05/10/2010. Have CLO structures performed as designed?

Announcements > 05/10/2010. Have CLO structures performed as designed?

Discussion Article on CLOs and Their Path Through the Recent Financial Crisis

The bank loan investment team at Deerfield Capital Management LLC has been managing CLOs for almost 10 years, and the senior members of the team managed CLOs at their predecessor firm for almost two years prior to moving to Deerfield. We thought it could be beneficial to discuss our views on how well these specific versions of CDOs have fared through the unprecedented financial and economic movements and changes of the last 24-30 months.

First, let us present a generalized event timeline for the CLO market:

The new issue CLO market saw tremendous growth in the period from 2004 through 1Q-2007. This growth helped to feed institutional loan demand, which the sell side was quite happy to fill. Rapid CLO growth, along with other leveraged fund vehicles, helped to drive rapid institutional leveraged loan growth: from the end of 2003 to the end of 2007, the Credit Suisse Leveraged Loan Index grew at a compound rate of over 45%, rising from $188 billion in 2003 to 841 billion at 12/31/07.
Roughly in the third quarter of 2007, the new issue CLO market slowed drastically; by mid-2008, the market essentially stopped cold.
Market prices for leveraged loans began to decline. Interestingly, the peak in the Credit Suisse Leveraged Loan Index was in February 2007, some months ahead of the end of the CLO issuance machine.
Credit fundamentals began to reflect the oncoming recession in late-2007 and into 2008.
As the recession's impact on the U.S. economy became evident and company results began to deteriorate, rating agencies began to downgrade credits.
Market prices for leveraged loans continued to drop throughout 2008. Correlations between all capital market asset classes began to increase towards 1.0.
CCC buckets in CLOs began to increase above the deals' haircut threshold, and CCC loan prices declined precipitously (as would be expected given their most risky nature).
O/C cushions began to quickly disappear.
The result of the timeline progression thus far was the shutting off of equity distributions and subordinate management fees from CLOs.
In the first quarter of 2009, Moody's changed their ratings criteria for CLOs, incorporating the new, much more severe macro environment, and many CLO debt tranches were downgraded.
By early-2009, loan market prices began to selectively rise; this trend was solidly established by 2Q-2009.
General credit fundamentals stopped getting worse and began to improve.
Rating agencies' credit downgrades stopped accelerating; selective upgrades began to be seen.
By the end of 2009, loan market prices, particularly CCC loan prices, had drastically improved. Total return for 2009 for the Credit Suisse Leveraged Loan Index was 45%; the total return for the CCC-rated portion of the index was 79%.
By the end of 2009 and into early-2010, some equity distributions and subordinate management fees from CLOs began to turn back on.
The loan market begins to contemplate a potential NEW new issue CLO market.

We would point out three additional items related to this generalized timeline:

Moody's projects that the trailing 12-month speculative grade default rate will drop to 3.3% by the end of 2010. JP Morgan's loan default forecast for 2010 is 4% -- this is down from an annualized default rate of over 20% in 1Q-2009.
The extremely high frequency of Events of Default that occurred in the ABS CDO market and were anticipated to occur in the CLO market never materialized.
Regarding the NEW new issue CLO market, we are hearing that one deal being shown to investors currently includes a higher senior fee, a shorter reinvestment period, and a AAA tranche price talk of L+175.

Admittedly, these last two and one-half years have been a stomach-churning roller coaster ride, and events have unfolded that would have been unimaginable and quite implausible in the investment approval processes of just about any CLO investor. Military organizations and sports teams tend to have post game/post battle action reviews to consider what happened and try to learn lessons from the past events. Most investment organizations would also profit by thinking about lessons learned after major market or significant individual security movements. Thus, what might CLO managers and CLO investors have learned from the crisis? We think there are at least three lessons to take away:

First, in our view, CLOs performed the way they were designed to perform. In times of stress, cash was redirected to the more senior notes. Clearly, investors are unlikely to just assume that deals will have the former standard metrics of 2% CDR and 70% recoveries. At the bottom of the market, it seemed reasonable to think about a potential multiyear CDR series that began at 16% (then ramping down), with recovery rates at 50%. In the closest capital markets analogy to the 1930s Great Depression period, the bottom line is that these structures proved to be robust and could weather an extremely difficult financial and economic period.

Second, there were many different strategies that were employed by CLO managers during the extreme stress. Some managers sold CCCs and realized losses to reduce the CCC haircut, some purchased discount obligations and some attempted to weather the storm by refusing to take losses they believed to be uneconomic. Only the passage of time will reveal which strategy or combination of strategies was better and how each portion of the capital structure was affected; this is particularly true for equity investors. Since these are long-lived vehicles, the game is not over.

