Minsky Conference “Gary Gorton” Debate

Last week, the Levy Economics Institute hosted the 20th Annual Hyman P. Minsky Conference, a wonkish discussion on all things Hyman Minsky. This year’s focus was on Financial Reform and the Real Economy.

For those of you who are not academic economists, Professor Hyman Minsky argued that stability eventually leads to instability. The stable economic backdrop causes people to become complacent and take on more risks than they might during riskier times.

For some background, see this discussion on Minsky by Prof. Steve Mihm. You can see our earlier guest posts on the Minsky Conference here and here.

The purpose of the Minsky Conference was to “address the ongoing effects of the global financial crisis on the real economy, and examine proposed and recently enacted policy responses. Should ending too-big-to-fail be the cornerstone of reform? Do the markets’ pursuit of self-interest generate real societal benefits? Is financial sector growth actually good for the real economy? Will the recently passed US financial reform bill make the entire financial system, not only the banks, safer?”

I was unable to attend, but several colleagues not only went, but reported back what they saw. The following discussion was art of a longer email thread on some of the emails; it is reproduced here with the permission of the authors.


Steve Waldman (Interfluidity)  writes about the original updates:

I find little to disagree with in your note, except that I find little resemblance between what you say and what I heard in Gorton’s speech.

You write “His key observation is that what is gained by having the Fed and other policymakers guarantee bank liabilities –- namely, financial stability –- is lost through the ensuing complacency which tends to spawn longer, more damaging crises.”

That’s just not what I took from the speech. What I heard was quite the opposite, that the crisis was basically a result of the financial system having evolved means of duration-mismatched finance that were NOT guaranteed, and that therefore the liquidity crises endemic to the system pre-Fed/FDIC had returned.

Gorton carefully avoided making specific policy recommendations, preferring instead to shelter in his self-aggrandizing evidence fetish and putting all hope in Dodd Frank’s Office of Financial Research.

But it seems to me that the clear implication of Gorton’s story — which described the crisis as an old-fashioned, individually rational but collectively destructive bank run — is to guarantee the shadow banking system. I heard nothing of a critique of, say, FDIC in his speech. (Gorton did point out that FDIC was something of a policy accident. Neither FDR nor the banks initially supported deposit guarantees but popular support forced Congress to act. But my sense was that he took this to be a happy accident, that despite an odd process we had stumbled into good policy.)

If you think the crisis was a run on the shadow banking system, AND you think that the sponsors and guarantors of the shadow banking system actually did an okay job in underwriting, AND you think that the right way to prevent “sunspot” bank runs is with deposit guarantees, then the logical policy response is to guarantee the shadow banking system, and not to worry so much about regulating or holding to account sponsors and guarantors, since market forces have in fact proven sufficient to enforce good-enough behavior.

This is almost a syllogism. Gorton set up all three assumptions quite explicitly. That he didn’t state the conclusion was rhetorically savvy, but doesn’t alter the implicit recommendation.

And yet what you heard was almost precisely opposite to what I heard. You see Gorton’s speech as a criticism of complacency due to guarantees, as a warning about moral hazard. I heard Gorton explicitly mock people for “jumping” to moral hazard as an explanation without “evidence”.

Maybe I misheard, and your description is a better characterization of Gorton’s view than my own. If I’m going to write so much about it, maybe I should give the speech another listen. Perhaps others can weigh on with their recollections.

I like your suggestion that lender of last resort activity should be provided, but carefully rationed to parties relatively distant from poor decisionmaking like money market funds, and that more comprehensive guarantees as were provided to several of the larger banks should be explicit and accountable rather than implicit and deniable, as they were via the “no more Lehmans / SCAP” approach.

I wish I had heard Gary Gorton use his considerable intellect and rhetorical skill to make that case. But that is not at all what I heard.


Note: There is an alternative policy recommendation consistent with Gorton’s set-up. Rather expanding guarantees to the shadow banking system, one could argue that duration-mismatched liquidity provision should be forced back onto the balance sheets of existing regulated and guaranteed institutions. Since Gorton’s account is consistent with this policy regime as well, I’ve overstated my case a bit in claiming that support of expanded guarantees was implied. The story Gorton tells implies either an expansion of guarantees to the shadow banking system, or banning the shadow banking system in favor of a remigration to bank balance sheets. I have a guess as to which of these options Gorton would prefer, but since he did not offer actual policy advice, I could be mistaken. However, neither option challenges the practice of guaranteeing duration mismatched liabilities, which practice I think that Andrew would like us to examine more critically (and I’d enthusiastically agree with Andrew on that).


Did you hear Gorton yday? I thought he was 1/4 right and 3/4 downright loony. 17bps of losses? The crisis had nothing to do with prop trading? How did no one challenge him on any of this?

I think that’s exactly right. Gorton tells an important story in his description of runs/liquidity crises that are out of line with the magnitude of any specific shock. But I thought his useful contribution was overwhelmed by the very Orwellian way in which his talk was framed.

Gorton is not content to describe an important piece of the crisis. He instead tried to make it the whole explanation of the crisis. That is not accurate.

