The fundamental problem in the banking sector is that bank creditors are not exposed to risk of loss. Who are the bank creditors? There are two key sorts: bondholders and depositors. So any useful reform of the banking sector must aspire to increase the risk of loss for (decrease the temptation for the government to bail out) these two classes of bank creditors.
The ICB Interim Report solutions are focused in precisely the opposite direction. They make it more likely that depositors and bondholders in the retail divisions of banks be bailed out by the government. The concept of separating retail and investment banking activities is that the retail unit can be bailed out by the government without also having to bail out the investment banking unit. But we don’t want to bail out the retail unit! We want the retail unit to be able to go bust like other businesses!
Furthermore, by isolating the retail activities of banks we increase the damage incentives to excessive risk-taking by banks – incentives created by bailout promises – cause to the real economy. Retail banking is (like almost all business) an intrinsically risky activity. Banks make loans to businesses and for mortgages and for personal loans, some of which will go bad and some of which are riskier than others. If the creditors of retail divisions of banks are guaranteed by the state (which is made more certain by the ICB proposals) then the banks will seek to engage in riskier retail activities – they will want to lend to riskier businesses, for riskier mortgages, for riskier personal loans. That will all look rosy in the boom times when it means more rapid growth, but then busts will be larger – we increase macroeconomic volatility.
The ICB hopes to shield retail units of banks from the larger losses on retail activities implied by this by requiring them to hold very high levels of capital. But it simply isn’t feasible to make them hold so much capital that no retail unit of a bank will ever go bust, and if it were possible to do it would be a terrible idea. Raising capital requirements for retail units is an expensive illusion – a cosmetic exercise to pretend that retail banks are unable to go bust when it isn’t true, which will cost banks an enormous amount in terms of expensive capital-raising.
That’s not to say there is nothing in the report to be commended. Bail-ins, whereby bondholders have semi-automatic conversion of their debt into equity when a bank becomes distressed – precisely what I and others called for in late 2008 instead of taxpayer recapitalisations – are usefully (albeit sometimes confusedly) discussed. And the competition sections will bear further reflection.
But the central proposal is just the wrong answer to the wrong question.