But someone has to do it!
So here goes:
By many, many historically predictive valuation meassures, stocks are overvalued to the tune of 75%-100%.
In the past, when stocks have been this overvalued, they have often "corrected" by crashing (1929, 1987, 2000, 2007, for example) . They have also sometimes corrected by moving sideways and down for a long, long time (1901-1920, 1966-1982, for example).
After long eras of over-valuation, like the period we have been in since the late 1990s (with the notable exceptions of the lows after the 2000 and 2007 crashes), stocks have also often transitioned into an era of undervaluation, often one that lasts for a decade or more.
In short, stocks are so expensive on historically predictive measures that the annual returns over the next decade are likely to net out to about 0% per year.
How we get there is anyone's guess.
But...
A stock-market crash of ~50% from the peak would not be a surprise. It would also not be the "worst-case scenario," by any means. The "worst-case scenario," which has actually been a common scenario over history, is that stocks would drop by, say 75% peak to trough.
Those are the facts.
Why isn't anyone talking about those facts?
Three reasons:
- First, as mentioned, no one in the financial community likes to hear bad news or to be the bearer of bad news when it comes to stock prices. It's bad for business.
- Second, valuation is nearly useless as a market-timing indicator.
- Third, yes, there is a (probably small) chance that it's "different this time," and all the historically predictive valuation measures are out-dated and no longer predictive.
The third reason is the one that everyone who is bullish about stocks these days is implicitly or explicitly relying on: "It's different this time."
Just so you know, every time there is a long bull market like the one we've had, people come up with lots of reasons to explain why it's different this time. (And understandably so! Everyone wants the bull market to continue, and no one wants to miss further gains.) They did that in the late 1920s. They did that in the late 1990s. They did that in 2007. Usually, however, it isn't different this time, and normality reasserts itself with a vengeance.
As for timing... Unfortunately, even if historical valuation measures are still valid, and stocks are poised to have another lousy decade, today's valuations won't help you predict what the market will do over the next year or two.
And it's almost as painful to miss further market gains by turning cautious too early than it is to get obliterated in a crash. (Almost.)
So that's why almost everyone's still bullish.
But what today's valuations do tell us is that no one should be surprised if stocks crash 50% or more or are still trading around the current level in a decade.
Now you know!
(None of this means that you should sell your stocks, by the way. I own stocks, and I'm not selling. It just means that you should be mentally and financially prepared for a major drawdown. You should also be diversified.)
Here are some of the valuation details...
Stocks are wildly overvalued on historically predictive measures
According to several historically valid measures, stocks are now more expensive than they have been at any time in the past 130 years, with the exception of 1929 and 2000 (and we know what happened in those years).
For example, the chart below is from Yale professor Robert Shiller. It shows the cyclically adjusted price-earnings ratio of the S&P 500 for the past 130 years.
As you can see, the current PE ratio of at least 26 is miles above the long-term average of 15. In fact, it is higher than at any point in the 20th century, with the exception of the months that preceded the two biggest stock-market crashes.
Does a high PE mean the market is going to crash? No. Sometimes, as in 2000, the PE just keeps getting higher for a while. But, eventually, the rubber band snaps back. And in the past - without exception - a PE as high as today's has foreshadowed lousy returns for the next seven to 10 years.
What about other valuation measures? Most of them paint the same picture.
Here, for example, are a few recent charts from Doug Short, one of the best market-chart makers around.
The first chart plots four valuation measures - the Shiller PE ratio above, another PE ratio (different calculation), the "Q ratio" (a measure of price to replacement cost), and a regression analysis for stocks themselves. Same message: Averaging the four suggests that stocks are ~80% overvalued.
The average:
And for good measure, here's another ratio - one that is fondly referred to as "Warren Buffett's favorite valuation measure." (Because he once said it was.)
This one charts the collective value of all stocks to the size of the economy (GDP). It recently hit its second-highest level ever.
You can quibble with any of these measures. But, collectively, they all say the same thing: Stocks are very expensive.
But isn't it 'different this time'?
Every time stocks get expensive, some people argue "it's different this time."
This time, they say, stock valuations like today's are justified, and stock prices will just keep going up.
Usually, however, it's not different.
Eventually, stock prices revert to the mean, usually violently. That's why the words, "it's different this time" are described as the "four most-expensive words in the English language."
But, yes, it's possible that it's "different this time." Sometimes things do change, and investors clinging to old measures miss big gains before they realize their mistake.
It's possible, for example, that Shiller's PE ratio is no longer valid. Shiller's friend, Professor Jeremy Siegel from Wharton, thinks that several things have changed and that stocks are still undervalued.
It certainly seems possible that the future average of Shiller's PE ratio will be significantly higher than it has been in the past 130 years. But it would take a major change indeed for the average PE ratio to shift upward by, say, 50%.
While we're at it, please note something else in the Shiller chart above: Sometimes - as in the entire first 70 years of the past century - PEs (blue line) can be low even when interest rates (red line) are low. That's worth noting, because today you often hear bulls say that today's high PEs are justified by today's low-interest rates.
Even if this were true - even if history did not clearly show that you could have low PEs with low rates - this argument would not protect you from future losses, because today's low rates could eventually regress upward to normal. But it's just not true that low rates always mean high PEs.
And in case some of your bullish friends have convinced you that Shiller's PE analysis is irrelevant, check out the chart below.
It's from fund manager John Hussman. It shows six valuation measures in addition to the Shiller PE that have been highly predictive of future returns. The left scale shows the predicted 10-year return for stocks according to each valuation measure. The colored lines (except green) show the predicted return for each measure at any given time. The green line is the actual return over the 10 years from that point. (It ends 10 years ago.)
Today, the average expected return for the next 10 years is slightly positive - just under 2% a year. That's not horrible. But it's a far cry from the 10% long-term average.
In conclusion ...
Stocks are priced to deliver lousy returns over the next seven to 10 years. I would not be surprised to see the stock market drop sharply from this level, perhaps as much as 50% over a couple of years.
None of this means for sure that the market will crash or that you should sell stocks. (I own stocks, and I'm not selling them.) It does mean, however, that you should be mentally prepared for the possibility of a major pullback and lousy long-term returns.
SEE ALSO:
- BLODGET: Okay, you can call me "Chicken Little." But I still think the sky will fall!
-BLODGET: This boom will end in a bust
-DEAR SILICON VALLEY: Here's your wake-up call