Edward L. Glaeser is an economics professor at Harvard.
Has the securitization of mortgages been a great boon or a terrible curse? Assets backed by American mortgages enable global risk-sharing, but securitization may also have weakened lenders’ incentives to screen out bad borrowers and to renegotiate bad loans. And should the government continue to subsidize securitization through entities such as Fannie Mae and Freddie Mac?
The pre-2006, conventional narrative about the rise of securitization ran something like this. Once upon a time, local bankers — all of whom looked and talked like George Bailey — received deposits from their neighbors and then lent this money back to other neighbors who wanted to buy homes. These bankers had a lot of local knowledge, but they were vulnerable to swings in interest rates or the local housing market. We should never forget what happened to the savings and loan industry when it was ensnared by rising interest rates.
According to the legend, this world was shattered by free-wheeling, guacamole-chomping financial entrepreneurs who figured out how to tie mortgage payments to securities and sell those securities worldwide. Michael Lewis’s “Liar’s Poker” provides us with an indelible picture of Salomon Brothers’ Lewis Ranieri, who coined the term securitization and did as much as anyone to make that market. Mr. Ranieri was hardly alone. Public agencies like Fannie Mae and Freddie Mac insured those mortgages against default, which helped make mortgage-backed securities far more palatable to investors.
Until 2006, securitization seemed to be a great leap forward, enabling borrowers to get cheaper loans and making banks less vulnerable to the vicissitudes of interest rates and housing prices.
Since the crash, a counter-narrative has emerged that emphasizes the dark side of securitization. If mortgage issuers passed along the default risk to Freddie Mac and Fannie Mae
or to the buyers of mortgage-backed securities, those issuers would have little incentive to screen borrowers properly. While issuers often do have some skin in the game, the enormous amount of both securitization
and sloppy lending during the boom made it natural to link the two phenomena.
Work by Benjamin J. Keys, Tanmoy K. Mukherjee, Amit Seru and Vikrant Vig, published in the Quarterly Journal of Economics this year (I edited the paper for the journal) found that mortgages that were more likely to be securitized also default more often. Securitization jumps at credit-score thresholds, like 620, and defaults are higher just above these thresholds, suggesting that the ability to resell the mortgages led to less scrutiny.
This view has been disputed by Ryan Bubb and Alex Kauffman (I played a role in advising them), who argue that defaults also rise above cutoffs for loans that are kept by their issuers. Their view is that these threshold effects represent banks’ internal use of cutoff rules, which leads loan officers to be more careful below specific thresholds. The debate continues, as it should. Evidence and counter-evidence are, after all, the best route to knowledge.
The anti-securitization view also argues that securitization makes it more difficult to renegotiate loans that have gone into default. Work, again by Mr. Seru and Mr. Vig, finds that loans that become seriously delinquent are more likely to lead to a foreclosure if they are securitized. On the other side, Manuel Adelino, Kristopher Gerardi and Paul S. Willen have made a strong case that renegotiation is often just not in the interest of the lender.
When the smoke clears, my bet is that the consensus will be that securitization did a bit to encourage lax lending and to discourage renegotiation, but that securitization was hardly the only — or even the prime — villain in the great housing convulsion. Lots of lenders, borrowers and home buyers did, or would have done, foolish things without the aid of securitization.
If we end up in that middle-of-the-road consensus, then securitization looks pretty good, because it did much to mitigate the costs of the crisis. Plenty of lenders would have gotten caught up in the exuberance of the boom even if they were holding onto the loans — as the savings and loans did 20 years ago. When the bubble burst, without securitization the banks would have been in far worse shape, because they would have been holding all the risk themselves. Securitization meant that the downturn in the American housing market was felt from Stockholm to Shanghai, which sounds bad except that it would have been a lot worse for us, and for our banking system, if the entire seismic shock had struck only banks in the United States.
That relatively positive view of securitization does not imply that the government should subsidize securitization, as it implicitly did through Fannie Mae and Freddie Mac. I didn’t like the Fannie and Freddie model before the bust, and I don’t like it any more today. If these agencies are to continue, they need to be far more conservative, charging high fees and taking few risks, and they need to be purely public agencies. Securitization has a role to play, but that role doesn’t merit vast public support.