In particular, Gorton mocks the important degree to which the crisis was a matter of ex ante insolvency. That is, he pretends it was only a “panic” so that ultimately, banks did nothing wrong and governments would have lost nothing if they had fully insured and guaranteed bank assets.

The most Orwellian bit was the panegyric to the importance of “evidence” in economics. He described basically all commentators other than himself as know-nothings — “journalists” who couldn’t possibly understand, economists who’d never heard of CDOs suddenly calling themselves experts, etc. He talked about the importance of data. And then he rose above his caricature of everybody else by offering up that all-important data, in the form of a single cherry-picked statistic. Rhetorically, he invited his audience to rise above the noise of all the know-nothings and join him in understanding his true scientific, evidence-based account.

I thought his statistic was 71 bps, but maybe it was 17 bps as you say. (I’m writing this from memory; I’ve little inclination to relive or relisten to the speech.) That difference doesn’t matter. The point is that he described the losses as trivial. That is a terrible lie.

Let’s unpack why. First of all, this statistic was realized losses on AAA tranches of subprime RMBS. Someone in the audience did try to call him on “realized”. Gorton mocked the gentleman by simply repeating “realized” in a tone intended to suggest that the interruption was ridiculous, and then moved on.

The word “realized” renders Gorton’s statistic very misleading. Losses are generally realized on securities when they are sold or at maturity. Many, perhaps the majority, of AAA tranches on subprime RMBS have not been traded post-crisis, but have remained on the balance sheets of the banks that held them before the storm. So, “realized” losses may dramatically underrepresent economic losses.

But that’s not all. Restriction of consideration to AAA subprime RMBS ignores a lot of losses, including losses on AAA securities that holders were not prepared to handle. A substantial fraction of an RMBS (usually between 5% and 20% of their value) was not AAA. Much of this non-AAA portion was mezzanine, not equity, and was itself recycled into CDOs or CDO-like securities, whose AAA portions took huge losses. So the economic losses on these deals might have been as high as about 15% [1500 bps!] before any AAA tranches of subprime RMBS would be impacted at all. Much of that value would have become losses to banks holding senior and “supersenior” tranches of CDOs. Also, as you’ve done a lot of work documenting, these mezzanine tranches of RMBS were often placed in the reference portfolio of synthetic CDOs, magnifying the potential exposures of banks and/or bond insurers, who often bore the risk of senior tranches of these deals.

[Gorton doesn’t tell us whether his loss statistic on plain RMBS losses incorporates or fails to incorporate reimbursements from insurers, whether so-called “monoline” insurers or other counterparties who wrote protection in the form of a CDS or similar agreement. If his loss statistic is net of reimbursements, then it ignores payouts made by insurers and other protection writers, who in general were themselves systemically important financial institutions.]

Finally, Gorton ignores the impact of government stabilization on the scale of losses realized in these securities. By government stabilization, I don’t mean the lender-of-last-resort activities that will ostentatiously be paid off without apparent losses to the public. By stabilization, I mean the massive deficit spending that has caused the US fiscal position to deteriorate by more than $4T over the past few years, increasing the total stock of Federal government debt by more than 40% and the stock of Federal debt held by the public by about 80% (since August 2007). If the government had held deficit expenditures constant, or limited them to at most the worst pre-crisis deficits (as a fraction of GDP) experienced over the past decade, losses on these securities would have multiplied dramatically. If this had truly been just a liquidity panic, as Gorton suggests, lender of last resort interventions would have been sufficient, and repayment of the TARP and Fed vehicles would have left the government balance sheet unimpaired (relative to prior projections) after that support was recovered. Yet even after TARP etc is paid back, the United States will still find itself several trillion dollars poorer than would have been projected in, say, 2006.

Gorton might argue that the Treasury’s losses were due to behavioral changes resulting from the public’s panic, that if the government had offered support earlier and unconditionally no panic would have occurred and no long-term deterioration of the balance sheet would have occurred. That’s an untestable counterfactual, so it cannot entirely be ruled out. But it’s pretty implausible. Fundamentally, the losses on RMBS and compilations & derivatives thereof are driven by losses on the value of the assets that secured them. If you really want to believe this was only a liquidity panic, you have to believe that there was no housing bubble, that housing prices would also have been sustainable had the government offered support sufficient to prevent the public from perceiving the hiccup. But home prices had been falling dramatically well before the financial panic occurred. I suppose you could believe that, absent a panic, defaults would not have occurred even despite falling housing prices, because borrowers would have eaten the losses and thereby sheltered the banks. Again, that strikes me as implausible.

Note that I don’t mean to say that Gorton’s liquidity run story is wrong, or that it is unimportant. On the contrary, it is right and represents a very important piece of what occurred. But Gorton manipulatively augments the story with a framing designed to suggest that this was _just_ a liquidity panic, that at worst the financial sector failed to manage liquidity adequately but that its lending and underwriting practices were ultimately not a big problem. There was, he suggested quite explicitly in his talk, no major problem of agency costs and moral hazard. (He describes those as stories that no-nothings jumped to without evidence.) Therefore, he absolves the financial sector of nearly all culpability. But his premise is untrue — this was a deep and serious solvency crisis that has been remedied only through massive government support.

