Half the subprimes were cash out refis. This isn't implausible. Freddie Mac reported that cash-out (meaning the new mortgage was at least 5% larger than the one it replaced) refis for its borrowers were 35% in the second quarter of 2007, and noted that refinancings as a proportion of total mortgages were declining, which is typical in a rising interest rate environment.
Now why is this so significant? It gives a completely different picture of the nature of the problem. It suggests that many of the people who took out subprimes weren't people who bought more housing than they could afford. It says they were already overstressed and overstretched financially. Using their home as a source of cash was a gamble to keep themselves out of bankruptcy, but in many cases, that bet didn't work out.
The high proportion of cash-out refis suggests that it would behoove someone to do some investigation to get a better grip on why people took these loans and what became of them. Were most, as Lee suggested, in bad shape and taking the one way they saw out, or were they merely foolhardy overspenders? If they needed the new mortgage to pay off other debts, how did they get in trouble in the first place?
The last large scale study of why people filed for bankruptcy (published in 2005 but looking at 2001 bankruptcies) found medical expenses were the top reason and job loss/interruption was number two. If these are the real reasons that a large proportion of subprime borrowers went that route, it suggests a completely different set of remedies than if it is say, primarily a housing bubble (too many people felt they could gamble on appreciation) or predatory lender problem.
Dean Baker pointed out that some borrowers defaulted before reset, which suggests that pre-existing financial stress may have played a role:[M]any of the subprimes were seriously delinquent or in foreclosure long before the mortgages reset to higher rates. In an analysis done early this year, the FDIC found that 10 percent of the subprime adjustable rate mortgages issued in 2006 were seriously delinquent (missed three or more payments) or in foreclosure within 10 months of issuance.
A parenthetical note: that bankruptcy study is horrendous and should never be cited by anyone. It counted any bankruptcy in which the parties had more than $1,000 in uninsured medical bills as having been caused by the bankruptcy, even when the respondents themselves cited other reasons. $1,000 is a nasty sum, but no matter how poor you are, it is not bankruptcy-level. Anyone who declares bankruptcy over a couple of thousand dollars worth of debt should sue their lawyer for incompetence. More rigorous studies find the commonest causes of bankruptcy are divorce and job loss; where medical problems are a cause, the main factor is usually income loss from lost work, not bills. Medical providers will almost always take a workout rather than bankruptcy.
But the broader point is interesting--though Yves, like most observers, is more interested than I am in assigning blame. I'm less interested in who was wrong, than in what broad forces pushed us in this direction. And he hints at an interesting one:
One other factor that may have contributed to the subprime frenzy: Lew Ranieri, the so-called father of mortgage backed securities, has stated that the overheated phase of subprime lending started at the end of the third quarter of 2005 and extended through most of 2006. When did the new bankruptcy law take effect? October 24, 2005. There is no ready way to prove a connection between the new law and the explosion phase of subprime growth, but consumers became much more cautious in taking on credit card debt after the law became effective. And the ones that had above median incomes which would force them into a Chapter 13 (meaning they'd have to repay their debts) might be even more eager to tap home equity if they saw themselves at risk.
One way to cast this: mean lenders got Congress to change the bankruptcy laws, which meant people desperte to get out of their credit card debt put their houses on the line. Another way: irresponsible borrowers confronted with the cost of their past profligacy, gambled their houses, too. The interesting explanation, though, I think is more value-neutral.
From what I hear, the evidence on bankruptcy reform is that all the actors involved behaved in a perfectly economically predictible way. Lenders, with more assurance that they would be repaid, became more willing to lend. But borrowers became less willing to borrow, so the amount of credit supplied to the market contracted. (Incidentally, people who think that we should protect the poor from credit cards and payday lenders should be glad about this.)
But there was a huge credit glut in this country, thanks in part to a torrential inflow of foreign capital. That credit had to go somewhere; if credit card borrowers wouldn't take it, they'd just offer cheaper rates to someone else . . . like mortgage borrowers. Mortgages aren't much affected by bankruptcy, because it's secured debt. Under the new law, as the old, borrowers either hold onto the house by committing to meet the payments, or surrender it to the lender. So falling interest rates in that market, especially combined with rising prices, were likely to produce the bubble conditions we saw.
This article available online at: