Comment: Measuring financial fragility

COMMENT BY CENTRALBANKNEWS

The measurement of financial stability remains elusive. Is financial stability the absence of a financial crisis or the presence of something - efficient financial intermediation perhaps? The unit of measurement is at present a "bank crisis". How to define a "crisis" and its duration are far from straightforward. In addition, looking only at crises disregards most of the available data as financial stability not instability is the order of the day in most economies. How policymakers would love something like an inflation target to concentrate their minds.

While such a measure may be a long way off, research by Goodhart and colleagues at the London School of Economics to develop a "welfare index of financial fragility" has produced what may be the first step in this direction. They find that two variables were most important in explaining effects on the real economy: the "probability of default" and the profitability of banks. In their model the former is measures by an index generated by the IMF, while the latter is proxied by a share price index for the banking sector.

By looking at the extent to which these two variables affect GDP, the researchers were able to generate an index which gives an indication of the build-up of risks in the banking sector. Although the measure does not quantify the risk to GDP resulting from banks' behaviour, the index can track how a country is performing over time. It can also be used to facilitate cross-country comparisons or, by entering new or potential data, to create forecasts.

Much work remains to be done, as Goodhart and his colleagues admit. The model and index incorporates the banking sector, but omits such obvious influences on financial stability as property prices or equity markets. From a policy perspective, the index does not give a definitive "danger level". Another concern might the measure's ability to escape the predictions of "Goodhart's Law", which says that once an economic indicator is made a target for policy, it loses the information content that qualified it for such a role in the first place. In what could prove to be a case of cruel irony, Goodhart's Law may just warn against the elevation of a potential "Goodhart index" to the level of a policy target.

Future work will no doubt look to address these areas, but the results from this tentative first step are "surprisingly good" and shows it can be done says Goodhart. Financial stability experts will no doubt follow these developments eagerly.

* Goodhart, C.A.E. et al. 2006. Towards a Measure of Financial Fragility, Financial Markets Group, London School of Economics.