Metro Bank, once the darling of Britain’s challenger lenders, has lost half its stock market value over the past six weeks or so. There has been no nasty loan loss, no vast money laundering scandal and no funding squeeze to rival the nightmares visited on other troubled banks.
Metro’s problem is far geekier. But as the share price slump shows, being found by regulators to have miscalculated your risk weighted asset tally — and overreported your capital ratios as a result — is a serious gaffe.
Last week, it emerged that the Bank of England’s Prudential Regulation Authority was probing how the mistake had happened, while the Financial Conduct Authority is to examine whether the miscalculation resulted in a false market in Metro’s shares.
A bank’s tally of risk-weighted assets, or RWAs, is calculated by assigning its various loans — mortgages, commercial loans and so on — into different buckets. Each carries a “risk weighting” that mitigates the gross value of the relevant loans. Metro had £8bn of RWAs — relative to £22bn or so of total assets — until regulators caught its mistake, whereupon the RWA number jumped 11 per cent.
With banks’ regulatory capital requirements calculated principally on the basis that it is a certain percentage of RWAs, a higher RWA number clearly implies a higher capital requirement. At the end of 2017, Metro’s capital, as defined by its core equity tier one ratio, was 15.3 per cent. In the numbers published last week, the ratio had slipped to 13.1 per cent: hence a swift £350m equity raising.
Metro’s mistake was surprising because its risk weighting system is a simple and standard regulator-defined one. Given Metro’s small scale (its total assets are less than 3 per cent of Lloyds’ and less than 1 per cent of JPMorgan’s), this issue is also systemically tiny.
But there have long been suspicions about the accuracy of RWA numbers reported by bigger banks, too. Many prefer to use more advanced — or “internal ratings-based” — models to categorise the riskiness of their loans, based on historic losses and other factors. Models need to be signed off by regulators, but their application is not routinely monitored.
In studies over the past six years, the Basel Committee on Banking Supervision has highlighted alarming discrepancies between similar banks’ so-called RWA densities in similar lending areas.
It found, for example, that models used by some banks rated apparently equivalent mortgage books to be far less risky than those of rivals. In one study, a median risk weighting of 16.9 per cent disguised a low of barely 5 per cent and a high of more than 80 per cent. Even within the same region there was a spread of more than 20 percentage points.
Capital levels demanded by RWA calculations have been falling. Many banks have responded to the higher post-crisis regulatory capital demands by engaging in “capital optimisation” programmes — a euphemism for tweaking models so that loans look less risky and thus attract as little capital as possible.
At the same time the ultra-low interest rates of the past decade have cut loan losses in many markets. That may be a welcome trend but it has fed through, via models, into a suggestion that the business is fundamentally less risky — just as the prospect of a cyclical downturn in credit markets is intensifying.
Some regulators have tried to offset this procyclical problem by imposing “counter-cyclical buffers” as add-ons to capital. Regulators also stress that the capital rules merely impose minimum levels that should be exceeded if bank bosses are concerned about mounting risks.
But underlying RWA calculations — the baseline for capital requirements — are clearly deeply flawed. Inspired by the Metro case, experts have begun lobbying for change.
In January, the Institute of Chartered Accountants in England and Wales called (a tad self-servingly) for RWAs to be audited. Now the secretary-general of the Basel committee, which sets the global rules, has echoed the appeal. Others believe that local and regional regulators should be policing the system more effectively, too.
One thing is certain: flimsy or spurious risk calculations make a nonsense of the precise and intricate rules on bank capital that they support. And that undermines the safety of the whole banking system. Reform is overdue.
Letters in response to this column:
Get alerts on Banks when a new story is published