The lost decade thesis has become one of the most weaponized market narratives in circulation. If you’ve read coverage on InvestorPlace or any of the financial news sites that picked up this story, you’ve seen the raw numbers paraded as certainty. Goldman Sachs says 3%. Bank of America says negative. Apollo says flat. The next ten years, we’re told, are a wash for U.S. stocks.
But what did these banks actually publish? Not what the promos and headlines say they published. The actual research.
The answer is more nuanced than the shorthand. And for investors trying to decide whether to shift their strategy, that nuance is the only thing that matters.
What Goldman actually put on paper
Goldman Sachs published three distinct forecasts on this topic over an 18-month window. They are not the same forecast. The gap between them tells you more than any single number.
In October 2024, David Kostin — Goldman’s chief U.S. equity strategist at the time — projected the S&P 500 would return roughly 3% annualized over the next decade. This was his “base case” and it was deliberately stark. Kostin argued the S&P 500 would rank in the 7th percentile of all 10-year returns since 1930. That is a deeply troubling number. It also came before the market ripped higher in 2025.
Thirteen months later, in November 2025, Goldman’s chief global equity strategist Peter Oppenheimer published the firm’s official 10-year outlook. His number: 6.5% annualized total returns. Oppenheimer’s base case breaks down as 6% annualized earnings growth, a 1% annual drag from valuation compression, and a 1.4% dividend yield. He also ranked U.S. equities dead last among all major global markets — behind emerging markets at 10.9%, Asia ex-Japan at 10.3%, Japan at 8.2%, and Europe at 7.1%.
That global ranking is the part the promos love. What they leave out is that Oppenheimer was careful to frame this as a relative call. U.S. stocks are expensive compared to the rest of the world — not necessarily overvalued into oblivion.
Then, in June 2025, Goldman’s new chief U.S. equity strategist Ben Snider revised the forecast upward to 7% annualized. Snider’s thesis was that structurally low interest rates and elevated profit margins (13% today versus 5.5% in 1980) may justify a permanent valuation premium. He argued that high valuations can persist longer than the models predict.
That’s a 4 percentage point spread across three Goldman forecasts over 18 months. Kostin at 3%. Oppenheimer at 6.5%. Snider at 7%. All below the historical 10% average. All within the same firm. If Goldman itself can’t agree on the number, the certainty you see in the promotional copy is a fiction.
The rest of Wall Street’s view
Bank of America went further in December 2025, projecting the S&P 500 would shed 0.1% over the next ten years. Not annualized. Total. A decade that ends with the index lower than it started. That is the most bearish major bank forecast on record.
Apollo’s chief economist Torsten Sløk also called for a flat decade. Sløk argued that elevated valuations, elevated rates relative to the 2010s, and reduced corporate buyback capacity would combine to produce near-zero returns for passive equity investors.
Neither BofA nor Apollo expects a crash. They expect a grind. Years of single-digit earnings growth getting consumed by multiple compression as the market normalizes from a CAPE ratio near the 99th percentile. No fireworks. No panic. Just returns that barely keep pace with inflation.
The CAPE ratio is doing all the work here
Every lost decade forecast ultimately traces back to one number: the cyclically adjusted price-to-earnings ratio developed by Robert Shiller. The CAPE takes the S&P 500’s inflation-adjusted price and divides it by the average of its real earnings over the prior decade. By smoothing out the business cycle, it gives a steadier read on valuation than the standard one-year P/E.
As of July 2026, the S&P 500’s CAPE ratio sits at 41.37. That is the 99th percentile of the full dataset going back to 1881. The historical average is 17.4. The median is 16.1. Today’s reading is 138% above that average.
The relationship between CAPE and subsequent 10-year returns is not a perfect predictor. It never has been. But it is not meaningless either. Every time the CAPE has entered these altitudes — above 35 — subsequent decade returns have been meaningfully below average. Not always negative. But consistently in the 0% to 6% range.
High CAPE alone doesn’t predict a crash. It predicts disappointing returns. There is a difference, and it matters.
The historical precedent matters more than the math
The U.S. has produced two lost decades in the modern era. Both followed setups remarkably similar to today’s.
