Prudent Lending Restored: Securitization After the Mortgage Meltdown

Structuring for Leverage:CPDOs, SIVs, and ARSs

79 Structuring for Leverage: CPDOs, SIVs, and ARSs 3 Creating less-risky portfolios through diversification is a fairly straightforward proposition. But while textbooks suggest that leverage can be used to increase risk, too, applying leverage is all too cursorily dismissed as being hindered by the practical inability to borrow at the risk-free rate. It is hard to disagree that despite the equity premium puzzle, we rarely see returns dear enough to cover the cost of the additional leverage. Furthermore, there typically are relatively few risk-averse investors who prefer leveraged risk-return combinations, naturally constraining the application of leverage. But in a well-established benign market with low borrowing costs and reduced risk aversion, markets for leveraged positions proliferate. In the seemingly benign credit environment of the last several years, markets saw low borrowing rates as an opportunity to increase leverage, in effect striving to make investments more risky in search of greater marginal yield. Sometimes the pursuit was masked by lowering issuance costs. The effect of both lower issuance costs and lower borrowing rates is the same: increased risk and return. In a benign market, the return is immediately realized while the risk remains latent—until the market turns. That is, until now. joseph r. mason I owe special thanks for constructive comments and criticism to Richard Herring, Robert Litan, Til Schuermann, Robert Eisenbeis, and other participants at the October 2008 Brookings–Tokyo Club– Wharton Conference. After the fact, we see that risk was masked in two principal ways. First, many highly leveraged instruments were sold on the basis of correlations with underlying collateral rather than the final highly leveraged structure. Auction rate securities (ARSs) were represented as “simple student loan asset-backed securities”; structured investment vehicles (SIVs) as “simple residential mortgage–backed securities (RMBSs); and constant proportion debt obligations (CPDOs) as “simple credit default swaps (CDSs).” Of course, the increased leverage embedded in the structures heightened the credit correlations of the underlying collateral of each, which themselves were only cursorily understood. Second, market arrangements for the new products were often artificially supported by seller institutions. ARS sellers supported auctions, SIV sellers repatriated debt, and CPDOs sold CDSs to their own sponsor institutions. Of course, when risk is managed through nonfundamental and nonreported means, the correlative properties of such investments will render them explosive; therefore they may fail unexpectedly. Indeed, that is what we have seen. The rest of this chapter briefly introduces the theory of leveraged returns and risk aversion, a theory that created both supply and demand for leveraged structured products. The following sections describe three applications of structured leverage: SIVs implemented term structure arbitrage to lower funding costs for typical consumer mortgage and other assets at the expense of liquidity risk; ARSs took structured leverage to the extreme, using monthly note auctions to determine rates for even safer underlying student loan collateral; and CPDOs sold credit protection into thin markets, boosting accounting returns as market spreads necessarily declined from increased supply. Of the three asset types, CPDOs are facing steep losses and the market for those structures no longer exists after rating agencies admitted that the statistical models for evaluating the structures were mathematically flawed; SIVs have all been dissolved; and ARSs are being supported by sellers in myriad legal settlements. Throughout the descriptions are three common themes. First, investment books’ dismissal of leverage on the grounds of practical borrowing costs is correct except in a financial bubble. There is no reason that a lender would lend at a riskfree or even appropriately risky rate if the risk premium on the use of funds exceeded borrowing rates enough to enable arbitrage. Second, therefore, the initial increase in leverage has to take place in a benign market environment with a safe underlying collateral or investment type. Third, once the arbitrage is in place and risk has been dismissed, there inevitably comes a push to increase the risk of the underlying collateral or embedded leverage to heighten profits. Of course, when risk manifests—as it inevitably does—dramatic losses follow. But while policy80 joseph r. mason makers ponder the complexity of the current generation of financial instruments, the underlying cause of the crisis is still just leverage. Review of Financial Literature on Equilibrium Risk Targets Academic journals contain a voluminous literature on equilibrium risk targets. In portfolio theory, investors choose their own individual risk-return preference. Risk aversion plays a part...