Two Stars for Shaky Ground

For the first time in many years, I wrote a review for Amazon. About Bethany McLean’s book on Freddie and Fannie, I say,

I was disappointed with this book, because I think that her earlier work, All the Devils are Here, co-authored by Joe Nocera, is probably the best journalistic account of the run-up to the financial crisis.

On “Shaky Ground,” here are my thoughts:

1. This book might have been titled “Sympathy for the Devils.” There is way too much sympathy expressed for the hedge funds that bought preferred stock in Freddie and Fannie. They were making a bet that the political process would come out a certain way, and they lost that bet, fair and square. End of story, as far as I am concerned. I should note that on several occasions representatives of the hedge funds have felt me out about doing some “research” or writing an article to support their position. I would not have done it for any amount of money. I am not accusing McLean of having succumbed to this, but I would not completely rule it out.

2. The other devil who gets a ton of sympathy is former Fannie Mae executive Tim Howard. McLean endorses all of his self-serving views, which include a claim that he did nothing wrong in Fannie’s giant accounting scandal. Also, his view is that had the Fannie management not been replaced, his team would have averted the crisis. Both claims may be true. In my opinion, Freddie and Fannie were better managed before both of their management teams fell in accounting scandals. But I think that more journalistic skepticism is in order. Regardless of who was in charge, there was pressure on Freddie and Fannie management to dive into high-risk lending, with shareholders seeing profits and regulators seeing a mission to expand home ownership opportunity.

3. She is no fan of Ed DeMarco, who was the only person in Washington working to gradually wind down the GSE’s, which is supposedly what everyone wanted. I think it is fair to say his approach was too unpopular with key players to be sustained. But he does not deserve to be dismissed by McLean with boo-words, like “free-market ideologue.”

4. She says that if you take it as given that the government is going to promote what the housing lobby wants, namely “home ownership” with little actual equity and a mortgage market dominated by the 30-year fixed-rate loan, then keeping Freddie and Fannie is better than the alternatives. If you accept the premise, then I agree. But there is a powerful case to be made against government caving into the housing lobby. The costs of this, including serial financial crises (the S&L crisis, the crisis of 2008) and misallocation of capital, are huge, and the social benefits are miniscule. (The private benefits can be enormous–just ask Tim Howard.) McLean does mention some of the evils of this housing-industrial complex, but her bottom line is, in effect “you can’t beat ‘em, so don’t try.”

Overall, this is not a terrible book. But if you read it, you should keep in mind that she gives the most favorable treatment possible to Freddie, Fannie, the hedge fund investors, and to policy makers who attempt social engineering using housing finance. Although the book is not completely one-sided, she does not give alternative points of view as much respect as I think they deserve.

How Bad was 2008?

Timothy Taylor writes,

my point here is not to parse the details of economic policy over the last seven years. Instead, it is to say that I agree with Furman (and many others) on a fundamental point: The US and the world economy was in some danger of a true meltdown in September 2008. Here are a few of the figures I used to make this point in lectures, some of which overlap with Furman’s figures. The underlying purpose of these kinds of figures is to show the enormous size and abruptness of the events of 2008 and early 2009–and in that way to make a prima facie case that the US economy was in severe danger at that time.

Taylor highlights the fall in house prices, the drop in bank lending, and the rise in the TED spread. However, if you look at just these indicators, the crisis ended relatively quickly. But employment just kept dropping (long after the official end of the recession). So it looks to me like the policies had a neutron bomb effect. The buildings (banks) were left standing but the people (workers) died.

As Taylor says, these are points that are not going to be settled. In my terminology, there are many frameworks that can be made consistent with observed economic performance. Some of these frameworks will be consistent with policies having made a positive difference, and others will not.

The Causes of Mortgage Defaults

The latest paper is by Fernando Ferriera and Joseph Gyourko. This article about the paper says,

Ferreira’s data show that even with strict limits on borrowing—say, requiring every borrower to put 20% down in all circumstances—wouldn’t have prevented the worst of the foreclosure crisis. “It’s really hard for certain regulations to stop the process [of a bubble forming],” Ferreira says. “I really wish my research had showed that it’s all about putting down 20% and all problems are solved, but the reality is more complicated than that.”

This analysis has both good points and bad points. The good point is that it goes against the “predatory lending” narrative. As a home buyer, you were better off with a predatory loan in 2002 (when prices were still headed higher) than with a prime loan in 2006 (when prices were near the peak). The bad point is the implication that there was nothing wrong with loans with low down payments. In fact, it was those loans that allowed speculation to get out of control.

Scott Sumner thinks that the finding that many of the mortgage defaulters were “prime” borrowers is enough to confirm that mortgage defaults were caused by a slowdown in nominal GDP growth. But mortgage defaults do not come from a lack of nominal GDP growth. They come from negative equity among mortgage borrowers.* And that comes from house prices falling, for which the main cause was the rapid rise in the first place. And both the rise in prices and the subsequent wave of defaults were much exacerbated by the fact that so many borrowers, “prime” or otherwise, had so little equity to begin with.

From part of the NBER coverage of the paper that Sumner does not quote:

The authors’ key empirical finding is that negative equity conditions can explain virtually all of the difference in foreclosure and short sale outcomes of prime borrowers compared to all cash owners. Negative equity also accounts for approximately two-thirds of the variation in subprime borrower distress. Both are true on average, over time, and across metropolitan areas.

Let’s assume that we can agree that the big drop in house prices caused the wave of mortgage defaults. Three possibilities:

1. The drop in house prices was a purely exogenous shock.

2. The drop in house prices was due to the slowdown in nominal GDP growth.

3. The drop in house prices was due to the internal dynamics of a housing market that had become saturated with speculative buying with little or no money down.

The stories about the study make it sound like it was (1). Sumner believes (2). I vote for (3).


UPDATE: See Megan McArdle for a similar point of view.

Greece and Representative Negotiation

John Cochrane writes,

So, the Drachmaized Greece that I see is not the cleanly devalued newly competitive powerhouse that some on the left seem to envision. Instead I see a two-currency economy. Pensioners and government workers and anyone unlucky enough to still have a Greek bank account get Drachmas. Hotel owners, restaurant owners, and exporters get euros, above or under the table.

My comments:

1. I agree with John that nothing real changes with a new currency. Instead, it is a way of arranging the government’s default. In addition to defaulting to bondholders, the government will default to other claimants, including pensioners. But the way it will default to the latter is by paying them in lower-valued currency.

2. I continue to believe that we will see an opaque bailout. What is happening now is pre-concession posturing on the part of the other European nations.

The classic example of pre-concession posturing is the labor union strike. One theory of strikes is that they take place because the union leaders are ready to make a deal, but they need to convince their membership that the union leaders bargained really hard. Going out on strike sends that message. Similarly, for the European leaders, engaging in table-pounding and other theatrics will help convince their constituents that they were really tough on the Greeks. Meanwhile, in the background, an opaque bailout will be arranged.

This theory of representative negotiation also holds for the nuclear negotiations with Iran. The theory predicts that there will be a deal, but in the meantime the negotiators will posture to indicate that they are being very tough with their opponents.

Speaking of Iran nuclear issues, I read Michael Oren’s new book about being Israel’s ambassador to the U.S. I found Oren credible, although for my taste he squeezes too much melodrama out of his experience. One of Oren’s points about the Obama Administration is that it has very tight message discipline, and I believe that we can see that in some of the negative reviews of Oren coming from Obama-linked writers.

Oren’s description of Obama amounts to saying that he operates using the oppressor-oppressed axis, which strikes me as accurate. Even so, it still requires some mental contortions to treat the leadership in Iran as oppressed, rather than as oppressors.

The Greek Crisis and the Subprime Crisis

Ana Swanson writes,

Matthijs compares the situation to the U.S. subprime crisis. Who was really at fault for the housing crisis in the U.S.: The subprime borrowers who bought houses they couldn’t afford, or the predatory lenders who encouraged them to take them out?

I, too, see parallels with the subprime crisis. However, I do not think that predatory lenders are to blame for either. In both cases, bank regulators were responsible for allocating credit. In the first instance, the regulators encouraged banks to treat mortgage loans as low risk. In the second case, they encouraged banks to treat all European sovereign debt as low risk. See The Regulator’s Calculation Problem.

The irony is that after messing up credit markets, the regulators ask for and receive more power. With the sub-prime crisis, the regulators were rewarded with Dodd-Frank. I presume that the ultimate outcome of the Greek crisis will be similar.

Anil Kashyap on Greece

Probably the best analysis so far. Mostly, it is a recap of the past. But in talking about the pending referendum, he writes,

if the public sides with Tsipras government, then there will be a very sharp recession over the next few months. Tax collection is likely to collapse. The Tsipras government is unlikely to survive the economic collapse.

He also writes,

Greece should have defaulted in 2010. Its debt burden then was unsustainable and nothing since then has changed this. It is true that financial markets were much more jittery at that time, but the money that was raised to pay off the creditors in that bailout could have been diverted to support Greece and other weak countries. Once the bad rescue of 2010 was undertaken, it was inevitable that some form of debt relief was going to be necessary.

Imagine how different the political dynamics in Europe would have been if the German and French banks had been explicitly bailed out.

Pointer from John Cochrane (and from Greg Mankiw and James Hamilton). Of course, I think that explicit bailouts are exactly what the political system will not allow. Even going forward, I still think that “opaque bailout” is the most likely outcome. But I also think that there are some lessons for us.

1. At some point, you do run out of other people’s money (that is actually more true for us than for Greece, because we are bigger and therefore harder to bail out).

