The vast system that provides home loans to millions of Americans has long been a strange place. A surprising development has made it even stranger.
Recently, interest rates on mortgages for expensive homes have fallen below those for smaller mortgages that the government promises to repay if the borrower defaults.
On Thursday, for instance, Wells Fargo, the nation’s largest mortgage lender, was offering to make the larger so-called jumbo loans at a fixed rate of 4.625 percent for 30 years. That compared with the 4.875 percent that the bank was charging on fixed 30-year loans that qualify for government backing.
On the surface, these moves in rates make little sense. The jumbo mortgages do not have a taxpayer guarantee of repayment. Anyone holding such loans relies solely on the creditworthiness of the borrowers to be repaid. Most of the jumbo borrowers are wealthy and have good credit scores, so they are not that high a risk right now. Still, their credit probably isn’t as strong as that of the federal government, which guarantees the smaller loans. As a result, those loans, often called conforming mortgages, should have lower rates than those on jumbo mortgages. Indeed, as far back as industry participants can remember, that has been the case.
The fact that jumbos are now cheaper points to dysfunctions in the mortgage market, which is going through a jarring adjustment that appears to be influencing guaranteed mortgages more than jumbo loans.
Since the financial crisis of 2008, the mortgage market has had two substantial sources of government support. First, government entities have been backing a far higher proportion of mortgages than in recent decades. Banks make these loans but don’t hold them. They package them into bonds and sell nearly all of them to investors, like pension and mutual funds. The second support has come from Federal Reserve. As part of its efforts to invigorate the economy, the Fed has been buying large amounts of those taxpayer-backed mortgage bonds. That helped bring about a big decline in mortgage rates in the grim years after the crisis.
But the Fed, believing the economy is gaining strength, has signaled that it may soon buy fewer of these bonds. Their price has fallen. This pushes up the yields on the bonds, which in turn drives up mortgage rates for people taking out new conforming home loans.
The rates on jumbo loans have also risen over the last few months. But because fewer jumbo mortgages trade in markets, they are less vulnerable to big swings in investor sentiment. That is not the case for guaranteed loans, which investors have sold heavily in recent weeks.
Of course, once uncertainty about future interest rates dissipates, and markets settle down, the rate on guaranteed mortgages could fall back below that of jumbo loans.
But policy makers have reason to be unnerved. If jumbo rates remain lower for a long time, it could mean that participants in the mortgage market have begun to believe they have a lower risk of default. That might seem a preposterous stance, given the government guarantee on conforming loans.
Still, there are circumstances under which the conforming loans might be riskier for the banks that make them â€" and recent efforts to overhaul the mortgage market may heighten that risk.
The government guarantees the conforming loans, but with one big caveat. If there are problems with those loans that lead to default â€" if the bank didn’t properly check a borrower’s income, for example â€" the government can effectively send them back to the bank that made them. Returning loans can saddle the lender with hefty losses, which has already happened after the government rejected many shoddy precrisis loans. As a result, mortgage experts say, banks have since the crisis charged borrowers extra interest to cover the risk of losses from taking loans back.
Amid all this, it’s possible to see why banks might believe that jumbo loans will result in fewer losses than conforming loans. From the outset, jumbo loans may simply experience fewer defaults than conforming mortgages, partly because their borrowers have higher credit scores. And when the conforming mortgages do default, the banks may not be able to predict what their losses will be because of the government’s ability to send back faulty loans.
If the situation persists, housing market officials might have to revisit some of the knotty questions at the heart of the mortgage system.
The government could loosen its standards and send fewer bad conforming mortgages back. The banks would stop fearing high losses on rejected loans, prompting them to charge borrowers less at the outset. Some people might ask why the government should relax its standards when it is the banks’ own inefficiencies that lead to loans that are sometimes sent back.
One way to reduce the risk of conforming loans is to require solid down payments, because these can reduce defaults and increase the amount recovered when a borrower does default. Right now, down payments are reasonably high on conforming mortgages, and maybe even in line with those on some jumbo mortgages.
But that could change.
In the housing overhaul, regulators had plenty of opportunities to draft rules that would enshrine sturdy down payments. They have gone out of their way to avoid doing that. The motivation was to maximize access to credit, which is another way of saying that the market has to be built to include borrowers who cannot afford sizable down payments.
That may be an admirable goal. But if down payments are allowed to fall on a large proportion of conforming loans, banks may respond by hoisting their rates to cover the risk that the government may send them back if they default in large numbers.
For prospective home buyers, access to credit would be diminished.
At that point, the government would have a stark choice. It could decide to change nothing, leaving some people unable to get mortgages. Or it could become more lenient, by signaling to the banks that it will send fewer loans back. While some in Congress would fight such a move, the housing lobby would likely support it â€" and it nearly always wins.
