This One's for Barry Ritholtz: The Real Problem With Originate-to-Distribute

Judging by his post today on the decline of lending standards, Ritholtz doesn't read Odd Numbers, which I'm very bummed about. He writes:

Underlying EVERYTHING -- housing boom and bust, derivative explosion, credit crisis -- is the enormous change in lending standards. I am not sure many people understand the massive change that took place during the 2002-07 period. It was more than a subtle shift -- it was an abdication of the traditional lending standards that had existed for decades, if not centuries.

After the Greenspan Fed took rates down to ultra-low levels, home prices began to levitate. More and more mortgage were being securitized -- purchased by Wall Street, and repackaged into other forms of bond-like paper. The low rates spurred demand for these higher yielding, triple AAA rated, asset-backed paper.

In this ultra-low rate environment, where prices are appreciating, and most mortgages were being securitized, all that mattered to the mortgage originator was that a BORROWER NOT DEFAULT FOR 90 DAYS (some contracts were 6 Months). The contracts between the firms that originated mortgages and the Wall Street firms that  securitized them had explicit warranties. The mortgage seller guaranteed to the mortgage bundle buyer (underwriter) that payments were current, the mortgage holders were valid, and that the loan would not default for 90 or 180 days

So long as the mortgage did not default in that period of time, it could not be "put back" to the originator.

This is a rephrasing of the critique against the "originate-to-distribute" model of securitization which has come under fire by many, including Ben Bernanke. Here is what the Fed Chairman said in May:

weaknesses in the application of the originate-to-distribute model became increasingly apparent last year....

at the point of origination, underwriting standards became increasingly compromised....The most notorious example is, of course, U.S. subprime mortgages. In this case, as in others, the incentives faced by originators were an important source of the breakdown in underwriting. The revenues of the originators of subprime mortgages were often tied to loan volume rather than to the quality of the underlying credits, which induced some originators to focus on the quantity rather than the quality of the loans being passed up the chain. 

But, as I've been talking about in a couple of recent posts, there is a good argument against this view. Yesterday, I pointed to one defense of this aspect of the orginate-to-distribute model put forth by Yale's Gary Gorton in a must-read paper for anyone who wants to have a better understanding of how subprime mortgages nearly destroyed the developed world's financial system. It's important to note that Gorton, who was a consultant with the now-infamous AIG Financial Products for 12 years, points out many problems with how subprime loans were created and sliced up, 

Here are Gorton's arguments for why it's unlikely that the incentives of subprime mortgage originators became catastrophically misaligned during the credit boom:

1. If you believe that incentives were misaligned, then you'd have to think that subprime originators would be decently protected when loans soured. But as the chart on this post shows, subprime lenders like Countrywide, Ameriquest, and Saxon were among those who bled the most. (Hat tip: Menzie Chinn) And here's another list of the subprime carnage.
2. When originators sell a loan they typically hold on to valuable servicing rights, so if the loan defaults the lender takes a hit. Gorton quotes Countrywide's 2007 annual report saying as much:

   When we sell or securitize mortgage loans, we generally retain the rights to service these loans.  In servicing mortgage loans, we collect and remit loan payments, respond to customer inquiries, account for principal and interest, hold custodial (impound) funds for payment of property taxes and insurance premiums, counsel delinquent mortgagors and supervise foreclosures and property dispositions. We receive servicing fees and other remuneration in return for performing these functions.

3. Retained interests such as interest-only, principal-only, and residual securities are also held onto by loan originators. Here is Merrill Lynch's 2007 annual report:

   Residuals: We retain and purchase mortgage residual interests which represent the subordinated classes and equity/first-loss tranche from our residential mortgage-backed securitization activity. We have retained residuals from the securitizations of third-party whole loans we have purchased as well as from our First Franklin loan originations...

   At the end of of 2007, Merrill's residuals exposure was 32 percent of the company's total subprime exposure.

4. Finally, if lending standards did decline sharply, then the originators would have opened themselves up to "first payment default" mortgages -- situations where the borrower defaults immediately. If this happens, most securitization contracts stipulate that the originator has to buy back the mortgage at a loss:

   It is difficult to see how a dramatic decline in underwriting would not result in a large number of first payment defaults that the originators would have to absorb. Since the originators would, in fact, absorb these mortgages, they have no incentive to make them in the first place.

I should say that a number of researchers have found that lending standards did decline during the subprime boom, but this paper makes a very persuasive counter-argument:

Given the multidimensional nature of ex ante credit risk, it is difficult to emphasize weakening in terms of some attributes as a decline in overall underwriting standards. The results show that while underwriting may have weakened along some dimensions (e.g. lower documentation), it also strengthened in others (e.g. higher FICO scores).

In other words, previous research has focused too narrowly when trying to gauge lending standards. A wider look at the factors that go into making a loan decision shows little evidence that standards fell off a cliff.

So what exactly is wrong with the "originate-to-distribute" model?

To start, one problem with subprime mortgages was that, unlike prime loans, they were very sensitive to housing prices. As home prices rose in the first part of this decade, not too many people cared about this fact, but once things turned south in 2006, reality set in.

But the much larger issue, Gorton notes, is that it was impossible for many holders of assets with subprime ties (like CDOs, CDO-squareds, and CDS's on subprime-backed securities) to actually figure out the health of the underlying mortgages:

In the current crisis there was a loss of information due to the complexity of the chain....investors do not have the resources to individually analyze such complicated structures and, in the end, rely to a lesser extent on the information about the structure and the fundamentals and more on the relationship with the product seller. Agency relationships are substituted for the actual information. To emphasize this is not surprising, and it is not unique to structured products. But, in this case the chain is quite long.