On 10 November 2014 the Financial Stability Board issued a consultative document on the “Adequacy of loss-absorbing capacity of globally systemic important banks in resolution”. The press release stated:
“The Financial Stability Board (FSB) has today issued for public consultation policy proposals consisting of a set of principles and a detailed term sheet on the adequacy of loss-absorbing and recapitalisation capacity of global systemically important banks (G-SIBs).
The proposals respond to the call by G20 Leaders at the 2013 St. Petersburg Summit to develop proposals by end-2014. They were developed by the FSB in consultation with the Basel Committee on Banking Supervision (BCBS) and will, once finalised, form a new minimum standard for “total loss-absorbing capacity” (TLAC).”
The purpose of the exercise is to ensure that failing G-SIBs can be resolved without either closing down and liquidating essential banking services, or calling on taxpayers for bail-out support. It aims to do this primarily by dividing G-SIBs into a holding company (hold-co) and an operating subsidiary (or subsidiaries) (op-co). The hold-co is then required to have a sufficiency of loss-absorbing capital (LAC), so that in the event of a major loss anywhere within the group, the external LAC of the hold-co can be written down and the hold-co liquidated (if necessary), with the capital thus released transferred, e.g. to the main operating subsidiary (via internal LAC), so as to recapitalize the op-co. The idea is that the op-co can soldier on after recapitalization without taxpayer support, providing necessary banking functions. Instead, the recapitalization is to be obtained by bailing-in the equity holders and bail-inable creditors providing eligible external T-LAC in the hold-co.
In this Consultative Paper on T-LAC there is a clear description of this process, leading up to the resolution of a failing bank under these new procedures. And there is no doubt that it is a clever and subtle idea. But there is no account of what might happen after this recapitalisation (of the op-co) has been put in place. In a game with many rounds, such as chess, the expert players are those that are trained to think many steps ahead. Within the bail-in process, the (main) operating subsidiary (the op-co) is meant to continue, so this is supposed to be a multi-stage exercise. Yet nothing is said in this paper about subsequent stages, nor the problems that might arise therein. I raise some queries about what might happen after the initial resolution is triggered.
What about liquidity?
The main focus of the Consultative Paper is on the provision of sufficient capital by a bail-in of creditors to support the continuing workings of the op-co. But there is not a word about ensuring that it will have sufficient liquidity as well. Sufficiency of liquidity is by no means assured.
The resolution process is bound to be newsworthy. The overall strength of the banking group will have become undoubtedly impaired when the hold-co is liquidated or drastically written down. The name and reputation of the bank will have been brought into question. The likelihood is that the initial reaction of both informed and uninformed investors (as with Northern Rock) will be to flee. In advance, no one can guarantee that this will not happen. Without protection from that eventuality in the guise of an associated commitment by the relevant Central Bank to provide sufficient lender of last resort (LOLR) support, the whole exercise stands at risk of failing disastrously at the first hurdle.
Since the op-co is solvent by design, there should be no ideological barrier to such LOLR support. Yet there are suggestions that Dodd-Frank (or other constraints) may restrict LOLR support in the USA. If true (and I hope not), this could put the whole approach at risk. If not, then what is needed for the final paper is the addition of a sentence such as, “With the op-co being thus clearly solvent, its Central Bank will certainly stand ready to meet any resultant temporary liquidity requirement”. In any case the FSB cannot burke the issue; liquidity provision must be openly discussed.
As noted earlier, with the hold-co liquidated or sharply written down, the overall valuation and strength of the banking group will have been much impaired. The op-co will have sufficient capital of its own, but it will no longer have the hold-co as a buffer above it. Thus the op-co will be significantly weaker than all the competitive banks around it. The op-co by itself will no longer have the T-LAC support deemed necessary for everyone else. What happens then?
Presumably the op-co cannot then pay out dividends or do buy-backs until it can fully reconstitute a hold-co that meets standard requirements? Is this the intention? If so, should the paper not say so? What additional constraints on the subsequent working of the op-co are envisaged? It will hardly be possible for the op-co (and/or a residual or new hold-co) to issue new equity or debt unless the constraints and terms under which the resolved entity will subsequently operate are fully and transparently spelt out. Yet the present draft consultative paper takes us in considerable detail up to the initial resolution, but is totally silent on life thereafter. What does the afterlife look like?
