New research from Neil Bhutta and Benjamin J. Keys finds that private mortgage insurers (PMIs) increased their risky lending due to moral hazard associated with the requirement that the government sponsored enterprises (GSEs) Fannie Mae and Freddie Mac insure mortgage securities they purchased.
In the U.S. mortgage market, private mortgage insurance (PMI) is mandated for high-leverage mortgages purchased by Fannie Mae and Freddie Mac to serve as a private market check on GSE risk-taking. However, we document that PMI firms dramatically expanded insurance on high-risk mortgages at the tail-end of the housing boom, contradicting the industry’s own research regarding house price risk. Using three detailed sources of mortgage and insurance data, we examine PMI application denial rates, default rates on PMI-backed loans, and growth rates of high-leverage lending around the GSE conforming loan limit, along with information extracted from company, industry and regulatory filings and reports. We conclude that PMI behavior during the housing boom in part reflects a “moral hazard” incentive inherent to insurance companies in general to underprice risk and be undercapitalized. Our results suggest that rather than providing discipline, private mortgage insurers facilitated GSE risk-taking.
Far from serving as a check on risky lending activities, the insurance policies issued by PMIs took advantage of the implicit guarantee associated with insuring a government-backed enterprise and engaged in riskier lending than other lenders in the market, as shown by lack of scrutiny when underwriting insurance policies.
[D]uring the housing boom PMI companies generally insured any loan approved by the automated underwriting systems of the GSEs, even if those loans did not meet insurer’s published guidelines. For example, MGIC reported that from 2005-2007, over 25 percent of their insured loans were approved through automated underwriting software despite falling outside of their published underwriting guidelines in terms of credit scores, LTV ratios, loan documentation, or other contract or borrower characteristics.
The paper also offers some revealing insights on how severely undercapitalized insurers were during the housing boom.
Rather than building up large reserves during the housing boom, when loan performance was relatively strong, industry reports suggest that about 60 to 65 percent of premium revenue went to total expenses, including claims. For instance, in 2006, the industry’s loss ratio (claims divided by net premiums earned) was 41 percent, while the expense ratio (all other expenses divided by net premiums earned) was 24 percent (MICA 2009). These values imply that PMI rms could reserve, at most, just 35 percent of premiums, well below the regulatory target of 50 percent. In fact, we estimate from the data reported by MICA (2009) that only about 18 percent of premiums went into reserves. Most of the remaining 17 percent of premiums likely went to shareholders via dividends and stock buy-back programs.
Insurance and other financial instruments like derivatives (which the authors propose as an alternative to GSEs purchasing insurance as a way to hedge risk) are useful tools that translate abstract concepts like risk into useable information in the form of prices. Unfortunately, implicit guarantees from GSEs distort this mechanism, encouraging firms to take on greater risks, including the risk of catastrophic downturns like the financial crisis 10 years ago.