Declining underwriting standards for mortgage loans played a major role in the housing boom and bust that culminated in the financial crisis of 2007-2009. During the housing boom in the years preceding the crisis, housing prices were rising rapidly and nearly all loans, irrespective of underwriting standards, performed well. In part due to this strong performance, underwriting standards became progressively more lax. Other reasons for lax underwriting standards included a major policy public objective to expand the rate of homeownership, hefty demand for mortgage securities by professional investors, and the securitization process itself in which the mortgage lender was generally not the final investor in the loan, or the security formed by a number of loans. In addition, there was a widespread belief that “traditional standards” (at least 20% down payment, the ratio of the monthly payment to income (known as “debt to income” or DTI) less than 28%, and FICO credit score greater than 700) were too strict.
The financial crisis provided a severe test of weaker underwriting standards and revealed weakness of mortgage loans created during this period. Falling home prices led an initial wave of defaults, particularly among owners of second homes. This triggered losses on complex mortgage securities that were held by financial intermediaries, including banks and hedge funds, often in highly leveraged accounts funded with short-term wholesale funds. Falling security values severely weakened these levered balance sheets eventually causing widespread financial failures and ultimately a very serious recession. This put massive further downward pressure on housing prices and led to a much more serious wave of defaults.
In retrospect, maybe the push toward lower underwriting standards was ill-advised. One effect is to increase demand for homes and therefore home prices. Other things equal, this means housing is less affordable. Further, if this effect is offset by liberal credit, the risk is that people take on financial burdens which are not feasible in the absence of continued housing price appreciation. If this argument holds water, maybe a useful move is to move in the other direction. By assuring high quality loans, the ability of private markets to fund these loans, either through portfolio lending by banks or private label securitization, is heightened.
One of the lessons many people took from the crisis is the need for mortgage lenders to retain a portion of credit risk on loans that they originate and sell. The Dodd-Frank Act (DFA) includes a requirement that lenders retain 5% of any securitization, unless the underlying loans meet the “Qualifying Residential Mortgage” (QRM) test. The QRM provides a safe harbor against risk retention by the lender. It would seem natural to establish conservative underwriting standards for the QRM.
So, what constitutes a reasonable standard for a QRM? This is a matter for regulators to address, but some very useful data have recently been released by Freddie Mac (the Federal Home Loan Mortgage Company). This data shows default rates for loans purchased or guaranteed by Freddie Mac from 1999 through 2011. Even through this very stressful and turbulent time period, loans that featured at least 10% down payment, a payment to income ratio of less than 36% and a borrower credit score north of 660, performed quite well with default rates less than 1%. Meanwhile, loans with lower down payment combined with a higher ratio of payment to income and lower credit scores performed much worse, as much as 20 times worse. One reason this data is interesting and revealing is that it shows that origination standards looser than the traditional 20% down, 28% payment-to-income ratio and 700+ FICO are consistent with excellent credit performance through a very difficult period.
Has this information provided a standard for a QRM? No, perhaps in deference to pressures from housing, lending and community organizing groups, regulators seem to be set on a path to equate the QRM standard with the looser “Qualifying Mortgage” (QM) standard that has been established to provide a presumption of meeting the ability to repay test, also specified by the DFA. QM standards include a maximum DTI of 43%, maximum origination fees of three points, and full documentation. The QM standard does not specify a minimum down payment or minimum credit score.
This is unfortunate. Easy lending standards, in particular low or zero down payments, helped create a positive feedback loop, or reinforcing cycle, that contributed to the boom/bust housing cycle. While apparently beneficial during the boom to those home borrowers who otherwise would not have been able to afford a home, the consequence of the bust was pretty negative for all homeowners. The effort to deflect conservative underwriting standards (which, by the way, largely prevail in many other countries that have home ownership rates equal or greater than in the US) is detrimental to development of an improved housing finance system.
If the QRM standard were more restrictive, that would mean that many loans currently being made would require credit risk retention by the lender, if put into a security and sold to investors (for loans that are made to be held in portfolio, lenders would only have to meet the weaker QM standard). The effect of this would be to reduce lender willingness to make such loans and would increase the note rate (mortgage rate) on such loans. This is the price to pay to dampen the boom/bust cycle.