What is the behavioral asset pricing model, and how does it incorporate investor sentiment into expected returns?

Question

AcadiFi · Accepted Answer

The Behavioral Asset Pricing Model (BAPM), proposed by Shefrin and Statman (1994), modifies the traditional CAPM by recognizing that market prices are set by the interaction of two types of traders: information traders (rational) and noise traders (sentiment-driven). The resulting market portfolio is not mean-variance efficient, and expected returns reflect both fundamental risk and sentiment-driven mispricing.

**BAPM Framework:**

In CAPM: E(R_i) = R_f + beta_i x [E(R_m) - R_f]

The market portfolio is efficient, beta captures all systematic risk, and no anomalies persist.

In BAPM: E(R_i) = R_f + beta_BM_i x [E(R_BM) - R_f]

where R_BM is the return on the 'behavioral market portfolio' — a combination of the mean-variance efficient portfolio and the noise trader portfolio. Beta_BM reflects covariance with this behavioral benchmark.

**Key Differences:**

| Feature | CAPM | BAPM |
|---|---|---|
| Market participants | All rational | Information traders + noise traders |
| Market portfolio | Mean-variance efficient | Tilted by sentiment |
| Beta interpretation | Covariance with efficient portfolio | Covariance with sentiment-influenced market |
| Anomalies | Should not exist | Expected and persistent |
| Price = Value | Always (in equilibrium) | Deviate due to noise trader demand |

**How Sentiment Enters Pricing:**

Noise traders create demand for stocks they perceive as attractive (glamour stocks, momentum winners, lottery-like payoffs) and avoid stocks they dislike (value stocks, recent losers). This shifts market weights away from fundamental-value weights.

Consider Waverly Technologies, a glamour tech stock:
- Fundamental value weight in market: 0.8%
- Noise-trader-inflated actual market weight: 1.4%
- The behavioral market portfolio overweights Waverly by 75%

Because the behavioral market portfolio tilts toward noise-trader favorites, the beta computed against this portfolio differs from the CAPM beta. Stocks favored by noise traders have lower behavioral betas (they move with sentiment, which dominates the benchmark) and therefore lower expected returns in the BAPM — consistent with empirically observed overpricing of glamour stocks.

**Empirical Support:**

1. **Value premium**: Value stocks (ignored by noise traders) earn higher returns — consistent with BAPM's prediction that unloved stocks have higher behavioral betas.

2. **Sentiment indices**: Baker-Wurgler sentiment index predicts cross-sectional returns. High-sentiment periods are followed by lower returns for speculative stocks.

3. **IPO underperformance**: Noise traders drive IPO overpricing; long-run underperformance follows as prices revert.

4. **Volatility puzzle**: Stock market volatility exceeds what dividend fundamentals justify — noise trader demand creates excess volatility.

**Limitations:**
- BAPM doesn't specify the noise trader portfolio precisely
- Behavioral betas are hard to estimate because the behavioral market portfolio is unobservable
- The model is descriptive rather than prescriptive (it explains anomalies but doesn't tell you how to exploit them)

Study behavioral asset pricing in our CFA Level III course.

What is the behavioral asset pricing model, and how does it incorporate investor sentiment into expected returns?

Master Level III with our CFA Course

Related Questions

Practice Questions