Banks are making more off mortgages than ever before, refusing to pass on lowered interest rates from federal policy, including the purchase of trillions in mortgage-backed securities by the Federal Reserve, to consumers. This isn’t really the enigma that the New York Times’ Dealbook makes it out to be. It’s simple collusion. Nobody offers 2.8% mortgage rates, so nobody gets them. As a result, the spread that banks capture on their mortgages widens.
Banks make mortgages, but since the 2008 crisis, they have sold most of them into the bond market, attaching a government guarantee of repayment in the process.
The metric effectively encapsulates the size of the gain that banks make on those sales. In September 2011, banks were making mortgages with an interest rate of 4.1 percent. They were then selling those mortgages into the market in bonds that were trading with an interest rate, or yield, of 3.36 percent, according to a Bloomberg index.
The metric captures the difference between the bond and mortgage rates; in this case it was 0.74 percentage points. The bigger the “spread,” the bigger the financial gain for the banks selling the mortgages. That 0.74 percentage point “spread” was close to the 0.77 percentage point average since the end of 2007. Banks were taking roughly the same cut on the sales as they were in previous years.
But something strange has happened over the last 12 months. That spread has widened significantly, and is now more than 1.4 percentage points. The cause: bond yields have fallen a lot more than the mortgage rates banks are charging borrowers.
So the natural rate for mortgages should be 2.8%, but banks aren’t passing the lower bond rates to their customers. Their middleman cut has grown.
The claim that banks are simply overwhelmed by mortgage demand is a crock. Mortgage lending fell to a 16-year low in 2011, and the mortgage purchase index shows a general sideways trajectory. Refinances have gone up somewhat with the tweaking of HARP and lowering of rates, but this is off a tremendously low bottom – lending fell 64% from 2006-2011. Anyway, banks were able to handle refi booms when rates dropped precipitously post-2008. And, smaller community banks seem to have no backlog whatsoever.
There may be a clogged pipeline, but you would have to see that as deliberate. Think about it, the less personnel that big banks hire to handle mortgage applications, the more they can constrict supply, the more they can charge for scarce mortgages. It’s a form of rationing, where banks make out of both sides – they don’t have to pay for labor AND they get to charge more for the product! And since they control so much of the market, there’s no real incentive for them to change their practices.
You would think that all this demand would get scooped up by someone who would rates. But that ignores the real issue of where the demand in housing is coming from. It’s not in mortgages.
Housing demand has improved this year, largely because investors and other buyers who have been paying in cash have scooped up quantities of foreclosed and other distressed properties. While lending to non-owner-occupants is down sharply from five years ago, it rebounded last year, rising 10% from 2010.
The money is being made in the sale of distressed properties to hedge funds and other investors, who are borrowing the capital to make these purchases from… the same big banks! Then, they plan to securitize the rental revenue, using the same banks as trustees to facilitate the sales. The big banks have no incentive to increase mortgage lending, because there is more money to be made on other pursuits, and because the real demand out there comes from the investor bubble.
This means that QE3 will have a hard time transmitting into the pockets of mortgage buyers, which is the ostensible goal, to boost consumer spending and mortgage purchases through lower rates. But the banks will sure be happy to increase their profits by doing nothing at all.
UPDATE: This Washington Post story gets at the same issues.
Photo courtesy 401K under Creative Commons License