Larry is also the author of Jimmy Stewart is Dead – Ending the World's Ongoing Financial Plague with Limited Purpose Banking, from 2010, which along with Anat Admati and Martin Hellwig's The Bankers' New Clothes is one of the central works outlining the possibility of equity-financed banking and narrow deposit-taking, and how it could end financial crises forever at essentially no cost.
Larry points out that the crisis was, centrally a run. He calls it a "multiple equilibrium." Financial institutions have promised people they can have their money back in full, at any time, but they have invested that money in illiquid and risky assets. When people all do that at the same time, the system fails. Such a run is inherently unpredictable. If you know it's happening tomorrow, you run to get your money out and it happens today.
This is a common view echoed by many others, including Ben Bernanke. What's distinctive about Larry's essay is that he pursues the logical conclusion of this view. If the crisis was, centrally, a run, all the other things that are alluded to as causes of the crisis are not really central. Short-term debt, run-prone liabilities are gas in the basement. Just what causes the spark, how big the firehouse is, are not central, as without gas in the basement the spark would not cause a fire.
Larry puts it all together nicely by starting with the 2011 Financial Crisis Inquiry Commission report:
"There was an explosion in risky subprime lending and securitization, an unsustainable rise in housing prices, widespread reports of egregious and predatory lending practices, dramatic increases in household mortgage debt, and exponential growth in financial firms’ trading activities, unregulated derivatives, and short-term “repo” lending markets, among many other red flags. Yet there was pervasive permissiveness; little meaningful action was taken to quell the threats in a timely manner. "Larry then takes apart each of these non-culprits, as below.
In my view, the understanding that the crisis was a run, that without a run there would have been no crisis, somewhat like the 2000 tech stock bust, and that lots and lots more capital is the only real answer, has emerged slowly over the last 10 years. Larry's essay is good for putting all the others to rest.
The insight is also optimistic. It is possible to fix one clear simple thing -- too much short-term debt, not enough capital. If all the long list of vague maladies named by the crisis commission need to be fixed by super-powerful and financially clairvoyant regulators, the job is hopeless despite the immense government expansion that would entail.
In many of these items, I think Larry oversteps a bit. The argument only needs to be that "these things might have been problems, indeed, they might have caused a recession, but without a run, induced by short-term debt, there would have been no financial crisis." Larry goes on to cast doubt whether any of them are problems at all, which is fun and provocative but more than necessary. Moreover, Larry really goes on to view recessions themselves as multiple equilibria, which is an interesting and provocative idea, but not necessary.
Larry's essay is even more to the point today. I'm at the Financial Cycles and Regulation conference at the Bundesbank. Every paper so far takes it for granted that there is such a thing as a "financial cycle," a buildup of "excessive" or "imbalanced" debt that precedes an inevitable round of default and crisis. It is the regulator's job to manage such "debt cycles" actively. Larry's essay disagrees from the word go. It's nice to have two diametrically opposed ideas in mind.
Larry's List follows. If you get bored, skim to the bottom for his more provocative ideas on runs.
1) Liar Loans, No Doc Loans, NINJA Loans and Other Subprime Mortgages
There just weren't that many subprime defaults, especially before the crisis hit. (And, I would add, defaults not insured by fannie, freddie, etc.)
2) The “Unsustainable” Rise in Housing Prices
House prices have risen and fallen before. And, I might add, many are sustained. If you're waiting for houses in Palo Alto or Manhattan to fall to, say Joliet IL levels, you may have a long wait ahead of you.
Larry makes an interesting comparison of real house prices with real GDP,
.. real house prices can rise for years, indeed, decades. They did so essentially every year for the 32 years between Q1 1975 and Q1 2007. The rise was both smooth and gradual with real house prices only 64 percent higher in Q1 2007 than they were in Q1 1975 – this despite real GDP rising by 170 percent over the same interval.
...between Q1 2003 and Q1 2007 ... real house prices rose by 22 percent. But over this period real GDP rose by 14 percent. Hence, real house prices rose only 2 percent faster per year than did the economy during the period of “unsustainable” house price increases.
