The Australian Prudential Regulation Authority (APRA) is currently undertaking a review of hybrid capital bonds (often referred to as 'bail-in bonds') issued by banks, with a focus on those known as Additional Tier 1 (AT1) instruments. This is hardly surprising since recent international experience (particularly the demise of Credit Suisse) has shown them not to work for the purpose for which they are designed. Their continued use, rather than just redesign of some of their features, is open to question.
Context to the review
Bail-in securities were introduced as part of the Basel 3 regulatory framework following the Global Financial Crisis, and introduced in Australia in 2013. AT1 securities are intended to provide the next line of protection, after common equity, against bank failure. This is meant to occur by them being 'bailed-in' to convert into equity should the bank’s equity capital base fall sufficiently for the bank to reach a point of “non-viability”.
APRA would make the decision to require the 'bail-in', but the exact features of what a bail-in would entail (exactly when, how many, and which of, the securities to be bailed-in) has been left purposely vague. Should a bail-in of all AT1 securities be insufficient to adequately recapitalize the bank, then another class of bail-in securities (Tier 2) would be bailed in.
It can be argued that the existence of bail-in securities (used to keep capital ratios above APRA-required levels) helps to reduce risk of depositor (or other creditor) runs by the psychological effect of them providing a larger buffer to absorb losses. But that could also be achieved by simply requiring higher levels of equity, and avoiding the question of whether a bail-in will work when it is implemented.
Problems with bail-in bonds
One reason for doubting the robustness of the bail-in approach is found by asking what happens when a bail-in occurs? The total capital (equity plus bail-in securities) ratio of the bank won’t be changed, just its composition.
So, the bank will likely need to issue new securities to increase its capital ratio - and what investors would want to subscribe to a new issue by a bank which has just been seen to be in difficulty? And larger depositors not covered by the Financial Claims Scheme will likely think it prudent to move their funds elsewhere.
Bail-in is meant, in theory, to enable the bank to continue operations, but in practice is more likely to sound the death-knell of the bank.
Other issues
There are further complications in the specific Australian situation. AT1 securities generally pay franked dividends, and have hence been attractive to Australian retail investors, and unattractive to foreign investors who do not value the franking credits. The foreigners prefer the AT2 securities which pay unfranked distributions.
It is difficult (impossible) to imagine John Lonsdale (CEO of APRA) not consulting the Australian Treasurer before doing a bail-in of a bank’s AT1 securities. And it is even more difficult to imagine the Treasurer being willing to have Australian retail investor holders of AT1 securities being bailed-in while foreign investors holding Tier 2 hybrids are not bailed in. The politics would prevent the bail-in regime from operating in the way it is designed to.
One way to overcome that problem might be to preclude retail investors from the AT1 market. The introduction of Design and Distribution Obligations (DDOs) powers to ASIC has meant that retail investors are no longer generally able to participate in new issues of AT1’s. But, paradoxically, once issued, the AT1’s trade on the ASX and so are available to retail investors.
Given the difficulties in pricing the bail-in risk of these complex AT1 securities, and thus identifying their appropriate yield, it is also paradoxical that they should be available to retail investors who really cannot evaluate the risks. If nothing else, legislation to prevent AT1 securities from paying franked distributions could be expected to diminish retail investor interest.
There are numerous other problems with the bail-in approach – most particularly that it is unlikely to work! But the fundamental question which should be asked is why require these complicated bail-in securities as part of the bank’s capital base rather than simply having a higher required level of common equity. It is by no means apparent that replacing bail-in securities with common equity would substantially increase the overall cost of bank funding.
Kevin Davis is Emeritus Professor of Finance at The University of Melbourne.