BY THE END of last year bank supervisors were supposed to have agreed on revisions to Basel 3, the international capital and liquidity standards devised after the financial crisis, which would then be all but complete. That did not happen, chiefly because some European authorities balked at the prospect of yet higher capital demands for the banks in their charge. Officials close to discussions at the Basel Committee on Banking Supervision, which draws up the standards, are still confident that agreement will eventually be reached. But further delay seems inevitable, if only because Donald Trump has yet to choose the American officials needed to complete the talks. And even if a deal is done, there are signs that the trend towards international regulation that gathered pace after the financial crisis may be going into reverse. “Post-crisis, there was a consensus for a global set of rules,” says Huw van Steenis of Schroders, a British fund-management firm. “That consensus has now broken.”

The crisis revealed that many banks had too little capital to absorb losses, were funded with too much debt and not enough equity, and were prone to illiquidity. Basel 2, the previous set of standards, completed in 2004, had required banks to maintain a minimum ratio of “tier-1” capital (equity plus qualifying debt) to risk-weighted assets (RWAs), with the weights determined either by banks’ own models or by a standardised approach. This had proved inadequate. Moreover, though Europe had adopted Basel 2 wholesale, American supervisors had applied it to just a dozen internationally active banks, fearing (with good cause) that Basel 2 would allow lenders to maintain dangerously low levels of equity. For the rest, they preferred to watch a simple leverage ratio, of equity to unweighted assets.

Basel 3, agreed on in 2010, was accepted on both sides of the Atlantic (and in many other countries). Its many and detailed provisions are being phased in gradually and have been reviewed and adjusted several times. The standards are far stricter than the previous ones in several ways, starting with tighter definitions of both tier-1 capital and RWAs. The equity component of tier-1 capital (common equity tier-1, or CET1) must amount to at least 4.5% of RWAs. Total tier-1 capital must be at least 6%. Stuart Graham of Autonomous, a research firm, reckons that the new rules in effect lopped around three percentage points from CET1 ratios. On top of this Basel 3 stipulated a “capital-conservation buffer” of 2.5% of RWAs and a “countercyclical buffer” of up to 2.5%, set by national regulators and intended to cool overheating. Both must be made up of equity (see chart).

Of G-SIBs and TLACs

The 30 institutions considered to be globally systemically important banks, or G-SIBs, incur an equity surcharge, ranging from 1% to 2.5% of RWAs. All these requirements will increase year by year until 2019. G-SIBs from rich countries must also have a “total loss-absorbing capacity”, or TLAC, comprising equity and convertible debt, of 16% of RWAs by 2019 and 18% by 2022. The handful of G-SIBs from emerging economies will have longer. The idea is that should they fail, they can be automatically recapitalised by bailing in investors, without troubling taxpayers.

Basel 3 also introduced a minimum leverage ratio, an idea European regulators resisted at first. Tier-1 capital must be at least 3% of assets, and more for G-SIBs. To deal with worries about liquidity, Basel 3 also requires banks to have enough high-quality liquid assets to withstand a month of outflows under stress and maintain sufficient “stable” funding, such as equity, long-term debt and retail deposits.

All this has meant that banks have become much better capitalised. Both CET1 and leverage ratios have climbed (see chart). Virtually all of the 100 big international banks and 110 others covered in the Basel committee’s latest monitoring report are keeping up with the Basel 3 standards.

A few tasks have not yet been tackled, such as whether the risk-weights of sovereign bonds should be raised. The latest kerfuffle stems from the committee’s attempts to bring more consistency to banks’ internal calculations of RWAs and to narrow the gap with the standardised approach, which typically yields higher RWAs and hence lower capital ratios. In 2013 the committee asked 32 lenders to work out CET1 ratios for the same hypothetical credit portfolio. The highest figure was four percentage points above the lowest.

That, the committee decided, was too much, so it proposed changes to close the gap, including minimum values for important parameters in internal models (such as the probability that certain types of loan will go bad). Most controversially, it suggested an overall “output floor”—a lower limit for the sum of RWAs—of between 60% and 90% of the number reached via the standardised method. If a bank’s internal model yielded a figure below the floor, the floor would be used instead.

The changes would have little if any effect on American banks, but some European (and Asian) lenders would see their RWAs and hence their minimum capital requirements go up. Unfair, said the Europeans: the changes, in effect, penalised them for keeping on their balance-sheets assets such as residential mortgages and loans to big companies, which American lenders are less likely to have. Not at all, retorted the Americans, who still mistrust risk-weighting: you should have put your houses in order sooner, as we did.

