Members of the NBER's Behavioral Finance Working Group met November 1 in Cambridge. Research Associate Nicholas C. Barberis of Yale University organized the meeting. These researchers' papers were presented and discussed:
Peter D. Maxted, Harvard University
A Macro-Finance Model with Sentiment
Maxted builds a general equilibrium macroeconomic model that combines diagnostic expectations and financial frictions. Diagnostic expectations are a forward-looking model of extrapolative expectations that overreact to recent news. Frictions in financial intermediation generate nonlinear spikes in risk premia and slumps in investment during periods of financial distress. The calibrated model is solved globally to characterize the full equilibrium dynamics generated by the interaction of sentiment and financial frictions. The model evaluates the causal role of over-optimism in triggering the amplification of financial distress into a full-blown crisis. Boom-bust investment cycles emerge endogenously out of a feedback from sentiment to financial frictions. Under the baseline calibration, the model predicts that financial crises are less likely to occur when expectations are diagnostic than when they are rational.
Francesco D'Acunto, Boston College; Ulrike Malmendier, University of California, Berkeley and NBER; Juan Ospina, University of Chicago; and Michael Weber, University of Chicago and NBER
Exposure to Daily Price Changes and Inflation Expectations (NBER Working Paper 26237)
D’Acunto, Malmendier, Ospina, and Weber show that, to form aggregate inflation expectations, consumers rely on the price changes they face in their daily lives while grocery shopping. Specifically, the frequency and size of price changes, rather than their expenditure share, matter for individuals' inflation expectations. To document these facts, the researchers collect novel micro data for a representative US sample that uniquely match individual expectations, detailed information about consumption bundles, and item-level prices. The results suggest that the frequency and size of grocery-price changes to which consumers are personally exposed should be incorporated in models of expectations formation. Central banks' focus on core inflation -- which excludes grocery prices -- to design expectations based policies might lead to systematic mistakes.
Samuel M. Hartzmark and Samuel D. Hirshman, University of Chicago, and Alex Imas, Carnegie Mellon University
Ownership, Learning and Beliefs
Hartzmark, Imas, and Hirshman examine how owning a good affects learning and beliefs about its quality. In a series of studies, the researchers show that after receiving a positive or negative signal about a good, ownership causes more optimistic or pessimistic beliefs, respectively, compared to receiving the same signal about a good that is not owned. This effect on beliefs impacts the valuation gap between owners and non-owners (i.e. the endowment effect), which is shown to increase with positive signals and disappear with negative signals. Moreover, the research demonstrates that people overreact to signals about goods that they own relative to normative benchmarks, but learning is close to Bayesian for goods that are not owned. In exploring the mechanism for this effect, it is found that ownership increases attention to recent signals about owned goods, exacerbating overextrapolation. The researchers demonstrate a similar relationship between ownership and over-extrapolation in survey data about stock market expectations. The findings provide a microfoundation for models of disagreement that generate volume in asset markets and have implications for any setting with trade and scope for learning.
Nicholas C. Barberis; Lawrence J. Jin, California Institute of Technology; and Baolian Wang, University of Florida
Prospect Theory and Stock Market Anomalies
Prospect theory was developed more than 30 years ago, but its predictions for basic aspects of asset prices such as the cross-section of average returns are still not well understood. Barberis, Jin, and Wang build a new model of asset prices in which investors evaluate risk in part according to prospect theory, and show how the model can be used to make quantitative predictions about average stock returns. They then examine the model's ability to explain 22 prominent stock market anomalies. They find that the model is helpful for thinking about a majority of the anomalies we consider, including the momentum, volatility, distress, and profitability anomalies. It performs most poorly on the value anomaly. The researchers present suggestive evidence that this is due to investors having incorrect beliefs about the distribution of value stock returns.
Lars A. Lochstoer, University of California, Los Angeles, and Tyler Muir, University of California, Los Angeles and NBER
Volatility Expectations and Returns
Lochstoer and Muir provide evidence that agents have slow moving beliefs about stock market volatility. This is supported in survey data and is also reflected in firm level option prices. The researchers embed these expectations into an asset pricing model and show that they can jointly explain the following stylized facts (some of which are novel to this research): When volatility increases the equity and variance risk premiums fall or stay flat at short horizons, despite higher future risk, these premiums appear to rise at longer horizons after future volatility has subsided, strategies that time volatility generate alpha, the variance risk premium forecasts stock returns more strongly than either realized variance or risk-neutral variance (VIX), changes in volatility are negatively correlated with contemporaneous returns. Slow moving expectations about volatility lead agents to initially underreact to volatility news followed by a delayed overreaction. This results in a weak, or even negative, risk-return tradeoff at shorter horizons but a stronger tradeoff at longer horizons (beyond where one can strongly forecast volatility). These dynamics are mirrored in the VIX and variance risk premium which reflect investor expectations about volatility.
Paul Goldsmith-Pinkham, Yale University, and Kelly Shue, Yale University and NBER
The Gender Gap in Housing Returns
Housing wealth represents the dominant form of retirement savings for most American households. Using detailed data on housing transactions across the United States since 1991, Goldsmith-Pinkham and Shue find that single men earn one percentage point higher unlevered returns per year on housing investment relative to single women, with couples occupying the intermediate range. The gender gap grows significantly larger after adjusting for mortgage borrowing: Men earn six percentage points higher levered returns per year relative to women. Data on repeat sales reveal that women buy the same property for approximately 2% more and sell for 2% less. The gender gap in housing returns arises because of gender differences in the (1) location and timing of transactions, (2) choice of initial listing price, (3) negotiated discount relative to the list price, and (4) length of holding period. Gender differences in upgrade rates, preferences for housing characteristics, and listing agents appear to be less important factors. The gender gap varies with demographic characteristics, but remains large in regions with high average education, income, and house price levels.