Third, many investment organizations with an undiversified business platform were staffed for continued exponential CDO asset growth, and thus suffered greatly when their subordinate management fees were shut off. These organizations also relearned the business lesson of diversification - a sole focus on one product line or one sales channel will get your business in serious trouble when the world inevitably changes.

We can probe down further to consider what characteristics of CLOs allowed them to weather the storm. A major factor would definitely be true diversification. Overall, CLO portfolios are well-diversified. In retrospect, many of the other CDO variants were not. A Diversity Score of 80 in a corporate loan backed CLO really means a score of 80, because the underlying cash flows being generated by corporate loans are not truly highly correlated. There is historical data on loan cash flows, defaults and recoveries, and this data is both robust and has a history over multiple economic cycles. Correlations are better understood in the corporate loan asset class. Obviously, some CLOs may have certain sector concentrations at times, and at the peak of the financial crisis period all capital market correlations increased, but for the most part, CLO corporate risk is a well diversified risk.

Another factor aiding the CLO structure is the manifest nature of excess spread. Unlike some other CDOs, corporate loan assets continued to produce a healthy amount of excess spread even when the credit markets and the economy took their downward plunge. Simplistically, 200bp of real excess spread can cure a lot of problems, e.g., 4% defaults in a year with 50% recovery rates can be cured by this 200bp of excess spread.

A third factor supporting the resiliency of CLOs is the fact that portfolio managers were able to purchase relatively "average" loan collateral, and the arbitrage between that loan collateral and the CLO liabilities actually worked. For the most part, mangers did not really need to gorge their portfolios with seemingly "spreadier" products like, e.g., 2nd liens, or bonds, etc. In other CDO structures with other asset collateral, e.g., investment grade bonds, high yield bonds, types of ABS, to make the initial arbitrage work, a manager needed to INITIALLY reach for more risk in some way, thus setting a portfolio up out of the box to be able to take less stress - think of having to operate much more closely to the edge of credit danger. One could generalize this concept to say that a structured product will tend to be more resilient if the underlying assets collateralizing the structure are right in the main part of the risk/return distribution - the more you are forced to play out on the tail of that distribution, the more inherently risky the structure will prove to be. Clearly, some CLO managers tried to add rocket fuel to significantly increase returns for the equity tranche, but one must remember that we are dealing in fixed income return distributions. If a credit is sufficiently risky (in a business or other way) to require equity-like returns, in our view you are almost never compensated enough for that risk, and are just imbedding that risk in the portfolio. Perhaps that kind of manager will prove to be more clever, or quicker or more prescient, but investors need to understand the imbedded risk involved.

Finally, we can offer our informed opinion as to what any possible new CLO structures might look like, as we have seen a few new deals in the market. Some common features that we have seen or heard about in these new CLOs include the following:

Much simpler cash flow structures and waterfalls
Shorter reinvestment periods to reduce reinvestment uncertainty
A higher portion of management fees skewed towards senior fees
A senior AAA tranche with pricing talk inside of L + 200 bps
Much less leverage than in the past (perhaps 5-6X)
Equity returns suggested to be in the 10-12% range

These new features are reasonable, we believe, given the recent experiences in the leveraged loan and CLO markets. Liquidity, at least in the short run, will clearly not be in the same overabundant supply as it was in the market boom. The new economic environment and the structure of "the Street" will both mean a challenging and different market. Also, CLO managers will have to be prepared to come to the table with more of an equity investment or partner closely with someone who is bringing equity money. This market may now look like other investment management markets (e.g., hedge funds) where the manager is required to have a significant commitment in the risky tranches, over and above their sensitivity to management fees.

In conclusion, the major events that have occurred over the past three years would have been extremely difficult to predict. Nevertheless, the CLO has proven to be a resilient structure that performed for the most part as it was designed. We believe that a new issue CLO market will develop due to the evidence of the structure's resiliency, however it will likely be in a different form as tentatively described above. We believe that DCM's long term credit approach is the correct approach to managing CLOs, and we will look forward to partnering with investors that know us and agree with our fundamental approach.

Some of the information contained in this document has been obtained from third parties; Deerfield believes, but cannot guarantee, that the data is accurate and makes no representations or warranties as to the accuracy of such information. This document is not an offer of any investment product described herein or otherwise. Any investment product managed by Deerfield will be offered solely pursuant to its offering memorandum or other offering document.

Back to Home page