The requirement that banks sell lots of bonds that can be bailed in during a crisis is a key part of the regulations to improve financial stability. But selling the bonds is half the story: someone has to buy them. And that could be an unrecognised problem.
In these inherently unstable times for financial institutions, covered bonds, deposits and other forms of ‘preferred’ debt all rely heavily on a nice thick cushion of bonds beneath them in the creditor hierarchy to bear the losses and keep the bank running when times get bad. The requirement for minimum amounts of these bonds and the resolution rules that turn those amounts in to real protection in practice are fundamental to the nice, cosy feeling that the markets enjoy at the moment. Nothing can go wrong with our bonds or deposits, MREL will ensure that.
This is of course true. But, and there is always a but, these bonds have to be bought by someone. I like to think of the bonds being owned by sophisticated asset managers analysing the credit, thinking carefully about relative value and being able to bear the loss - if it happens - then passing that loss on to their investors. Never pleasant but nothing that will destabilise the financial system.
Then I looked at the data from the ECB’s securities database which gives an idea of who owns which bonds.
Firstly, unlike covered bonds, the vast majority of bail-inable debt is held domestically. In the case of Greece and Portugal it is almost all of those bonds. Not only is that contrary to the ideal of capital markets union, it is also a bit odd. All in the eurozone, why should there be such a vast difference between a German and a Greek investor’s perception of the relative value of a Greek bank’s bail-in debt? Could it be that national prejudices trump yield considerations more in the case of bail-inable debt than they do in, for example, covered bonds? Despite the fact that yields are more compressed in covered bonds (so there is less incentive to go abroad to buy)?
Does this matter? Massively. Bail-inable debt must be rolled over when it falls due (with a few exceptions like callable tier two bonds). If investors and issuers are both subject to the same economic downturn at the same time, rolling that debt domestically could become a real problem when times turn bad.
Then there is the type of investor. According to the ECB database, the vast majority of bail-inable debt is owned by other banks except in Italy, where it is mainly owned by households. I’m not sure which worries me more.
Retail investor holdings of covered bonds hugely complicate the - supposed to be smooth – resolution process with the ugly detail of politics. We have a problem as long as there is even the suspicion that a bail-in might not operate in the predictable way that it is supposed to – retail favoured over other classes of bond holder. Arguably this has already happened in a couple of cases. Equally as bad is the possibility that the entire resolution process could be delayed because a bail-in would be politically unacceptable (for example, in the case of an imminent election).
Even worse, what about banks holding each other’s bail-in debt? A lot has been said about the systemic importance of reducing the bank-sovereign credit nexus, less about the correlations of credit within the banking sector. Banks have long been discouraged from holding one another’s capital. Now that bail-in debt is capital in all but name, perhaps we need similar discouragements for bank’s helping each other meet their MREL targets. Without that, the benefits of that safety cushion may be totally undermined.