The comparison between real house price and real GDP is unusual. Larry writes
One can write down models with a fixed supply of housing in which house prices will rise pari passu with output, at least in the long run. One can also write down models in which there is a variable supply of housing and the price of housing stays fixed, while the quantity of housing rises with output.
In any case, it's not a financial crisis without short-term debt.
Certainly, a temporary drop in house prices could have produced a contraction in construction.... Moreover, a decline in a given sector doesn’t augur an economy- wide recession.
And, an important point
a drop in the price of homes does not adversely impact most homeowners. Yes, the value of their asset falls. ...
But you still get to live in the house. If you could pay the mortgage before, you still can.
if we’re talking about a nationwide decline in house prices, as we are with the GR, even those who moved experienced no economic harm because their ability to buy at a lower price offset their need to sell at a lower price.
The house price drop was great news for young people who live in apartments, as the stock price drop was great news for their retirement investment opportunities. Yes, there are theories in which the losers impact the economy asymmetrically, such as Mian and Sufi's, but we're straying away from the point here. A house price "bubble" and "burst" is not per se a reason for a financial crisis.
As in many of these "causes" it's important to distinguish events before October 2008 from those afterwards. Yes, there was a huge recession, and that caused house price declines, job losses, mortgage defaults, and so forth. But causes of the crisis have to come first.
3) Ratings Shopping
"overrating affected less than one half of one percent of the U.S. bond market. Furthermore, this small figure surely overstates the importance of ratings shopping as many of the downgrades were caused by the GR itself,"
4) Increased Bank Leverage
"Sky-high bank leverage is another part of the standard GR explanation....Bank leverage actually fell over the period 1988 through 2008.16 Equity rose from 6 percent of bank assets in Q1 1988 to 10 percent in Q1 2008."
Be careful here. Larry's point is that there was no voluntary increase in leverage, and especially as measured and monitored by regulators, and no such increase was a key cause of the crisis. That doesn't mean the overall amount of leverage is fine. Larry's main point, as mine, is that the banking system has way too much leverage overall.
5) Too Little [Regulatory] Capital
"According to Cox [ Christopher Cox, Chairman of the U.S. Securities and Exchange Commission (SEC)..], Bear Stearns was well capitalized when it failed, with a capital ratio over 13 percent and a debt-equity ratio of 6 to 1. Indeed, it appears that Bear Stearns could have easily passed the current Dodd-Frank stress test immediately prior to its demise. Consider this statement from Chairman Cox.
"The fate of Bear Stearns was the result of a lack of confidence, not a lack of capital. When the tumult began last week, and at all times until its agreement to be acquired by JP Morgan Chase during the weekend, the firm had a capital cushion well above what is required to meet supervisory standards calculated using the Basel II standard. "
Lehman was also well capitalized prior to its demise. It had tier-1 capital of 11 percent when its creditors pulled the plug. An 11 percent capital ratio is close to the current banking system’s tier-1 capital ratio of 12.3 percent, calculated based on the Federal Reserve’s recent stress tests. This indicates that today’s banking system is no safer than was Lehman Brothers when it was driven out of business."
Again, in this draft, Larry doesn't emphasize enough that the point is a decline in capital, a weakening of regulations, or a decline in regulatory capital. His and my overall point is that capital needs to be much, much larger overall, and that will stop runs. But the event of the crisis was a combustion of the regular old gas in the basement, not an addition of lots of new gas.
6) Egregious and Predatory Lending
"adjustable-rate mortgages, mortgages with balloon payments, interest-only mortgages, piggy-back, and so-called pay-option ARM loans."
