Key Takeaways From Banks' First-Quarter Earnings
We're maintaining our call on the U.S. banks and will know a lot more by the second quarter.
The traditional U.S. banks we cover have finished reporting first-quarter results, and we have a lot more information than we did heading into earnings season. This remains an extremely difficult situation to analyze. Nobody knows exactly how the coronavirus and the economy will develop, and nobody knows what future credit losses will be. Even so, after reviewing the latest disclosures and rerunning our scenario analysis, we still think the traditional U.S. banks are undervalued. While much of the market has generally recovered to some degree, many of the banks have not. In a market where bargains are more difficult to find than several weeks ago, we think the banks are still worth considering.
As we argued in our previous piece, we think today's valuations are a debate about capital adequacy, not about earnings per share for the next quarter or year. We still calculate that something worse than the U.S. Federal Reserve's severely adverse stress test scenario would have to occur for today's valuations to make sense. As such, we think the odds of superior returns for investors at today's valuations are still constructive. Nobody knows exactly what credit losses will look like in the future, or how long the COVID-19-related lockdowns and economic downturn will last, which makes investing in the banks tough. The risks are real, but on a probability-adjusted basis, we think many of the banks offer an attractive margin of safety today.
There has been a partial recovery in bank stocks since the bottom in mid-March, but the recovery for the banks we cover has been much smaller than the recovery seen for the overall market. For the banks under our coverage, price/tangible book value multiples are still roughly on par with levels seen in 2008.
Earnings Projections and Fair Value Estimate Updates
Net interest income is coming in better than we expected. Hedging programs, deposit repricing, and higher balance sheet growth are all helping. The overall trend for net interest margins is still down, but many of the banks are supporting margins better than we anticipated, and better balance sheet growth has led to positive revisions for net interest income in our projections across the board. Higher balance sheet growth, driven by commercial clients drawing down on undrawn credit lines, was a factor we had called out in our previous piece and was one of the primary areas where we thought the banks had room for upside compared with our original estimates. This has indeed played out.
Most fee items weren't too bad in the first quarter. As we expected, trading, mortgage, and portions of investment banking were all strong. However, we started to see the initial strain on payment-related fees, and wealth fees will also come under more pressure in the second quarter. After updating our projections, we now expect fees for the remainder of the year to be down around 10% for our coverage.
The latest payment volume data from companies such as Visa (V), Mastercard (MA), Fiserv (FISV), and JPMorgan (JPM) show that there may be some light at the end of tunnel. The initial drop in volume had already started to level off by late March to early April, and some volume even showed some recovery by late April through early May, although it was still nowhere near pre-COVID-19 levels.
Expenses are coming in better than we expected. We originally thought that the increased hiring, increased minimum wages, and an inability to focus as much on previous expense-saving initiatives would cause us to negatively revise our expense projections. However, the latest results and guidance suggest that the banks are still able to actively manage their expenses.
For provisioning, we now expect a second, even larger step up in reserves in the second quarter. Essentially all of the banks admitted that the economic data had deteriorated since they made their original reserve estimates as of March 31. As such, we think another increase in reserves will occur in the second quarter, after which we expect provisioning to begin to level off in the third and fourth quarters.
Putting all these updates together, we now expect earnings per share overall to fall around 57% in 2020 and to remain below 2019 levels in 2021, as provisioning remains elevated and rates remain low. We have also lowered our fair value estimates by around 2% on average compared with our estimates from April 7.
Most of the banks provided more details on their COVID-19-specific exposure (mainly restaurants, lodging, transportation, and retail) and energy exposure. These are likely to be some of the highest-risk exposures for the next several years. COVID-19-specific exposures averaged around 10% of loans for our coverage, which we view as manageable. Energy exposure did not change much from 2019 fourth-quarter levels. Cullen/Frost (CFR), Zions (ZION), and Comerica (CMA) remain the most exposed to the energy sector. When combining energy and COVID-19 exposures, Regions Financial (RF) appears the most exposed at 155% of common equity Tier 1 capital, Zions is a close second at 146%, and Fifth Third Bancorp (FITB) is third at 126%. Most of the other banks were at 100% or less. Unless these balances turn out to be as toxic as the worst mortgages from the 2008 downturn (that is, subprime/alt-A Florida and California types of mortgages), we don't think these portfolios will be unmanageable.
One of the key concerns we think investors have today is, "What if good underwriting doesn't matter this time?" In other words, if the government forces whole sections of the economy to shut down, it may not matter if a bank was a good underwriter or a bad underwriter; all companies in that sector will suffer equally. While there is an element of truth to this, we think underwriting will still matter this time around. How conservatively a bank structured its loans, what the collateral requirements were, and so on, will all still play roles in what losses ultimately are for each bank. We also think that the U.S. Treasury and Federal Reserve are focused on not letting companies that would have otherwise been fine go bankrupt. There should be more support from the government for the higher-quality companies that more conservative banks would have been lending to.
