Cross-sectional pricing is the comparison of asset valuations across many securities at a single point in time, rather than tracking one asset through time. It asks which firms or assets are rich or cheap relative to peers given shared risk factors, revealing dispersion, mispricing, and the premia the market assigns to common characteristics.
How it works
Where time-series analysis tracks one asset's path, the cross-section holds time fixed and ranks the universe by valuation metrics (earnings yield, multiples, factor loadings). Asset-pricing models then test whether differences in expected returns map onto exposures to common factors; residual gaps flag relative mispricing or an embedded premium not yet arbitraged away.
Why it matters now
In the 2025-2026 AI-concentration regime, the debate is whether mega-cap dominance is a genuine repricing of fundamentals or a concentration premium. If that premium is real it should show up cross-sectionally — in wide dispersion of multiples across firms — and its absence there is itself a signal worth watching.
Example
As of late 2025 the Magnificent Seven traded at forward multiples far above the median S&P 500 constituent, yet if the broad cross-section of P/E ratios shows little widening between AI-exposed and non-exposed names outside the mega-caps, the supposed AI concentration premium is concentrated rather than diffused — visible in the index level but not yet in cross-sectional pricing across the broader market.
Frequently asked
- What is cross-sectional pricing?
- Cross-sectional pricing compares asset valuations across many securities at a single point in time, asking which names are rich or cheap relative to peers. It holds time fixed and ranks the universe by metrics like earnings yield, multiples, or factor loadings, then tests whether expected-return differences map onto shared risk exposures rather than tracking one asset's path through time.
- How does cross-sectional pricing differ from time-series analysis?
- Cross-sectional pricing holds time fixed and ranks many assets at once, while time-series analysis tracks a single asset's path across many dates. The cross-section answers 'which firms are cheap relative to peers right now?'; the time series answers 'how has this asset moved over the cycle?'. Factor models like Fama-French were built to explain the cross-section of returns.
- Why does cross-sectional pricing matter in the AI-concentration regime?
- Cross-sectional pricing reveals whether mega-cap dominance is genuine repricing or a narrow concentration premium. If an AI premium is real and diffuse, it should show up as wide multiple dispersion across the broader universe, not just at the index level. Flat cross-sectional dispersion outside the Magnificent Seven signals a premium that is concentrated rather than market-wide.
- What does cross-sectional dispersion tell you?
- Cross-sectional dispersion measures how widely valuations or returns spread across the universe at one moment, and high dispersion signals a market discriminating sharply between winners and losers. Narrow dispersion suggests crowded, index-driven moves; wide dispersion rewards stock-selection and relative-value desks. It is the raw material from which cross-sectional pricing extracts premia and mispricing.
- How do factor models relate to cross-sectional pricing?
- Factor models are the formal tool for cross-sectional pricing: they regress expected returns across securities onto exposures to common factors such as value, size, and momentum. Residual gaps not explained by factor loadings flag relative mispricing or an unpriced premium. The Fama-French three-factor model (1992) is the canonical framework for explaining the cross-section of returns.