How to Measure Financial Fragility

We may not have a high degree of success at predicting precisely when a financial crisis will occur or exactly how big it will be, but what we can and should do, says Éric Tymoigne, is develop effective ways of detecting and measuring the growth of financial fragility in a system.  “[S]ignificant economic and financial crises do not just happen,” he writes, “there is a long process during which the economic and financial system becomes more fragile.”

One of the purposes of the Financial Stability Oversight Council (FSOC) that was created by Dodd-Frank is to provide regulators with an early warning system regarding threats to financial stability.  In light of this, Tymoigne provides his latest contribution to the construction of a measure of systemic risk and identifies specific areas in which we need better data.  With the aid of Hyman Minsky’s theoretical framework, Tymoigne has developed an index of financial fragility for housing finance in the US, the UK, and France.  The point is not to attempt to predict when a shock to the system is likely to occur, but to measure the degree to which such a shock would be amplified through a debt deflation.  From the abstract of his latest working paper:

… instead of focusing on credit risk … financial fragility is defined in relation to the means used to service debts, given credit risk and all other sources of shocks. The greater the expected reliance on capital gains and debt refinancing to meet debt commitments, the greater the financial fragility, and so the higher the risk of debt deflation induced by a shock if no government intervention occurs. In the context of housing finance, this implies that the growth of subprime lending was not by itself a source of financial fragility; instead, it was the change in the underwriting methods in all sectors of the mortgage markets that created a financial situation favorable to the emergence of a debt deflation. Stated alternatively, when nonprime and prime mortgage lending moved to asset-based lending instead of income-based lending, the financial fragility of the economy grew rapidly.

Tymoigne also distinguishes between measuring financial fragility in this way and detecting bubbles:

… the point of detecting financial fragility is not to detect asset-price bubbles. Indeed, the three degrees of financial fragility [Minsky’s famous distinction between “Hedge,” Speculative,” and “Ponzi” finance] are defined independently of the accuracy of the asset pricing mechanism. Ponzi finance does not require the existence of a bubble (i.e., that PA be above its “fundamental” value, however defined); it just requires rising prices of collateral or other assets held by the entity involved in a Ponzi process. The latter condition is required for cheap refinancing to occur and/or to liquidate at a price that covers debt services. More broadly, for Ponzi finance to occur, net worth should be rising, but that does not tell us anything about the existence of a bubble. This is, however, not a problem because this is not what financial fragility aims at detecting; the aim is to detect debt deflation risks that result from the interaction between debt and asset value on the upside. Thus, an economy may be highly fragile even if there is no bubble, and a bubble may exist, but financial fragility may be limited because of the limited recourse to external funding. In that case, a debt deflation is not possible given that debt is not involved, or is of limited size. An example of this would be the current housing bubble in China, which is mostly self-funded.

Read the working paper here.