期刊-NBER美国国民经济研究局-2010no2_40页_1mb

NBER Asset Pricing Program Summary

Core Content

The NBER Asset Pricing (AP) Program, established in 1991, has over 130 members who meet annually in three regions (Midwest, East Coast, West Coast) to discuss their research. The program has been significantly influenced by the 2008 financial crisis, leading to a re-evaluation of conventional asset pricing theories and an increased focus on models that incorporate frictions, investor heterogeneity, and time-varying risk.

Main Research Areas

1. Quantitative Structural Asset Pricing Models

• These models aim to provide unified explanations for various financial market phenomena.
• Before the financial crisis, they helped explain financial institutions, market frictions, liquidity, and investor behavior.
• Post-crisis, the analysis of these models has intensified, especially in explaining return predictability and risk premiums.

2. Time-Varying Risk Aversion

• Campbell and Cochrane developed a model where investors' risk aversion varies over time based on past consumption.
• Wachter extended this model to explain stock and bond returns, showing that real rates are countercyclical.
• Chen, Collin-Dufresne, and Goldstein applied the model to corporate bonds, noting that Baa-Aaa spreads are influenced by the timing of defaults.
• Verdelhan showed that time-varying risk aversion can explain the currency carry trade.
• The model also explains why "growth stocks" with backloaded dividends have a growth premium rather than a value premium.

3. Long-Run Risk

• Bansal and Yaron proposed that investors worry about persistent, small changes in expected consumption growth.
• In the long-run-risk model, value stocks are seen as more risky due to higher covariance with consumption shocks, leading to higher risk premiums.
• Hansen, Heaton, and Li found that the covariance between cash flows and news shocks depends on time trends.
• Beeler and Campbell simulated the model and found that long-run risk in the Bansal and Yaron model may be overestimated.
• Schneider and Piazzesi extended the model to nominal government bond prices, showing that higher inflation is bad news for future consumption, leading to higher returns on long bonds.

4. Disaster Risk

• Rietz suggested that rare disasters in consumption explain the equity premium.
• Barro documented historical disasters and developed models where disasters affect both stocks and bonds.
• Burnside, Eichenbaum, and Rebelo and others used options data to measure disaster risk.
• Martin and Gabaix introduced new techniques for solving models with disaster risk, such as higher-order cumulants and linearity-generating processes.

5. Time-Varying Risk

• Kandel and Stambaugh showed that time variation in consumption growth variance (heteroskedasticity) affects asset pricing.
• Bansal and Yaron combined long-run risk with heteroskedasticity to explain return predictability.
• Chen demonstrated that time-varying risk increases the likelihood of firm defaults and explains credit spreads and capital structure.

6. Financial Intermediation

• He and Krishnamurthy explored models where households invest through financial intermediaries.
• They introduced an "intermediation constraint" that limits the amount of funds households can invest, leading to higher risk premiums in bad times.
• Vayanos and Woolley examined active portfolio management and how it can lead to higher returns for funds.

7. Heterogeneous Expectations

• Schneider and Piazzesi used the Michigan survey to show that young households expect higher inflation than older ones, leading to lower real mortgage rates and increased house prices.
• Malmendier and Nagel found that investors' expectations are shaped by their lifetime experiences, with those who experienced low stock returns being more pessimistic.

8. Other Heterogeneity

• Krueger and Lustig showed that uninsurable income shocks do not affect the equity premium.
• Kiyotaki, Michaelides, and Nikolov found that relaxed collateral constraints do not lead to higher house prices.
• Parker and Vissing-Jorgensen noted that rich households have more volatile consumption, which may explain average premiums.
• Chien, Cole, and Lustig developed a model where households do not rebalance portfolios in response to aggregate shocks, leading to time variation in risk premiums.

Key Findings

• The financial crisis has prompted new research into asset pricing models that incorporate frictions, heterogeneity, and time-varying risk.
• Time-varying risk aversion and long-run risk provide more consistent explanations for the size and variation of risk premiums across asset classes.
• Financial intermediation plays a crucial role in shaping market outcomes, especially during periods of economic stress.
• Investor expectations and behavior are influenced by personal experiences and macroeconomic conditions.
• Heterogeneous agent models offer insights into market behavior, but their impact on aggregate asset prices may be limited.

Conclusion

The financial crisis has stimulated significant research in the Asset Pricing Program, leading to a deeper understanding of market dynamics. The focus has shifted from explaining isolated phenomena to providing unified models that account for multiple financial market behaviors. Future research is expected to explore the interaction between different types of risk, such as disaster risk and time-varying risk, and to refine models that better explain asset pricing in real-world conditions.