Recently there has been discussion of retaining the Term Asset Lending Facility (TALF) beyond the crisis as a permanent fixture of the safety net. While I have supported the general notion of a lender of last resort for securitization funding since the mid-1990s – as securitization grew to provide significant levels of bank funding – the TALF still needs to fundamental improvements before it can address the needs of banks and the economy in this crisis, or any other.
The big problem is that the TALF makes artificial distinctions among collateral types that it will support without any thorough systematic decisionmaking process to justify its reasoning. With the discount window, we have a process that determines which collateral is acceptable that can adapt over time to new financial instruments. The TALF, however, has made arbitrary distinctions in collateral it will accept and can be expected to do so in the future.
We have already seen those distinctions affect bank strategic decisions and skew industry funding. For instance, TALF rules exclude collateral originated by issuers with bond ratings below a cutoff and also exclude collateral judged “too risky,” particularly subprime credit cards. Such arbitrary distinctions risk imposing arbitrary value judgments on the industry and separating the “haves” from the “have nots” in today’s tight funding world.
Moreover, it can be argued that collateral distinctions have no justification in the world of securitization. In securitization, engineering matters, not collateral. If you properly engineer a deal, regardless of collateral, the resulting, say AAA, securities should all bear identical risk. In the simplest form of tradeoff, low credit quality collateral is securitized with high credit enhancement and high credit quality collateral is securitized with low credit enhancement, so the risk to the investor in identically-rated bonds from each deal is identical.
So what the Fed is really saying when making distinctions among collateral types is that some deals are properly structured and some are not. The problem with their ranking, however, is that they exclude established collateral sectors like subprime credit cards – which did not contribute substantially to the crisis – and include less established collateral sectors like subprime mortgage loans – which did contribute to the crisis. That does not make sense.
A better way to distinguish among collateral classes is whether the collateral is too new (meaning too little is known about performance) to properly structure the securitization. On that basis, pay-option ARM and subprime mortgage loans should be excluded for the foreseeable future (unless they are willing to hold a prohibitively high credit enhancement, which they cannot afford or they would already be securitizing). But seasoned collateral types – including subprime credit cards) would be eligible to the extent that they have experienced prior crises and demonstrated their relative stability.
With such a facility available, the Fed could advantage or disadvantage new less-established securitization paradigms relative to established types to affect the rate of growth of the new sectors, thereby affecting the rate of innovation depending upon its contribution to risky growth.
Right now, however, the artificial distinctions have outlived their usefulness and are beginning to draw lines in the industry that will not be able to be erased if the situation lasts much longer. Bank of America’s hobbled credit card securitization ability is just a canary in the coal mine. Most diversified institutions are not greatly affected by the distinctions because of their broader earning and capital bases. But smaller, less diversified, monolines see the TALF distinctions as crucial to their business lines. Combined with the new rules that will cut credit card lender margins to the bone, those artificial TALF constraints become critical many credit card lenders’ futures.
Nonetheless, even the TALF paradigm is insufficient for a true lender of last resort for securitizations. What is really needed is a way to subsidize the mezzanine and sub-AAA investment grade securities, that make the AAA, AAA. Right now, the reason AAA aren’t selling is that they require a higher yield to reflect credit enhancement risk. Subsidizing credit enhancement for plain vanilla securitizations, therefore, could motivate a small sum to help the market help itself. Hence, a small sum of money, well placed, can do more to restore orderly markets than a large sum meant to replace those markets. But in this credit crisis where securitizing banks have been unnecessarily demonized, such crucial policy seems politically beyond reach.
† Hermann Moyse, Jr./Louisiana Bankers Association Professor of Finance, Louisiana State University, Senior Fellow at the Wharton School, and Partner, Empiris LLC. Contact information: [email protected]; (202) 683-8909 office. Copyright Joseph R. Mason, 2009. All rights reserved. Past commentaries and testimony are blogged on http://www.rgemonitor.com/financemarkets-monitor/bio/626/joseph_mason.