Banks Might Be Undercapitalized, Depending How You Count

Author:
Publish date:

If you think that capital regulation is a good way to make banks safer then a pretty good form of capital regulation would go something like:

Regulator: You should have a lot of capital relative to your assets.
Banker: Okay, got it. Now how are you measuring assets?
Regulator: Oh, you know, different ways. Whatever makes sense.
Banker: Umm. Well ... what counts as capital?
Regulator: I dunno. Capital-y things. Maybe stressed capital-y things.
Banker: I see. And what counts as a lot?
Regulator: Some number. Definitely a percentage. I guess a pretty high one?

Does that sound dumb? The thing is that most games have explicit clear rules, so some people just assume that any set of explicit clear rules constitutes a game, and then try to game it. Those people often work at banks. Kling's Law of Bank Capital Regulation holds "that the capital measure used by regulators will, over time, come to be outperformed by a measure that the regulators are not using," which suggests that the best capital measure for regulators to use is "all of them."0 Actually using an infinity of measures is impossible but can be approximated through confusion.

I don't really think that international bank capital regulators got together this week and were like "let's release a bunch of reports, and some unofficial rumors, stressing different tweaks to capital measures to confuse everyone into conservative un-game-y responses," but wouldn't it be fun if that's what happened? Yesterday the Swiss National Bank released its 2013 financial stability report praising CS and UBS for their work on Basel risk-weighted capital but recommending that they build more capital to improve their simple leverage ratios. Also yesterday, the Prudential Regulatory Authority of the Bank of England released its capital shortfall exercise, finding an aggregate capital shortfall among Barclays, Co-op, Lloyd's, Nationwide and RBS of £27.1bn, due in part to stressing their capital positions "to reflect expected future losses and an assessment of future cost of conduct redress,"1 and in part to re-weighting their risk-weighted assets to weightings that the BoE found more congenial.

And today, U.S. capital regulators ... talked to Bloomberg? Or something? Anyway:

U.S. regulators are considering doubling a minimum capital requirement for the largest banks, which could force some of them to halt dividend payments.

The standard would increase the amount of capital the lenders must hold to 6 percent of total assets, regardless of their risk, according to four people with knowledge of the talks. That’s twice the level set by global banking supervisors.

U.S. regulators last year proposed implementing the 3 percent international requirement for what’s known as the simple leverage ratio. Now the Federal Reserve and Federal Deposit Insurance Corp., under pressure from lawmakers, are weighing increasing that figure for some of the biggest banks, according to the people, who asked not to be identified because the discussions are private. ... U.S. banks have had to comply with a simple leverage requirement of 4 percent for two decades.

That 6% simple leverage ratio would be in addition to Basel risk-based capital requirements. Bloomberg notes there are alternatives:

FDIC Vice Chairman Thomas Hoenig has called for scrapping risk-based rules entirely in favor of a simple 10 percent leverage ratio, calculated to include even more off-balance-sheet assets than allowed under Basel and define capital more narrowly. ... A bipartisan Senate bill introduced in April by David Vitter, a Louisiana Republican, and Ohio Democrat Sherrod Brown would set the leverage ratio at 15 percent.2

So: will the U.S. settle on a simple leverage ratio of 3%, 4%, 6%, 10%, or 15%? Will the denominator be GAAP assets, IFRS-ish assets, Basel assets, or some sort of hybrid? Will the numerator be tangible common equity, or Basel tier 1 capital, or something involving long-term unsecured debt too? Will the simple leverage ratio be used in addition to a (higher) risk-based capital ratio, or on its own? Maybe!

Bloomberg notes a KBW research report on the potential impact of the 6% simple leverage rule; here's how KBW calculates current U.S. bank capital:

Which one matters? Who knows!3 Better work on improving all of them?

Maybe? One approach to perhaps-pending regulation is to ignore it until it happens; eventually there'll be clarity and bankers will be able to get back to the hard work of optimizing their assets and capital positions through gamesmanship. But for now this haze of potential requirements at least suggests a more rigorous, and harder to game, capital regime than whatever will ultimately be put in place. And it's at least a little foolish for any bank to ignore these potential rules entirely; now is probably not the time, for instance, to get big in a business that is total-asset-intensive but RWA-light.

It's a cliché that bankers will always manipulate rules to their advantage because they're smarter, better paid, and more motivated to game the rules than the regulators are to stop them. This recognizes that regulators and banks each have certain advantages in their, let's say, contest with each other, and that one of the banks' main advantages is piles of money. It's possible that one of the main advantages of the regulators - numerous and divergent and competitive and populist and beholden to special interests and so forth as they are - is disorganization. Right now they're making good use of it.

U.S. Weighs Doubling Leverage Standard for Biggest Banks [Bloomberg]
SNB lauds UBS, Credit Suisse capital steps, warns on leverage [Reuters]
Financial Stability Report 2013 [SNB]
Prudential Regulation Authority (PRA) completes capital shortfall exercise with major UK banks and building societies [BoE]
BOE: U.K. Banks Need More Capital [WSJ]

0.To be perhaps more technical, it suggests that the best capital measure for regulators to use is "the one they're not using," which, again, is best accomplished through misdirection.

1.I.e. just guessing what Libor, etc. will cost them. The guessing, though, comes not from the PRA - the capital regulator - but from the Financial Conduct Authority, the regulator that'll actually send them (part of) the bill, so it's an informed guess: "for conduct costs, the FSA (now FCA) provided analysis on the potential future costs that firms may incur."

2.And also basically scrap risk-based rules though in a less clear way.

3.Credit markets, maybe? KBW note that "The market appears to accept the non-Basel III Tier 1 Common ratios as Wells has the narrowest CDS spreads ..."

Related