Now banks face bail-in, should we rethink deposit insurance?

The current deposit insurance system in many countries is the result of rapid responses to a major unexpected crisis, made worse by major deficiencies in the ability to handle such crises. Now the problems of crisis management are being addressed, with resolution plans for systemically important banks involving the bailing in of creditors, perhaps it is time to review what deposit insurance should look like in the new normal environment, rather than lock in more expensive arrangements that were only meant as a stopgap. This question is particularly important for the EU, which is proposing to gear up the system further.

Before the global financial crisis, the perceived wisdom on deposit insurance was that ordinary people could not be expected to understand and monitor the risks their banks were taking and hence ought to be protected from loss should their bank fail. Beyond some relatively low sum, it was thought that deposits constituted a form of deliberate investment and hence holders of these deposits should be subject to the risks they were taking. Some countries disagreed and offered more and even total insurance; others, such as New Zealand, offered none at all.

Attitudes then changed. To maintain confidence during the financial crisis, insurance was extended and is now at much higher levels. Many depositors have priority among unsecured creditors and hence can expect to get back all, or at least most, of their money even if they are not insured; although they may have to wait up to a decade for the final payment when the resolution proceedings are completed. Additionally, it is no longer thought appropriate that the taxpayer should bail out a systemically important bank that is in trouble, but that it should be bailed in by its shareholders and then, if that is not enough, by its creditors in increasing order of seniority. In the EU, no creditor should be made worse off by this process than they would have been had the bank gone through normal insolvency. Restitution is to be made from a resolution fund contributed to by all banks in advance and in proportion to their deposits.

It makes sense to react vigorously in a crisis to stop the system imploding, but once the crisis is over surely the system should return to normal, where bank failures are rare and most banks that fail are small and not a threat to the financial system? Not doing so would be an admission that previous deposit insurance systems were mistaken, as were any ensuing depositor losses at the time. As far as I know there is no science behind the €100,000 or $250,000 US caps. They are just big numbers intended to make sure that the whole of almost all deposits are protected and remove any social justice arguments from resolutions.

One argument in favour is that this will reduce the chance of any bank run as depositors are protected. However, by the time depositors realise they ought to run, the smart money will have already left the bank, the prices of other securities that might be bailed in will have already begun to fall and access to wholesale funding will have largely dried up, as with Northern Rock in the UK in 2007.

The position is complicated by the existence of contingent convertible bonds (CoCos) and related hybrid securities that can be written down or converted into shares at some trigger point short of insolvency so that the bank may 'recover' without the authorities intervening. Because insured deposits form a reliable funding base that will not disappear early in a crisis, this helps the authorities 'buy' enough time to be ready with an orderly resolution if the bank does fail.

Nevertheless, with the triggering of CoCos, there is a danger that the whole process of resolution is brought forward as other creditors seek to exit before making losses. Hence the depositor cushion becomes less valuable. That makes a strong argument for countries like New Zealand that have neither deposit insurance nor depositor preference to introduce them. Without deposit insurance, a bank run would make sense as soon as the CoCo market becomes obviously volatile. Indeed, one might expect a general run on the banking system, as if one major bank is in trouble it will lead people to worry that the others might be the same, whatever they say.

This suggests such countries should actually have implicit deposit insurance, as some sort of general guarantee will be needed to restore confidence. This was the case for Australia and New Zealand in 2008, and Australia has kept the arrangement. It is clearly debatable whether it is more costly to have permanent deposit insurance, which incurs ongoing costs, or simply to introduce a guarantee in the rare event of a major crisis. There were some major problems with the temporary New Zealand scheme largely because it was introduced in a hurry.

The EU issued a proposal for a comprehensive deposit insurance system in December 2015 which would cover the eurozone and increase the costs for some countries. While one might question whether it is fully necessary, it has one aspect one might have expected to see in all insurance schemes: namely, reinsurance. That is normal practice in other areas of insurance and would help heavily indebted countries that suddenly face a banking problem.

Perhaps now is the time to consider reducing the cost of deposit insurance for normal times and focus on having a resolution scheme for large banks in systemic crises, rather than doubling up deposit insurance and resolution funds and providing insurance for such large deposits. Since most of the failures in normal times will be of small banks, this is simply subsidising their deposits. It is not clear to me that the system we are progressively moving into makes as much sense as it might. Perhaps a more fundamental rethink is called for?