Monday, September 8, 2008

Top Market Cap US Banks: A Long Term View on Value

Contrarian value investors following methods prescribed by David Dreman in his book "The New Contrarian Investment Strategy" (book review notes here) might be tempted today to take a closer look at some US banks. Dreman doesn't believe in market timing but he does believe that market participants are prone to making systematic errors in judgment due to inherent biases present in human behaviour. Essentially, when times are bad for the market, a sector or individual stocks, market participants have the tendency to be overly pessimistic and price in a perpetually bad scenario. Buying good quality companies with lower than average P/E ratios and higher than average dividend yields form the core of Dreman's contrarian strategy.

One problem with investing in the best banks is that it's extremely hard to know which ones they are! Managers of banks haven't been able to clearly identify all of the exposures and risks, so it's unlikely as individual investors that we will fare much better. In these type of circumstances, where superior knowledge of a company is not possible, it's better to spread risks out by buying a basket of sector stocks either individually or through a vehicle such as an ETF. This is a similar strategy as was advocated in the pharmaceutical article posted here.

Looking at three top market cap US banks, we observe the behaviour of dividend yields over time. The long term average dividend yields for Wells Fargo (WFC), Bank of America (BAC) and U.S Bancorp (USB) across the entire time period of 1970 to 2007 is 4.07% (2007 data points are present on the graph but not labelled). The current 2008 average dividend yield for these three banks is 5.98%. One caveat to these higher than normal dividend yields is that the banks are at higher payout ratios than normal. However, all three banks have had higher payout ratio's in the past so we are not in un-chartered territory.
Below is a graph of P/E ratios for the same three banks over the 1970 to 2007 time period (WFC had a -ve P/E in 1987 but for graph legibility reasons this is not shown). It's interesting to observe the generally inverse relationship between P/E ratios and dividend yields. Around 1974 when inflation was rampant with sky high oil prices, the P/E ratios averaged 6.7 times and dividend yields were at 6.4%. During difficult banking conditions in the early 80's and in 1990, one observes a return to very low P/E ratios and high dividend yields similar to levels found in 1974. Looking at current 2008 data, the combined average P/E ratio for these banks is ~15 times, which is still high by historic standards and yet the dividend yield is significantly higher than average. This data does not paint a consistent picture of value. If you believe in regression to the mean, then you would want to buy bank stocks when their P/E ratios are significantly under the long term average. The combined long term average P/E ratio for these three banks over the given time period is 12.7 times. When examined through the lens of P/E ratios, these three banks can not be bought (yet) at historically cheap levels.

No comments: