Since 2013, when the G20 called on the Financial Stability Board to devise a proposal to ensure the world’s biggest banks could be recapitalised in the event of a crisis, the FSB has set to the gruelling, and, in the words of its chair Mark Carney, “often tedious” work of creating the Total Loss Absorbing Capacity (TLAC) framework.
TLAC will likely be adopted by the G20 at this year’s summit in Antalya, Turkey, which begins on Sunday. But while the FSB can be credited for hammering out a piece of global regulation, the ambiguities inherent in TLAC may render the rules all but impracticable, right at the point when it will need to work.
That rules are so complex and take effect across jurisdictions with different laws and politics further confuses implementation and practice; so much so that no banker GlobalCapital has spoken to yet knows how TLAC will work exactly, although the final proposal is broadly the same as the earlier consultation draft.
The final proposal, released on Monday, did include some needed, though mostly minor, clarifications and changes. Among them, the FSB pointed to concerns that the concept of a Pillar II TLAC requirement was too opaque. The FSB's solution was to no longer refer to it as Pillar II. Rather, it is now called “additional firm-specific requirements” for external TLAC.
Whew! So that’s settled.
Nice kit, but how do you plug it in?
The most important clarifications were the setting of the minimum requirement for TLAC — at a lower end of the range than many expected last year, when the FSB issued its consultation paper — and a timeline for compliance.
But when it comes down to it, the final proposal still lacks the most important details: when a bank is in crisis, TLAC sets out no firm procedures for who gets final say on how a bank is resolved.
Rather, the FSB puts a lot of weight on “agreements” between local and home authorities on matters such as these. It is by agreement between those two parties, for example, that an excess of overall TLAC due to a bank taking a multiple point of entry (MPE) approach, will be eliminated or minimised at either the parent or material subgroup level. Let the bickering commence.
To assume that an international bank’s host authority and the local authority of one of its global subsidiaries will be inclined to work together when it comes to capital planning or resolution processes is to assume that both authorities have the same things at stake. That is an error.
For example, then-US Treasury secretary, Henry Paulson blamed the UK regulator for blocking Barclays’ purchase of Lehman Brothers during the collapse; an act he claimed “screwed” the US. It is a claim the regulator later denied, but added at the time that it would probably not have signed off on it without shareholder approval, had Barclays come to them with a proposal.
And when US regulators levied a $9bn fine on BNP Paribas for violating US sanctions, it nearly sparked off a diplomatic crisis between France and the US, which some feared at the time could complicate talks of a transatlantic trade and investment agreement between the European Union and the US.
Eyebrows were also raised on this side of the Atlantic when Deutsche Bank and Santander failed the US Federal Reserve’s stress tests earlier this year, while all US-based banks passed.
Authorities are bound to approach resolutions in a way that impacts their jurisdictions the least, and which result in the least political blowback. And that could well be at the expense of another jurisdiction’s welfare.
The fact is, right at the point when a disagreement between jurisdictions is probably most likely, that is where TLAC fails to provide any detail.
But what about the reality?
For an example, look at the FSB’s thinking on banks that plan to come into TLAC compliance via a single point of entry (SPE) approach. Under that setup the parent bank holds the minimum TLAC requirement of 16%-18% of risk weighted assets, and each of its “material sub-groups”, defined by a number of measures, must hold 75%-90% of the internal TLAC — TLAC not sold to third party investors and readily available in the jurisdiction of operation — that they would have to hold if they were resolution entities themselves.
The FSB stipulated in its press conference that the host authority for such material subgroups is not limited in its ability to impose additional requirements, though it expects that requirement to be in the 75%-90% range.
We will see how that works out. Some jurisdictions could easily decide to impose as much as will be tolerated by the parent bank without causing it to pull the plug on a given foreign operation. The reason is simple: why let a group in France, for example, have any power over the resolution process of a significant banking group operating in your jurisdiction? Of course, local authorities are going to want the final say in what happens to the banks that represent a significant portion of their respective financial sectors.
The FSB does acknowledge that “technical issues and questions will arise in the course of implementation” and banks’ crisis management groups (CMG) will review such problems. The FSB will also help coordinate discussions among CMGs, “relevant committees and working groups” and the FSB itself, and the board will have a technical review of implementation by the end of 2019.
It’s also true that banks are just better capitalised than they were before the crisis, making major banking failures less likely in the first place.
But for all its complexities and nuances, TLAC remains a plan that relies on what appears to be a fundamental error: an assumption of the ability of international authorities to work together efficiently and in a disinterested fashion when the pressure rises.
What the FSB needs to do is come up with examples of how it thinks TLAC would work in practice — something it has not done yet, to bankers’ frustration, presumably because it doesn’t want to be responsible for directly influencing banks’ compliance strategies. But it could help finish the conversation, because who really knows? By the time a review comes around in 2019, it may be too late.