Despite Warren Buffett’s selloff, bank stocks look like great buys in this market
To grasp why beaten-down value stocks look like the place to be in this super-pricey market, just compare the bargain bank stocks to the high-flying S&P 500.
Given today’s valuations, the overall big-cap market can’t hand you a strong future return without working wonders. Banks, by contrast, just need to go from garnering the lowest of low expectations, to posting profits that are mediocre by past standards. Looking at banks and the glamour-tech- dominated S&P side by side provides a case study in how what’s cheap is most likely to win the recovery. (For more, see “The champ’s big comeback: Why beaten-down value stocks are poised to thrive.”)
It’s somewhat surprising, then, that Warren Buffett, the greatest “value” investor of all time, just dumped a huge chunk of his holdings in what looks like one of the market’s cheapest pockets, commercial banks. It’s possible that the Berkshire Hathaway chairman turned negative on this longtime favorite because he foresees a long, rocky recession caused by COVID-19. Such a recession would swamp the stalwarts with wave after wave of defaults on credit card, small business, and commercial real estate loans, not to mention the threat of more credit losses from such ravaged big corporate sectors as airlines, hotels, and commercial real estate.
In its 13F filing for the second quarter, posted on Aug. 8, Berkshire disclosed that it had sold large portions of its holdings in five commercial banks during the quarter. Buffett shed 26% of his shares in Wells Fargo and a stunning 62% in JPMorgan Chase, as well as stakes in U.S. Bancorp, M&T, and PNC, for a total of $6.7 billion.
Keep in mind that Buffett remains a big owner of bank stocks. He kept all of his 925 million shares in Bank of America, a $22 billion investment that’s Berkshire’s largest after Apple. Still, given his big sale, it’s natural to question whether the Great Lockdown is posing such a dire threat to the lenders’ futures that you risk poor returns even at what looks like steep discounts. The best view: Prices are so low that the banks are providing one of the best “margins for safety,” ironically one of Buffett’s favorite metrics, of any group of stocks. Even if a lot of things go wrong, the banks can still deliver for investors. By comparison, the S&P 500 is teetering at such rarefied heights that a sudden blast of bad news could send it toppling.
How banks became a bargain
The beauty of banks is that they were cheap at the peak of the market, and the pounding from COVID-19 has made them a whole lot cheaper. By comparison, the S&P 500 was richly priced before the outbreak, and it’s just as expensive now—more so if you assume the recession will depress profits for a couple of years. On Aug. 17, the index closed at 3382, within a whisper of the all-time high of 3386 set on February 19.
By comparison, the big banks haven’t come close to regaining their February levels––and if you’re hunting for deals, that’s a good thing. Around Valentine’s Day, the combined market capitalization of the five commercial banks with the highest valuations—J.P. Morgan, Bank of America, Citigroup, Wells Fargo, and U.S. Bancorp—was $1.177 trillion. To be conservative, I compared that value not to their 2019 earnings––which were higher––but their average profits from 2017 through last year. By that metric, their “normalized” annual GAAP net earnings, after preferred dividends, over three years were $94.6 billion. Hence, the five had an overall price/earnings ratio of 12.44.
Since then, the group has lost one-third of its value as its total market cap shrank to $792 billion. As a result, its P/E based on those “normalized” profits dropped from 12.4 to 8.4.
Let’s compare that dive to the course of the overall S&P 500. For normalized profits, we’ll use the $139.40 in earnings per share that the 500 posted for 2019, an all-time record. At the market peak in February, the S&P’s P/E was a lofty 24.3, or almost exactly twice the multiple for the banks.
Today, the S&P’s 35% resurgence from the March lows has planted the flag at the previous summit, and restored the 24-plus P/E. So now, the market multiple is almost three times the combined P/E for our five banks.
Of course, the banks’s earnings have suffered a big hit, dropping to a combined $6.7 billion for the first half of 2020 or $13.4 billion on an annual basis, a fall of 85% from the three-year norm. That’s a lot more than the substantial 37% hit the analysts polled by S&P are expecting by Q3 for the overall S&P.
The beauty of low (earnings) expectations
What matters is what the banks need to do from here to provide gains that, in all probability, will beat the overall market. Let’s forecast only a partial comeback for the banks, combined with continuing skepticism about their future. We’ll also assume that the economy returns to 2019 levels of GDP by the end of 2021— in other words, back to even.
But we’ll posit that the banks don’t rebound nearly as fast. Let’s assume that in 18 months or so, their profits return to just 75% of the 2017–2019 average of $94.6 billion—or to $71 billion. We’ll also assume that their P/E is only 13, which is around half the current market multiple. In that case, their combined market cap would rise from today’s $792 billion to $923 billion. That’s a gain of 16.5% over less than a year and a half, or 11% annualized.
How is the S&P likely to fare? We’ll make our projections for the overall large-cap market a lot more optimistic. We’ll assume that profits regain their 2019 level along with national income. That’s a stretch, because to get there, companies would need to roar back from COVID-19 and return to their highest earnings-per-share ever, on margins on sales that are at least 20% above the averages of the pre-pandemic decade.
Even if that happens and the S&P’s multiple stays steady at 24.3, the S&P would deliver no gain from where it is today. You’d just get a puny dividend yield in the 1.5%-a-year range.
A P/E of over 24 will be hard to sustain, since it’s well above the long-term average and means that even at high levels of EPS, investors would expect more big jumps in profits. To get a more reasonable, measured forecast, I like to use my what I call the “Tully-20” rule. It assumes that the P/E will revert to 20 when GDP goes back to 2019 levels, which is still a formidable number.
If the P/E does return to 20, then the S&P would need to earn $170 a share just to maintain its current level of 3385 (3385 divided by 20). That’s over one-fifth above its record EPS in the flush times of 2019. On the other hand, if Tully-20 reigns and earnings go back to $140, a much more probable picture, then the S&P would retreat to 2800 (20 times $140). That would mean the S&P would fall 17% in the next 18 months or so.
The S&P would need a great performance to earn $140 again by the end of 2021, and still flatline at a 23-plus P/E. Or it could perform a near miracle at a more reasonable multiple of 20 and still deliver a zero capital gain. And the chance of a big drop is high.
Just compare those hurdles with what the banks must accomplish to beat the odds. A lousy P/E of 13 combined with profits 25% below the three-year average gets you a 17% gain by the end of 2021 (and over 10% on an annual basis).
Big banks are a proxy for the economy, more so now because they’re far better capitalized than in the Great Recession. It’s hard to imagine a recovery where the banks don’t rebound in lockstep with the rest of American business. If that script plays out, they’ll do much better than our 17% over the next year and a half. Investors think that banks are dogs that can barely waddle. If they just manage to stroll, they’ll rank among the market’s big winners.