To support our view that the banks have been more conservative this time around, during the earnings calls, many banks gave details on their energy and COVID-19-specific exposures. The banks have structured their energy portfolios differently since 2015, with Fifth Third even having large portions of its book hedged all the way to $4-$5/barrel oil through 2020. Restaurant balances for many of the banks often have over 50% of exposure tied to quick-service restaurants. Loan/value ratios and client selection for commercial real estate clients appear conservative for M&T Bank (MTB). Multiple banks also commented on the fact that their leveraged lending portfolios have minimal exposure to higher-risk covenant-lite or term B loans. While we don't want to minimize the risks, and we expect losses to increase, we think the latest results and information reaffirm our thesis that it will be much harder to break the banks this time around and that underwriting has been much more conservative.
Losses Still Appear Manageable; Some Banks Appear More Vulnerable
After updating our projections, we still see minimal deterioration in capital levels for the banks we cover. We see core preprovision net revenue dropping around 12% and provisioning jumping significantly, causing an average drop in net income of 58%. We see dividend payout ratios jumping to over 100% for many of the banks. We currently model a downturn with losses that are approximately half as bad as what a severely adverse stress test scenario would imply. Even under our projections, common equity Tier 1 ratios only deteriorate around 20 basis points from 2020 first-quarter levels.
One of the hardest aspects of analyzing the banks today is that no one knows what losses will be. We may be underestimating the severity of the upcoming downturn; it's very hard to know for sure. We've rerun our scenario analysis using the vantage point of a "typical quarter" to estimate how many quarters it would take for a bank to breach its minimum common equity Tier 1 ratio requirements. Even using the maximum loss rates from the most recently available stress tests, almost all of the banks would last for over two years before even breaching their first common equity Tier 1 requirement. Under this scenario, Citigroup (C), Wells Fargo (WFC), M&T, and KeyCorp (KEY) appear the most vulnerable while JPMorgan, Bank of America (BAC), and PNC (PNC) look the strongest. Given the disparity between KeyCorp's and M&T's company-run stress tests and the Federal Reserve's version, we're inclined to think this may be too harsh on both banks. Using their company-run test loss rates has them appearing to be much stronger.
In addition, we compared current reserves with past stress test loss rates, to get a better feel for which banks might be more vulnerable to relatively higher reserving going forward. Based on our analysis, Zions and Fifth Third appear the ripest for higher reserving. They have some of the lowest relative reserve coverage and some of the highest relative exposure to energy and COVID-19. Wells Fargo may also be a bit light on cards, and Regions may be a bit light in commercial and industrial, considering its higher COVID-19-specific exposure.
Stock prices are also an important part of any final analysis. In our price/tangible book value thought experiment, we use our current assumptions for long-term return on tangible equity and cost of equity to back into an implied P/TBV ratio. We then compare this with the current stock price and back into how much tangible book value would have to be destroyed for today's prices to make sense. This gives investors a general sense of what they’re paying for at these prices. We use three scenarios. Scenario 1, our base case, uses our current assumptions for ROTE and COE. Scenario 2 assumes that long-term ROTE is 100 basis points below our base case. Scenario 3 is based on our analysis of what happens to long-term ROTE for each bank if rates fall to zero and stay there forever. This typically results in a long-term ROTE that is even worse than under scenario 2. For each scenario, we calculate how much tangible book value would have to be destroyed, and what loss rates would have to look like to bring about these losses.
In our base case, almost all of the banks appear cheap today and would have to experience losses worse than past stress tests for today's valuations to make sense. Maybe this will occur--nobody knows for sure--but we think this implies a pretty high margin of safety as a severe bear case would have to play out. In scenario 2, the results begin to get a bit more mixed but still imply that the majority of banks would be cheap. In scenario 3, Citigroup, Wells Fargo, and KeyCorp are the only three that still appear cheap to us under both stress scenarios.
How to Think About Timing and Catalysts
The second and third quarters of 2020 will probably be the most important ones for the banks. We think we are much more likely to know whether our call on the banks has worked out by the end of the third quarter. This is because we will have a much better idea of how the economy and credit losses are playing out as the second and third quarters develop. We'll have a much better sense of what is working and what isn't for the government programs and if we've made any material progress on the healthcare side of the equation. We'll also have a much clearer idea of what it means to try and reopen the economy.
Many of the initial 90-day deferrals will also start to run out during this period. Credit costs are the key unknown for the banks, and after the second quarter and third quarters, we'll have a much better idea of what credit costs will probably look like. This will be the first catalyst, removing some of the paralyzing uncertainty around credit costs. At this point, we'll likely either have been wrong or right about our call on the banks. Just removing uncertainty around credit costs probably won't get the banks back to pre-COVID-19 valuations, however. The second catalyst will be a robust enough economic recovery to allow room for rates to rise.
We've rerun our dividend safety analysis, and we think JPMorgan, Bank of America, and PNC appear to have the safest dividends under our coverage. Citigroup stands out as the weakest. We would expect Citi to be the first to cut dividends if it came to that, and we don't think it would take a "severely adverse stress test" level of losses to cause the bank to do so. Wells Fargo appears weaker than most, but not as weak as Citi, while the rest of the banks we cover are somewhere in between. While KeyCorp and M&T look weak under the Fed's stress test scenario, we are inclined to weigh the company-run stress test scenarios more, which make them appear safer.
Eric Compton does not own (actual or beneficial) shares in any of the securities mentioned above. Find out about Morningstar’s editorial policies.