What We Still Don’t Know About the Mortgage Crisis

Ian Ayres

On Tuesday, September 16, at a rally at the Colorado School of Mines, Barack Obama criticized John McCain, saying:

Just today, Senator McCain offered up the oldest Washington stunt in the book: you pass the buck to a commission to study the problem. But here’s the thing; this isn’t 9/11. We know how we got into this mess.

It’s one thing to criticize McCain for inaction, but I disagree with Obama’s claim that we know how we got into this mess. In fact, if pushed, I would say I knew a lot more about the causes of 9/11 than I do about the causes of the mortgage crisis.

I just read Robert Shiller’s excellent book, The Subprime Solution, and he makes a powerful case that the end of the price bubble in residential real estate was the crucial triggering factor.

Indeed, the graph, based on the Case/Shiller U.S. National Home Price Indices, is prima facie evidence of a bubble that expanded and popped:

The dramatic downturn in housing prices was the key trigger — causing an increase in mortgage defaults. But the next step in the meltdown seems to have been the byproduct of 1) high leverage at investment banks and other financial intermediaries and 2) uncertainty over how much of the default risk was held by particular intermediaries.

Assessing a firm’s exposure is complicated by the massively complex web of derivatives and tranches oftentimes unreported in firms’ balance sheets. Firms like Lehman Brothers suddenly had trouble borrowing when lenders had trouble assessing the value of the security they offered. When you’re massively leveraged, just small increases in uncertainty over your asset value can dry up your ability to refinance your debt.

But to my mind, there are still dozens of important unanswered questions.

How much of the crisis was caused by subprime borrowers who made mistakes by borrowing (i.e., would not have borrowed if they had better information)?

Some proportion of the loans had very low down payments and low interest rates for one or two years. It might have been a rational choice for someone to take a chance on homeownership — putting very little money at risk and thinking (correctly): “If housing prices continue to go up, I’ll be able to refinance my house when the two years are up.”

In 2005, Elizabeth Warren gave a paper at Yale’s Legal Theory Workshop, and I called her on a claim in one of her footnotes that down payments for first-time homeowners were minuscule.

I was sure this claim had to be wrong. There is no way banks can loan money with virtually no equity cushion. But guess again.

According to The Washington Post, “four out of 10 first-time buyers used no-money-down mortgages in 2005 and 2006, according to surveys by the National Association of Realtors.” And the median down payment for first-time buyers in those years was just 2 percent.

With so little of their own money at risk, it shouldn’t be a wonder that many borrowers default when housing prices decline. Would you want to keep paying on a $200,000 mortgage when the house is only worth $150,000?

If you want to know why the mortgage system is so fragile, you should look to the drop in home equity — which was occasioned by low down payments and second mortgages that pulled equity out of houses where the owners paid off some of their outstanding principle.

But the problem of high borrower leverage was compounded by the high leverage of many of the firms that ended up holding the mortgage papers. Some investment firms were leveraged 30 to 1 (or more).

Years ago, Shiller called upon Freddie and Fannie to conduct “stress tests” to see whether they could survive a downturn in real estate prices. One of the chiefs had concluded that they could survive a 10 percent downturn in prices, but didn’t think it was plausible that prices would fall more than that.

I like Shiller’s “stress test” idea; but I’m also attracted to the New Orleans levee metaphor: Should our mortgage system’s levees be able to withstand a 20-year flood, or should we design them to withstand a 100-year flood? Levees are not costless. Neither are financial-safety measures.

We also don’t know the extent to which specific terms of subprime loans contributed to the spike in home foreclosures. Some of the loans (particularly those going to minorities) were at high interest rates not justified by the risk of borrower default. Some mortgages were interest-only loans, where the borrowers were not building up additional equity overtime. And many of the loans had teaser rates, which effectively required refinancing after two years.

My tentative take is that the low down payments were more important than any of these other factors as far as adding to systemic risk. Second place would be the effective refinancing risk (because after the house prices started to decline, lenders were not willing to lend $300,000 on a house that was only worth $250,000).

Third place would be the interest-only loans. If the down payments had been higher, interest-only loans would not have been much of a problem at all. There is a systemic benefit with amortizing loans: in a system where people borrow money at different times, there will be different amounts of home equity in the system. But this benefit is dramatically reduced by the ease of taking out second mortgages to pull out any built-up amortized equity.

I’ve done a lot of work on the problem of high borrowing interest rates, and marking up interest rates can exacerbate the probability of default. But while more needs to be known, I believe that this impact is likely to be less than the impact of these other terms. (These inflated rates, however, are a problem in and of themselves — and remain an action item for enlightened regulation.)

But I’m still not sure how many of the defaults are caused because borrowers can’t pay (for example, because of interest resets and the resulting need to refinance), and how many are caused because borrowers are choosing to walk away from mortgages that are seriously underwater.

We also don’t know the answers to parallel questions concerning the lending side. It’s easier to see the possible rationality in the behavior of the loan originators; to the extent that they were flipping the mortgages in the securitization market. But what we still don’t know is why the ultimate buyers were willing to buy.

Was this a failure of Super Crunching? Was it a failure of corporate governance (in that the managers of the buying firms had incentives to unprofitably grow their empires)? Was the failure caused by originator fraud (or the moral hazard of substituting bad-doc loans for what historically had been high-quality loan pools)?

I think it is probably some mixture of all three — with poor corporate-governance incentives particularly explaining the failure of the rating agencies to start downgrading the debt earlier.

Knowing the answers to these causal questions is important if we are going to craft useful policy responses. But that will be the subject of another post — which will more directly focus on Shiller’s subprime solution(s).