Yesterday I posted a chart of the S&P 500 Price to Peak-Earnings multiple going back to 1871. I noted that relative to history the current multiple is high. Since 1871, this multiple has averaged 11.9x, and today the multiple is 18.7x. Simple arithmetic implies that the multiple would need to contract by 36% to get back to the historical average.
Truthfully though, the average multiple depends a lot on how long of a historical period that you consider. Since 1871 the average is 11.9x, but over the last 50 years it’s 15.5x and over the last 20 it’s 18.9x. The historical period that you choose to use has a big effect on the conclusion. If you only look at the last 20 years, equities are downright cheap.
People who gravitate towards long term history will tend to write-off the 20 year average as skewed by the dot-com bubble. I tend to agree with that analysis, so I should warn that I don’t think I believe what I’m about to write, but it’s worth considering that maybe we’re in a different paradigm and the last 50 years are irrelevant let alone the last 150 years. Maybe the 20 year average is the right one to use. Maybe equity multiples are at a permanently higher plateau because innovation has permanently decreased the cost of equity.
To help illustrate this theory, I inverted yesterday’s chart to make the PE multiple into an Earnings Yield on Peak-Earnings. This makes for easier comparison to 10 year interest rates. Looking at the chart I can see an argument for three distinct periods. 1) 1871-1950 2) 1950-2000 3) 2000-present. Simplistically, the key traits of the three periods are 1) equities riskier than bonds 2) equities and bonds at risk parity 3) A very low interest rate environment.
My argument that we are in a new paradigm (which, again, I don’t think I believe) is that over this long history, the equity risk premium has been erased, and the cost of capital has been permanently lowered. In other words the spread between the earnings yield and interest rates is no longer valid and interest rates will be permanently low. If you believe these two factors, I think you can justify a permanently higher range for equity multiples with an average of 19x+.
The argument that the equity risk premium is irrelevant is that the S&P 500 is really not any riskier than a long term treasury bond. It’s true that an individual stock could go bankrupt or see a meaningful contraction in its business. However, there is almost no risk that a basket of the 500 largest publicly traded US companies will be impaired in the same way. The S&P 500’s earnings stream is not meaningfully riskier than the coupon payments of a 10 year treasury note. True, the bond does have absolute certainty of repayment, but the S&P 500 has inflation protection. I think the riskiness balances out. So does the S&P 500 really deserve any risk premium to treasuries? With the proliferation of ETFs and passive investment funds, perhaps the spread between the earnings yield and interest rates deserves to go back to near zero.
Furthermore, maybe today’s low interest rates are justifiable for the long term. Today’s capital markets are deeper and more competitive than ever in history. In the 1920s capital had to be physically transferred from location to location. Armored cars, trains, etc. actually had to physically move gold and securities between locations to fund investments. Today all the savings of all the people of most of the world can buy whatever security or basket of securities they desire with the click of a button. The competition between savers for investment returns is intense because barriers to competition have been torn down. The abundance of capital drives the cost of capital lower, perhaps permanently.
It’s hard for me to see a reason why the above arguments aren’t true, but I do know that a) it’s way easier to make these arguments after a five year bull market than at the bottom of a bear market b) markets will naturally swing from euphoria to despair. These are the type of arguments that justify euphoric prices. c) “this time” is almost never different and I don’t think that it is. But I think it’s worth considering that it could be.
Source: Robert Shiller, Avondale