The Nifty Fifty period of the late 1960s and early 1970s produced a cluster of high-quality growth stocks that investors deemed unbreakable. Xerox, Polaroid, Avon, Coca-Cola — all trading at extreme multiples. The top cohort hit 42 times forward earnings. From December 1964 to December 1974, the largest stocks returned negative 0.44% annualized. The Dow crossed 1,000 in 1972 and didn’t reclaim it for a decade.
The dot-com bubble produced the same outcome with different actors. From March 2000 to March 2010, the S&P 500 returned negative 0.99% annualized. The top ten tech stocks lost an average of 74% peak to trough. The index took 13 years to regain its 2000 high.
Both lost decades had one structural feature in common: extreme concentration at the top of the index. In 1972, the five largest stocks made up about 25% of the S&P 500. In 2000, the top seven tech names represented 22% to 24%. Today, the Magnificent Seven make up roughly 34% of the index — the highest level in modern history. The CFA Institute published research in 2025 demonstrating that after periods of extreme concentration, the largest-cap stocks systematically underperform over 1, 3, 5, and 10-year horizons. The mechanism isn’t a mystery. When one cohort of stocks dominates the index, the cap-weighted index becomes a bet on that cohort. If those stocks revert, the index goes nowhere.
Where the bull case punches back
This is the part that tends to get lost in the promo coverage.
The 2000 tech bubble had top stocks trading at 56 times forward earnings. The Mag 7 trade at roughly 28 times. That is expensive but not insane. Their earnings quality is higher. Their balance sheets are stronger. Their businesses are actual cash-generating machines with pricing power, not speculative stories about the internet changing everything.
Ben Snider’s argument at Goldman — that structurally lower rates and permanently higher profit margins justify a higher CAPE — is not obviously wrong. If the equilibrium CAPE has moved from 17 to, say, 30, then the market is only 38% overvalued instead of 138%. That changes the return math substantially.
Profit margins at 13% versus 5.5% in 1980 are a real difference. Corporate America is less capital-intensive, more global, and more profitable than it was four decades ago. If those margins persist, earnings growth alone can deliver 5% to 6% annualized even without multiple expansion.
The bull case is not weak. It is incomplete. It assumes that margins stay elevated, that interest rates stay structurally low-ish, and that the Mag 7 maintain their competitive advantages without regulatory disruption. Each of those assumptions is plausible. None is guaranteed.
What the lost decade thesis actually means
The lost decade narrative has been distilled into a promotional tool. It appears in ads for market timing services, for active management pitches, and for anyone selling an alternative to buy and hold. The framing is always the same: passive investing is broken, here is our solution.
That framing obscures the real texture of the research.
Goldman Sachs projects 3% to 7% annualized returns over the next decade, depending on which strategist you ask. Bank of America projects a flat decade. Apollo projects a flat decade. These are all below the 10% historical average. But they are not zero in most cases. The range matters.
A 5% annualized return over ten years turns $10,000 into $16,289. That is not exciting. It is also not a disaster. The problem is not that you lose money. The problem is that it forces a different planning assumption than the 10% backtest that underpins most retirement calculators.
For younger investors — people with 20-plus-year horizons — this is noise. A below-average decade in the middle of a career is a footnote in a lifetime of compounding. For someone five years from retirement, it is a material planning risk.
The lost decade thesis is not a prediction. It is a probability-weighted scenario analysis produced by the largest sell-side institutions on Wall Street. Those institutions have been wrong before — systemically too bearish in 2010, too bearish in 2016, too bearish in 2020. They tend to miss structural bull markets because their models are built on mean reversion, and mean reversion has been delayed repeatedly by monetary policy.
But being wrong about timing does not make the underlying math wrong. The CAPE is at the 99th percentile. Concentration is at an all-time high. The historical parallels are real even if the outcome is not predetermined.
What separates useful analysis from promotional noise is honesty about the range of outcomes. The promos tell you the lost decade is certain and buy and hold is dead. The banks’ actual research tells you returns are likely below average, with a wide confidence interval, and that the range includes everything from flat to 7% depending on assumptions.
The difference between certainty and probability is the difference between marketing and analysis. If you are reading coverage of this thesis, check whether the source treats it as a forecast or a scenario. That one distinction will tell you more than any single percentage point. The future is not written. But it is constrained by the math of today’s valuations. Smart portfolios respect those constraints without being paralyzed by them.