Gorton has written quite a lot, and I’ve read only a bit of his work. Perhaps he has tried to address my concerns elsewhere. But standing on its own, despite a lot of insight and even some brilliance, Gorton’s speech was at best poorly argued. I hate to be uncharitable, but I already have been and I’ll stick with that. The talk struck me not only as weak due to constraints of time or format, but as purposefully manipulative.


Scott Frew responds:

Satyajit Das in a review of Michael Lewitt’s Death of Capital on nakedcapitalism makes the argument you seem to want to make against Gorton, Andrew. In particular, paragraph 3 below.

Satyajit Das: Dead Hand of Economics

In “The Death of Capital”, Michael Lewitt, an investment professional and editor of the HCM Market Letter, explores the ultimate effect of “zombie economics”. Mr. Lewitt’s thesis focuses on the outcome of these economics, in particular, the rise of financialisation, debt and speculation and its effect on the real economy.

“The Death of Capital” is robust in its arguments, especially in it denunciation of Wall Street practices which he views as unproductive and morally reprehensible. The book makes the case that financialisation ultimately has the effect of undermining the fundamental role of capital in societies and financial markets as a mechanism for saving and channelling funds into real businesses. Drawing heavily on the work of Adam Smith, Karl Marx, Keynes and Hyman Minsky, Mr. Lewitt outlines his thesis clearly.

His solution seems curious, in the light of the trajectory of his critical argument – greater regulation, imposition of a tax on speculative transactions (in effect, a Tobin tax) and “principle based” reform. It is unclear why the proposed reforms can or will work, given that the very forces that propelled the financialisation that Mr. Lewitt criticises would be charged with implementing them. As Scottish philosopher David Hume knew: “All plans of government, which suppose great reformation in the manners of mankind, are plainly imaginary.”

The reality is that economics and economic relations are an adjunct to a larger process – the process of broad social control. Marx wrote about the fetishism of money, arguing that “the money-form of the world of commodities … actually conceals, instead of disclosing the social character of private labour, and the social relations between individual producers”. Human beings and societies are unable to see their own products and social relationships for what they are and become slaves to powerful forces.


Rajiv Sethi adds:

Steve and I have been discussing the Gorton speech at length, first at
the conference and then by email over the past couple of days. I have
very little to add to his last message, since I agree with him both on
the substance of the speech and on the policy implications stemming from

Here, in a nutshell, is my response. Gorton really must have known that
he was cherry-picking the data in order to make a dramatic but very
misleading claim. But even if you grant him the dubious hypothesis that
this was a crisis of liquidity and not insolvency, this does not
necessarily warrant a policy of providing whatever liquidity was needed
to keep these institutions solvent. They were engaged in massive
maturity transformation without an explicit government guarantee. They
should have been allowed to go under as long as the real economy could
have been insulated from the damage. One of the things that has been
bothering me most about testimony and speeches by Geithner, Bernanke,
Dudley and others is the suggestion that there was no way to insulate
the real sector from the collapse of the shadow banking system. But this
seems to be an article of faith, not based on a serious examination of

As Perry Mehrling has recently noted, the original mandate of the
Federal Reserve was to “channel credit preferentially to productive
uses. Section 13(2) makes clear who was supposed to get the credit:
“Discount of Commercial, Agricultural and Industrial Paper”, not
speculative financial paper and not Treasury paper.” It’s still not
clear to me why we could not have insulated the real economy from a
collapse of the shadow banking system by direct lending to non-financial
firms. Maybe there’s an argument to be made that this would have been
too little too late, but I don’t see any of the policy architects even
addressing the point. All I hear is that the policy saved us from a
second Great Depression, as if no alternative policy could have done so.

Here are links to Perry’s post, and my response:


Finally, Andrew, I’m not sure where you got the idea that Minsky was
optimistic! He thought we were caught between a rock and a hard place. A
small government sector, balanced budget, and no lender-of-last-resort
interventions would result in severe recessions following periodic
crises. Large government, counter-cyclical deficits, and vigorous
lending would preserve an inefficient industrial structure and encourage
imprudent financial practices. Details here:


Thanks to all of you for copying me on this interesting discussion.



Steve Waldman:

Rajiv and I have had a side conversation on this. Rajiv is a person of unusually generous spirit, and was willing to contemplate the possibility that Gorton is simply so self-assured that what seemed manipulative was natural and unconscious. I felt compelled to defend my nastiness. Rajiv also pointed to a quite interesting piece by Gorton on the CDS market and the informational efficiency of bond pricing, which asked the provocative question of whether we might not want bond markets to be so informationally efficient, in the same way as we might not want medical insurance markets capable of discriminating between good and bad risks ex ante. Though I’d ultimately answer “no”, that we want informationally efficient bond markets, I found the piece very useful and thought-provoking. When he’s not acting as a sly advocate for certain causes, Gorton is a brilliant guy with a lot to contribute.


20th Annual Hyman P. Minsky Conference on the State of the US and World Economies
Financial Reform and the Real Economy
April 13–15, 2011
Ford Foundation
320 East 43 Street, New York City

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:

Posted Under