Once a failing bank is put into resolution, a forensic auditor will no doubt be sent in to estimate the necessary size of the creditor haircuts required to recapitalise the op-co sufficiently. There can hardly be two bites at this, if only because transactions in such assets will begin again once the initial write-downs have been established. For this and other reasons – such as uncertainty about the effect of the resolution process on market values – my own expectation is that the forensic auditor will aim to err on the side of austere caution in such valuations; i.e. to propose larger haircuts than subsequently turn out to have been necessary.
It is of course possible that this is not the case, and that the auditor underestimates the necessary haircuts. This would be particularly likely if the initial resolution should turn out to be the start of a systemic and major financial crisis. What then if op-co should fail to repair profitability, and need further recapitalisation? In the Eurozone this possibility is met within the Banking Union by the SRM and ESM. Should there be a similar back-stop world-wide?
Let me revert to the more probable outcome – that the proposed haircuts go well beyond what eventually turns out to be necessary. Will this not open up the authorities to legal suit under the ‘no creditor worse off’ principle? This possibility can be mitigated by giving the bailed-in creditors warrants against such upside recovery. But the more that the upside recovery option would go to reimburse prior creditors, the less the attraction of the equity in the op-co (or the reconstituted hold-co) to new buyers. How is, or should be, this balance set? What will be the form and detail of such warrants?
Pro-cyclicality and contagion?
As argued earlier, the valuations of the forensic auditor will tend to err on the austere side, and the haircuts will be larger than previously expected on the basis of the last, pre-resolution accounts. The market will be shocked. The bail-inable debt will, almost by definition, have previously been bought by optimists (and the ill-informed), and their optimism will suddenly appear misguided. There must be a good chance that, after the first newsworthy resolution, the market for such bail-inable debt will completely dry up for some period of time before re-opening at a much higher (and to bank CEOs unattractive) yield. What then happens as bail-inable debt rolls over? Will there not be an incentive on all other banks to deleverage sharply to avoid the penalties imposed on banks falling short of T-LAC? All regulation tends to be pro-cyclical. Is this not another example where the T-LAC approach would reinforce the cycle and makes the financial sector even more prone to systemic crisis?
There is some (but not much) mitigation of this danger from the one-year criterion for (eligible external) T-LAC (Box 11, p. 16). This states that
“Eligible external TLAC must have a minimum remaining maturity of at least one year.
In addition, the appropriate authority should ensure that the maturity profile of a G-SIB’s TLAC liabilities is adequate to ensure that its TLAC position can be maintained should the G-SIB’s access to capital markets be temporarily impaired.”
This does mean that if another bank is forced into resolution during the period during which refinancing roll-overs is impossible – assumed to be less than one year – it will still have enough T-LAC to recapitalise the op-co. But just as much bail-inable debt will pass through the one-year gateway as will need roll-over refinancing, so measured T-LAC will still be falling.
How can one protect against the contagion and procyclicality that this regulatory procedure – alongside most other regulation – is likely to enhance? Why not give Central Banks the power to vary the maturity time limit for eligible external T-LAC, so that in a systemic crisis (preferably with such a crisis being called on the basis of a pre-determined metric rather than just discretion), all banks can treat all bail-inable debt right up to final maturity against the T-LAC requirement? Could or should the percentage amount of T-LAC be made state-contingent? Has any thought been given to the question of how to counteract the pro-cyclical potentiality of T-LAC, and to the possibility that the exercise might worsen rather than improve the systemic stability of the financial system?
I hope that such thought has been applied, but there is no sign of this in the consultative document. It takes us up to the occasion of the initial resolution but is completely silent on what happens afterwards. While getting the first step right is important, if one cannot foresee the likely course of the subsequent stages of the game, one is likely to lose the subsequent match, and the financial system and the economy too.