CLO Yawn


There's a big story in the NY Times about how the financial structures being used to finance many corporate loans—so-called Collateralized Loan Obligations or CLOs—look very similar to those used to finance mortgages during the housing bubble.  Yup.  That's true. CLOs are a securitization structure, like MBS.  (If you want to know more gory details, see here.)  But that's really where the similarities end.  While the financing transactions are similar, the asset class being securitized is fundamentally different in terms of the risk it presents, and that's what matters.  The financing channel might be more vulnerable to underpricing than other financing channels because of opacity and complexity, but is the underlying asset class that matters in terms of societal impact.  This is for (at least) four reasons. 

First, corporate debt values are not correlated with each other in the way housing values are. Underpriced housing financing results in an asset price bubble because of the serially and geographically correlated nature of housing prices. When housing finance is underpriced, borrowers can bid up home prices. That in turn increases the ability of other borrowers to get credit against other properties because loan amounts are based on property valuations, and valuations are based on comparable properties.  This mechanism also works in reverse, when housing prices fall, and gets exacerbated because when a borrower defaults, the borrower usually ceases to care for the property, which pushes down the value of nearby properties.

There's no equivalent mechanism with corporate borrowing. If Firm X defaults on its loan, it doesn't affect the value of a loan to Firm Y or of Firm Y itself, any more than your default on an auto loan affects the value of my auto loan or my car.  Housing is just different.  Thus, bad loans to one company don't have the same spillover effects as bad residential mortgage loans. 

There could still be spillover effects from corporate defaults if lots of financial institutions take huge losses and are decapitalized and can't lend, but that would take a lot of losses happening at the same time, and post-Dodd-Frank there is better (and to some degree more) capital in the financial system. Doesn't mean that losses won't sting, but the system is less fragile.

Bad mortgages metastasized into a financial crisis in 2008 not just through their financialization through securitization, but also because MBS were being widely used as repo collateral, and their value as a class became uncertain.  I don't have a sense of how widely CLOs are being used as repo collateral, but there are a lot less of them than private-label MBS, so I suspect the issue is less here.  More importantly, the problem in the fall of 2008 wasn't really realized losses--only a tenth of ultimate losses on mortgages had been realized by then.  It was the uncertainty of the extent of future losses.  Those should be more predictable on CLOs.  There are many fewer underlying borrower companies, and the lack of correlation on defaults and recoveries among the underlying loans makes it easier to predict losses, even with complicated cashflow waterfalls.

Second, housing is of macroeconomic importance in a way that corporate debt is not. Part of it is the size of the market--the $12.0 trillion residential mortgage market is much bigger than the corporate debt market ($8.6 trillion total). [As an aside, this raises the question of why corporate finance is a standard law school offering, but not housing finance...]

But it's not just about size. Home equity is the major asset for most households, whereas household exposure to corporate debt markets is limited and indirect. Most households have limited investments (if any) in securities, and most of those are in equities or government debt, not corporate debt. This matters because it means that we are unlikely to see a wealth effect recession from problems in the corporate debt market.  Mian and Sufi's work has shown that the impairment of household balance sheets played a critical role in the Great Recession--as household's saw their wealth diminish, they pared back on spending. That's not a mechanism that's going to happen with corporate debt. 

Third, there's a much better market clearing mechanism for corporate debt than for mortgages:  bankruptcy.  Corporate debt obligations get cleared in bankruptcy.  If there's a discrepancy between collateral value and debt amount for corporate debt, bankruptcy's got a ready solution, namely stripping down the secured claim to the value of the collateral. But thanks to Nobelman and Caulkett, that's not possible for home mortgage debt.  Instead, the clearing mechanism for housing is foreclosure, which is a woefully inefficient process that overcorrects.  Bankruptcy does a reasonably good job of market clearing. 

Fourth, there's also the possibility of private workouts—the loan sizes in CLOs make it worth the time for CLO managers to take the time to negotiate workouts, to the extent the CLO structure allows it.  (Limitations on changes of CLO payment terms--a sort of Trust Indenture Act § 316(b) issue, but done through contractual terms—may be an issue here, however.)  The problems that existed (and exist) with residential mortgage loan servicing simply don't have an analog with CLO restructurings of underlying loans. 

To the extent I've got a systemic concern about CLOs, it is about the risk of sponsor-support resulting in insured deposits being used to bail out CLOs.  This is something that regulators will have to police--CLOs that are getting support from their sponsors should be getting it on arm's-length terms.  But this is about policing regulatory borders, not about a fundamental problem threatening the system.

All of this is to say that there might be elevated defaults, even a bloodbath in corporate loans. But the systemic consequences will be much milder than 2008. The system should be able to handle things, and if bad investments result in investors taking losses, well, isn't that how the game is supposed to work?  As long as we don't see systemic spillovers, let the losses fall where they will.