By Nicholas Colas of Convergex
Last Time US Stocks Were So Expensive, This Happened
Summary: The best argument for avoiding US stocks is simple: valuation. Using the Shiller PE Ratio (price divided by a 10 year lookback at earnings), domestic equities trade for 29.9x earnings versus a long run average of 16.7x. The last time they were this expensive was early 2002, or 15 years ago. So how have they done since? The S&P 500 has appreciated 116% (a 5.3% CAGR) since February 2002 on a price basis. With dividends reinvested, that return jumps to 175% (a 7.1% CAGR). Not bad, but not the 9.5% average return from 1928 – 2016 either. As for how we got here, look to Consumer Discretionary (up 197% on a price basis), Health Care (+172%) and Energy (+170%). The largest drag was Financials, up only 10.0% since February 2002. But if the long term is a pretty good story, consider that 1) equities were down 22.0% in 2002 and 2) part of the appreciation over the last 15 years is due to exceptionally low long term interest rates. To get a similar outcome over the next 15 years, earnings growth may be the only driver.
According to the US Government, I will live to the ripe old age of 82.4 years. If I can make to age 62, however, that buys me 3 more years and I will expire halfway through 85. And if I make it to 70, I can tack on almost another year on top of that, within sight of 90. If you would like to see your own expected life span, just click here for the Social Security Administration’s “Life Expectancy Calculator”: https://www.ssa.gov/cgi-bin/longevity.cgi
My goal, therefore, is to get to 62 years old in decent health. Statistically, that is the easiest way to tack on a few more years of life. And it fits with my mantra on such matters: “The purpose of life is to stay alive.” Or, as my doctor tells me, “You’re a guy. If you can make it to 60, you can make it to 80”.
Equity market valuations play a similar role in estimating future returns – essentially they are one measure of the potential longevity of any given bull or bear market. High valuations point to lower future returns; lower valuations hold the promise of better future returns. Given that low valuations tend to occur during periods of economic and capital markets turmoil when asset prices (and investors) are depressed, this all makes good intuitive sense. Or, as the old trading saw goes: “Instead of crying, you should be buying.”
One measure of current US stock market valuation is the Shiller PE, which is simply the price of the S&P 500 divided by the average earnings of the index over the prior decade. Here’s how it looks right now:
- As of the close today the Shiller PE stands at 29.89x earnings.
- This is well above the long run (back to 1880) average of 16.7x.
- It is also quite close to the 30x multiple of the Black Tuesday crash of 1929, and looks quite a bit like the valuations of the dot com bubble of the late 1990s.
- You can see the chart here and also get the data in tabular form: http://www.multpl.com/shiller-pe/
There are two problems with crying “Fire” in the theater over this number. First, it is not exactly a clean data set. The accounting rules over earnings have changed a lot in the last 100 years, after all. Second, interest rates play a critical role in equity market valuations, and the Shiller PE ignores them entirely. With long rates as low as they are you would expect to see very high equity market valuation. That’s just math.
Still, there is some good work out there on how Shiller PEs can inform our perspective on future returns. Cliff Asness of AQR wrote one piece back in 2012 where he looked at 10 year returns on US stocks based on the starting Shiller PE. Here is what he found:
- There is a direct and linear correlation between the Shiller PE and future equity market returns. If you bought US stocks when that measure was below 12, your average 10 year real rate of return was just over 10% (going back to 1926).
- Conversely, if you bought when the Shiller PE was over 25 (as it is now), your average real 10 year return was just 0.5%.
- You can read the paper here, by clicking “Download” (it is 9 pages long, but very readable): https://www.aqr.com/library/aqr-publications/an-old-friend-the-stock-ma…
We can take this analysis one step further, and look at what happened the last time the Shiller PE was close to 20; conveniently, it was 15 years ago – right at the start of 2002. What’s happened since? A few points:
- The S&P 500 is up 116%, or a 5.3% compounded annual growth rate (CAGR) over the last 15 years.
- If you reinvested dividends as paid, those returns jumps to 175%, or a 7.0% CAGR.
- While nicely positive, a 7.0% total return over the last 15 years is below the 9.5% CAGR of US stocks from 1928 – 2016 or the 11.5% CAGR from 1967 – 2016.
- It therefore seems that Shiller’s PE worked as expected in forecasting returns from February 2002 to now; they were lower than expected.
- You can see a detailed record of annual returns for the S&P 500, 3 month T Bills and 10 Year T Notes here, back to 1928: http://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html
- If you want to see a very user friendly historical total return calculator for the S&P 500, check this out: https://dqydj.com/sp-500-return-calculator/
So if you had gone all-in on US stocks the last time the Shiller PE was this high and waited 15 years, you would have more than doubled your investment. Yes, you would have had to live through some pain first. The S&P 500 was down 22.0% in 2002, after all. And then there was 2008, down 36.6%. On the plus side: over the last 15 years those (2002 and 2008) are the only down years for the S&P 500.
So how did we get a 7.0% averaged compounded growth rate from a relatively high starting point for US stock valuations? Here’s how:
- Three sectors really killed it, but probably not the ones you think. They were Consumer Discretionary (+197% over the last 15 years), Health Care (+172%), and Energy (+170%). Yes, Technology did fine (+142%), but not as well as you’d think given the starting point of 2002.
- One did not: Financials, only up 10.0% on a price basis.
- Volatility over the last 15 years has, on average, been no different from the last 27 years. The average daily close for the VIX since 2002 has been 19.7. Over the longer run, that average is 20.
- The sharp eyed reader will say “Tell me about Netflix, Amazon and Priceline – they are technically Consumer Discretionary stocks, even though they are probably more correctly Tech companies”. And you’d be right. Netflix went into the S&P 500 in December 2010 and is up 411% since then, adding almost $50 billion of market value over that time. Amazon joined the S&P 500 in November 2005, when its price was $44. It is now $845/share, adding just over $350 billion of market value since its inclusion. Priceline went into the S&P 500 in late 2009, and is up 697% since then, adding $75 billion in market cap.
Can history repeat itself over the next 15 years? The Shiller PE of today is exactly where it was 15 years ago, after all, and things worked out fine. A few final thoughts:
- The missing piece of this puzzle is the direction of interest rates over the next 15 years. Back in March 2002, the US Treasury 10 year note yielded 5.42%. Now, that yield is 2.31%.
- The effect on valuation between those yields is profound. Take a perpetuity of $100 as an example. At a discount rate of 5.42%, it is worth $1,845 ($100 divided by that rate). At 2.31%, it is worth $4,329, or more than twice as much. Yes, equity valuation math has some wrinkles to it that make the math fuzzier, but you get the idea. Lower rates inflate financial asset prices a lot.
- The other factor to consider is earnings growth. S&P 500 earnings have been flat for 3 years at $117-$119/share. If the “Trump Trade” has one fundamental effect on US corporations, it must be to break that stagnancy. Over the 15 years we’ve tracked in this note, S&P 500 earnings grew by a compounded rate of 7% annually ($46/share in 2002 to this year’s expected $130/share).
That, along with lower rates, is what allowed US stocks to break the drag of a high Shiller PE back in early 2002. For the next 15 years, however, earnings will almost certainly have to go it alone if stocks are to replicate the performance of the last decade and a half.