Shiller’s Subprime Solution(s)

In my last post, I focused on what we still don’t know about the causes of the subprime crisis.

But here I’ll tell you about six solutions proposed by Robert Shiller in his book The Subprime Solution. (He has also recently published an op-ed in The Washington Post and an op-ed in The Wall Street Journal.)

Shiller separates the short-term need for a bailout from the need for long-term solutions. Much is being written now about short-term bailouts, and Shiller presciently has a chapter in his book on the inevitable need for cleaning up the past.

But this post responds to Shiller’s suggestions about what we should do prospectively to make sure this doesn’t happen again. For the long term, Shiller proposes “six major ways of improving the information infrastructure.” Three of his proposals are directly related to providing the system with more information (quoting, but reordering, from The Subprime Solution, P. 122):

1) promoting comprehensive financial advice;

2) improving disclosure of information regarding financial securities; and

3) creating large national databases of fine-grained financial data.

The other three proposals are, to my mind, substitutes for providing more information:

4) establishing a consumer-oriented financial watchdog;

5) adopting default conventions and standards that work well for most individuals; and

6) creating a new system of economic units of measurement.

I generally am more attracted to the three informational substitutes than to the first three proposals.

Shiller’s suggestion that government should subsidize unbiased financial advice for poor people is particularly weak. There isn’t good empiricism that shows that financial advice helps retail consumers make better decisions.

Many people are averse to thinking about their finances (even some of my colleagues at Yale Law School). And in this part of the book, Shiller too strongly embraces the notion that the subprime crisis was caused by consumers who would not have borrowed if they had been given better information.

He views borrowers through a complicated emotional lens, describing them alternatively as “complacent” (P. 6), prone to “excessive optimism” (P. 54), and “blandly accept[ing]” inappropriate terms (P. 133). But, as emphasized in my last post, I’m more open to the idea that some borrowers were making rational decisions about risks and rewards. Providing more information would not change these people’s decisions.

I am intrigued, however, by this suggestion of Shiller’s:

that every mortgage borrower have the assistance of a professional akin to a civil law notary. … In Germany, for example, the civil law notary is a trained legal professional who reads aloud and interprets the contract and provides legal advice to both parties before witnessing their signatures. (P. 134)

It might be worthwhile to test whether a point-of-purchase civil law notary would avoid some ill-advised borrowing.

Shiller also thinks that the ultimate buyers of the mortgages were making systematic errors. He wants improved disclosure about financial securities to help the ultimate mortgage buyers from making ill-advised purchases:

The subprime residential-mortgage backed securities were grossly misjudged because no one outside the rating agencies understood the information to correctly gauge the soundness of the mortgages on which they were based.

The stage was perfectly set for unscrupulous mortgage originators to lend to low-income people who were likely to default, and for mortgage securitizers to sell the soon-to-default mortgages to unsuspecting investors. (P. 136)

I’m skeptical about Shiller’s claim that the ultimate buyers lacked sufficient information or the sophistication to understand the data. Or even if they did, this is not an error they are likely to make again. Once bitten, twice shy.

They will demand, and originators will have incentives to offer, additional information. There might be a role for government in mandating standardized reporting so that comparisons can be more easily made; but I think it’s more likely that mortgage buyers simply underestimated the likelihood of a fall in real estate prices.

This is a failure of interpretation, not really a lack of data about the particular mortgages. (It’s also a reason that I’m skeptical of providing borrower advice. Apparently, Shiller would not have wanted Alan Greenspan and Paul Krugman to give financial advice because they underplayed the risk of a housing bubble. But where would the army of unbiased advisers come from in a world where, by assumption, we are caught in a bubble mentality?)

I’m similarly not persuaded that we need to create “large national databases of fine-grained financial data.”

To my mind, Shiller hasn’t made the case that failures in the current FICO-score system substantially contributed to the subprime crisis. His best justification for fine-grained data is, I think, to support his call for “continuous work-out mortgages” — mortgages whose repayment streams would be made contingent on information about both the state of the economy and the state of the borrower.

The problem with this proposal is that it is at war with the book’s embrace of transparency and simplicity. Automating the negotiating when housing prices fall makes a lot of sense, but I think you probably get 95 percent of the systemic benefit by making the payment simply contingent on this single price variable.

There is a certain disjunction between Shiller’s diagnosis of the problem and the solution. The problems are that:

1) we had a housing bubble that popped;

2) we had high systemic leverage at both the borrower level and the investment-bank level; and

3) we had asymmetric information about which financial intermediaries were exposed to the toxic debt.

Only the last of these is really an information problem, and providing more fine-grained information about borrower financial data doesn’t help solve the problem that we don’t know which institutions are exposed to bubble risk.

Shiller’s other three proposals might better respond to these problems. His embrace of Elizabeth Warren‘s Financial Product Safety Commission could be used to limit consumer leverage.

I’m attracted to the idea that we require 5 percent owner equity both at time of original sale and with regard to any subsequent refinancing; or government might allow a bit more flexibility by promoting less leveraged defaults.

Shiller appropriately “calls for the authoritative assertion of [a] new standard boilerplate for … mortgages,” and backs up the boilerplate by suggesting that the government only accept conforming mortgages as collateral for loans to mortgage lenders.

Viewed through the lens of the three problems (the bubble, the leverage, and the opaqueness), what are the best regulatory responses?

Deterring Bubbles: Economists don’t have a lot of proven tools to stop bubbles; but I like Shiller’s success in creating derivatives on city-level real estate prices. We might have fewer real estate bubbles if informed investors could more easily short the Los Angeles market.

Reducing Leverage: As mentioned above, we might require that homeowners maintain at least 5 percent equity at times of borrowing and refinancing. Systemically, there is also an advantage to requiring amortizing loans and discouraging second mortgages that pull out built-up data — because we’re macroeconomically safer when there is a distribution of home equity with 5 percent being the bottom.

We could also cap the amount of leverage allowable at financial institutions; but there are very real costs to constraining bank leverage, and we should proceed cautiously, or maybe not at all, on this dimension.

Reducing Opacity: This is, in some ways, the hardest. The problem of not knowing exposure to mortgage risk is serious in a world where mortgage pools are divided and subdivided again into esoteric products. In insurance markets, we tend to think that this atomization helps spread risk, but I don’t think there are good suggestions on how to get the opacity genie back in the bottle (or whether we should).

On the second page of his book, Shiller recalls that in 1919, John Maynard Keynes predicted that the punitive German reparations would result in disaster. Then, with one of the more chilling uses of the word “quite,” Shiller says: “A comparable disaster — albeit one not of quite the same magnitude — is brewing today.”

Shiller has been called a Cassandra, but he is Keynesian when it comes to his predictions and his diagnosis of the problem. His long-term informational solutions aren’t uniformly as convincing — but then again, I haven’t heard others that I like much better.