It’s reasonable to assume that the banking system in the North Atlantic region is insolvent and would be bankrupt but for the reality of recent government bailouts and the expectation of future government bailouts. Certainly, for the system as a whole, the marked-to-market value of its assets is way below that of its liabilities. I strongly suspect that even the hold-to-maturity value of its assets is well below that of its liabilities. Although the system as a whole is broke, there are no doubt individual banks that are solvent. We may not, however be certain as to which banks are solvent and which banks are not.
I also take it is given that it is desirable – essential even – to preserve the core of the banking system and to keep it operating without interruption, because it fulfills an essential role in the intermediation of funds between financial surplus units and financial deficit units – a role for which no substitute can be found or created in the short and medium term. The bulk of the banking system therefore needs to be bailed out. In practice this means that most of the large banks need to be bailed out in the first instance. Consolidation through mergers, acquisitions or liquidations will mostly have to wait until order has been restored in the global financial markets.
The main remaining question then becomes who will pay for the bail out, the tax payers or the existing creditors of the banks (including the shareholders and other providers of equity). I have a strong preference for putting much of the cost of a bailout on the existing creditors. This is in part for reasons of equity and fairness: the existing creditors made bad investments/loans; they ought to pay for their failures. They earned a risk premium while the going was good. They ought to eat the risk when it materialises. It is also for incentive reasons. Future lending to banks and future purchases of bank obligations will be undertaken with a better appreciation of the credit risk involved. Another massive over-expansion of the banking sector will be less likely.
Conventional insolvency procedures are of course one mechanism for ensuring that the creditors pay. Typically, the shareholders get wiped out and the debt gets restructured. Conventional bankruptcy as applied to banks is, however, too disruptive of essential intermediation and portfolio management. So we need a special resolution procedure (SRR), which creates a condition called regulatory insolvency, which allows the financial structure of banks to be modified radically in little or no time and at little if any transaction cost.
Regulatory insolvency can be invoked by the bank regulator or some other duly constituted body, before the conditions for normal balance sheet insolvency (liabilities in excess of assets) or liquidity insolvency (inability to meet ones financial obligations) apply. When regulatory insolvency is invoked, an Administrator takes over the running of the bank. This Administrator has full powers to manage the assets, liabilities and employees of the bank. The ordinary shareholders lose their voting power and indeed any governance rights. Their other ownership and income rights can be diluted or extinguished. The claims of the other unsecured creditors can be subjected to a charge (or haircut) or be extinguished completely. The management and board can be sacked without golden parachutes. The assets and/or the remaining liabilities of the bank can be sold or transferred to other parties, or the bank can be managed by the Administrator as a going concern until further notice.
The Conservator appointed for Fannie and Freddie is an example of such an Administrator. The FDIC appoints such Administrators for federally insured deposit-taking banks in the US. There was no SRR for investment banks in the US, nor was there or is there one in the UK for any banks. This is about to change for the UK. In the US the issue has become moot with the disappearance of the independent investment bank.
In the case of deposit-taking banks, one class of creditors – individual depositors and holders of saving accounts – has been declared a sacred cow by most governments, partly to avoid the systemic externalities of bank runs, partly because individual monitoring of bank creditworthiness by holders of small bank accounts may be inefficient, but mainly for political reasons. I will take this government guarantee of individual deposit and saving accounts as a binding constraint. That leaves the other creditors, secured and unsecured.
I propose that the special resolution regime for banks about to be considered by the UK Parliament be designed in such a way that it achieves any desired increase in the capital ratios of a bank entering the SRR (or any desired reduction in the leverage ratio) to the maximum possible exent through a mandatory debt-to-equity conversion. Unsecured creditors would be first in line for a haircut (in reverse order of seniority). Secured creditors would retain their claim to the collateral securing their loan or bond, but would share with the unsecured creditors the haircut in their claim on the bank. New capital should only injected by the government if there either is not enough debt in the balance sheet or if the balance sheet of the bank is considered too small. While it is conventional for existing equity holders to be wiped out before debt is converted into equity, I don’t consider this essential. Dilution may provide adequate punishment and incentives for future equity investment decisions.
It may be the case that, when the authorities reflect on the lessons of the crisis, they may decide that some or all banking activities should be provided through state-owned banks and that some or all existing banks ought to be nationalised (have majority state ownership). It would not be right, from an efficiency or an equity perspective, if such nationalisations were to cause the orginal shareholders and creditors to get a better deal than they would have had without government assistance. The dilution or extinguishing of the existing shareholder stakes and the haircut on the existing creditors (and/or the mandatory debt-to-equity conversion) should therefore take place before new public sector capital goes in.
I hope that the new UK special resolution regime will put the tax payer before the existing shareholders and the existing creditors of the banks. The worst of all possible worlds would be the Irish (and now also Danish) approach where all creditors of the banks are guaranteed by the government (for a fee that undoubtedly will not cover the government’s opportunity cost) and the tax payer is left without any upside.