Value vs Growth: 50 Years of Factor Data Decoded

Key Takeaway

Over the past half-century, value stocks have outperformed growth stocks by roughly 3 to 5 percentage points per year on average. Yet this premium is far from constant. It vanishes for years at a time, occasionally reverses violently, and has been declared dead at least three times since Fama and French first documented it in 1992. Understanding what drives the cycle between value and growth is essential for any investor making factor allocation decisions.

The Academic Foundation

The value premium entered the academic canon with Fama and French (1992), who showed that stocks with high book-to-market ratios (value) earned substantially higher returns than stocks with low book-to-market ratios (growth). Their follow-up paper in 1993 formalized this into the three-factor model, with HML (High Minus Low) capturing the value premium alongside market risk and the size factor.

The finding was not entirely new. Benjamin Graham had advocated buying cheap stocks decades earlier. But Fama and French provided the first rigorous, large-sample evidence that value outperformance was systematic, persistent, and not explained by market beta alone.

Lakonishok, Shleifer, and Vishny (1994) offered a behavioral interpretation. They argued that investors systematically overextrapolate past growth rates, bidding up glamour stocks to unjustifiable valuations while neglecting firms with poor recent performance. When these expectations mean-revert, value stocks earn excess returns. Their evidence showed that contrarian strategies earned roughly 10 to 11 percentage points per year more than glamour strategies, even after controlling for risk.

The debate over whether the value premium represents compensation for risk (the Fama-French view) or exploitation of behavioral errors (the Lakonishok et al. view) has never been fully resolved. In practice, both mechanisms likely contribute.

50 Years of Data: Decade by Decade

The HML factor, measured as the return spread between the top 30% of stocks ranked by book-to-market and the bottom 30%, tells a story of long cycles rather than steady compounding.

HML Factor Returns by Decade

The full-sample average conceals enormous variation. The 1980s and 2000s were golden decades for value investors, while the 2010s delivered the worst sustained value drawdown in recorded history. Understanding why requires examining the macro and market forces that drive the cycle.

Cumulative Performance: Two Divergent Paths

A dollar invested in U.S. value stocks (top book-to-market quintile) in January 1975 grew to approximately $185 by December 2025, compared to roughly $65 for growth stocks (bottom book-to-market quintile). Both figures are nominal, before transaction costs.

Cumulative Growth of $1 (1975-2025)

The late 1990s and the 2010s represent two episodes where growth overtook value in cumulative terms. Both periods share a common feature: they were dominated by a narrow group of mega-cap technology and growth stocks that attracted concentrated capital flows.

What Drives the Cycle?

Four primary forces drive the rotation between value and growth dominance.

Interest Rates and Duration

Growth stocks behave like long-duration assets. Their value derives primarily from earnings expected far in the future, making them highly sensitive to discount rates. When interest rates fall, the present value of distant cash flows rises dramatically, benefiting growth stocks disproportionately.

Value stocks, by contrast, tend to have higher current earnings yields and shorter implied duration. They are less sensitive to rate changes. The relationship is not mechanical, but the broad pattern is clear: falling rates favor growth, and rising rates favor value.

The period from 2010 to 2020 saw the 10-year Treasury yield decline from roughly 3.8% to below 1.0%, creating a powerful tailwind for growth stocks and a headwind for value. The rate hiking cycle that began in 2022 reversed this dynamic and contributed to a significant value recovery.

Investor Sentiment and Extrapolation Bias

Lakonishok, Shleifer, and Vishny (1994) documented that investors systematically overestimate the persistence of past growth rates. During periods of rapid technological change or economic optimism, this tendency intensifies. Growth stocks receive increasingly stretched valuations while value stocks are neglected.

The dot-com bubble of 1997-2000 is the textbook example. Growth stocks returned 33% annualized while value stocks earned only 5%. When the bubble burst, value dramatically outperformed for the next seven years.

Sector Composition Shifts

The composition of value and growth indices has changed substantially over time, which matters for interpreting historical returns.

This concentration means that the value vs. growth debate has increasingly become a debate about technology sector valuations. When tech stocks rise on genuine innovation (cloud computing, AI), growth indices benefit mechanically. When tech valuations compress, value indices outperform partly because they have minimal tech exposure.

Economic Regime Dependence

Value stocks tend to outperform during economic recoveries and periods of above-trend growth, when cyclical sectors and beaten-down companies experience earnings mean-reversion. Growth stocks tend to outperform during economic slowdowns when scarce growth commands a premium.

The strongest value returns have historically occurred in the 12 to 24 months following a recession, as beaten-down cyclical and financial stocks recover.

The Value Drought: 2017-2020

The period from mid-2017 to late 2020 represents the deepest and most sustained drawdown for the HML factor since data begins. Value underperformed growth by roughly 40 percentage points cumulatively over this stretch.

Several forces converged. Ultra-low interest rates reduced the discount rate applied to growth company cash flows. The dominance of FAANG (Facebook, Apple, Amazon, Netflix, Google) and later mega-cap tech concentrated index returns among a handful of growth stocks. Quantitative easing compressed risk premia across markets, reducing the mean-reversion mechanism that traditionally benefited value.

The COVID-19 pandemic accelerated these trends, as lockdowns boosted digital economy companies (growth) while devastating traditional economy companies (value). By September 2020, the valuation spread between the cheapest and most expensive stocks reached levels not seen since the dot-com peak.