"...in 2007, before the GR, the foreclosure rate was 5 percent. Its lowest value, between 2002 and 2007, was 3 percent, which was observed in Q3 2005.If one assumes that all of the 2 percentage-point increase in subprimes involved predatory lending, we’re still talking about predatory lending causing, at most, 0.3 percent more mortgages to definitely default, namely, enter foreclosure. This is simply too small a figure to matter to the overall economy. Indeed, given the size of the 2007 mortgage market, it represents just $32 billion. In 2007, U.S. GDP was $14.4 trillion. The economy’s 2007 total net wealth was $68 trillion."
The Dodd-Frank act piled every suggested fix to every perceived financial problem in one place. Even if one regards predatory lending as a problem, even if one does not regard competition as the best disinfectant and guarantor of good treatment, even if one thinks it needs fixing, Larry's point is that such a fix has nothing to do with stopping crises.
7) Dramatic Increases in Household Mortgage Debt
Surely, the addition of over $750 billion in mortgage debt in the course of 6 short years must represent a priori evidence that a massive recession was in the works. Not so. ... The increase in borrowing to purchase homes was not associated with a massive spending spree on the part of the American public. Indeed, the share of GDP consumed by the public remained fixed at roughly 67 percent between early 2002 and late 2007.
What about household debt payment service as a share of disposable personal income? There was an increase prior to the GR, but nothing extraordinary. Between Q4 2001 and Q4 2007, the ratio troughed at 12.1 percent in Q2 2004 and peaked in Q4 2007 at 13.2 percent. A 13.2 percent ratio is small and the increase from trough to peak is only 9 percent
As with houses, your debt is my asset. Larry is verging here into causes of the recession rather than the crisis. There is a case for asymmetries, but the first-order mistake is to think that just because I am in debt we all are in debt.
8) Exponential Growth in Trading Activity by Financial Firms
Here, again, we have a supposed reason for the Great Recession that has no counterpart in economic theory. If Joe and Sally sell the same share of stock back and forth to each other an infinite number of times in, say, a second, nothing real will happen to Joe and Sally or the economy.
9) Unregulated Derivatives and the Repo Market
The reigning narrative – that derivatives were misunderstood and over rated by compliant rating companies – has been questioned in a recent study by economists Juan Ospinal and Harald Uhlig. They examined 8,615 residential mortgage-backed securities (RMBS) over the period 2007-2013, almost all of which were rated AAA. Through 2013, the cumulative loss on these “toxic” securities was only 2.3 percent. Some three quarters of the AAA-rated RMBS had essentially zero losses through 2013. On a principal-weighted basis, the average loss rate was only 0.42 percent.
Yes, losses were far higher for non-AAA rated segments of the RMBS market. But that’s what one would expect from a “great” recession. However, these securities represented a small fraction of the RMBS market.
I think the logic here is that if securities were overpriced, then they should have fallen. That they did not is interesting. Of course, if they had fallen that is not necessarily evidence of overpricing, as there was a huge recession after the run!
What about REPOs? Did they cause the GR? Well, they certainly increased in the run up to the GR. But short-term financial-company borrowing has been growing far faster than the economy for decades. The fact that some economic variable rose rapidly prior to the GR is not evidence that it caused the GR. Smart phone sales tripled between 2005 and 2008, but no one would link that to the GR. Of course, Repos would be implicated in causing the GR had they been part of excessive leveraging by financial intermediaries. But, as discussed above, overall financial- company leverage fell, not rose prior to the GR.
Here again I think a big qualification is in order. The run on repo was a central part of the crisis, and I think Larry and I agree that way too much such run-prone financing was the key cause of the crisis. Again, I think the point Larry is making is that the financial system as constructed is always vulnerable, that the crisis was not brought on by some sudden and preventable increase in debt.
10) Investors Mispriced/Ignored Risk
11) Unaligned CEO Incentives
Yet another explanation for the GR is that CEOs of financial institutions had too little “skin in the game.” Jimmy Cayne, former head of Bears Stern, would surely disagree. Cayne lost close to $1 billion as his bank collapsed. Ken Lewis, CEO of the Bank of America, had $190 million to lose by making wrong decisions and succeeded in losing $142 million. Lehman Brothers’ Dick Fuld received most of his 2007 compensation in the form of Lehman Brothers’ stock.
