For most investors, the forecasting power of valuation metrics is pure gospel. With the price-to-earnings ratio of US stocks hovering near 30, many are starting to taking it on faith that these are the last days of the bull market.
Yet valuations don’t predict stocks’ direction – and they never have. Heresy? More like history – more than a century of it – proving that today’s lofty PEs mean little.
This isn’t my 2026 forecast (that will come in late December or January), but valuations won’t sway it. Pundits tout many valuation measures for timing. In addition to PE, there’s dividend yield, price-to-book and price-to-sales ratio (the latter I created more than 40 years ago).
The logic: Low valuations indicate “cheap” stocks – so “buy low.” Lofty ones supposedly signal froth – foretelling meager returns or worse. Better “sell high.” Hence the current fear.
Acrophobia – not facts – underpins this. It was hardwired into us when our ancient ancestors roamed the wild. Big falls can risk injury or death, so avoid heights – good survival advice. Bad investing practice.
Consider the S&P 500 and a complex but top-drawer statistic called R-squared, which shows how much one recurring phenomenon explains of another. Readings of zero mean A never causes B, while 1.00 signals A always causes B.
Since 1872, the S&P 500’s annual starting PE (using trailing 12-month earnings) and forward 1-year returns have an R-squared of 0.01—essentially no causality! R-squared using three- and five-year returns is 0.03 and 0.02. That means just 3% and 2% of US stocks’ forward three- and five-year returns, respectively, even possibly stem from PEs. Randomness. Believing the reverse is just dumbness.
Why? First, valuation metrics are widely known – hence already widely priced into stocks. Also, stock prices look forward while earnings look backward. Even projected earnings, used by many, stem from current expectations, already baked into sentiment and prices (and often wrong).
This can make P/Es look nosebleed high at great buying opportunities. Recall early 2009, when stocks soared, anticipating recovery from the financial crisis. But recession-decimated earnings didn’t yet reflect it. The S&P 500’s Dec. 31, 2007 PE of 17 soared to 60 by 2009’s start. Overpriced? No…a generational buying opportunity.
Yes, bad returns can follow high PEs. At 2000’s start the S&P 500’s PE topped 30 and annualized -2.3% for the next five years. It began 2022 with a PE near 30 and slid -18.1% that year.
But high PEs preceded great returns, too. 2003 started with a 32 PE. US stocks leapt 29%, annualizing 12.8% the next five years. 2021 launched with a 38 PE. Yet stocks returned 29% that year, annualizing 10% returns over three years … despite 2022’s bear market.
US PEs were high throughout most of 2009-2025. Yet stocks soared 944% through October – even with two bear markets in route.
What of the famous “Cyclically Adjusted PE (CAPE)?” It aims to mute cyclical factors by averaging a decade’s inflation-adjusted earnings. It wasn’t designed to time stocks, but many use it that way anyway.
Some cherry pick intervals where it seemingly worked – its R-squared with 10-year future returns approaches 0.30. (Though this still means CAPE can’t explain 70% of 10-year returns.) But it fizzles down to randomness levels over shorter—and longer—periods. Why tie your retirement to random-like metrics?
Valuations aren’t entirely useless. They can help pick stocks within value categories (what I created price-to-sales ratios for). But for broad market predictions? No.
Faith in valuations’ predictive powers sets you on a road to market PErdition. Be a valuation heretic.
Ken Fisher is the founder and executive chairman of Fisher Investments, a four-time New York Times bestselling author, and regular columnist in 21 countries globally.
This story originally appeared on NYPost
