First posted on:
Finance: Research, Policy and Anecdotes, 2 March 2018.
Consuelo Silva-Buston, Wolf Wagner and I have been working on our paper, ‘The economics of supranational bank supervision’, for quite some time, but finally have a version that we feel comfortable with to publish as a CEPR Discussion Paper. It is the third (but not last!) of a series of papers that Wolf and I have co-authored on the tension between national bank supervision and cross-border banking, a tension that comes out especially during the failure and resolution of banks, as we show in our 2013 paper with Radomir Todorov, published in Economic Policy.
Does that imply that supranational supervision is a better solution? Not so fast, we argue in our theory paper, published in the International Journal of Central Banking, where we formalise the trade-off between cross-border externalities from bank failure and heterogeneity in the cost of bank failure, where the former makes cooperation, if not supervisory integration, welfare-improving and the latter, welfare-reducing.
In the latest paper with Consuelo, we take this model to the data. We first hand-collected data on supervisory cooperation among a global sample of countries over the period 1995 to 2013. The information is at the country-pair level and was gathered from the supervisory bodies' websites and official documents available online. Supervisory cooperation can take many different forms and degrees of intensity. In our work we distinguish between four (and increasingly intense) forms of cooperation:
- A Memorandum of Understanding for information sharing and onsite inspection;
- A College of Supervisors;
- A Memorandum of Understanding on crisis management and resolution; and
- A supranational supervisor.
We then regress these gauges of supervisory cooperation on measures of cross-border externalities (including foreign bank share, financial market integration, and currency unions or peg) and heterogeneities (structure and level of financial development, political structure, historic links, geographic proximity, etc.).
Consistent with theory, we find that higher cross-border externalities between two countries increases both the likelihood of cooperation and its intensity. Distinguishing between different dimensions, we find that it is all three – cross-border externalities through bank ownership links, spill-over effects through financial markets, and linkages within a currency union – that increase the probability and intensity of cooperation. We also find that higher heterogeneity between countries decreases the likelihood and intensity of cooperation, again consistent with theory, though the economic effect of heterogeneity is less prominent than that of externalities.
In summary, economics matters! Countries are not only driven by politics and history when agreeing to cooperate among regulators, but also by the net benefits of doing so.
We can also use the exercise to predict which countries are likely to cooperate with each other. Take the European Union. The banking union is currently limited to euro area countries, but is – in principle – open to non-euro countries of the EU. Using our model to consider externalities and heterogeneity of non-euro EU countries vis-à-vis the banking union countries, we would predict most of them not to join the banking union, maybe with the exception of Bulgaria and Denmark, countries with a currency board and a peg to the euro, respectively.