The scramble by the three main federal housing agencies, Fannie Mae, Freddie Mac, and the Federal Housing Administration (FHA), to attract ever more borrowers is reaching a dangerous new phase. Congress’ goal is to boost homeownership, whatever the risks—and its implicit deal with the agencies has long been more funding in exchange for more lending. But given recent strong demand, the low-hanging fruit has all been picked. AEI’s National Mortgage Risk Index (NMRI), which is based on a near-census of agency loans, shows that agency mortgage originations increased 13 percent in 2015 over 2014—and 19 percent for first-time buyers. In order to tap new customers, Fannie, Freddie, and FHA, which together account for over 85 percent of total agency lending, are turning increasingly on each other. Although AEI would normally applaud more competition, the usual rules do not apply here. Without any market penalty for reckless lending, increased competition is leading the agencies into a dangerous bidding war that will only further open the debt spigot and undermine lending standards.
Fannie and Freddie fired the opening salvos in December 2014 by announcing, at the behest of their regulator, the Federal Housing Finance Agency (FHFA), plans to extend credit by purchasing loans with combined loan-to-value (CLTV) ratios of up to 97 percent in order to meet their congressionally mandated affordable housing goals for low-income borrowers. This signaled to FHA that they were once again going to compete for the riskiest borrowers, FHA’s bread and butter. Fannie, which implemented the program much faster than Freddie, saw its market share in that segment jump at the expense of FHA’s.
It didn’t take FHA long to return the fire, and it did so with a much bigger gun. The next month, January 2015, FHA announced a major cut of 50 basis points to its mortgage insurance premium, which amounted to a declaration of all-out war. While up to then FHA had been primarily focused on higher risk loans, the result, intended or not, was that it turned to actively pursuing the middle of the risk spectrum, which is critical to Fannie’s and Freddie’s affordable housing mission. The NMRI, which measures the likely level of default under stressed conditions similar to the 2007/08 housing crisis, shows that the majority of FHA’s additional business carries a stressed default rate of between 8 percent and 16 percent. FHA now accounts for over one-third of this middle-risk spectrum, up from just one-quarter before its premium cut–while maintaining a market share of over 90 percent in its core business, high-risk loans with stressed default rates greater than 16 percent. [1]
For FHA, the premium cut has paid off handsomely with a massive increase in market share. It now accounts for 28 percent of the total agency business, up from 23 percent just before the cut. It can also claim to look safer on paper. Not only did the new business raise FHA’s mutual mortgage insurance (MMI) fund, which is financed through premiums that backstop losses on its mortgages, to the 2 percent funding level required by Congress, it also raised FHA’s median FICO score from 672 to 678 over the past year. By poaching less risky borrowers with a FICO score greater than its own median, FHA was able to improve its numbers.
Fannie, on the other hand, is bleeding. It has lost 4 percentage points from its peak market share in February 2015. True, it also looks safer because it lost some of its riskier borrowers to FHA, but in a business model that requires volume to please its patrons in Congress, this is probably little consolation. By contrast, Freddie has been remarkably successful at holding its ground. Its market share is actually slightly up since FHA’s premium cut. Why? First, because of its smaller share of high CLTV loans, and second, because it has loosened its lending standards more than Fannie has. Today, Freddie’s loan mix has shifted away from the low-risk category, with increases in both the medium-risk and high-risk groups, which has allowed it to avoid the FHA poaching that has so affected Fannie.
If all this amounted to was risk shuffling between different federal agencies, one might be tempted to shrug it off. But in fact, the bidding war could easily lead to a huge spike in overall housing risk. For example, FHA’s higher median FICO score and greater MMI fund could be used to justify even more and riskier lending. Or else Fannie and Freddie could venture even deeper into risky territory, where they are prone to mispricing risk by insufficiently compensating for lower down payments, as shown by FHFA’s own analysis. Unfortunately, this is the direction in which we are headed. The NMRI shows that the stressed default rate for the three main agencies is up 0.6 percentage points over the last year and up over 1.1 percentage points for first-time buyers, where competition is fiercest.
It remains to be seen whether Fannie and Freddie will dare to provoke FHA a second time, or if FHA will further escalate by cutting its mortgage insurance premium again. The lesson is that the way the current incentives are structured, Congress all but ensures that the federal housing agencies will find ways to increase leverage and loosen lending standards further, thereby exposing homeowners and the public purse to greater risk of another housing crisis.
NOTES
[1] By way of collateral damage, FHA has also poached from Rural Housing Services (RHS). Like FHA, RHS aims to promote homeownership for low- and moderate-income people. RHS’ share of agency lending is down by almost 40 percent to just 3 percentage points of the total today.