Developing the ‘Financial Instability Hypothesis’: More on Hyman Minsky’s Approach

(cross posted at EconoMonitor)

Since Paul Krugman kicked-off a heated discussion about Minsky’s views on banks, and because the annual “Minsky Conference” co-sponsored by the Ford Foundation and the Levy Economics Institute occurred this past week, I thought it would be useful to run a couple of posts laying out what Minsky was all about. This first piece will detail his early work on what led up to development of his famous “financial instability hypothesis.”

Minsky’s Early Contributions

In his publications in the 1950s through the mid 1960s, Minsky gradually developed his analysis of the cycles. First, he argued that institutions, and in particular financial institutions, matter. This was a reaction against the growing dominance of a particular version of Keynesian economics best represented in the ISLM model. Although Minsky had studied with Alvin Hansen at Harvard, he preferred the institutional detail of Henry Simons at Chicago. The overly simplistic approach to macroeconomics buried finance behind the LM curve; further, because the ISLM analysis only concerned the unique point of equilibrium, it could say nothing about the dynamics of a real world economy. For these reasons, Minsky was more interested in the multiplier-accelerator model that allowed for the possibility of explosive growth. In some of his earliest work, he added institutional ceilings and floors to produce a variety of possible outcomes, including steady growth, cycles, booms, and long depressions. He ultimately came back to these models in some of his last papers written at the Levy Institute. It is clear, however, that the results of these analyses played a role in his argument that the New Deal and Post War institutional arrangements constrained the inherent instability of modern capitalism, producing the semblance of stability.

At the same time, he examined financial innovation, arguing that normal profit seeking by financial institutions continually subverted attempts by the authorities to constrain money supply growth. This is one of the main reasons why he rejected the LM curve’s presumption of a fixed money supply. Indeed, central bank restraint would induce innovations to ensure that it could never follow a growth rate rule, such as that propagated for decades by Milton Friedman. These innovations would also stretch liquidity in ways that would make the system more vulnerable to disruption. If the central bank intervened as lender of last resort, it would validate the innovation, ensuring it would persist. Minsky’s first important paper in 1957 examined the creation of the fed funds market, showing how it allowed the banking system to economize on reserves in a way that would endogenize the money supply. The first serious test came in 1966 in the muni bond market and the second in 1970 with a run on commercial paper—but each of these was resolved through prompt central bank action. Thus, while the early post-war period was a good example of a “conditionally coherent” financial system, with little private debt and a huge inherited stock of federal debt (from WWII), profit-seeking innovations would gradually render the institutional constraints less binding. Financial crises would become more frequent and more severe, testing the ability of the authorities to prevent “it” from happening again. The apparent stability would promote instability.

Extensions of the Early Work

With his 1975 book, Minsky provided an alternative analysis of Keynes’s theory. This provides his most detailed presentation of the “financial theory of investment and investment theory of the cycle.” The two key building blocks are the “two price system” that he borrows from Keynes, and the “lender’s and borrower’s risk” often attributed to Kalecki but actually also derived from Keynes. Very briefly, Minsky distinguishes between a price system for current output and one for asset prices. Current output prices can be taken as determined by “cost plus mark-up,” set at a level that will generate profits. This price system covers consumer goods (and services), investment goods, and even goods and services purchased by government. In the case of investment goods, the current output price is effectively a supply price of capital—the price just sufficient to induce a supplier to provide new capital assets. However, this simplistic analysis can be applied only to purchases of capital that can be financed out of internal funds. If the firm must borrow external funds, then the supply price of capital also includes explicit finance costs—including of course the interest rate, but also all other fees and costs—that is, supply price increases due to “lender’s risk.”

There is a second price system, that for assets that can be held through time; except for money (the most liquid asset), these assets are expected to generate a stream of income and possibly capital gains. Here, Minsky follows Keynes’s treatment in Chapter 17 (the most important chapter of the General Theory, according to Minsky). The important point is that the prospective income stream cannot be known with certainty, thus is subject to subjective expectations. We obtain a demand price for capital assets from this asset price system: how much would one pay for the asset, given expectations concerning the net revenues that it can generate? Again, however, that is too simplistic because it ignores the financing arrangements. Minsky argued that the amount one is willing to pay depends on the amount of external finance required—greater borrowing exposes the buyer to higher risk of insolvency. This is why “borrower’s risk” must also be incorporated into demand prices.

