Policy makers and lawmakers are tossing this notion of "risk retention" around as if it were a magic pill, but it could be a fatal potion for any future funding of essential consumer and commercial lending via the securities markets - particularly if taken in combination with amendments to securitization accounting and bank regulatory capital requirements. Risk retention has been at the center of the debate over financial system reforms at least since the Administration released its white paper last March.
One of the most significant problems in the securitization markets was the lack of sufficient incentives for lenders and securitizers to consider the performance of the underlying loans after asset backed securities (ABS) were issued. Lenders and securitizers had weak incentives to conduct due diligence regarding the quality of the underlying assets being securitized. This problem was exacerbated as the structure of ABS became more complex and opaque. Inadequate disclosure regimes exacerbated the gap in incentives between lenders, securitizers and investors. The federal banking agencies should promulgate regulations that require loan originators or sponsors to retain five percent of the credit risk of securitized exposures. The regulations should prohibit the originator from directly or indirectly hedging or otherwise transferring the risk it is required to retain under these regulations. This is critical to prevent gaming of the system to undermine the economic tie between the originator and the issued ABS.
This is an over simplification. In particular, it puts securitizations of consumer assets like credit cards, student and auto loans or commercial assets like equipment loans and leases in the same bucket as subprime and alternative underwriting criteria MBS or securitizations of securitizations like CDOs and SIVs. The former are widely understood to play an important role in reopening credit to households and business while the market for the latter continues to be dead. It also overlooks the fact that in many many transactions and in entire ABS subsectors, sponsors did/do retain a portion of the risk in the form of a first loss, residual or other substantive interest. Sometimes because it is to the sponsor's economic advantage to do so, sometimes because enough investors in the specific market segment demand that they do. But I am not writing today to argue that the question of risk retention should be settled by the marketplace. That might be as naive as demanding across the board that sponsors "retain an economic interest in a material portion of the credit risk of securitized credit exposures." And it's probably a losing argument. A provision of H.R. 4173, Wall Street Reform and Consumer Protection Act, passed last week by the House, calls for a 5% risk retention requirement and includes language that would permit retention requirements to be customized to reflect loans with lower credit risk. (This, as we discuss below, is on the right track.) The Senate is expected to begin debating a Restoring American Financial Stability bill in the new year that calls for 10% risk retention. Another shoe dropped this week when the FDIC issued an advance notice of proposed rule making that could require sponsors of securitizations begun after March 31, 2010 to retain a 5% economic interest in order to qualify for safe harbor in event of an FDIC takeover. (The safe harbor would protect investors from the FDIC going after assets in securitizations. A new rule was required by the loss of sale treatment and potential consolidation of securized assets on sponsor balance sheets as a result of implementing FAS 166 and 167. Securitizations completed or begun before March 31, 2010 are grandfathered in. My Kitchen Sink column in the upcoming January issue of Housing Wire Magazine details this and other outcomes of the new securitization accounting rules.) At the same time, the FDIC finalized rules proposed in August by banking regulators to require banking institutions to reflect newly consolidated securitizations in calculations of measurement of risk-weighted assets. Ergo, risk-based capital requirements are expected to rise. Game Changer If assets are consolidated, the economics of securitizing for sponsors changes dramatically. Matt Jozoff and team in J.P. Morgan's US Fixed Income Strategy group have taken a closer look at this effect in their "Securitization Outlook," a special topic report published December 11. In fact, capital charges are at least eight times higher for a credit card ABS given on-balance sheet treatment. Explains Jozoff: before the accounting changes, a bank might sell the triple-A and retain the single-A and triple-B tranches. Given a 50% risk weight on single-A and 100% on triple-B and an 8% capital charge, the all-in capital charge for issuing would be about 1%. By contrast, consolidating the credit card receivables results in 100% of the 8% capital charge. The reserves required against those credit card receivables add incremental capital costs as well. Under FAS 167, not all securitizations will be consolidated on their sponsors' balance sheets (for background on this curious result, see my August and January HW Magazine columns). Most analysts expect credit card securitizations, SIVs and other bank managed funding conduits to be consolidated. By contrast, whether private MBS come back onto the balance sheet or not will depend on whether the banks services the loans and has a significant risk position such as the residual. (Note, this is not a complete list of the types of bank-sponsored securitizations that must be evaluated for consolidation under FAS 167.) However, risk retention could force consolidation, particularly if it is specified as a first loss risk (the FDIC proposes a vertical slice of the securitization or a representative sample of the assets). Jozoff observes that while there seems to be broad agreement risk retention is a good idea, little thought has been given to what an optimal amount might be. I agree - 5% and 10% seem arbitrary levels, plucked from the air. According to Jozoff this offers significant risk of "regulatory overshoot, where policies are implemented without regard to consequences." A particular consequence Jozoff fears is the continued shutdown of the private label MBS market. He's not alone in fearing this outcome. Jozoff cites independent articles from the IMF and BIS both of which found that the optiminal risk retention amounts are dependent on the asset being securitized. In particular, Jozoff quotes from the IMF report
... the optimal retention scheme, defined in terms of which tranches are retained and their thickness, depends critically on reasonable assumptions about the quality of the loan pool and the economic conditions during the life of the securitization.
... a securitizer that is forced to retain exposure to an equity tranche backed by a low-quality loan portfolio when an economic downturn is highly probable will have little incentive to diligently screen and monitor the underlying loans, because the changes are high that equity tranche holders will be wiped out irrespective of any screening and monitoring.
The take-away for Jozoff is that it can be argued that, "had the 5% retention scheme been in place as early as 2006, the subprime boom would have happened exactly as it did, as losses have far exceeded the 5% threshold." Implementing a retention policy defined by the type and quality of the assets would be far more effective, but it would also be more complex and time consuming to devise. But there is time for this - the market response has effectively shut down for now the market excesses retention policies seek to prevent.