After my recent foray into a moderately obscure banking merger, I also happened upon a variety of other depositories -- specifically a group of Canadian banks. Now this focus is likely just a momentary sidestep from my tried and true dividend growth spiel, but I believe I can add some insight while I'm here in unfamiliar banking waters.
The companies that caught my eye included: Bank of Montreal (BMO), Bank of Nova Scotia (BNS), National Bank of Canada (NA), Royal Bank of Canada (RY) and Toronto Dominion Bank (TD). I'd like to preface the following by suggesting that my knowledge of these companies remains limited. However, I did what anyone might do when they want to learn more about a company and went to research resource -- specifically F.A.S.T. Graphs.
Instantly I was struck by what appeared to be a group of companies that were worthy of further research. All five traded at a price-to-earnings ratio around 12 or lower, had a sizable current dividend yield over 3.5%, a relatively strong historical growth rate and debt levels, along with reasonable estimates of future growth that appeared to yield the possibility for outsized returns.
From this initial starting point, I moved on to the various graphs and began educating myself on what the market has been willing to pay for a Canadian bank ownership in the past. I could talk about it or walk you through some charts, but I believe the best use of your time would be to simply observe a summary of my findings in the historical "normal" P/E ratio table below:
Here we see that the market was willing to capitalize these five banks' earnings with a multiple of anywhere from about 10 to 14 over historical 5, 10, 15 and 20-year periods. If you average everything out, you come up with an aggregate average normal P/E of about 12.5. Now when I first noticed this, I was a bit let down. By understanding the mechanics of the previously referenced total return estimate, I knew that the default P/E within the calculator was 15. In addition, I appreciate the fact that a good bit of the total return is dependent on capital appreciation. Thus while the companies might be quite solid, I realized that the total return expectations would have to be scaled down.
Next I wanted to get a feel for how these companies -- along with their past market pricing -- compared to comparable companies that I was more familiar with; banks such as J.P. Morgan (JPM), U.S. Bancorp (USB) and Wells Fargo (WFC). By running the same process and skipping to the punchline, I present the historical "normal" P/E ratios for those three American banks:
Here we see a similar story as the Canadian banks, with perhaps a bit of a premium priced in; overall the market multiples range from about 12 to 15, with an average aggregate normal P/E of about 13.7. Additionally, it occurred to me that JPM -- with an average P/E of about 12.6 -- had been valued more closely to Canadian banks rather than its American counterparts. In viewing USB and WFC together, the average normal P/E turned out to be about 14.25 against the previously quoted 12.5 amongst the five Canadian banks.
Admittedly it might be largely unfair to lump all of the banks together. But then again -- given that I didn't have particularly good clarity into the industry -- I believe it works for a thought experiment. Realizing that the group of Canadian banks I selected seem to trade at a lower valuation than the USB and WFC counterparts, I was curious as to whether or not the seemingly stronger fundamentals would make up for this fact. Below I have listed the 5-year expected earnings growth rate given by consensus analyst's estimates, the current dividend yield and the current P/E.
|Exp. Growth||D. Yield||Current P/E|
Once again ignoring the folly of lumping all the banks together, these five companies held in equal parts would aggregate to a 7.9% expected 5-year growth rate with a 4.16% yield and 11.1 current P/E ratio -- we'll call this combination "CA." Here's a look at USB and WFC:
|Exp. Growth||D. Yield||Current P/E|
Aggregating these two U.S. banks, we find a 5-year expected growth rate of 6.75%, a 2.65% dividend yield and a current P/E of 12 -- we'll call this "US." In other words, the Canadian bank combination has a higher expected growth rate, higher dividend yield and lower price-to-earnings ratio. It seems that the only thing that the US combination has going for it is a higher historical P/E ratio.
Finally, to test the importance of this slight perceived premium, I wanted to compare hypothetical returns amongst our made up "CA" and "US" securities. Additionally, it's imperative to note that historical P/E ratios in no way suggest potential outcomes. However -- in lieu of additional valuation insight -- looking to the past as a semblance of a guide toward the future isn't an altogether unreasonable place to start. Here is what the "CA" security might look like:
Notice that I arbitrarily selected $100 as beginning stock price, but the $8.98 in EPS and $4.16 in dividends are based on the previously cited 4.16% current dividend yield and 11.1 P/E. Earnings and dividends are grown by the 7.9% expected growth rate each year and the final price is based on the 12.5 average P/E multiple. In total -- given the assumptions -- each $100 investment in CA banks would compound at about 10.4% a year. In addition, you would receive just over $26 in dividends for a total annualized return of roughly 13.7%.
In completing the same exercise for the "US" security, we find the following:
Again, I arbitrarily set the beginning price to $100, but the $8.33 in EPS and $2.65 dividend is based on the 2.65% average current yield and P/E ratio of 12. Earnings are once more grown at the average expected growth rate -- in this case 6.75% -- and the final price is based on EPS in the 5th year multiplied by the historical average P/E of 14.25. Interestingly, each $100 investment in US banks would compound at about 10.5% a year, or just a touch quicker than the CA example. Said differently, the change in relative P/E expansion would make a difference, despite the higher initial P/E ratios of US banks. Although to be fair, the US security example only yields about $16 in aggregate dividends -- for a total return of 12.6%. But that's not necessarily the point. A small decline in the US price or an incremental increase in the CA price would effectively make the two scenarios equal. Or notice that the US payout ratio is much lower and thus those banks might be able to increase their dividends at a quicker pace to make up for the lack of initial yield.
The point was what while the Canadian banks might have immediately appeared more favorable -- when viewing current income, expected growth and relative valuation -- the analysis actually came out closer than what you might expect. This is almost solely due to a higher ending P/E ratio. Now obviously one could argue with great reason that historical P/E ratios don't equate to future multiples. But the underlying message of the thought experiment remains: how a company's fundamentals are valued can be just as important as the fundamentals themselves. With similar banks this might be hard to see, but the ideology becomes even more practical when comparing companies in different industries. Take the market's continuous "premium" valuation of Coca-Cola (KO) and the permanent "discounted" valuation of say Chevron (CVX), or really any one of the banks I mentioned in this article, for example. In short, the fundamentals of a company are the quintessential building block to a good analysis -- but don't lose consciousness of what reasonable valuation multiples might look like either.