Week 7 - Theories, stock-level returns Flashcards
Four Types of Models
- Preference-based models
- Belief-based models
- Models of cognitive limits
- Models with frictions
Preference-based models
Barberis and Huang (2008) - Stocks as Lotteries
- Based on prospect theory
- Single period model -> risk-free asset and J risky assets with multivariate Normal Payoffs
Investors:
- Identical expectations about security payoffs
- Identical CPT preferences -> defined over gains/losses in wealth
- reference point is initial wealth scaled up by the risk-free rate
Utility defined over: W = W1 -W0*Rf
Barberis and Huang (2008) Model
- CAPM Holds - prospect theory gives the same prediction as the Expected Utility model
- To make more interesting predictions, remove assumption of multivariate normal assumption of returns -> introduce small, independent, positively skewed security into economy
- Prediction -> the new security earns negative excess return -> skeweness is priced, in contrast to concave EU model where only coskewness with market matters
Barberis and Huang Model Equilibrium
- Equilibrium involves heterogenous holdings - some investors hold large, undiversified positions in new security whilst others hold no positions in new security
- Heterogenous holdings aris from non-unique global optima - not from heterogenous preferences
- Since new security contributes skewness to the portfolios of some investors, it is valuable and so earns a low average return
- Only works if security is highly skewed - otherwise would need too undiversified a position to add skewness to the porfolio
Applications of Barberis and Huang (2008)
- Low average return on IPOs - IPO returns are highly positively skewed
- Low average return of distressed stocks, bankrupt stocks, OTC stocks
- Overpricing of Out-of-the-money options on individual stocks
- Low average return on stocks with high idiosyncratic volatility
Belief-Based Models
- Daniel, Hirshleifer, Subrahmanyam -> overconfidence, self-attribution bias
- Barberis, Shleifer, Vishny -> Representativeness, conservatism
Overconfidence
- 1980s -> models where rational investors observe different information may not generate much trading volume
- Each investor infers other’s signals from prices, or from willingness to trade -> reduces own willingness to trade
- Overconfidence -> overestimation of the precision of one’s own information signals
- Dismissiveness -> underestimation of precision of others’ signals
- overconfidence and dismissiveness can generate significant trading volume
Empirical tests of Overconfidence and Dismissiveness
Grinblatt and Keloharju (2009):
- Data from Finland shows that overconfident people trade more
- Overconfidence is self-reported confidence minus how confident individual should be based on performance on aptitude tests
Glaser and Weber (2007)
- Use data from German online brokerage to measure two types of overconfidence: Overplacement and overprecision
- Found that overplacement predicts trading, but overprecision does not
Barber and Odean (2001)
- Argue that, since men tend to be more overconfident than women, they will trade more and have worse returns
-confirm this using brokerage data
DHS Model
- DHS present a model of misvaluation based on overconfidence
- A risk-neutral, representative investor is overconfident about the private information he gathers - stress biases in the interpretation of private, rather than public information
- If the private information is positive, overconfidence means that investors will push prices up too far relative to fundamentals
- This leads to long-term reversals and value premium
Self Attribution bias in DHS
- When public information confirms the private signal, the investor becomes even more confident in the private signal
- When public information disconfirms the private signal, he does not lose much confidence in the private signal
- This leads to momentum in addition to a value premium
Representativeness
- individuals focus more on first terms and often neglect the base rates
Law of small numbers
- Kahneman and Tversky (1971) propose that individuals have an incorrect belief in law of small numbers - they think that even a small sample will reflect characteristics of the population
- Gambler’s fallacy - When we dont know the model generating the data, LSN leads us to over-infer from small sample - generating over-extrapolation of recent trends
BSV Model
- Representative, risk-neutral investor
- earnings on all assets follow a random walk
- investor thinks that earnings at any time are driven by:
mean-reverting regime: earnings are more mean reverting that in reality
trending regime: earnings trend more than in reality
BSV Biases
Conservatism - tendency to underweight new information relative to priors
Representativeness - Motivated by LSN - people expect even short samples to reflect properties of parent population
BSV example
When a company announces surprisingly good earnings, conservatism means that investors react insufficiently, pushing the price up too little
- since the prices is too low, subsequent returns will be higher on average -> PEAD and momentum will be higher
After a series of good earnings announcements, representativeness causes people to overreact and push the price up too high
- Law of samll numbers leads investors to believe this is a firm with high earnings growth, so forecast higher earnings in the future
- since the price is now too high, subsequent returns are too low on average -> long term reversals + value
