FIN 355 Behavioral Finance Class 3. Individual Investor Behavior Dmitry A Shapiro University of Mannheim Spring 2017 Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 1 / 27
Stock Market Non-participation A lot of people did not invest at all in stock market; Mankiw and Zeldes (JFE, 1991): in 1984 only 28% had invested into stock market; All theories say they should; Heaton, Lucas (2000): Take background risks seriously (housing risk, labor risk, etc.) Still should invest stock since stocks are not correlated with these risks. Transaction cost explains some non-participation but not all of it: There are wealthy and sophisticated people who do not participate. Note : with plain risk-aversion you still should invest something. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 2 / 27
Behavioral Hypothesis Bernatzi and Thaler (QJE, 1995) show that if agents has prospect utility over annual gains/losses then they are reluctant to participate. If you are loss averse over annual gains and losses in stocks then no participation; Your value of (110, 0.5, 100, 0.5) is 0.5 110 2 (100)/2 < 0 so do not want to invest. Loss aversion is NOT enough either because most people have background risks and that the stock market would diversify. Say other risks are (30000, 1/2, 10000, 1/2) Utility is 1/2(30000) + 1/2(10000)( 2) = 5000 where 2 comes from loss aversions. With that additional lottery the risk structure becomes: (30110, 1/4, 29900, 1/4, 9890, 1/4, 10100, 1/4) Its utility is approximately 5007. Need loss aversion and narrow framing to explain the rejection Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 3 / 27
Non-participation Why narrow frame the stock market? As before: either non-consumption utility, e.g. regret or intuition. The background risks are not immediately accessible while stock information is. Test of the model is Dimmock and Kouwenberg (JEF 2010) Loss Aversion and Household Portfolio Choice. Measures loss aversion from household survey data; Shows that more loss averse households are less likely to participate. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 4 / 27
Ambiguity Aversion/Competence Stock market distribution is ambiguous so stay away, OR Stock market distribution is ambiguous and the agent does not feel competent to evaluate it. Formalization: Epsten and Schneider Learning Under Ambiguity, 2005. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 5 / 27
Underdiversification a) home bias French/Poterba, AER 1991 showed that 94% into domestic portfolio; b) home-bias at home : within a country people buy stocks located nearby US data: Ivkovich & Weisbenner (JF, 2005); US data: portfolios of local holdings do not generate abnormal performance (Seascholes & Ju, JF 2010); Scandinavian data: Greenblatt & Kelojarhn (JR, 2001); c) excessive allocation to own company stock in 401(k) plan; d) significant positions in a few stocks. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 6 / 27
Underdiversification Loss Aversion/Narrow Framing applies to a) and d) By framing narrowly you ignore diversification benefits Risk-seeking, e.g. via π(p) in PT If some stocks are positively skewed, a PT investor may want to take a large undiversified position in a few of them. Intuition: by doing that the investor gets a lottery like payoff and PT investors like it. A nice paper on the topic Polkovnichenko (2005); Allocation between T-bills, market and stock (positively skewed). PT investor puts a lot on the stock which makes his position undiversified. Mitton and Vorkink (RFS, 2007): underdiversified investors hold more positively skewed stocks and they underperform. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 7 / 27
Underdiversification Ambiguity Aversion (applies to a), b), c)). How to vary it? Competence Hypothesis (Heath, Tversky (1991), Fox and Tversky QJE (1995)) Competence is what you feel you know relative to what could be known. This is testable: AA will increase for a bet if you are reminded about bets you know more about. This is what happens in Ellsberg paradox Another prediction: AA can be increased if people are reminded about other more competent people. When for students of San Jose University it was mentioned that particular bets for stock movements were also observed by stock analysts then WTP decreased. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 8 / 27
Underdiversification Ambiguity Aversion/Competence (applies to a), b), c)) Foreign countries, distant stocks are less familiar so I avoid them; Test: Graham, Huang and Harvey (2005): investors who self-report(!) to be more confident trade more often and have more diversified portfolio. Loyalty (a,c) Cohen (2004) Loyalty Based Portfolio Choice Larger allocation of company stock in stand-alone companies not conglomerates; Increase amount of investments after spin-off and decrease after merge Patriotism In more patriotic countries there is more home-bias; People living in more patriotic states have higher home bias. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 9 / 27
Underdiversification Optimism (a,b, c) If you optimistic about your prospects you will be optimistic about your country, etc. Test: Kilka, Weber (2000): US students higher return and smaller confidence interval on US stocks versus German stocks. German students are the opposite. Strong/Xu (2003): local fund managers assign higher returns to home country Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 10 / 27
Underdiversification. Effect of Attention Barber and Odean (JFE, 2008): All that Glitters: The Effect of Attention and News on the Buying Behavior of Individual and Institutional Investors. Attention is a scarce resource; Attention affects buying more than selling: There are tons of potential candidates to buy; Most individual investors hold relatively few common stocks (mean 4.3 and median 2.61). Individual investors sell only stocks that they already own; they do not sell short (less than 1% of positions are short). This buy-sell asymmetry is smaller for institutional investors. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 11 / 27
What Grabs Attention? Three indirect ways: news; unusual trading volume; extreme returns. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 12 / 27
Summary When someone buys - someone sells; Professional investors should exhibit a lower tendency to buy, on high-attention days and a higher tendency on low-attention days; For example, individual investors make nearly twice as many purchases as sales of stocks experiencing unusually high trading volume (e.g. the highest 5%). Similarly individual investors are net buyers of stocks with extremely good or extremely bad last day returns; Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 13 / 27
