Disclaimer: The views expressed here are mine and may change without notice. Past performance is not indicative of future results. All investments carry risk, including financial loss. This analysis is for educational purposes only and does not constitute investment advice or recommendations of any kind. Conduct your own research and seek professional advice before investing. Please see important disclaimers here and here.
________________________________
Banks buy and sell money which is a relatively simple business. However, we continue to see bank failures whether it was the savings and loan crisis of the 1980s or the 2009 financial crisis, to name two. The primary culprit has always been the asset side of the balance sheet, meaning, what does the bank DO with the money is of prime importance. Thus, assessing the lending-related downside risk for a bank is far more important than upside when assessing opportunities.
Consider the case of large US banks that are currently priced cheaply versus history. While large banks are better capitalized compared to past crises, the recent coronavirus crisis has put pressure on their commercial loans leading to increased risk. The underlying cause is the uncertainty regarding the long-term future of hotels, malls, retail, and office properties; their valuations have become hard to assess. We don't see their lower valuations yet on the asset side of the bank balance sheets because of accounting-related protection from the CARES act and lower transaction activity. And banks aren't talking about this. For the accounting part, banks are allowed to book revenue even if they haven't received the revenue from a COVID-impacted property. However, there are multiple indications suggesting significant destruction in commercial property values. Here are a few examples (emphasis is mine):
In total, we saw that the values of lodging properties were down 28.9%, in aggregate, for
those that had been re-valued lower and 28.5% for retail loans.(source) S&P Global Ratings in July downgraded the entire Starwood commercial mortgage-backed
security to speculative grade after a reappraisal of the four regional malls backing the debt
valued them 66% lower than when the bond was issued.(source) 
Vornado (large office landlord) shows a substantial decline in profits
As you see, a few areas of the CRE market are under severe stress which leads to uncertainties regarding their values. This is not a good setup for an investor specifically when it comes to investing in banks. This is because the wide range of outcomes on the asset side translates into the risk of large and permanent capital loss. Remember this risk from uncertainty is on the top of risk that a bank investor faces around not knowing how the bank has underwritten its numerous loans.
A careful analysis of the large banks recent financial statements and financial reports leads me to conclude that Bank of America (BAC) is best positioned given its good underwriting culture and least exposure to the CRE loans. The latter happened because it purposely reduced its commercial real estate (CRE) exposure post-2009 crisis after realizing that CRE loans are usually the culprit of huge losses in the prior banking crisis (1980s savings & loan, 2009 crisis). On the contrary, Wells Fargo has become the largest CRE lender in the US. Its CRE exposure is the highest with those loans comprising roughly 17% of the loan book. That compares to roughly 12% for JP Morgan, 10% for Citi, and 8% for BAC. To be specific, Wells Fargo's office, hotel, retail, and mall loans alone make 8%+ of the total loan book compared to 3%+ for BAC.
Banks typically are leveraged companies. When assets are ten times the equity, a 10% loss on the assets will wipe out the equity if earnings are not enough to close the gap. This is where Bank of America shines as well. Apart from lower exposure to risky loans, it has much larger earning power that comes from its well-diversified sources compared to other banks such as Wells Fargo. BAC generates a lot of fee-based revenue from its retail, asset management, and investment banking businesses.
Let me quantify the downside scenario. We saw the data from private market valuations and other sources quoting a decline of little under 30% for hotels. Let's assume a draconian scenario where the value of all COVID impacted CRE loans declines by 30%. This is a lower probability event though you need to think about it when you are dealing with major uncertainties in a leveraged financial institution.
Consider this: In 2019, Bank of America earned well over $30 billion pre-tax annually after reserving roughly $3.6 billion for loan losses. If 10% of all $1 trillion of the BAC's loans were hit by problems from COVID, resulting in total losses of 30%, the company would still nearly break even.
I think BAC is not only better protected, but I also expect it to earn in mid-teens on growing equity with benefit from one of the lowest cost of deposits. The current market price provides a good opportunity to buy a superbly managed bank run by an underappreciated management team led by Brian Moynihan. He is a true banker given his focus on making the bank simpler. One crude measure is the length of the annual report which fell almost 25% in the past decade compared to 14% growth for Wells Fargo, 6% growth for Citi, and flattish for JP Morgan!
In summary, when evaluating a bank's attractiveness, I think a strong consideration should be given to its downside risk from the asset side of the balance sheet. Given banks' high leverage, I would rather own a well-run bank at a fair price, such as Bank of America, versus a risky bank at a cheap price.