12) Democratization of Finance
Under this theory, government sponsored enterprises (Fanny and Freddie) and government regulators were too permissive with banks in their quest to help the poor get into affordable housing. ...if this were the chief or even a major cause of the GR, subprime mortgages would need to have played a much larger role than they did.
13) The Federal Reserve Kept Interest Rates Too Low
Thirty-year mortgage interest rates were certainly lower between 2000 and 2007 than in the prior quarter century. But they weren’t that low especially adjusted for inflation. In the 1990s, the real 30-year mortgage rate averaged 7.91 percent. It averaged 6.27 between January 2000 and December 2007. This decline is hardly something to write home about, let alone pretend is the underlying GR culprit.
..the foundational bank-run models --- Bryant (1980), Diamond-Dybvig (1983), Pech and Shell (2003) and related models – admit multiple equilibrium in which financial-market collapse arises absent any fundamental financial- or real-sector problem... from the perspective of these models, the question is not whether the banking system will fail, but when. Hence, it’s passing strange that the FCIC report makes no mention whatsoever of either paper, let alone the theory underlying bank runs.
In completely ignoring the theory of bank runs... the FCIC pretended that what happened wasn’t intrinsic to how the financial market is structured. Instead, the commission, for whatever reasons, appears to have rounded up the usual suspects and held a sham trial.
(My emphasis.) Larry goes on for several pages documenting spreading panic. An earlier quote is good here
SEC Chairman, Mary Schapiro’s, 2010 testimony to the House Financial Services Committee...includes this statement.
The immediate cause of Lehman's bankruptcy filing on September 15, 2008 stemmed from a loss of confidence in the firm's continued viability resulting from concerns regarding its significant holdings of illiquid assets and questions regarding the valuation of those assets. The loss of confidence resulted in counterparties and clearing entities demanding increasing amounts of collateral and margin, such that eventually Lehman was unable to obtain routine financing from certain of its lenders and counterparties
Unsafe at Any Speed, and the limits on bailouts
The banks failed because they could. And they could fail because they were leveraged. They falsely promised to make repayments regardless of the circumstances.
The overall level of leverage is way too high. This reinforces my earlier interpretation of Larry's comment about leverage not being the problem -- there was way too much leverage, but its increase did not directly cause a crisis.
The next part is really interesting.
The Federal Reserve is also leveraged. In the aftermath of Lehman’s collapse, the Fed effectively insured not just checking and saving accounts, but also money market funds. These obligations were officially and, respectively, FDIC and Treasury obligations. They ran to some $6 trillion. But neither institution had $6 trillion in ready cash to make good on its insurance. Hence, the Fed would have been on the hook. Indeed, had things gotten worse, there would surely have been a run on the life insurance industry’s cash-surrender value policies, which, at the time, also totaled roughly $6 trillion.
Now imagine, as discussed in Kotlikoff (2010), that the government’s explicit and implicit pledges of insurance had been called by the public. I.e., suppose the public had, despite the promises of government insurance, headed straight to the banks, money market funds, and insurance companies to empty out their accounts and cash out their cash-surrender value policies. In this case, the Fed would have had to print $12 trillion virtually overnight. The M1 money supply at the time was just $1.5 trillion. Hence, this would have produced fully-justified fears of hyperinflation leading everyone to run for their money before prices soared
The U.S. has yet to experience a run on its central bank. But this is common in countries like Argentina, ...
Larry is, of course, an expert on all the explicit and implicit credit guarantees our government offers. I was unaware that $6 trillion of "cash surrender value policies" existed, and given the bailouts of other insurance policies we would certainly have seen them bailed out too. Fannie and Freddie guarantee most mortgages.
Though it does represent promises of payment, I don't think this really is "leverage" of the Federal Reserve. The government has, in essence, written a lot of put options, which is a different thing.
