Bank deposits have, at least in the past, provided a source of cheap finance for other activities of banks, particularly investment banking. The value of those guarantees are uncertain, especially in the light of recent regulatory changes, and it is unclear who benefits from them; customers, debt holders, shareholders or managers?
The guarantee has had two effects. One is excessive risk taking by banks, involving significant leverage and own position taking. The second is bad lending decisions (or bad acquisition decisions), for example, HBOS in the UK, Spanish Banks, Irish Banks and Landesbanks in Germany.
I see two factors that have exacerbated these problems. First the governance of banks was particularly poor. Shareholders of UK banks are highly fragmented even by the standards of dispersed ownership markets- it’s an extreme form of free riding or ‘ownerless companies’ described by Lord Myners or ‘weak owners strong managers’ described by Professor Mark Roe of Harvard. And of course debt holders are passive either because they know they will receive a full pay-out or because they don’t get a vote until the bank is already distressed or broke.
The second effect is the interplay of high leverage and deposit guarantees. Debt is relatively cheap to a bank compared with equity because of those guarantees. The tax advantages of debt and the mispricing of guarantees lowers the cost of debt to a bank. It should come as no surprise that banks don’t like raising equity and have a ‘love affair’ with debt?
The high leverage encourages boards to focus on the return on equity and make decisions that increase those returns even if the bank does not earn its cost of capital i.e. its return on assets. This makes for a classic conflict between debt and equity. Unfortunately depositors and many bondholders feel that in the event of distress they are either guaranteed through deposit insurance or they will receive all their money back because government does no dare to allow banks to go broke.
Five possible solutions to the crisis in banking
One obvious solution is to set capital requirements so high that shareholders take far more of the risks, and the value of guarantees are lowered. The problem with this solution is that setting capital requirements at say 15% or 20% will involve raising huge amounts of equity. For some banks which have a debt overhang problem this will impose large losses on existing shareholders. Such a solution will take some time to implement.
A second solution, although much easier said than done is to improve the governance of banks. We could broaden the fiduciary duty of banks to be responsible not only to shareholders but also to depositors and even bondholders. That is, get them off this opium of return on equity and onto return on assets. Also ensure their remuneration is related to return on assets. However, (large) banks are very complex organizations and bank boards do not have the information to pinpoint problems early enough. Problems of fraud, mis-selling as well as excessive leverage should tell us that with the best of directors some banks are simply too complex for boards to manage with confidence.
A third solution is to provide a level playing field between equity and debt. We could price guarantees better to reflect leverage and risk but I think this is a complex and difficult task and I see little prospect of doing this well, because the leverage and risk of banks can change so quickly, and our models for pricing risk are not very robust. We could also remove the tax advantages of debt for banks although this is not as easy as it looks.
A fourth solution is to improve resolution regimes. Here we can copy UK utilities which have had special resolution regimes that ring fence the utilities, lock up dividends when credit ratings fall below a certain level, and have special administration which can take decisions that favour consumers over shareholders. In addition you can have ‘bail-ins’ (e.g. cocos).
Resolution regimes, on their own, are unlikely to constitute the full solution. Banks are different from utilities. The latter are assumed to be making an operating profit. That is leverage does not destroy the underlying business. With a bank high leverage affects the riskiness of the assets and can impair the underlying assets and the profitability of the business’ thereby destroying value. This upsets one of the fundamental assumptions of the Modigliani and Miller theorems that leverage and the value of assets are independent.
I believe resolution regimes should be viewed as complements not substitutes to the other remedies. Innovations such as cocos are untested. Also, losses of the banks have been so large that the size of the bail in might have to be very large to prevent a disaster. In a recent fraud case in our courts, the size of losses at one stage exceeded I believe £7 billion, although they were later reduced to about £1.3 billion. Literally, a middle ranking trader can ‘sink’ a bank’.
Finally, we might improve the governance rights of certain classes of bondholders. Provide them with incentives to better monitor the banks. It can make little sense that a class of security holders that has say 10 times the liabilities of equity holders only has voting rights when the bank is insolvent or when its covenants are breached; by then it is probably too late. It may be useful to investigate allowing greater control rights by certain classes of debt holders for example when the rating of a bank falls below a certain level.
Unless we can be confident that these solutions will work we will have to go back to the sledgehammer option of raising equity requirements to very high levels. That is if you are too big to fail, we will impose such high equity capital requirements so as to reduce the risk of insolvency to virtually zero.
One solution I have not discussed is ring-fencing the lending bank from the investment bank, as suggested by the Vickers commission. I think this recommendation is absolutely right but it is not a silver bullet. Many of our banks have failed because of bad lending decisions and separation will not reduce the costs of these failures. However, it clearly provides a mechanism for shielding core activities from excessive risk taking in a number of other activities (eg trading in derivatives) by banks and reducing interconnectedness.
I suspect that some of these solutions will force banks to slim down and even break themselves up. I suspect this is desirable. Little is known about the value of universal banks. We do know however, that some of them are too big to fail and their failure could threaten the public finances.
Julian Franks spoke at the Global Leadership Summit #gls2013. He also led the new Corporate Governance Programme on 3 June 2013.