A sweeping 40-year meta-analysis finds that growth stocks' long-run premium is far more concentrated, and fragile, than most investors assume. The implications for portfolio construction in 2026 are hard to ignore.
Key Takeaways
For a generation of portfolio managers trained on the post-2008 bull market, the debate between value and growth investing has felt settled. Growth won, spectacularly. But a landmark meta-analysis published this month by the Meridian Institute for Capital Research is complicating that narrative in ways that demand attention from anyone managing a long-duration portfolio in 2026.
The study, which synthesised return data across U.S., European, and Asia-Pacific equity markets from 1984 to 2024, found that the growth premium, while real, is startlingly concentrated in time. Specifically, 73% of the cumulative excess return attributable to growth stocks over value stocks occurred in just 11% of all trading months. Strip out those narrow windows of outperformance and the 40-year advantage evaporates entirely.
"Most investors have been drawing the wrong lesson from the data," said Dr. Caroline Fitch, lead author of the study and a former quantitative strategist at Vanguard. "They see growth's long-run track record and assume it's a durable, reliable premium. It isn't. It's episodic, it's rate-dependent, and it clusters in very specific macro environments." The research identified three distinct macro conditions that reliably ignite growth outperformance: falling real rates, expanding corporate profit margins, and elevated investor risk appetite, a combination that described much of the 2010–2021 period.
Since 2022, all three conditions have been in reverse. Real rates are meaningfully positive for the first time since the mid-2000s. Profit margins across the S&P 500 have contracted from their 2021 peak of 13.2% to 10.8% as of Q1 2026. And while risk sentiment has recovered from the 2022 trough, it has not reached the speculative extremes that historically catalysed growth's best years.
The Meridian paper includes a detailed analysis of historical value-growth rotation cycles, defined as sustained periods of at least 18 months in which one style meaningfully outperforms the other. The researchers identified seven complete cycles since 1984, with an average duration of 6.8 years. The current value cycle, which began in earnest in late 2021, is now roughly four years old. If historical patterns hold, that suggests we may still have two to three years of value tailwinds remaining, though Fitch is careful to note that cycle lengths are highly variable and macro shocks can truncate or extend them.
Marcus Albright, chief investment officer at Calloway Capital Management, finds the rotation data compelling but counsels against binary positioning. "The mistake investors make is treating this as an either-or decision," he said. "The data is very clear that factor-blended approaches, owning quality value alongside selective growth at reasonable prices, have outperformed pure-style portfolios in 14 of the last 20 calendar years. The outlier years where pure growth crushed everything tend to anchor people's memories."
The practical question for portfolio managers is how to translate this research into positioning. The Meridian team's back-tested optimal allocation for the current macro environment, positive real rates, normalising margins, moderate risk appetite, is a 55/45 blend tilted toward value, with an emphasis on free cash flow yield rather than traditional price-to-book metrics. That distinction matters: book value has become an increasingly unreliable valuation anchor as intangible assets have come to dominate corporate balance sheets.
"The growth premium is real, but it's a once-in-a-decade trade, not a set-and-forget allocation. Investors who confuse the two are going to find the next decade very sobering." , Dr. Caroline Fitch, Meridian Institute for Capital Research
The research also resurfaces the ongoing debate about factor investing frameworks. Smart-beta strategies that systematically harvest the value factor have attracted significant flows over the past three years, but the Meridian data suggests that naive factor tilts underperform dynamic allocation approaches that adjust exposure based on prevailing macro conditions. The implication is that investors need more than a static tilt, they need a framework for knowing when to lean in and when to fade.
For long-term institutional investors, the study's most sobering finding may be this: if the next decade looks anything like the historical periods that followed prolonged growth dominance cycles, which the researchers identify as 1972–1977 and 1999–2006, then value could outperform growth by 40 to 60 percentage points cumulatively before the next rotation begins. That is not a prediction. But it is a data point that should inform every strategic asset allocation review happening in 2026.
Passive investing dominance reaches a new milestone as ETF assets surpass active mutual fund AUM for the first time in history.
An analysis of 12,000 earnings events reveals that markets punish minor EPS misses far more severely than the fundamentals warrant.
A survey of 600 finance executives finds AI is reshaping capital allocation decisions at a pace that few organisations were prepared for.
