|Bank of England governor Mark Carney on Tuesday|
There’s an art to conducting stress tests if you’re a banking regulator that wants to take credit for bolstering financial system stability.
You want the tests to look really tough, but for most banks still to pass them. If too many large ones fail then that will simply spread investor panic and invite criticism of the regulator for not knocking banks into shape, but if the tests look too easy then both their pass marks and the examiner lose credibility.
You have to show some teeth.
So, the Bank of England (BoE) sailed through its own stress test of UK banks with flying colours.
On Tuesday, it declared that in the event of an even more severe global and UK recession than through the financial crisis – with a 4.7% fall in UK GDP, investor flight, a sterling and equity market collapse, interest rates up at 4%, with a 40% fall in commercial real-estate values and more than 9% unemployment – UK banks could absorb £50 billion in losses in the first two years and still keep lending to creditworthy borrowers.
That’s because they are three-times better capitalized than 10 years ago. In 2007, that kind of hit would have wiped them out.
It must be fun plucking these figures out of the air, especially when you can say that just two banks – Barclays and RBS – would see their common equity tier 1 (CET1) depleted just below the examiner’s pass mark, the so-called systemic reference point.
Barclays’ CET1 ratio would fall to 7.4% in such an economic and market calamity, so 50 basis points below its systemic reference point of 7.9%. RBS’s CET1 would fall to 7%, so just 40bp below its 7.4% systemic reference point. The test also includes hits to capital from further misconduct charges, which Barclays and RBS are each now striving to resolve.
But the test uses as its starting point end-2016 CET1 ratios. And because both banks have already boosted capital their narrow failure in these tests do not require them to raise additional capital now.
Time for a celebration, then?
Not too fast. Peter Richardson, analyst at Berenberg, says the BoE’s message that banks don’t need to raise equity as a result of the tests is not the full story. He notes: “Despite positive headline results, signs of cyclical risk continue to mount.”
He adds: “With realized losses on UK lending at an extreme cyclical low, we remain concerned that risks from the UK cycle remain underappreciated and underpriced.”
The canary in the coalmine is a bank that passed the test. Analysts at Credit Suisse point out: “Lloyds is the biggest shock in the UK stress tests since its adjusted drawdown of 4.9% is almost two times last year and implies a 14.7% end-state CET1 requirement, versus latest guidance of ‘upwards pressure’ on its 13% target.”
In the 2016 stress test, Lloyds was judged likely to suffer a wipe-out of 2.5% of CET1 at the worst point. This time round, its hit came in at 4.9%.
“This appears to be driven by interest-rate shock given the consumer (including from MBNA) losses added around 1 percentage point versus 2016,” judge the Credit Suisse analysts.
Mark Carney stresses that the BoE has a close eye on the rapid build-up of consumer credit and for that reason the stress test “incorporated a severe consumer credit impairment rate of 20% over three years across the banking system as a whole”.
Lloyds is not a global systemically important bank (G-SIB) and so does not have to conserve capital above a higher barrier to meet its systemic reference point. It passed the latest test with just 40bp of capital to spare.
The bank itself released a statement pointing out that: “This scenario is the most severe stress for the group since inception of the BoE stress tests.”
The BoE has not asked Lloyds to provide a revised capital plan, noting its CET1 was already at 14.1% in its third-quarter results.
However, the BoE has raised its countercyclical buffer requirement for all UK banks from 0.5% to 1% with binding affect from November 28, 2018.
Worse, Carney acknowledges that the stress test did not overlay the deep economic recession and asset price falls that might accompany a bust in present elevated global debt levels and collapse in financial markets onto a disorderly Brexit as well.
Carney says that losses from a disorderly Brexit probably wouldn’t be worse than those from the recession and market falls in the stress test. He says: “We view the overall risk environment – apart from Brexit – as standard.”
But what if both happen at the same time? To misquote Sherlock Holmes, a disorderly Brexit is the rather large and fierce dog that might have been talked about non-stop, but did not actually bark in the 2017 stress test.
Carney says that for banks an 18- to 24-month transition period is the minimum necessary to adjust to an orderly Brexit with a full agreement on a new trade relationship between the UK and EU. Perhaps to protect himself from charges of mutiny, Carney tries to paint disorderly Brexit as unlikely. It would seem to be a risk UK banks should certainly prepare for.
Carney concedes: “In the extreme event in which the UK faced a disorderly Brexit combined with a severe global recession and stressed misconduct costs, losses to the banking system would likely be more severe than in this year’s annual stress test.
“The FPC [Financial Policy Committee] will therefore reconsider the adequacy of a 1% UK countercyclical capital buffer rate during the first half of 2018, in light of the evolution of the overall risk environment.”
Worries over collateral values behind secured debt and particularly over unsecured consumer loans will continue to weigh on UK banks’ cost of equity, despite the latest test pass.
Carney sees some evidence of the kindness of strangers waning in regard to foreign investment in UK assets in the higher equity risk premium now attaching to purely UK focused businesses, when equity risk premia are falling in other developed equity markets.
Isabelle Jenkins, banking and capital markets leader at PwC, says: “With £67 billion of debt – and rising – held on UK credit cards, the forecasting exercise is a useful test run. However, this is just the tip of the iceberg.
“Banks have a real challenge looming in 2018 when they will need to include the ‘stressed’ IFRS 9 [international financial reporting standard] projections in their internal capital assessments, regulatory submissions and ad-hoc stress testing.”
There will be more scenarios to cater for, more complex modelling and more detailed data requirements.