A few weeks ago I asked Twitter:
For those that think U.S. stocks are currently overvalued, when was the last time that U.S. stocks were FAIRLY valued?
Immediately the trolls were out in full force:
Though these first responses were mostly jokes, the tweet evolved into a great discussion about what it means for a market to be fairly valued. My favorite response to this question came from Drew Dickson:
They’ve never been fairly valued. The price is always wrong. It’s just that we don’t know how wrong they are.
Now, are there times when certain stocks (or markets) are extremely unfairly valued? I think yes.
Can we predict WHEN things might change? I think no.
Though Drew’s answer is the most accurate, I wasn’t satisfied. I still wanted a date. I still wanted to test something.
I looked through the other replies and came across one from Christopher Bloomstran which read “2010/2011”. Despite his lack of knowledge surrounding Solana (lol I love you Chris), Bloomstran is one of the best thinkers in FinTwit. So I took the bait. If 2010/2011 was the last time markets were fairly valued, I wanted to know what that meant for markets today.
Market Performance & Fair Value
To make this thought experiment simpler we will assume that January 2011, right in the middle of Bloomstran’s range, was the last time that U.S. stocks were fairly valued. So how have markets performed across history up until January 2011? Let’s see.
I analyzed data going back to 1920 and found that both earnings and prices (when including dividends and adjusting for inflation) grew at a similar pace through January 2011 across most of history.
To visualize this I plotted some bars below representing the annualized growth in earnings and prices (with dividends and adjusted for inflation) from the start decade listed until January 2011. For example, from January 1920 to January 2011, earnings grew at 6.5% per year while prices grew at 7.1% per year, and so forth:
As you can see, while there is some variance between earnings and price growth, they are pretty similar in most decades. Nevertheless, there are two notable exceptions—1980 and 2000.
In 1980 stocks were coming out of a bad slump (remember the Death of Equities cover?), so prices ended up growing faster than earnings (7.6% vs. 4.9% annually) over this time period. And in 2000 stocks were so overvalued from the DotCom Bubble that prices had to come down. Therefore, prices contracted by 1.6% annually even as earnings grew by 3.7% annually over the next decade.
Outside of those two periods, earnings and price have mostly moved in lockstep over time. But what have earnings and prices done since then? From January 2011 to September 2021 (the latest data available), U.S. stock earnings grew by 8.2% per year while prices grew by 13.0% per year. This means that, to get back to the same price-to-earnings ratio (P/E) as in January 2011, U.S. stock prices would need to fall by roughly 40%.
“Aha!” you might be thinking, “There’s the proof. The market is extremely overvalued!” Well, it’s not that simple.
First, things are a bit different today than they were in January 2011. Back then the 10-year Treasury was paying 3.36%. Today it’s paying roughly 1.95%. With fewer ways to generate a return, investors have bid up U.S. stock prices accordingly.
Second, if we change the time period we use, our interpretation of the results would change as well. For example, from January 1990 to September 2021, earnings grew at 6.4% per year while prices grew at 8.2% per year. This might seem alarming, but in January 1990 markets were undervalued by about 25% (relative to January 2011). Therefore, we should expect prices to grow faster than earnings through September 2021, but how much faster is anyone’s guess.
Lastly, if you are still worried about U.S. stocks being overvalued, you have a little less to worry about today. With the S&P 500 nearly 10% off of its highs, a good chunk of any such market overvaluation has already been removed by investors.
But let’s play devil’s advocate and assume that the market is still overvalued today. What’s an investor to do?
What To Do When the Market is Overvalued
While investing in an overvalued market might seem scary, knowing that a market is overvalued provides almost no useful information to investors. For example, I remember so many investors who claimed that U.S. stocks were overvalued back in January 2017 when I first started writing. And, by many metrics, they were right.
To get back to the same valuation levels as January 2011, U.S. stock prices would have needed to drop by 20%. But what did they do instead? From January 2017 through February 2022, U.S. stocks prices grew by 13% per year (including dividends and adjusted for inflation).
Now imagine sitting on the sidelines because you knew the U.S. stock market was somewhat overvalued back in 2017. You would still be waiting to get back in five years later.
In fact, even if you had perfectly timed the COVID-19 bottom and bought on March 23, 2020, you still would have purchased U.S. stocks 6.6% above their January 2017 prices (including dividends):
This is why using valuation as a timing tool can be so dangerous. Because valuation is a relative game where the rules (and the players) are always changing. What made sense in one context may not make sense in another. And with everything in constant flux, comparing markets across time can be extremely difficult.
This is why the only options you have when investing in an overvalued market are:
(1) Ignore the overvaluation altogether, or
(2) Make minor adjustments to your portfolio and sin a little
For most of my investment career I followed option (1), but after recent events I did a little of option (2).
Does this mean that U.S. stocks are currently overvalued? It seems very plausible. However, they may stay “overvalued” longer than we anticipate.
After all, a lot has changed since January 2011. Maybe our thinking about valuation needs to change with it. Thank you for reading!
If you liked this post, consider signing up for my newsletter.
This is post 282. Any code I have related to this post can be found here with the same numbering: https://github.com/nmaggiulli/of-dollars-and-data