Asset Pricing with Heterogeneous Consumers and Limited Participation: Empirical Evidence

Alon Brav, George A. Constantinides, Christopher Géczy2002 · DOI

Abstract

This study provides evidence that the equity premium and the value premium (the excess return of value stocks over growth stocks) can be explained during the 1982-1996 period using a stochastic discount factor (SDF) derived from individual household consumption. By moving beyond the "representative agent" model and accounting for incomplete consumption insurance and limited market participation, the authors demonstrate that these premiums are consistent with economically plausible levels of relative risk aversion (RRA).

Introduction: The Representative Agent Paradox

In classical asset pricing, the "Representative Agent" model assumes that the economy behaves like a single, aggregate consumer. However, this model famously fails to explain the Equity Premium Puzzle: the historical 6% annual outperformance of stocks over risk-free bonds. To bridge this gap using aggregate data, models require a Relative Risk Aversion (RRA) coefficient exceeding 50—a level far beyond the "economically plausible" range of 1 to 10.

This paper addresses this discrepancy by focusing on two real-world frictions:

  1. Incomplete Consumption Insurance: Individuals face idiosyncratic risks (e.g., job loss) that cannot be fully insured away. These shocks make individual consumption more volatile than national averages.
  2. Limited Participation: A significant portion of the population does not hold stocks. If only a subset of the population trades in the market, the consumption patterns of the entire population should not be used to price those assets.

Methodology and Data

The authors utilize the Consumer Expenditure Survey (CEX) to reconstruct quarterly consumption of non-durables and services for individual U.S. households from 1982 to 1996. They construct a Stochastic Discount Factor (SDF) as the weighted average of individual households' marginal rates of substitution, rather than using per capita averages.

To isolate the drivers of their results, they employ Taylor expansions of the SDF, allowing them to analyze the "moments" of consumption growth—specifically mean, variance, and skewness. They also test the model against different "asset holder" thresholds, filtering the sample to include only households with significant financial assets (ranging from $0 to $40,000) to identify the behavior of the marginal investor.

Key Insights: The Power of Skewness

The study’s most significant finding is the role of cross-sectional skewness in consumption growth. While aggregate data smooths out individual shocks, household-level data reveals a higher probability of "disastrous" drops in consumption. This downside risk makes equity appear much riskier to an individual investor than it does to a theoretical representative agent.

When accounting for these individual "bumps," the equity premium is explained with an RRA coefficient of approximately 3.0 to 4.0. This is a major resolution to the puzzle, as these values align with standard economic expectations of human behavior.

Limited Participation and the Value Premium

The research also sheds light on the Value Premium—the tendency for "Value" stocks to outperform "Growth" stocks. The results indicate that value stocks are particularly sensitive to uninsurable individual consumption shocks, explaining why they command a higher risk premium.

Furthermore, as the definition of "asset holder" is tightened to include only wealthier households, the standard per capita model's performance improves (requiring lower RRA), though it remains inferior to the heterogeneous agent model. This suggests that while limited participation is a factor, the heterogeneity of consumption among those who do participate is a primary driver of asset prices.

Conclusion

By shifting the focus from national averages to individual household experiences, this paper resolves long-standing puzzles in finance. The evidence suggests that the "Representative Agent" is a flawed abstraction for asset pricing because it ignores the uninsurable risks and skewed consumption distributions faced by real investors. With an RRA of 3, the equity premium is no longer a mystery but a rational reflection of individual risk.

Summary updated

Abstract

We present evidence that the equity premium and the premium of value stocks over growth stocks are explained in the 1982-1996 period with a stochastic discount factor (SDF) calculated as the weighted average of individual households' marginal rate of substitution with low and economically plausible values of the relative risk aversion (RRA) coefficient.Household consumption of non-durables and services is reconstructed from the CEX database.Since the above premia are not explained with a SDF calculated as the per capita marginal rate of substitution with low value of the RRA coefficient, the evidence supports the hypothesis of incomplete consumption insurance.We also present evidence is that a SDF calculated as the per capita marginal rate of substitution is better able to explain the equity premium and does so with a lower value of the RRA coefficient, as the definition of asset holders is tightened to recognize the limited participation of households in the capital market.
BibTeX
@misc{brav2002asset,
  author = {Alon Brav and George A. Constantinides and Christopher Géczy},
  title = {{Asset Pricing with Heterogeneous Consumers and Limited Participation: Empirical Evidence}},
  year = {2002},
  doi = {10.3386/w8822},
}

Loading related content...

Figure