At the start of February, Europe’s largest banks began to report their full-year 2020 results. With the half-way point of the first quarter of 2021 already looming, investors wanted to hear more about the outlook, particularly for credit costs.
Here is the good news: the executive summary from bank CEOs is that the worst is past. And the tantalizing possibility is that write-backs from conservative provisions taken in the first half of 2020 could nudge up reported profits in 2021.
José Antonio Alvarez, chief executive of Santander, says: “We expect this year the cost of risk to go down from the 128 basis points we had [in 2020] and start the convergence to 1%.”
That would be a return to pre-pandemic levels.
At BNP Paribas, cost of risk in 2020 rose to €5.7 billion, an increase of €2.5 billion from 2019. Compared with Santander, that €5.7 billion is a more modest 66bp of the loan book, with 16bp of that counting against still performing stage 1 or 2 loans.
The NPL peak may have been flattened and the worst may have passed. But don’t take anybody’s word for it
The French bank guides that cost of risk has passed the worst point.
“We do not anticipate any further deterioration of the economic scenario,” chief executive Jean-Laurent Bonnafé told analysts on the bank’s earnings call. “A progressive normalization of the cost of risk is anticipated in 2021, after the peak in 2020, and we expect it to decrease to a level close to the cycle average.”
For BNP Paribas, that was 39bp during the four years from 2016 to 2019.
Here is the bad news: you can’t rely on any of this. It all depends on the vaccines outpacing the virus, and the reopening and recovery now being forecast by the IMF and anyone else with a degree in economics, proceeding just as expected in the second half of this year.
Experience suggests it might be wiser to expect the unexpected.
Credit rating agencies
Step outside individual banks’ modelling of their own portfolios for a moment, and credit rating agencies are more worried about the system in aggregate.
Standard & Poor’s has reduced its overall estimate for pandemic-related credit losses across the global banking industry to $1.8 billion in 2020 and 2021. This is below the €2.3 billion it forecast in July 2020. But while it now sees lower losses, it thinks they may persist for longer.
Overall, it sees $123 billion of credit losses this year among banks in western Europe, compared with $135 billion in 2020. That would be a decline of just 9%. S&P also expects $88 billion of losses in 2022.
“Our expectation for losses has shifted more into 2021,” says Osman Sattar, director, European financial services ratings at S&P Global. “2019 marked the end of a multi-year, very benign period for credit losses and even into 2022 it will remain elevated.”
S&P expects that, in aggregate, big banks’ pre-provision earnings will be able to absorb these credit losses, with some headroom.
However, Sattar warns: “Still, further upticks could continue to weigh on banks’ ratings, and inevitably, some banks will incur net operating losses. Moreover, credit losses over 2021 and 2022 could still be volatile and higher than our base case.”
This is not the kind of talk investors want to hear. They want to know when the ratings agencies will shift their outlooks from negative to positive and help bank credit default swap (CDS) spreads to grow into their current rather optimistic levels at close to historic tights.
But S&P is sceptical.
“Visibility is quite poor,” says Giles Edwards, senior director, financial services ratings. “We are not likely to move ratings before the end of the year when the reckoning on credit losses becomes truly apparent.”
He explains: “Fiscal support is still in evidence as is forbearance, though that has come off a little bit. Until you get a normalization in those aspects, perhaps in the middle of the year, it will take a while for weaker loans to become non-performing.
“It is at the end of the third quarter and start of the fourth when you will see a significant pick-up in non-performing loans (NPLs). And we expect them to continue to rise through next year.”
Andrea Enria, chair of the supervisory board of the European Central Bank (ECB), is another sceptic about banks’ credit models. He should be one. That is his job.
Enria said, when introducing the most recent supervisory review and evaluation process: “What we notice is that they have flattened in their models the responsiveness of their probabilities of default to decreases in GDP.”
Of course, we have never seen such a swift and sizeable monetary and fiscal policy response.
Enria admits: “So, there could be justifications for this adjustment, but it is so massive that we think we need to look more into that.
“The key point on which we have been zooming in recently is the ability of banks to produce reliable projections on asset quality and therefore on capital. And this is where the focus of our teams will be particularly centred in the coming months.”
Moratoria concern
Enria is suspicious that moratoriums and forbearance make bank forecasts on bad debts, which are now supposed to be more forward-looking, less reliable.
He says: “Moratoria is a peculiar area of concern, because of course, the delays in payments are one of the main indicators the banks use to classify their customers, and moratoria generate a sort of blindness, a lack of visibility in terms of payments.”
Supervisors have told banks that they should have developed additional indicators and looked through the moratoria to be able to reclassify customers if they are unlikely to pay back their loan.
“We have not seen a lot of that happening,” Enria reports. “We have not seen sufficient loans being reclassified as forborne, and we were concerned because the indicators showed that this was not reflecting the reality on the ground.”
European supervisors have been asking banks for more information on this.
The NPL peak may have been flattened and the worst may have passed. But don’t take anybody’s word for it.
They all know exactly what you want to hear.
