little) collateral and use the proceeds to purchase loans quickly and cheaply. This approach also mitigates the risk of getting stuck with under-collateralized warehouses.

In this regard, we believe that print-and-sprints may become more popular, though still have their challenges because investors may be reluctant to fund a CLO, which starts with a relatively small asset pool given current market conditions.

The Economics: Engineering Debt Issuance Relative To Asset Value To Better Understand The Economic Benefits To All Investors

The current market dislocation has made it more difficult to differentiate between cheap assets that may still be fundamentally attractive, and those that are distressed (see chart 2 below, which compares various European Loan Index prices against the ITraxx Crossover). As such, manager asset selection is becoming increasingly important. In either scenario, par can be built by simply buying underlying collateral at the prevailing market prices, meaning that fewer funds are required from the issuance proceeds to purchase such assets.

Lower loan prices present a particularly attractive opportunity for equity investors, who as a result may only contribute limited upfront proceeds compared to times when assets are trading near par. In the event that the economy rebounds, the returns for such investors can be significant.

We observe that proposals for structures such as the above are becoming more popular, and are paying careful attention to the sources and uses of funds to better understand the economic benefits of the CLO to all investors.

In particular, our analysis focuses on the difference between the price of the assets purchased plus the money retained in the transaction relative to the proposed amount of rated debt that is being issued. In such instances, we may cap the amount of rated note issuance to the economic value retained in a transaction, barring any other mitigant (see "Global Methodology And Assumptions For CLOs And Corporate CDOs," published June 21, 2019). Our analysis also takes into account additional features, such as structures that allow a large leakage of principal proceeds to equity investors and those that issue more debt than the amount of proceeds required to purchase the assets.

Chart 2

Frequency Switch Mechanisms And Interest Smoothing Accounts May Finally Be Put To The Test

If history truly does repeat itself, then we may want to take note.

CLOs with liabilities that pay quarterly are potentially exposed to liquidity risk given the embedded optionality that leveraged loan issuers have to reset their interest periodicity. The CLO notes are scheduled to pay interest quarterly, while the underlying assets may reset their interest period from quarterly to semi-annually, potentially exposing the transaction structure to reset risk (see "European CLO 2.0 Structures Evolve In Response To Changing Market Conditions," published June 5, 2014).

At no point has reset risk been more evident than during the 2008-2009 global financial crisis. During this period, we noted that a significant proportion of the underlying assets in several pre-crisis CLOs switched to pay less frequently than the CLO liabilities. In our view, during times of stress, borrowers are likely to look for ways to create breathing space and minimize liquidity concerns when it comes to servicing debt payments. One way of doing this is to reduce their payment frequency, especially when benchmark rates for different tenors are trading close to one another, meaning minimal opportunity cost for floating-rate payers (see chart 3 below).

Chart 3

We are observing a similar trend in the corporate space today, where corporate borrowers are notifying their lenders that they will switch to semi-annual interest payments in addition to drawing on their revolving credit facilities (see chart 4).

Chart 4

In order to mitigate the effects of these timing mismatches, all S&P Global Ratings-rated European CLO structures have incorporated interest-smoothing accounts (ISAs) and embedded triggers to switch the payment frequency on the rated notes to semi-annual under certain conditions.

ISAs effectively retain a portion of the interest proceeds received from assets that pay interest less frequently than CLO liabilities in order to help pay quarterly interest on the CLO notes.

Overall, ISAs should help smooth out interest cash flows. At its extreme, a payment frequency switch mechanism (FSM) supplements the ISA in the event that a significant majority of assets reset their interest payments to semi-annually in any single quarter. If the FSM is triggered, the CLO's payment frequency will permanently switch to semi-annual too.

So far, we have not observed any CLOs triggering a FSM.

Current CLOs Are Likely To Pull Their Effective Dates Forward

As highlighted earlier, the current leveraged loan environment appears conducive to purchasing fundamentally sound credits that are trading cheap. In such instances, this may provide CLO managers who are in their ramp-up phase with an added advantage over their peers to build their portfolios at a low price, encouraging them to declare their portfolios effective sooner than they have in the past. The other motivation, in our view, is to avoid any risk of an effective date event due to CLOs missing the opportunity to take advantage of falling asset prices, and at the same time being hit by a double whammy of downgrades and defaults. An effective date event (if triggered) will typically result in the CLO structure deleveraging.

From our perspective, all S&P Global Ratings-rated CLOs have been rated using a stable quality approach. Therefore, as long as the S&P CDO Monitor test is passing, the CLO may be declared effective any time between the closing date and the last day of the ramp-up period, as defined in the transaction documents.

In New Issue CLOs, Managers Mull Tweaks To Concentration Limits

As new issue CLOs are being redesigned, we find that changes to concentration limitations or portfolio profile tests are growing in prominence. Not surprisingly, the tradeable allowance of 'CCC' obligations is widening in CLOs, (e.g., anywhere up to 15%-35 of the portfolio balance). Participants may note that these concentration limits may not always be aligned with the 'CCC' excess threshold as discussed earlier, which in our view is a deviation from how CLOs have traditionally been structured. Combined with other developments we are seeing, such as changes in industry and single-obligor concentrations, we view these developments as a way for CLOs to navigate through the pressures in the corporate credit markets and incorporate some flexibility when testing these limitations on the effective date and subsequently avoiding a scenario of deleveraging the CLO notes (see chart 5 below, which represents the average SPWARF for the most prevalent industries in the SPWARF 100).

Chart 5

Whatever the changes, and however they are being constructed, CLOs are clearly preparing for what may prove to be an opportunistic, yet turbulent, time in the leveraged loan market.