“A journey of a thousand miles begins with a single step.”
Chinese proverb from Tao Te Ching.
The recent international agreement to improve the loss-absorbing capacity of globally active banks is an important move in the right direction. But financial regulators should go significantly further to make these banks and the global financial system resilient.
On November 10, the Financial Stability Board (FSB) proposed standards for global systemically important banks’ Total Loss Absorbing Capacity (TLAC). The idea behind TLAC is that, in addition to capital, large banks will be required to issue long-term debt liabilities that can be converted into equity to preserve financial stability without government support if the bank experiences losses that threaten its solvency.
The proposed level of TLAC is notable: 16%-20% of risk-weighted assets. That is roughly twice the Basel III capital requirement. In other words, subject to important caveats and after adjusting for double-counting, the combination of equity and TLAC debt increase the losses that a bank can absorb without recourse to public assistance by roughly 2½ times compared to the Basel III standard.
And, that’s not all. The FSB proposal also seeks to manage the thorny issue of cross-border resolution. Without an effective solution to this problem, national regulators will continue to have incentives to bail out big cross-border banks, perpetuating the too-big-to-fail problem.
We welcome these FSB efforts. In the remainder of this comment, we briefly outline the new TLAC standard and then discuss its pitfalls. Like many academics, we conclude that regulators should focus more on raising requirements for common equity, rather than equity-like debt instruments.
As in our recent primer on bank capital, let’s start with the simple accounting identity: Assets = Fixed Liabilities + Net Worth. (Net worth is the same as bank capital.) Now, in addition to capital (which we have drawn in green), banks will have to issue what we labeled “TLAC eligible debt” (in yellow).
Simple Bank Balance Sheet: Assets = Fixed Liabilities + TLAC Eligible Debt + Net Worth
The primary motivation for this new requirement is the widely held view that the baseline Basel III capital requirement is inadequate. That is, while Basel III is much tougher than its abysmal Basel II predecessor, it is not sufficient to make the financial system safe. Because global banks’ total assets are typically on the order of 2 to 3 times their risk-weighted assets, the 10% Basel III requirement for large global banks means that they are only required to finance between 3% and 5% of total assets with equity capital, rather than debt. [See, for example, the leverage ratios of selected global banks here.] Our view is that this is far too slim a buffer to insure against the need for a public bailout.
The principal reason for our doubts about TLAC is uncertainty about its effectiveness in absorbing losses to the financial system as a whole, not just those of a particular bank. The key question is who will be allowed to hold TLAC, which – like common equity – is composed of contingent liabilities issued by the bank. As one of us has argued here, the loss-absorbing liabilities of a systemic bank should be held solely by long-only investors. If, instead, leveraged entities were allowed to own TLAC debt, then the recapitalization of one insolvent bank could lead to a cascade of failures of leveraged intermediaries. Importantly, it is not sufficient merely to constrain TLAC ownership by other systemic banks (as Basel/FSB rules appear to do).
Put differently, credible insurance for a rainy day in the financial system can only be provided from outside the leveraged sector. Otherwise, it would be like an insurer who sells earthquake insurance and invests all the proceeds in real estate in the same region. This practice may pay off handsomely when the ground is stable, but it provides no protection when the earthquake hits. In the case of the financial system, once the rain arrives, TLAC becomes equity. Other than a government, who can stop a run on leveraged holders of TLAC in a banking crisis?
Why, then, did the FSB propose TLAC debt requirements rather than simply raising the requirement for common equity? The answer has to be that banks find debt less costly to issue, prompting them to argue that higher equity requirements will harm the economy. However, that perceived cost difference at least partly reflects government debt subsidies (especially through public guarantees for bank liabilities and through the tax deductibility of interest payments). As a result, the social cost of requiring banks to issue common equity is lower than the private cost, and potentially quite small.
How about the FSB’s measures to facilitate cross-border resolution? Large, globally active banks are incredibly complex. (In an earlier post, we noted that several have more than 1,000 legal subsidiaries around the world.) Resolving such an entity without use of public funds and while maintaining global financial stability would require an incredible level of international cooperation – cooperation to determine rapidly who has to pay for what losses and which counterparties are protected.
But authorities usually have a legal obligation to protect the counterparties in their own jurisdictions. These duties may conflict with efforts at international cooperation and can lead to what we would call an official sector bank run. That is, host-country authorities would have the incentive to seize all the resources that they can to protect their own depositors and investors. And there would be a destabilizing first-mover advantage: a regulator who waits may have nothing left to seize. The resulting grab race by the authorities hosting the subsidiaries of a large bank group could destabilize the entire group, with a knock-on to the global system as a whole.
Domestic bankruptcy codes are designed to prevent just such a grab race by a firm’s creditors in order to maximize the potential value of an insolvent firm as a going concern. That is what the various legal stays on creditor claims and the rules establishing priority of new claims aim to achieve. Unfortunately, we don’t see a clear path to a cross-border analog. Other than 100% capital backing of all claims, what level of locally pre-positioned capital would always be sufficient to satisfy local claimants?
It is this problem that led the FSB to propose that loss-absorbing capacity be pre-positioned at host-country subsidiaries in an amount equivalent to 75-90% of the TLAC requirement that would apply on a stand-alone basis (what is called “internal TLAC”). What this means is that subsidiaries need to be able to stand on their own to a substantial degree. And, importantly, they will need to have a liability structure that allows for recapitalization without reliance on decisions at the level of the banking group or the group’s home supervisor. This approach should diminish the incentives for a grab race among regulators, but it will not eliminate them.
In addition to complaining bitterly about capital and TLAC requirements in general, we fully expect that banks will protest the pre-positioning of TLAC geographically. They will say that the new rules will force them to forgo the benefits of global banking. The economies of scale and scope that come from the centralized management of capital and debt are being destroyed, they will say. There will be an element of truth to these claims. Every regulation has both costs and benefits. However, the benefits of lowering cross-border costs of banking would have to be amazingly large (far beyond what seems plausible) to compensate for the risks associated with another banking crisis triggered by, say, a grab race among national regulators.
In the past, global banking groups have been structured to take maximum advantage of government guarantees. One way or another, this has to stop. It would be far better to force banks to bear the social costs of the risks they create now, as part of an effort to promote financial stability, rather than wait until another crisis does it as a result of populist vengeance.
We encourage the FSB and national bank regulators to push common equity requirements substantially higher, and to make sure that systemic risks don’t migrate from better-capitalized banks to less resilient intermediaries. But, until they can do that, the new TLAC requirement is much better than nothing.