Are Stocks Undervalued?
The month of August 2010 was one of the worst Augusts ever for stocks. The S&P 500 lost 5%. After a big drop in the value of stocks, it’s common to hear the question, “Are stocks undervalued?” We will endeavor to answer that question.
But first, we should examine the question of value more carefully. Questions of value are always subjective and relative. When someone asks if stocks are undervalued, we are compelled to answer with another question, “Compared to what?” Some would want to know if stocks are cheap relative to their own history, and this is a valid analysis. However, historical value comparisons are helpful, but not extremely useful for making investment decisions today. All we can derive from such an exercise is to determine if stocks are cheap relative to history. This may or may not be relevant in today’s markets.
We believe that, in addition to considering the value of stocks to their historical levels, stock market valuation should be measured against other investment options available to us today. So the question of value in the stock market can be posed as “Are stocks undervalued relative to Treasuries?” Of course, investors can invest in any number of assets, but we will restrict our investigation to a two-asset universe of stocks and Treasury bonds for the sake of readability.
Specifically, we will want to compare the value of stocks, as measured by the S&P 500 stock index, versus bonds, as measured by the 10-year US Treasury Note.
Price to Earnings Ratio
Most investors are familiar with a measure of value known as the “price-to-earnings ratio,” or “P/E.” In simple terms, this ratio tells us how much investors are willing to pay for every $1 of a company’s profit. This ratio can also be derived from an index comprised of many stocks, like the S&P 500.
Let’s look at historical P/E levels for the S&P 500 for some context.
There is a glaring problem with using P/E as a measure of value and the chart above makes this apparent. As economist and mutual fund manager Dr. John Hussman has written, stocks are a claim on a very long-term stream of cash flows. Therefore, using temporarily depressed earnings (the “E” in “P/E”) in the middle of a recession to measure value can be misleading. For example, the chart above shows that at the bottom of the sell-off in March of 2009, it appeared as though stocks were as overvalued as they have ever been. This absurd result is because the price of the stock market, in the depths of a recession, is looking forward to better days. Considering stocks overvalued in the recession would have led to missing the 70% recovery in the S&P 500 that ensued over the next several months.
Price to Peak Earnings
Hussman recommends using an alternative that would effectively correct distortions in P/E at important turning points in the stock market. Instead of looking at current earnings of the S&P 500, “price-to-peak earnings” or “P/PE” will take the fluctuating price of the S&P 500 and compare it to the previous high in earnings per share. P/PE answers the question, “How much are investors willing to pay for every $1 of peak earnings.”
The following chart highlights Hussman’s adjustment to traditional P/E.
Note first that the P/PE gets to “Undervalued” territory at the market bottom in March 2009. Also note that while the current P/PE is below average, it is not yet into the “Undervalued” territory. Overall, Hussman’s adjustment appears to work better as a measure of value compared to the traditional P/E.
Adding Relative Value
Investors are not facing a decision of whether to invest in stocks or hoard cash. Instead, they have many asset classes to choose from: stocks, corporate bonds, Treasury bonds, gold, foreign currencies, etc. To keep things from becoming overly complex and laborious, let’s assume the array of investment choices facing investors is limited to the S&P 500 or 10-year Treasury Notes. We now wish to find if stocks are a value relative to Treasuries.
First, let’s go over the rationale. Buying stocks will allow the investor to earn whatever dividend is paid by stocks and the potential for capital appreciation. On the other hand, Treasuries will pay the investor a rate of interest and a guarantee from the US government. Stocks feature no such guarantee and are, therefore, riskier.
Theoretically, there exists a price level on stocks low enough (and a dividend yield high enough) to warrant taking the additional risk in lieu of Treasuries. Or, conversely, there exists a yield on Treasuries so low that the additional risk in equities is warranted. So the theory is that P/Es and P/PEs are inversely correlated to the yield on Treasuries. And this can be empirically verified by the downward sloping linear regression line in the following scatter chart, which plots Hussman’s P/PE versus the yield on the 10-year Treasury Note from 1962.
Note that as Treasury yields drop, investors have been historically willing to take on equity risk, as measured by the P/PE ratio.
If we were so inclined, and we are, we could regress P/PEs on Treasury yields to back into a “fair value” for the S&P 500. For those of you familiar with statistics, interest rates would be the independent variable and P/PEs would be the dependent variable. The price investors are willing to pay for every $1 of peak earnings depends on interest rates.
The results of the regression are as follows: given the recent 10-year Treasury yield of 2.50%, stock investors should be willing, on average, to pay 20.69 times peak earnings. The most recent peak earnings figure on the S&P 500 is $85.11. The historical relative fair value for the S&P 500 given current interest rates is 1,760.92. The most recent closing value of the S&P 500 was 1,050.00. Truth be told, the regression was a sloppy one but even if it was tight, we would only want to use it as a guide. However, what we can conclude from our analysis is that (a) stocks are indeed quite undervalued, (b) Treasury yields are too low (and Treasury prices overvalued), or (c) a combination of the two. We are long some stocks and short Treasuries for many reasons, but this valuation anomaly is a significant part of our reasoning.
And in case you were wondering what the long term behavior of “fair value” versus “actual value” of the S&P 500 was, the following chart shows exactly that.