RMBS Should Pool Loans from Different Originators
October 1, 2013
While Qualified Mortgage and Qualified Residential Mortgage Rules are designed to increase mortgage quality, investors must be cautious about privately issued securities in a post-GSE era where the underlying loans are subjected to a particular lender’s underwriting and sourcing processes.
Private-label securitization exposed investors to both borrower credit risk by way of significant relaxation of underwriting standards and a more insidious lender process risk. It was only a matter of time before poor loan manufacturing processes ill-equipped to handle riskier loans would lead to heavy losses for investors. With momentum building in Congress for a dominant role for private issuers in mortgage secondary markets, controls must be designed into any new securitization process to mitigate excessive exposure to lender process deficiencies.
While a borrower’s ability to pay, willingness to pay and equity stake (down payment) are critical indicators of the likelihood that a mortgage will default, how that loan is made also factors prominently into the outcome. During the years leading up to the mortgage crisis, private-label securities were created from pools of mortgages originated by individual lenders or their affiliates applying their underwriting guidelines. Over the passage of time, lenders not only relaxed their underwriting standards, but also undertook efforts to “streamline” their originations processes in order to place themselves in a more competitive position on pricing and service to the market.
As a result, critical tools in assessing borrower and collateral quality became less effective in identifying risk while providing significant cost efficiencies. Appraisals performed by qualified appraisers independent from production incentives increasingly gave way to automated valuation models, drive-by rather than full appraisals and appraisers effectively captured by production units.
Recently I conducted a statistical analysis of mortgage default, using actual loan-level data representative of industry prime originations during the years 2002-07. This analysis showed that, after controlling for borrower credit, ability-to-pay, down payment and a number of other loan, borrower, geographic and economic risk factors, loans originated by the 10 largest originators during the period were zero to five times more likely to default, depending on the originator. In other words, mortgage default is not simply a function of the borrower’s risk profile. And in fact, a lender’s process can severely distort the underwriting picture.
Take for example, a stated income loan where a borrower told the loan officer that they had an annual salary of $150,000 when in reality it was $100,000. The debt-to-income ratio, a key risk factor in the underwriting process in this case, was not informative of the borrower’s risk. The result was that the private-label securitization process exposed investors to massive loan manufacturing defect risk that could not be teased out of a poor external ratings process or from any cursory due diligence performed. Default risk was therefore amplified by poor underwriting processes.
To be sure, QM strengthened regulatory oversight and, eventually, QRM will reduce such effects but cannot eliminate them entirely. Capping debt-to-income ratios at 43%, limiting loan-to-value ratios, requiring full documentation of income and assets and holding appraisers to industry standards of objectivity and performance mask deficient loan origination processes. It is relatively easy to originate high-quality mortgages even with poor processes and maintain low default experience on these loans. However, eventually, as the imprint of the pain from the crisis recedes years from now, there will once again be a call to relax underwriting standards in order to promote a more “robust” housing market and ensure broader access to credit markets.
As a result, investors should worry about how much lender process risk they carry in a security. One way to address this concern would be to pool loans across different originators in a way that diversifies lender process risk. Under either a utility or government corporation model, the securitization process could add such a step that creates balance across lenders in loan pools. And lender issuers would remain on the hook for credit losses experienced in the underlying pool from their loans. Moreover, such a process would still work under various risk-sharing arrangements, such as senior-subordination structures.
It is easy to blame much of the damage from the mortgage crisis on loose underwriting standards, and while it is certainly a major contributing factor, significant defects in the loan manufacturing process made things worse.
Clifford Rossi is the Professor-of-the-Practice at the Robert H. Smith School of Business at the University of Maryland. This article originally appeared in the American Banker.