Value vs Momentum: Cliff Asness on the Trade-off

Cliff Asness, founder of AQR Capital Management, has spent decades investigating a puzzle in factor investing: value and momentum factors are negatively correlated over long periods, yet both have earned premiums historically. Value outperforms in recoveries from deep recessions; momentum dominates in trending markets. The obvious question: why hold both if they pull in opposite directions? Asness’s answer: the asymmetric payoff and diversification benefit justify the drag in some periods.

The Negative Correlation Puzzle

Over the past century, both value investing and momentum investing have delivered excess returns—premiums above what basic market-risk models would predict. Yet the two factors move in opposite directions much of the time.

Value works because investors overpay for growth and popularity (rich stocks get richer in price), creating a subsequent correction. Meanwhile, stocks that have been beaten down have overshot on the downside. When reality catches up to pessimism, they rebound. This payoff peaks in recovery phases: recessions flatten momentum and compress valuations, then rebounds favor depressed value stocks as optimism returns.

Momentum works because markets trend. Stocks that have done well continue to outperform for quarters or years, not because fundamentals strengthen monotonically, but because the narrative persists and investors chase. Momentum dominates in bull markets and goldilocks regimes—when growth is steady, fear is moderate, and trend-following trades flow. It also thrives in crises (panic flight to quality) where high-quality winners keep winning.

The correlation between them turns negative because value’s moment of glory—the recovery—is often the worst time for momentum. The recovery favors cheap, lagging stocks, exactly what momentum investors have shunned. Conversely, when momentum peaks (late in a bull market), value is worst-off because the most expensive, most-beaten-down stocks have already lagged for years.

Why Hold Both Despite the Tension?

Asness’s core argument rests on two pillars: diversification and asymmetric payoffs.

Diversification: If you hold only value, you capture its premium in recovery periods but suffer through trend-driven bull markets and momentum-led rallies. A portfolio that holds both value and momentum smooths returns across market regimes. You sacrifice some upside in the 40% of periods when only one factor shines, but you reduce volatility by holding an uncorrelated (or negatively correlated) hedge in the other.

The mathematics are stark. A barbell of 50% value + 50% momentum has lower total variance than 100% of either factor alone, even though they move opposite. Asness has documented this repeatedly: the diversification benefit of holding conflicting factors outweighs the drag from drawdowns in factor-specific environments.

Asymmetric payoffs: Value and momentum have different return distributions. Momentum delivers frequent, moderate wins—beating the market by 1–3% in many quarters. It compounds steadily. But it has catastrophic tail risks: when a trend reverses hard (e.g., 2000, 2008, 2020 for growth-heavy momentum), drawdowns can exceed −30%.

Value delivers less frequent wins. Many years, value lags badly. But its tail risk is rightward-skewed: when recessions hit and fear peaks, value has spectacular reversals. A value portfolio down −20% in a 2-year bear market often rebounds +40% in the recovery year, clawing back losses and then some.

Asness argues that this asymmetry—infrequent huge wins for value, frequent modest wins for momentum—justifies holding both. You capture momentum’s steady compounding and value’s rare tail recoveries. Neither alone offers the full distribution of opportunity.

The Factor Timing Temptation and Its Danger

A natural follow-up: if value and momentum are negatively correlated, why not time them? Hold value when momentum is set to underperform, then switch to momentum when value is set to lag?

Asness is deeply skeptical of factor timing. Predicting which factor will outperform next is nearly as hard as predicting which stocks will outperform. The costs of switching—taxes, transaction fees, market impact, and the very real risk of switching at precisely the wrong moment—typically exceed any benefit.

Instead, Asness advocates a static allocation to both factors, rebalancing periodically. If you allocate 60% to value and 40% to momentum and hold for years, you reduce the psychological and operational burden of timing. You also lock in the rebalancing benefit: when value lags and falls below 60%, you buy it at lower prices. When momentum surges above 40%, you trim it after it has outperformed.

This is not a buy-and-forget strategy, but a disciplined, infrequent rebalancing approach that avoids the timing trap.

Empirical Evidence and The “Cliff’s” AQR Approach

AQR’s research—published in academic papers and popular books by Asness—has shown:

1. Both premiums are real and significant over long periods (decades to centuries). Value and momentum are not statistical artifacts.

2. Negative correlation is persistent. Across countries, asset classes (stocks, bonds, currencies), and time periods, value and momentum move opposite at correlation levels ranging from −0.2 to −0.6.

3. Diversification benefit is material. A 50/50 or 60/40 portfolio of value and momentum has lower Sharpe ratio volatility than concentration in either factor, even accounting for the cost of rebalancing.

4. Timing is unreliable. Simple rules (e.g., “switch to value when value is cheap relative to momentum”) backtest well but fail in live trading, likely due to transaction costs and overfitting.

AQR’s practical solution: build multi-factor portfolios that hold multiple premiums (value, momentum, carry, quality, low volatility) with disciplined weights and rebalancing. The idea is that no single factor leads all the time, but a diversified collection captures multiple premium sources and reduces the drag of any single factor’s drawdown periods.

The Skeptical Pushback

Critics argue that Asness’s justification for holding conflicting factors amounts to “paying for diversification you don’t need.” If value and momentum have negative correlation, standard portfolio theory says mixing them reduces variance—but it also reduces expected return unless the factors have sufficiently high individual premiums relative to their correlation drag.

In periods of extended momentum dominance (like 2015–2020, when growth stocks soared), a value anchor dragged returns noticeably. Investors who trusted Asness’s multi-factor approach suffered relative underperformance for years. Some asked: was the diversification benefit worth it?

Asness’s reply: yes, if your time horizon is decades. The answer depends on your beliefs about both the magnitude of each premium and your tolerance for single-factor drawdowns. If you can only stomach a 10% drawdown, holding only momentum in 2000 or 2008 is ruinous. The value hedge, even if it lags in bull markets, caps the damage.

Extensions Beyond Stocks

Asness’s logic extends to other assets. Corporate bonds have value (higher spreads = better protection) and momentum (recent outperformers trend). Currencies have value (interest-rate differentials) and momentum. A globally diversified portfolio that blends value and momentum across asset classes captures the benefit of the tension without being hostage to one factor’s regime.

Some of AQR’s most interesting work explores how value and momentum interact with macroeconomic regimes: inflation, real growth, risk-on/risk-off. The takeaway is that the tension between the two isn’t random—it’s systematically related to economic cycles. Understanding those cycles helps explain when each factor shines.

The Human Element

Beyond the math, Asness emphasizes that holding both value and momentum requires emotional discipline. When momentum dominates and value lags for 3+ years (as happened in 2015–2019), most investors abandon the value allocation. They switch to a pure momentum portfolio at the peak, only to suffer when momentum crashes. Asness argues that pre-commitment to a balanced approach—and trusting the diversification argument—prevents this emotional blunder.

He often cites the parallel to market timing: most investors know buying the market bottom and selling the top is impossible, yet they try anyway, with predictably poor results. Similarly, most investors know diversification works, yet they concentrate in the factor that has recently dominated, and suffer for it. Discipline and humility are features, not bugs.

See also