As equities reach new highs we are increasingly being asked what seems like a very basic yes/no question: “Are stocks expensive?” The question is simple, but there are at least two pitfalls that prevent a simple answer. First, there are countless ways to calculate how expensive or cheap stocks are, and they will often provide conflicting conclusions. Second, once you have reached a valuation conclusion, we believe it would be foolish to make an investment decision on the basis of that alone. Let’s dig a bit deeper so that the next time you read or hear comments on this topic you can put them in the proper context.
There is no single, correct way to value the stock market. Some valuation metrics are backward looking, while others are forward looking. They can be based on earnings, assets, equity, sales, cash flows, etc. Some incorporate adjustments for current or expected inflation, growth, and real rates, while others make no such adjustments. To get a broader sense of how expensive or cheap stocks are, we look at three different metrics and continually consider others. This is what they are telling us now:
Debating the pros and cons of each of these metrics is beyond the scope of this paper, but you can see that with just three metrics we get conflicting readings. In one reading, it would be factually correct to claim stocks are more than 30% overvalued today (current P/E versus historical P/E). Which begs the question, why do we remain modestly overweight stocks in our portfolios?
The reason lies predominantly in the predictive power of valuation measures and future returns. The graph to the right plots the starting P/E ratio and the annualized return over the next five years, going back to 1900. There is evidence of a small statistical relationship between lower (higher) P/Es and higher (lower) five-year returns. The key word being small. In fact, P/Es explain less than 15% of the variability of stock returns over the following five years! This is why we spend so much time seeking to understand what drives the other 85% of future asset returns — by focusing on the impact of the economy and policy.
Furthermore, even if the explanatory relationship between P/Es and returns were stronger, given the current P/E of 19.7, simple regression analysis would still predict an acceptable five-year return in excess of 5%.
Valuation is always a good starting point, but it should never be the ending point for making an investment decision. And we continue to believe that while valuations are modestly expensive, economic and policy conditions remain supportive of further upside.
© Copyright 2016. BMO Financial Corp. All Rights Reserved.