The Recovery: 2021-2025

The value recovery began in late 2020 with the announcement of effective COVID vaccines, which triggered a rotation from pandemic winners into economically sensitive stocks. This recovery intensified as inflation rose and the Federal Reserve began raising interest rates in March 2022.

From January 2021 through December 2025, the HML factor delivered approximately 4.2% annualized returns. This recovery was driven by several reinforcing factors.

Rising rates. The Federal Funds rate rose from near zero to above 5%, compressing growth stock valuations while increasing the earnings of financial sector value stocks.

Valuation mean-reversion. The extreme valuation spread of late 2020 created a mechanical tailwind for value as the cheapest stocks repriced upward.

Energy sector strength. The value index had significant exposure to energy companies, which benefited from the commodity price surge of 2021-2022.

Tech multiple compression. High-growth technology companies that had traded at 30 to 50 times sales saw multiples contract to 10 to 20 times, dragging growth index returns lower.

International Evidence

The value premium is not unique to U.S. markets. Data from developed international markets confirms the pattern, although the magnitude varies.

Japan stands out with the strongest value premium, partly reflecting the persistent undervaluation of Japanese companies during the post-bubble decades. European value has also been relatively strong, supported by a financial sector heavy composition.

The correlation of value returns across regions is moderate (0.4 to 0.6), which means that a globally diversified value strategy has experienced smaller drawdowns than a U.S.-only approach.

Risk-Adjusted Comparison

The value premium is often described as compensation for risk. If so, what additional risks do value investors bear?

Value stocks have modestly higher volatility and larger maximum drawdowns, consistent with the risk-based interpretation. However, the Sharpe ratio improvement is meaningful: 0.52 versus 0.38 represents a substantial edge over half a century. The HML factor itself has low market beta (0.10), confirming that the value premium is largely orthogonal to broad market direction.

Is the Value Premium Dying?

The value drought of 2017-2020 revived a recurring question: has the value premium been arbitraged away? Asness et al. (2000) addressed this question early, examining whether the documented premium was likely to persist.

Arguments that the premium has diminished include widespread awareness since Fama and French (1992), the growth of value-oriented ETFs and smart-beta products that may have compressed the spread, and structural changes in the economy that favor asset-light growth companies over asset-heavy value companies.

Arguments that the premium persists include the behavioral mechanisms (extrapolation bias, overreaction to bad news) that are deeply embedded in human psychology, the career risk of holding ugly, out-of-favor stocks that limits institutional adoption of deep value, the cyclical nature of the premium which guarantees extended periods of underperformance that shake out impatient capital, and the 2021-2025 recovery that demonstrated the premium had been dormant rather than dead.

The most balanced interpretation is that the unconditional value premium has likely compressed from its historical average of 4 to 5 percent to something closer to 2 to 3 percent, but that the conditional premium (buying value when spreads are wide) remains substantial.

Timing Value vs. Growth

Research by Asness et al. (2000) found that the valuation spread between value and growth stocks has modest predictive power for subsequent relative returns. When the spread is unusually wide (as in late 2000 or late 2020), value tends to outperform over the following three to five years. When the spread is narrow, future value returns are lower.

This is not a trading signal. The timing is imprecise, and the spread can widen further before reverting. But it does suggest a rebalancing discipline: maintain a strategic value tilt at all times, and increase it when valuation spreads are historically wide.

As of early 2026, the value-growth spread sits near its long-run median, suggesting neither a strong tactical signal in either direction nor the extreme conditions that preceded the 2000-2007 or 2021-2025 value rallies.

Practical Considerations

For investors deciding between value and growth allocations, the evidence supports several conclusions.

The long-term edge favors value. Over 50 years, the value premium has been positive, statistically significant, and present across geographies. Ignoring value means accepting lower expected risk-adjusted returns.

Patience is required. Value can underperform for 3 to 5 years at a stretch, as the 2017-2020 period demonstrated. Investors who abandon value after a drawdown typically miss the subsequent recovery.

Diversification across styles reduces regret. A 60/40 or 50/50 blend of value and growth captures most of the value premium while avoiding the worst periods of relative underperformance.

Global diversification helps. Because value cycles are imperfectly correlated across regions, a globally diversified value tilt experiences smaller drawdowns than a single-country approach.

Monitor the valuation spread. While not a timing tool, the spread between value and growth valuations provides useful context for portfolio construction decisions.

The Long View

The debate between value and growth investing is as old as financial markets. Graham advocated for value in the 1930s; Fisher championed growth in the 1950s. Half a century of systematic data confirms that both styles generate positive long-term returns, but that value has maintained a persistent, if cyclical, edge.

The mechanism behind this edge is likely a combination of rational risk compensation and behavioral mispricing. Value stocks are genuinely riskier during economic downturns, when their cyclical earnings and higher leverage amplify losses. But investors also systematically overpay for growth, extrapolating recent trends and underweighting the base rate of mean-reversion in corporate profitability.

For quantitative investors, the value factor remains one of the most robust and well-documented sources of excess return. The key is recognizing that the premium operates on a business cycle frequency, not a quarterly one. The investors who capture the value premium are those who can maintain exposure through the inevitable periods when value looks like it will never work again.