Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
Some unintended consequences can take decades for their full force to be felt. So it was for the global package of bank capital reforms introduced in 2004.
One ripple has been seen in the recent tribulations of midsized UK lender Metro Bank. The bank was founded by Vernon Hill, an entrepreneur known for his retail showmanship to woo depositors, sometimes with the aid of a dog. He is also the founder of Commerce Bank, acquired by Toronto-Dominion Bank in 2007 for $8.5bn, and the tiny Philadelphia-based Republic First Bancorp, which has suffered its own financial difficulties.
Metro applied five years ago to move to an advanced capital regime introduced under the so-called Basel II package of global banking reforms. The framework allowed sophisticated banks to use their own models to assess the riskiness of loans, rather than employing the standardised risk scores from the banking rule book. Those assessments of risk then fed into banks’ capital requirements.
The idea was that banks would manage risk better if they had to consider and capitalise for probable losses, rather than using a blunt instrument that encouraged riskier lending by applying the same treatment to a high risk loan as to a safe one.
By 2018, when Metro began its application to use internal models, the approach had already fallen out of favour with global regulators, after the financial crisis showed that sophisticated models did little to insulate banks from unforeseen and ruinous losses.
But Metro was heavily incentivised to persist with its application all the same. Even as the global regulatory world turned against internal models, Metro’s biggest competitors can still use them to set aside an average of less than half the capital for an identical home loan as the so-called challenger bank would have to.
But the wheels came off in mid-September when Metro, having described how moving to internal models would turbo charge its growth, admitted its application was delayed indefinitely and might never be approved. The bank’s shares shed more than 50 per cent in three weeks, forcing it into a £325mn capital raise and £600mn of debt refinancing last weekend.
Internal models were never supposed to distort competition in the way they have in the UK mortgage market. But Metro’s experience suggests this isn’t going away any time soon.
One Metro insider says the bank was frustrated in its efforts by long waits for feedback from the UK’s Prudential Regulation Authority. He adds that even if its models are eventually approved, there will be so many added buffers that the benefit will be far less than originally envisaged. In other words, even in the best-case scenario, the gap with the big UK banks will remain. Metro publicly said on Monday that it might pause the application.
Metro arguably faced uniquely high hurdles because of a 2019 scandal where it was found to have applied the wrong risk scores to loans, an error that did little to put the bank in the PRA’s good graces. But other banks have had similar frustrations over their models. Another challenger recently privately abandoned its multiyear bid to move to internal models because the cost outweighed the likely benefit. Senior executives at other banks say they have spent more time and money on approvals to move to internal modes than they spent on their banking licences.
UK regulators’ reluctance to swiftly approve challengers’ bids stems partly from concerns that smaller and newer banks cannot prove they can predict losses on loans better than the standardised models developed by global experts.
And some regulators’ gripe more generally is that models are being used primarily as a way to cut banks’ capital requirements, referring to various pieces of research, including this 2014 paper, pointed to gaming by banks.
The next package of global capital reforms, dubbed Basel 3.1 by the British and the Basel endgame by the Americans, reduces the potential for this by setting “floors” for how low capital charges can go in business lines and banks overall, no matter what the model says.
The US has gone one further, proposing to ban even its biggest banks from using internal models to calculate the risk of their loans from mid-2025. Some in regulatory circles are quietly asking if the UK should follow suit if it truly wants to promote competition in retail banking — the PRA’s longstanding secondary mandate.
After all, there are two ways to solve the competitiveness issue that has driven the challengers into expensive and usually fruitless processes — lower the bar for all, or raise it. Following a spate of recent banking crises, appetite for the former is not high.