What happens in an even more massive run, with more massive bailouts is an interesting question. It's not as simple as "print [ing] $12 trillion overnight." The Fed issues reserves, convertible to cash, but always in return for something else. So, this would have put a big strain on the Fed's legal limitations of what collateral it can accept and from who.
The Fed mostly deals with commercial banks. Imagine a massive run on commercial banks, perhaps stemming from a rumored cyberattack that emptied one of them out. The Fed would have to lend against the entire portfolio of bank assets, not just liquid securities. Goodbye Badgehot.
As Larry points out, really the FDIC and Treasury are the ones guaranteeing non-bank debts. The Treasury would have to borrow $12 billion overnight, sell it to the Fed, and then use it to bail out here and there. The ban on direct Fed-Treasury purchases would make this very hard, and would probably have to be scrapped.
But the huge increase in money would clearly be a temporary increase in money demand, and not obviously inflationary. Moreover, the Treasury, Fed, FDIC, etc. would take on assets. If these operations could be reversed after the panic passes -- if there is not a tremendous amount of actual lost value, as there was not last time, the money could be soaked up again. Even if not, the operation would not be inflationary if people thought the government could retire the debt and soak up the interest-paying reserves by future surpluses. We get inflation -- and Argentina gets inflation -- if and only if this nightmare involves a large fiscal transfer, that the US government cannot or will not pay off, that is financed by a permanent increase in non-interest-paying money.
US Federal debt is about $10 trillion larger than in 2008, and we're running $1 trillion deficits, with no end in sight. The reliability of the fiscal resources to make good on all these put options is, I think, a serious problem, and the heart of the potential inflation Larry describes.
The Role of Opacity
Bear Stearns was among the first to be picked off by those who stood to gain by a financial collapse because it was viewed as particularly opaque
The fact that Bear’s stock was valued at $60 per share one week before JP Morgan bought it for $2 per share (less a $29 billion sale of Bear’s troubled assets to the Fed valued at far less than $29 billion) tells us that no one knew anything about Bear’s assets, either before it died or when it died. Its valuation was, it seems, purely a matter of conjecture. Before it didn’t, the market apparently though Bear’s assets were worth something because everyone else thought its assets were worth something. This too is the stuff of multiple equilibria
Gary Gorton likens financial crises to a salad bar, where someone says "there's a news report of e-coli in the (inaudible)". So what do you do? Absent information on which ingredient has the e-coli, and the time or inclination to investigate, you go order a hamburger.
Now ask yourself, where is there a mountain of debt that can't be repaid, much of it very short term, phoney-baloney accounting, and opaque off-balance-sheet exposures (put options)? Sovereigns.
Bad/greedy/lazy/irresponsible actors, we’re told, engaged in all manner of financial malfeasance, risk taking, negligence, theft and greed. And what we’re told is true. There were plenty of bad actors... But the story of these bad actors is not the real story of the Great Recession. The real story is that both the economy and the banking system are inherently unstable. ... If enough people think enough people think a bank is going down, that bank will go down regardless of its true condition. If enough people think the economy is going down, the economy will go down, also regardless of its true condition.
The first conclusion seems to me spot on. The second one is more provocative. Here Larry signs up with multiple-equilibrium theories of recessions, as well as simple multiple equilibria associated with run-prone assets like overnight debt. It's verbally plausible. If you think there will be a recession tomorrow, you fire workers and do not invest, and there is a recession today. But you can see a crucial derivative needs to be greater than 1.0 for that to work. Lots of formal models have multiple equilibria, but I've spent 10 years putting nominal multiple equilibria back in the new-Keynesian model, and real multiple equilibria are harder to get going.
One approach to addressing the problem of financial multiple equilibrium is to replace Dodd- Frank with more fundamental financial reform, such as Kotlikoff (2010)’s Limited Purpose Banking (LPB). LPB would transform all financial corporations into 100 percent equity-financed mutual fund holding companies subject to full and real-time disclosure supervised by the government.
Choose your equity-financed banking flavor, and we can end financial crises forever. Just why we don't do it seems an important -- and sadly forgotten -- question.