HOWEVER long a storm lasts, clearing up takes longer. On December 7th Mario Draghi, president of the European Central Bank and head of the committee that approves global bank-capital standards, declared that revisions to Basel 3, the version drawn up after the financial crisis of 2007-08, were complete. The overhaul of the previous rules, which were blown away in the tempest, began eight years ago. The revised set, informally called Basel 4, will not take full effect until 2027.
That lengthy period of adjustment is one way in which Basel 4 is less demanding than banks, notably in Europe, had feared. Several other tweaks mean that the standards banks must eventually meet will be less exacting than first proposed. Already forced to bolster their balance-sheets with lots more equity—of which the crisis showed them to be woefully short—banks may deny that they have got off lightly. But they probably have.
Basel 4 was supposed to be settled a year ago. It wasn’t, because of a row over proposed limits on banks’ use of internal models to calculate their risk-weighted assets (RWAs), which may also be worked out from a “standardised” approach. The ratio of common equity to RWAs, known as the CET1 ratio, is a key indicator of banks’ capital strength. Low risk-weights mean lower RWAs, which in turn mean banks need less equity for a given CET1 ratio. To limit the potential advantage from using models, Basel’s standard-setters proposed setting a floor for the minimum ratio of RWAs calculated from models to the answer from the standardised approach.
Because European banks tend to hold a lot of assets with low risk-weights, such as mortgages and corporate loans, they wanted a low floor or none. Because American banks have fewer such loans and their domestic rules already contain a floor, they wanted a higher one, arguing that the Europeans’ habits were unfair and unsafe. Basel’s drafters first suggested a floor of between 60% and 90% of the amount under the standardised model. After much haggling—with the French the most stubborn holdouts—the floor was set at 72.5%.
The fine details make life a little easier for banks. Residential mortgages, “specialised” lending, such as infrastructure loans, and some corporate loans will incur lower risk-weights in the standardised approach than first suggested. Basel 4 does forbid using fancier models for loans to large companies. But banks had worried that their use for smaller firms would be curbed too.
Basel 4 also revises the capital requirement to cover operational risk—such as fines for bad behaviour or the cost of computer hacks. Banks must use a standardised method, outlawing an alternative that gave them some discretion. This too looks less stringent than it might have been. In principle, the operational-risk requirement may be multiplied to reflect past transgressions (because banks with a bad record may sin again). National supervisors, however, may choose not to bother. Standard-setters also postponed by two years, until 2022, the implementation of stricter coverage of trading losses. Several countries had already announced delays.
Analysts at UBS estimate that Basel 4 leaves European banks with a capital shortfall of €40bn ($47bn), but this dwindles to €5bn if supervisors make offsetting changes—Nordic countries may unwind increases in requirements made ahead of Basel 4. The impact will vary hugely across banks. And banks point out that Basel 4 is not their only burden. On December 12th UniCredit, Italy’s biggest bank, estimated that Basel 4, new guidelines from the European Banking Authority, a supervisor, and other rule changes would lop 4.7 percentage points, or one-third, from its CET1 ratio between 2017 and 2027. (It expects to earn more than enough to offset this.)
Banks also say they have plenty of equity, should disaster strike again. They may be right. But if interest rates stay ultra-low, central banks will have little scope to cut them after another crash. The financial system will therefore be more reliant on banks’ shock-absorbers. So yes, cheer the completion of Basel 4, but hope that those are thick enough—or not put to the test.