In the post-crisis market, the Federal Housing Administration (FHA) has become a major player — and the only player for borrowers with limited cash and not-so-great credit. FHA’s share of the single-family mortgage market is about 30 percent today, compared with 3 percent in 2007. Unfortunately, FHA has suffered heavy losses on loans it insured during the go-go years prior to 2007, which have eaten into its reserves, and these are now below the level mandated by law.
As a result, it must either increase income, curb losses or both.
In April, FHA increased the upfront mortgage insurance premium from 1.75 percent of the loan amount to 2.25 percent. Against intense political opposition, it also banned the widespread practice of having home sellers contribute to buyers’ downpayments through the intermediation of nonprofit corporations, who took a piece of the contribution for themselves. These loans had extremely high loss rates.
Proposals to curb seller concessions: On July 15, FHA’s proposals for additional loss reduction measures appeared in the Federal Register. One proposal is to reduce the allowable concession that home sellers can make to purchasers from 6 percent of the purchase price to 3 percent, which is the cap on conventional mortgages. When home sellers agree to pay buyers’ settlement costs, they try and often succeed in getting it back in the sale price. The resulting price inflation depletes borrowers’ real equity.
For example, if a home sells for $100,000, a buyer who borrows $96,500 and pays $3,000 in settlement costs in cash has $3,500 of equity in the property. If the seller pays the $3,000 and raises the price to $103,000, the loan amount will increase to $99,400 and the borrower’s real equity will fall to $600. If the seller pays $6,000 in settlement costs, the real equity of the buyer will be negative.
The reduced borrower equity associated with seller concessions increases FHA’s losses. As an illustration, on loans made in 2005 without seller concessions, lender claims for loss reimbursements amounted to 6.9 percent of sale prices. Where there was a concession up to 3 percent, the claim rate was 7.9 percent. Where the concession was above 3 percent up to 6 percent, the claim rate was 10.9 percent. These data are drawn from FHA’s proposal in the Federal Register.
Other proposals to curb losses: FHA would also limit the current maximum loan-to-value ratios (96.5 percent on purchases and 97.75 percent on no-cash-out refinancing) to borrowers with credit scores of 580 or higher. Borrowers with scores below 580 would be limited to 90 percent, and those with scores below 500 would be rejected. Claim rates in past years on loans that did not meet the new specs were roughly 2 1/2 times as high as those that did meet the specs.
FHA also proposes that cash reserve requirements be increased, and that maximum debt-to-income ratios be based on the credit score and other factors. As with the other proposals, these are modest and the need for them is well documented.
Lenders don’t like the proposals, of course; they don’t like any changes that force them to say "no" to more potential borrowers. But these changes are designed to protect the integrity of the FHA system, in which lenders have a vital stake. They also are designed to avoid the specter of taxpayers eventually having to foot the bill if FHA losses continued at an unsustainable level.
Underwriting restrictions imposed by Fannie, Freddie and Congress: When I finished drafting this article, it occurred to me that not too long ago I criticized Fannie Mae, Freddie Mac and Congress for imposing stricter underwriting rules. Was I being inconsistent? I don’t think so. The restrictions proposed by FHA are designed to reduce its losses from defaults whereas the restrictions imposed by the other agencies and by Congress are designed to prevent borrowers from making foolish mistakes. The first kind of restriction is sensible, the second is not.
Historically, underwriting rules have been designed to protect lenders against excessive loss. They were never designed to prevent borrowers from making foolish mistakes, and with good reason. Marginal borrowers with shaky credentials, whose loan requests are nonetheless granted, are better off as a group, even though their default rate is high and they impose large losses on lenders.
An example: Assume a marginal group of home-purchasing borrowers for whom the default rate is 20 percent and that the lenders lose 50 percent of the balance on every loan. Lenders lose $1 for every $10 they lend. There is no way that lenders can cover a loss rate of this magnitude in their pricing because it would require that every borrower in this group be charged a rate 10 percent above the rate charged to a group of very low-risk borrowers. This is not possible because the higher rate would make the default rate even higher. Exactly the same logic would apply if an insurer bore the losses and had to cover them with insurance premiums.
While this group of borrowers is a loser for the lenders, the borrowers are not losers. Eighty percent of them purchased and retained houses that would otherwise have been unattainable. Even the 20 percent who failed are not necessarily worse off for the experience — in all probability, some are and some aren’t. The lenders and/or insurers lose, but the borrowers as a group come out ahead.
This will be true even with the subprime loans made during the go-go years, on which the loss rates will be well above the 10 percent in my example. More subprime borrowers benefited than were hurt by the excessively liberal underwriting climate that allowed them to become homeowners or refinance.
Underwriting rules designed to protect borrowers hurt more borrowers than they help: Such rules bar some loans to borrowers that carry low risk to lenders, and therefore should be made. This is clearly the case with the rules adopted after the crisis to require that mortgages be affordable to the borrower — now enshrined as the law of the land by the Restoring American Financial Stability Act of 2010. These rules prevent lenders from trading off weaknesses in affordability measures against larger downpayments and better credit, which means that perfectly good loans are not being made. The self-employed, comprising the small-business owners that Congress purports to favor, are particularly disadvantaged by this inane rule.