Investment can proceed only if the demand price exceeds supply price of capital assets. Because these prices include margins of safety, they are affected by expectations concerning unknowable outcomes. In a recovery from a severe downturn, margins are large as expectations are muted; over time, if an expansion exceeds pessimistic projections these margins prove to be larger than necessary. Thus, margins will be reduced to the degree that projects are generally successful. Here we can insert Minsky’s famous distinction among financing profiles: hedge (prospective income flows cover interest and principle); speculative (near-term income flows will cover only interest); and Ponzi (near-term receipts are insufficient to cover interest payments so that debt increases). Over the course of an expansion, these financial stances evolve from largely hedge to include ever rising proportions of speculative and even Ponzi positions.

Even in his early work, Minsky recognized that desires to raise leverage and to move to more speculative positions could be frustrated: if results turned out to be more favorable than expected, an attempt to engage in speculative finance could remain hedge because incomes realized are greater than anticipated. Thus, while Minsky did not incorporate the now well-known Kalecki relation (in the simple model, investment determines aggregate profit), he did recognize that an investment boom could raise aggregate demand and spending (through the multiplier) and thus generate more sales than projected. Later, he explicitly incorporated the Kaleckian result that in the simplest model, aggregate profits equal investment plus the government’s deficit. Thus, in an investment boom, profits would be increasing along with investment, helping to validate expectations and encouraging even more investment. This added weight to his proposition that the fundamental instability in the capitalist economy is upward—toward a speculative frenzy.

In addition, in the early 1960s he argued that impacts on private sector balance sheets would depend on the stance of the government’s balance sheet. A government-spending led expansion would allow the private sector to expand without creating fragile balance sheets—indeed, government deficits would add safe treasury debt to private portfolios. However, a robust expansion would tend to cause tax revenues to grow faster than private sector income (with a progressive tax system and with transfer spending falling in a boom) so that the government budget would “improve” (move toward surplus) and the private sector balance would deteriorate (move toward deficit). Once he added the Kalecki equation to his exposition, he could explain how this countercyclical movement of the budget would automatically stabilize profits—limiting both the upside in a boom, and the downside in a slump.

With the Kalecki view of profits incorporated within his investment theory of the cycle, Minsky argued that investment is forthcoming today only if investment is expected in the future—since investment in the future will determine profits in the future (in the skeletal model). Further, because investment today validates the decisions undertaken “yesterday,” expectations about “tomorrow” affect ability to meet commitments undertaken when financing the existing capital assets. There is thus a complex temporal relation involved in Minsky’s approach to investment that could be easily disturbed. Once this is linked to the “two price” approach, it becomes apparent that anything that lowers expected future profitability can push the demand price of capital below the supply price, reducing investment and today’s profits below the necessary level to validate past expectations on which demand prices were based when previous capital projects were begun. The margins of safety that had been included in borrower’s and lender’s risk prove to be inadequate, leading to revisions of desired margins of safety going forward.

Minsky continually developed his financial instability hypothesis to incorporate the extensions made to his investment theory over the course of the 1960s, 1970s, and 1980s. The Kalecki equation was added; the two-price system was incorporated; and a more complex treatment of sectoral balances was included. Minsky also continued to improve his approach to banks, recognizing the futility of Fed attempts to control the money supply. He argued that while the Fed had been created to act as lender of last resort, making business paper liquid, the Fed no longer discounted paper.  Indeed, most reserves supplied by the Fed come through open market operations, which greatly restricts the Fed’s ability to ensure safety and soundness of the system by deciding which collateral to accept, and by taking a close look at balance sheets of borrowers. Instead, the Fed had come to rely on Friedman’s simplistic monetarist view that the primary role of the Fed is to “control” the money supply and thereby the economy as a whole—something it cannot do. The problem is that attempts to constrain reserves only induce bank practices that ultimately require lender of last resort interventions and even bail-outs that validate riskier practices.  Together with countercyclical deficits to maintain demand, this not only prevents deep recession, but also creates a chronic inflation bias.

Next week we will get into more of his writing on “what banks do” along with his recommendations for reform.