Attention Allocation Theories Home bias and under-diversification can be rationalized using attention allocation theories (Mondria and Wu, 2011): investors need to decide how to allocate their attention; assume small informational advantage about local stocks; optimally to process information about local assets and hold a greater portion of local stocks (as compared to market portfolio); thus initial and small asymmetry gets magnified further. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 14 / 27
Test. Mondria and Wu (2011) Attention can be measured using Google Trend website that shows the number of searches across different regions and different time periods. Using tickers for the stocks in S&P500 authors document that investors search 43% as much information about local stocks. Stocks with abnormal asymmetric information are defined as those for which locals allocate more attention than nonlocals. Authors claim that there is a difference in returns between high asymmetry stocks and no-asymmetry stocks. High asymmetry stocks have higher return. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 15 / 27
Naive Diversification Bernatzi and Thaler (AER, 2001) Even if people diversify they allocate 1/n wealth to each n investment options regardless of what they are. Experimental test. Three groups: Bond fund vs. stock fund; Bond fund vs. balanced fund; Stock fund vs. balanced fund; In all three cases 50-50 was the most popular but that means very difficult allocation. Field test: Plans with more equity funds should see more equity allocation. Look at 170 big pension plans; If equity funds fraction is 37% then the allocation is 49%; If equity funds fraction is 65% then the allocation is 60%; If equity funds fraction is 81% then the allocation is 64%; Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 16 / 27
Naive Diversification The 1/n heuristics is motivated by marketing diversification heuristics. When make multiple decisions at the same time go with different options Simonsa (90, JMR) a candy example. Hubermang and Jiang (JF, 2005) have better data. Prediction: If more funds are offered then people will hold more funds. NO Prediction: If more equity funds are offered then people will allocate more to the equity NO Overall: people do allocate 1/n over those funds that they choose. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 17 / 27
Excessive Trading Barber/Odean (JF, 2000) have a huge database from 1991 to 1996 of individual trading Average returns of individuals is below all sorts of benchmarks such as: portfolio they had in the beginning of the year, market portfolio, CAPM benchmark, 3 factor model benchmark. This in net returns (!) after transaction costs. No strong underperformance in gross returns. The more you trade the worse you do (gross is higher than S&P500 the net is lower). Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 18 / 27
Excessive Trading Why people trade a lot rationally? Liquidity or need for cash; If they need liquidity why would they put money in stock to begin with Information Then they should do well Rebalancing/Hedging More trading is going on Tax reasons Happens on tax-deferred accounts; Applicable only when you sell stock at loss. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 19 / 27
Excessive Trading Behavioral view: Overconfidence Test: Barber/Odean (QJE, 2001): overconfidence varies with some characteristics. Men are more confident so they trade more. Barber/Odean (RFS, 2002): switch from phone trading to online trading. People now have more information so they feel more confidence and trade more. Glaser/Weber (2007) have dataset of online traders in Germany 1997-2001. Use a questionnaire to the people from their database Optimistic people trade more, overconfident do not. Optimistic people do not do better. They are not rationally optimistic. Another behavioral story is entertainment Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 20 / 27
Selling Decisions. Disposition Effect Disposition effect: people are reluctant to sell at loss relative to purchase price. Odean (JF, 1998) have a 1987-1993 data from discount brokerage. Find that people tend to sell stocks at a gain rather than at a loss. Rational explanations: Information Subsequent returns of sold stock is higher than that of the losers that individuals hold Liquidity (cannot explain it) Taxes: No they should predict the opposite. However, there is a December effect when people sell losers Disposition effect is weaker in taxable accounts but still it dominates (Ivkovich/Poterba/Weisbenner AER, 2005) Genesove/Meyer (QJE, 2002) disposition effect in housing markets. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 21 / 27
Disposition Effect Behavioral Hypothesis: Conservatism: Form the opinion and too slow to update the belief. Assume good news is released after you formed the opinion Ignore it, think that the price is too high and sell. Prospect Theory/Narrow Framing If in gains you are risk-averse and prefer to sell; If in losses you are risk-seeking and prefer to keep; Barberis/Xiong (JF, 2009) formalize the idea. Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 22 / 27
Disposition Effect Linnainmaa (JF, 2010) used Finnish individual investor trading data to show that the disposition effect can be explained in large part by investors use of limit orders: Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 23 / 27
Disposition Effect Chang, Solomon (JF, 2015): Hypothesis: investors avoid realizing losses because they dislike admitting that past purchases were mistakes; Delegation reverses the effect since its managers who made the mistake; Show the disposition effect applies only to nondelegated assets (e.g. individual stocks); Delegated assets, like mutual funds exhibit robust reverse-disposition effect; Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 24 / 27
Trading and Learning Individual investors: underperform standard benchmarks. e.g. low cost index fund; sell winning investments while holding losing investment; are heavily influenced by limited attention and past return performance; tend to hold underdiversified portfolios; In addition, individual investors learn wrong: engage in naive reinforcement behavior learning by repeating past behaviors that coincided with pleasure while avoiding past behaviors that generated pain; Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 25 / 27
Trading and Learning. Investors are more likely to purchase a stock they previously sold for a profit than one sold for a loss (Strahilevitz et al. 2011); Investors more likely to buy a stock in an industry if their previous investment in this industry have earned a high return (Huang, 2010); Investors whose 401(k) experienced greater return or lower variance increase their savings rate (De, Gondhi and Pochiraju, 2010); Investors whose age cohorts have experienced high stock market returns throughout their lives are less risk-averse and more likely to invest in stocks (Malendier and Nagel, 2011); Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 26 / 27
Dmitry A Shapiro (UNCC) Individual Investor Spring 2017 27 / 27