|
|
The committee had proposed restricting the use of banks’ internal models for calculating risk-weighted assets—which in turn help determine how much capital banks must have at hand. Models varied too much, it said; low risk-weights were flattering some banks’ ratios. But European bankers and officials had complained for months that the proposals would penalise banks that have lots of (low-risk) corporate loans or mortgages (eg, in Germany or Sweden). They sniffed an American plot: American banks, holding fewer such assets, would be untouched.
Mr Ingves gave few details, but said that the new set-up would “largely retain” internal models, though with minimum values for important parameters (such as the probability of default). A “standardised” approach will replace alternatives based on banks’ models for estimating operational risk (big fines, say, or cyber-security breaches).
Not surprisingly, the thorniest topic remains open. The committee had proposed an “output floor”, a lower bound for the risk-weighted sum of a bank’s assets, of 60-90% of the answer yielded by a standardised method. Mr Ingves said that he “expected” a floor to be in the final deal. But there is still work to do, and what happens next isn’t yet clear. (The committee had set itself a deadline of the end of the year.) Central-bank governors and chief supervisors must endorse the new rules, once they are agreed on.