Profitability Works Best When Analysts Agree

In empirical finance, few relationships are as robust and time varying as the link between profitability and returns. This analysis explores how analyst information can improve the signal quality of the profitability factor. The evidence shows that profitability’s predictive power is conditional on analyst disagreement. When analysts broadly agree on fundamentals, profitability predicts returns cleanly. When they do not, the signal fades into noise.

To test this idea, we divided the U.S. stock universe (applying a few straightforward liquidity filters) into quartiles based on analyst forecast dispersion, measured as the standard deviation of analysts’ current-year EPS estimates relative to their mean. In simple terms, it captures how much analysts disagree about a company’s expected earnings, providing an intuitive proxy for informational uncertainty.

There are certainly more sophisticated ways to measure analyst disagreement, but refining the metric is not the focus here. The goal is to see whether conditioning on any reasonable measure of disagreement helps explain when profitability works and when it fails.

We used the bottom quartile (low dispersion) and the top quartile (high dispersion) from January 2005 through October 2025, and within each group ranked stocks by profitability (ROA), defined as operating profits over book equity, into 20 portfolios.

Low Dispersion: Where Profitability Shines

In the bottom quartile of analyst dispersion, the profitability factor behaves exactly as expected. Returns rise almost monotonically across the 20 profitability portfolios, and the top bucket compounds at 12.22% annualized, outperforming both the 8.9% universe and the 10.6% benchmark.

Equally interesting is the spread between the top and bottom portfolios. The relationship is not only strong on the long side but also informative on the short side, the lowest-profitability bucket delivers a -9.73% annualized return. In other words, profitability does not just identify winners, it effectively flags the market’s weakest performers.

Once we’ve examined the behavior of the factor across sorted buckets, we can move to a more practical application: building a tradable long/short portfolio. The long/short strategy goes long the 50 most profitable stocks and short the 50 least profitable, based on the factor rankings. The portfolio is rebalanced monthly to reflect updated signals. Transaction costs are not included, so results should be viewed as illustrative rather than investable performance.

Even under these simplified assumptions, the strategy produces a stable premium over time. As a standalone long/short portfolio, it earned an annualized return of 15.8% versus 10.6% for the market (VTI), with a Sharpe ratio of 0.81 and an alpha of 19.5%. Despite drawdowns similar to the market (–50% vs. –55%), the profitability strategy’s compounding advantage more than doubled total returns over the 20-year period.

This analysis is intended to highlight the economic intuition and behavior of the factor rather than represent a fully implementable strategy. Future research notes will focus on design and execution aspects under more realistic trading conditions.

High Dispersion: When Noise Dominates

In the top 75 percent dispersion group, the pattern deteriorates noticeably. The profitability gradient flattens, and the spread between high and low portfolios weakens substantially. The high-profit portfolio’s annualized return falls to 5.36 percent, underperforming both the benchmark and its low-dispersion counterpart.

This pattern is consistent with the argument of Miller (1977), who showed that when investors hold divergent beliefs and short-selling is limited, market prices tend to reflect the optimism of the most bullish investors rather than the consensus view. In such settings, pessimists are effectively sidelined, and disagreement pushes valuations upward, leading to weaker subsequent returns.

When uncertainty dominates, profitability ceases to signal expected cash flows and instead reflects belief heterogeneity, consistent with the empirical findings of Diether, Malloy, and Scherbina (2002).

Why disagreement matters

Analyst dispersion captures disagreement about fundamentals, a reflection of uncertainty, asymmetric information, and behavioral noise.

Miller (1977) famously argued that when investors hold divergent beliefs and short-selling is constrained, prices tend to reflect the optimism of the most bullish investors. The result is systematic overpricing and lower future returns.

Diether, Malloy, and Scherbina (2002) later confirmed this empirically, showing that stocks with greater dispersion in earnings forecasts earn significantly lower subsequent returns. High dispersion signals uncertainty about future cash flows, which weakens the link between profitability and expected returns.

When dispersion is low, however, market participants broadly agree on the earnings process. In such an environment, profitability becomes a credible signal of economic productivity, consistent with the framework of Fama and French (2015). Investors can more easily identify and price high-quality firms, leading to a reliable and persistent profitability premium.

Hong and Stein (2007) further explain that disagreement, combined with limits to arbitrage, prevents pessimistic investors from correcting overpricing. The result is a market driven by belief heterogeneity rather than fundamentals — precisely the environment where the profitability factor loses its edge.

Recent Underperformance and Market Context

Notably, the long–short profitability strategy has struggled since mid-August 2025, though the bulk of the drawdown traces back to the sharp market selloff in April. The pattern coincides with broader weakness across equity long/short hedge funds. Over the past quarter, the space has experienced significant factor crowding, deleveraging, and position unwinds, amplifying short-term volatility in otherwise stable fundamental signals.

Our interpretation is not that the profitability factor is broken, but that it is temporarily suffering from correlated de-risking within the long/short equity space, a recurring phenomenon during liquidity-driven rotations.

Curious to dig deeper? If you’d like to discuss this research further or explore potential collaborations, feel free to reach out at [email protected]

Visit our website here: www.noaxcapital.com

References