Officials say that pretty much all the disagreements have been sorted out except, crucially, for the output floor. But for now there is no one to talk to on the American side; and the isolationist mood among some Republicans may work against a deal. In a letter to Janet Yellen, the Fed’s chairman, in January, Patrick McHenry, vice-chairman of the House Financial Services Committee, said it was “unacceptable” that the Fed “continues negotiating international regulatory standards for financial institutions among global bureaucrats in foreign lands without transparency, accountability or the authority to do so”. In other words, stay out of Basel until you get new orders.

Would it matter if there were no deal? Politically, very much so: it would be one sign among many (in trade, security and elsewhere) that global co-operation can no longer be taken for granted. In practical terms, perhaps not so much. Apart from this point, Basel 3 is largely done; most banks use the standardised approach anyway. But failure may give some European banks an excuse to put off adding extra equity or reducing risks.

Perhaps delay may be no bad thing: given more time, European economies and banks should become stronger, making agreement easier. And once a new American team is in place, a way forward may emerge. Mr Trump’s nationalist streak could even help. Given that the proposals scarcely affect American banks but may burden foreign competitors, an agreement could widen the lead the Americans already enjoy over their rivals. On the European side, some fierce critics have softened their tone. Recently Andreas Dombret, a Bundesbank official who had opposed the output floor outright, moderated his stance to “any output floor should not be too high.”

Stronger non-American banks are also keen for a deal. Bankers are annoyed about the delay over completing Basel 3, because delay means uncertainty. “Do you ever want to stop?” sighs one banker, reeling off a list of changes to regulation over the past few years. Even if signed off tomorrow, the amendments would not take full effect until the mid-2020s.

Beware a new transatlantic divide

A change in American supervision along the lines proposed by Mr Hensarling or Mr Hoenig would pull America back towards a system of supervision based primarily on minimum leverage and away from the CET1 ratio and the Basel liquidity constraints, as well as dismantling much of America’s domestic regulatory apparatus. If Mr Hensarling’s bill became law and the big banks rejected the option of a 10% leverage ratio in exchange for regulatory relief, something resembling the old transatlantic divide in supervision could emerge: Basel standards for the big American banks and all European ones; domestic, leverage-based supervision for smaller American lenders.

Smaller banks tend not to operate outside their home countries. What worries international banks more is that domestic supervisors are tightening requirements for foreign lenders on their turf. If a parent bank abroad gets into trouble, they do not want capital to be siphoned away from a local subsidiary to prop up the parent; and if the subsidiary stumbles, they want to be sure it has enough equity of its own so it can be allowed to fail without imposing on local taxpayers. Since last July foreign banks in America have had to create separately capitalised “intermediate holding companies” for their local subsidiaries. In November Valdis Dombrovskis, the European Union’s commissioner for financial services, proposed something similar for the EU. The details have not yet been worked out, but he wants G-SIBs in the EU to maintain enough loss-absorbing capacity locally to permit orderly bankruptcy.

Even cautious banks worry that supervisors may trap capital that could be better deployed elsewhere

Britain’s departure from the EU, scheduled for 2019, may give the ratchet another turn. Banks in the EU have “passports” that allow them to serve the entire union from any member state, without local branches or subsidiaries. Most have chosen London as their base. After Brexit, London-based banks will lose their passports. No one yet knows which operations and how many jobs will move away, but regulators in France and Germany, say, may insist that hubs in Paris and Frankfurt are separately capitalised.

Some banks, such as HSBC and Santander, have long capitalised subsidiaries in different places separately. And it is not hard to understand why national supervisors might insist that others do the same: a little sand in the wheels of globalised financial institutions may be a good thing. Nonetheless, even cautious banks worry that supervisors may trap capital that could be better deployed elsewhere. The effect would be to raise lenders’ minimum capital requirements even further.

“If you have a very fragmented approach to banking regulation,” says David Strachan, a British ex-supervisor who is now working for Deloitte, a consulting firm, “the cost this imposes on banks trying to operate globally is material.” If some countries lose confidence in supervisors elsewhere in the world, ring-fencing may increase. If others follow the lead of America and Europe in asking for intermediate holding companies, “then it becomes quite damaging.”

Even cautious banks worry that supervisors may trap capital that could be better deployed elsewhere