Hyman Minsky

McGraw Hill 2008

Early on during my ultimately-to-be-abandoned Harvard PhD studies in climate science, I remember talking to fellow graduate students about the long-term prediction ability of global atmospheric computer models.

These models divide the atmosphere (and sometimes the oceans too) into a three-dimensional grid covering the globe. The detail of real sky and water gets replaced with boxes perhaps a hundred miles square and half a mile high. The computer is programmed with the laws of physics and chemistry and applies them to each box and its boundaries with the others. The modeller can tweak things like the concentration of greenhouse gases in the atmosphere and the computer predicts what the climate will be like in thirty years’ time.

For all the approximations that go into these models (each one of which can be a PhD topic in its own right) the models do a reasonable job at replicating the behaviour of the atmosphere in the recent past, and hence offer at least a plausible stab at the future. The only caveats are the inputs, particularly those that involve human behaviour such as carbon emissions and land use. Twenty years ago, the inputs had to be specified in advance before the model could run.

My question to my colleagues was that this surely was unrealistic. If global warming took place as the models seemed to say it did, then people would modify their behaviour. They might cut back on greenhouse emissions, or more drastically, would change land use as different parts of the planet became more or less viable for habitation and farming. Any genuinely predictive climate model would have to incorporate these feedbacks otherwise it would be worthless. But how could you do that? ‘Economics’ was the answer. That was the province of another academic department, where experts understood the mechanisms of price and demand and for whom the outputs of our models were their inputs.

Twenty years ago, these two academic departments weren’t talking, although they might be now. But I was reminded of this discussion when I read the recently reissued 1986 book by the late Hyman Minsky on financial stability. Outsiders can learn a lot about the mysteries of economics by reading Minsky, in particular how much of it depends (like climate science) on holding certain things fixed in order to reach conclusions about particular models of reality. Mainstream economists have long treated banking and finance in this way, as a kind of invisible transmission mechanism connecting important things like money and output.

Flying in the face of fashion, Minsky zeroed in on these neglected areas. He saw how banking and finance could mutate into much more than a mere linking mechanism, and become a source of instability for the entire economy. It was obvious to him (if not others) that banks would seek to maximise profit by financial innovation, and that these innovations would undermine stabilising tools such as the control of money supply by central banks. Minsky sketched out how these innovations could lead to bubbles with a taxonomy of financing mechanisms. ‘Hedge finance’ was the safest (investment cash flows paid back interest and principal), then came ‘speculative finance’ (principal had to be refinanced with new debt or equity), followed by ‘Ponzi finance’ where income was insufficient even to cover interest payments.

It is hard to overemphasise just how prescient Minsky was. The growth of commercial paper (CP) markets and non-bank deposits in the 1960s and 70s provided him one example of innovation undermining bank supervision (Until I read his book, I didn’t know that CP backup lines were instigated by US regulators in 1970 after a meltdown in the market). But the emergence of a trillion dollar market in asset-backed commercial paper and structured investment vehicles (SIVs) during the decade after Minsky’s death in 1996 would have fascinated him. And what better example of Ponzi finance than collateralised debt obligations composed of subprime mortgage bonds, that were wiped out when subprime mortgage borrowers failed to refinance after their teaser rates expired?

Minsky saw himself as a Keynesian, but he went further with he called the ‘financial instability hypothesis’. When I ran a peer-reviewed journal, I would routinely encounter the mainstream economics device of equilibrium in submitted papers on subjects such as derivatives pricing or portfolio theory. Minsky persuasively argues that equilibrium is an illusion, a mere ‘period of tranquility’, during which financial innovation ‘validates’ increasing risk-taking and asset bubbles. What better example of Minsky’s validation mechanism than Value-at-risk (VAR), which enabled banks to use quiescent periods of low market volatility and ample liquidity to justify low capital requirements for supposedly hedged trading positions?

If Minsky was prescient about the problems that could occur, he was no less prescient about the solutions that would ultimately be needed. Here he is, writing in 1986 on financial reform:

Financial reform needs to confront the public nature of much that is private. Big or giant organisations carry an implied public guarantee (i.e. contingent liability) on their debts. This introduces a financing bias favouring giant banks, for the implicit public liability leads to preferred market treatment. Government intervention to validate the cash flow commitment takes place even if investments are inept.

Today, the ‘too-big-to-fail’ problem is at the heart of post-crisis banking reform initiatives. Other Minsky ideas, such as allowing central banks to discount eligible assets (i.e. lend money against them as collateral) not just to banks but all financial institutions, became part of the central bankers’ toolkit during the crisis. Janet Yellen, herself an avowed Minskian, has become vice-chairman of the Federal Reserve—no better sign that a once-neglected economist has got the recognition he deserves.