Arnold Kling is incredulous when it comes to Pimco’s Bill Gross, who said yesterday that “without a government guarantee, mortgage rates would be hundreds — hundreds — of basis points higher, resulting in a moribund housing market for years.”
“The standard estimate in the literature is that Fannie and Freddie reduce mortgage rates by 25 basis points or less,” says Kling, and he’s correct: that is indeed the standard estimate. But the standard estimates were all calculated back in the day, when no one worried much, if at all, about mortgage default risk. Here’s Gross again:
It’s an $11 trillion secondary mortgage market. Agencies are about a half of it. The other half, or a good portion, were financed when people thought housing couldn’t go down in price. We know that’s not the case now. So to suggest there’s a large place for private financing in the future of American housing finance is unrealistic.
The 25bp figure dates back to the time when Frannie accounted for roughly half of all mortgages: back then the best estimate was that Frannie-conforming loans were about 25bp cheaper than private-label loans, all other things being equal. That 25bp wasn’t only a function of the government guarantee, so much as it was a function of the fact that Frannie’s capital requirements were much lower than banks’ capital requirements, and the fact that Frannie had huge economies of scale when it came to things like managing the risks associated with prepayment and negative convexity.
Now, however, everything has changed: Frannie accounts not for half of the market, but for essentially all of it. And while private money is happily pouring into corporate bonds and other credit instruments, it’s still shying away from mortgages, partly because no one really knows how to price them any more.
In order to price a mortgage, lenders and investors need to have an idea of what the borrower’s default risk is. And right now, they haven’t got a clue. We know that default risk is highly correlated to house prices, but we don’t know whether or when or by how much house prices might fall. We also know that default risk is a function of societal norms: historically, people paid their mortgage first, before other debts like credit cards, or other household expenses. That’s changing dramatically.
And beneath it all is the fact that the US housing default rate isn’t something that any lender can calculate: rather, it’s something controlled mainly by US government policy. The government has a huge range of ways in which it supports the housing market, and it can remove those supports at any time, sending defaults spiking. Alternatively, it can throw even more support at the housing market, as Gross would like to see:
Gross, speaking at a Treasury forum on the future of housing finance giants Fannie Mae and Freddie Mac, said a massive refinancing program to slash monthly mortgage payments could boost consumer spending by $50 billion to $60 billion and boost home prices by 5 to 10 percent.
“Policymakers should quickly re-engineer a refinancing opportunity for all mortgagees that are current on payments and are included in GSE-securitized mortgages,” said Gross, PIMCO’s co-founder and co-chief investment officer.
This is a crazy idea: the last thing we want is to send large checks to anybody lucky enough to have a conforming mortgage. Nor do we want house prices to rise even further away from the market-clearing level which we’d see were there no artificial government support. To the contrary, we want houses to become more affordable and to account for a much lower proportion of households’ budgets than they did during the housing boom.
But the point is that if it wanted to, the government could throw so much money at the housing market that defaults would no longer be much of a problem. And so the key analysis of anybody trying to price mortgages has to be primarily political rather than economic or financial.
Bond investors are by their nature very cautious folk, and they hate massive uncertainty about something as important as future housing default rates. So they simply avoid the housing market altogether, absent a government guarantee. That’s why Gross says that the guarantee is worth hundreds of basis points. I don’t know whether he’s right, but I’m sure the number is more than Kling seems to think.