In this post, readers will discover that total bank returns and valuations are not correlated in the short term but highly correlated long term. Current bank valuations should be attractive to investors with a long-term buy-and-hold investment style.
In addition, readers will gain a better understanding why Warren Buffett and Berkshire Hathaway (NYSE:BRK.A) (NYSE:BRK.B) have made an outsized investment in bank stocks in 2018-19. Patient investors like Buffett can have a reasonable expectation that total returns for bank stocks over the next 10 years should be in the range of 200% to 300%.
Of course, investing in bank stocks is not without risk which explains why investors are currently shunning bank stocks and why valuations are favorable to historic trends.
Total returns (dividends plus price appreciation) for 40 banks have been studied since 1990. For the purpose of this analysis, bank valuation is defined as the median Price to Tangible Book Value (P:TBV) of the 40 banks for the five time periods studied. The smallest of the 40 banks is $13.6 billion Mississippi-based Trustmark Corp. (TRMK). The largest is $2.7 trillion JPMorgan Chase & Co. (JPM).
Chart 1 shows the median Price:TBV of the 40 banks by month since January 1990. The 40 banks have had an average month-end P:TBV of 2.12 over the past 40 years. The August 12 median ratio is 1.65, down from a post-Financial Crisis high of 2.25 in January 2018.
To be clear, a 1.65 median P:TBV is not downright cheap. Since 1990, banks plumbed their lowest valuations during the great bank recession of 1990-91 (when Buffett made his first investment in the old Wells Fargo which was eventually taken over by Norwest) and in 2011 (when Buffett went long on Bank of America). Since 1990, bank stocks have been valued below 1.65 about one-third of the time.
At the end of Q2 2019, banks comprised 42.6% of Berkshire Hathaway’s investment portfolio as reflected in chart 2. During the past 18 months, Berkshire has taken positions in JPM and PNC Financial Services Group (PNC) as well as added to holdings in Bank of America Corporation (BAC) and US Bancorp (USB).
Chart 3 reinforces a message I conveyed in my May 4 article about the Regional Bank ETF (KRE): Short-term returns in bank stocks are highly random, and therefore, it is folly to buy a bank stock or bank ETF if an investor does not plan to own the equity for a long time.
Note in Chart 3 that the R-square for 3-month total returns is .002. In other words, short-term bank returns are random regardless of bank valuations.
While investors improve the odds of getting a favorable total return by increasing their holding time of bank stocks to 1 year, as Chart 4 shows, the R-square is still low at .1975. 1-year bank investors cannot have great confidence of a positive total return at virtually any valuation entry point, although odds of a favorable outcome are better at lower valuations than higher.
Bank investors with a 3-year time horizon can with some confidence expect to see a relationship between valuation and 3-year returns as shown in Chart 5. However, an R-square of .4817 (correlation of roughly 70% squared), while clearly superior to investment horizons less than 3 years, is still not statistically significant for the cautious investor seeking to manage downside risk. A 3-year bank investment horizon does appear to insulate investors to a fairly large degree against an actual loss of principal provided the bank stocks are purchased at a median P:TBV of no greater than 1.8.
By moving the bank investor horizon to 5 years, the R-square improves significantly to .6641. Even a casual eyeballing of the data in Chart 6 shows a clear relationship between historic return and valuation. Bank investors with a 5-year horizon can with some confidence expect a positive total return provided banks are bought at a P:TBV of less than 2.0. In contrast, investments made when bank stocks have a P:TBV greater than 2.5 have failed about 50% of the time since 1990 to have produced a positive total return.
Chart 7 is the money chart as it shows a statistically strong R-square of .8073 (a correlation of roughly 90% squared) between valuation and 10-year total returns. Over 10 years, statistically speaking, about 81% of the total return of bank stocks can be explained by valuation at the time of the bank stock purchase.
When the median P:TBV of the 40 banks has been 1.6 to 1.7 since 1990 as it is in mid-August 2019, total returns over the next 10 years have historically been between 200% (tripling the initial investment) and 300%. It is worth noting that the odds of negative total returns appear negligible at today’s valuation.
This article was written by
Disclosure: I am/we are long BRK.B, JPM. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: Having retired from a 31-year career at Bank of America in 2011, though I currently have no equity holdings, i do have certain financial interests in BAC.