2. When you run out of other people’s money, political tensions rise considerably. See my essay Lenders and Spenders.

When Economists Were Right, Allegedly

Richard Baldwin writes,

Barry Eichengreen added specificity to this in January 2009 with his insightful column “Was the euro a mistake?”, noting: “What started as the Subprime Crisis in 2007 and morphed in the Global Credit Crisis in 2008 has become the Euro Crisis in 2009. Sober people are now contemplating whether a Eurozone member such as Greece might default on its debt.” He wrote that the alternative to default was “fiscal retrenchment, wage reductions, and assistance from the EU and the IMF for the cash-strapped government.”

He predicted – again dead on – that “[t]here will be demonstrations against the fiscal cuts and wage reductions. Politicians will lose support and governments will fall. The EU will resist providing financial assistance for its more troublesome members. But, ultimately, everyone will swallow hard and proceed … In the end, the EU will overcome its bailout aversion.” The farsightedness is astounding. In January 2009, few knew the Greeks had a problem serious enough to require debt restructuring.

Pointer from Brad DeLong, via Mark Thoma.

That sounds impressive. He also cites other economists. But a couple of cautionary notes.

1. The best way to develop a reputation as far-sighted is to make many vague, conditional predictions. Later, you call attention to those that sound correct, and if necessary you wiggle out of those that sound incorrect by pointing out the conditions or taking advantage of their vagueness. I am not accusing Eichengreen of doing this. I have others in mind. But what might Baldwin have found had he had searched through past articles and looked for bad predictions?

2. How best to generalize this point? My guess is that “Economists’ predictions should always be taken as gospel”

3. Is the correct lesson that we should pay attention when economists warn about sovereign debt issues? Consider that many of us have issued warnings about the United States.

Opaque Defaults

James Hamilton writes,

The bottom line for me is that Greece’s current debts and any new loans that get extended from here are not going to be repaid in the present-value sense defined above. The current debt load and its associated interest bill are simply too big. The only solution is default on a significant portion of outstanding Greek debt. Indeed, significant debt restructuring is a necessary step for Greece to move forward, whether within the euro or based on a new currency. My view is that any realistic negotiations at this point simply have to take those facts as given.

My bottom line is to suggest that in forecasting the outcome of sovereign debt crises, assume that everyone’s goal is to make defaults as opaque as possible. For a small country like Greece, there ought to be many ways to do this. When the crisis hits Japan, it will be more difficult. When it hits the U.S., it will be more difficult still. But by then the international technocrats will have had a lot of practice.

Axel Leijonhufvud vs. Calomiris and Haber

Leijonhufvud writes,

The financial structure inherited from the 1930s divided the system into a number of distinct industries: commercial banks, savings and loan associations (S&Ls), credit unions, and others. It also divided it spatially. Banks located in one state could not branch across the line into another. This structure of the financial sector gave it great resilience. On another occasion I used the metaphor of a ship with numerous watertight compartments. If one compartment is breached and flooded, it will not sink the entire vessel.

This is directly the opposite of what Calomiris and Haber argue. They say that American cultural hostility to large banks produced an overly fragmented system, and that this fragmentation is the root cause of the peculiar instability of American finance.

Leijonhufvud elaborates,

At the time, the abolishment of all the regulations that prevented the different segments of the industry from entering into one another’s traditional markets was seen as having two obvious advantages. On the one hand, it would increase competition and, on the other, it would offer financial firms new opportunities to diversify risk. Economists in general failed to understand the sound rationale of Glass-Steagall. The crisis has given us much to be modest about.

In other words, economists championed open competition without thinking about how this would turn idiosyncratic risk into systemic risk.

The thing is, I was following these regulatory issues at the time, and I do not think that economists were as influential as we would like to believe. I think that the driving factors were computer technology, inflation, and massive lobbying. With high inflation, the regulatory ceilings on deposit interest rates that were a vital part of the regulatory structure became untenable. Moreover, with computer technology, it became easy for Wall Street to “disintermediate” banks using money market funds and securitization. These forces produced an inevitable turf battle between commercial banks and investment banks, which took years to resolve, as everybody lawyered up on the lobbying front. There were multiple possible political/regulatory outcomes, but hanging on to Glass-Steagall was not one of them.

Another Leijonhufvud quote:

Deregulation. . .allowed the great investment banks to incorporate and one by one they all did so. . .Incorporation meant limited liability for the investment bank and no direct liability for its executives. The incentives for executives in the industry changed accordingly. . .Now they are seen as jet-setting high rollers. Economists in general failed to predict this change in bankers’ risk attitudes. We have much to be modest about.

I cannot disagree with that.

However, I would instead put most of the emphasis on the regulations that directly affected housing finance. The pressure to lend with little or no money down and the designation of highly-rated mortgage securities as low risk for bank capital purposes are the main villains in the story as I tell it.

Public Availability of Freddie, Fannie Loan Performance Data

Todd W. Schneider has a write-up and some analysis.

I decided to dig in with some geographic analysis, an attempt to identify the loan-level characteristics most predictive of default rates, and more. As part of my efforts, I wrote code to transform the raw data into a more useful PostgreSQL database format, and some R scripts for analysis. The code for processing and analyzing the data is all available on GitHub.

I recommend reading the entire post.