Behavioral Finance Economics 437 Behavioral Finance Predictability Apr
Behavioral Finance Economics 437 Behavioral Finance Predictability Apr 11, 2017
General Theme: Predictability n Fama French: “Cross Section” Article n Book/Market and Size => Stock Returns n “Value Investing” or “Contrarian Investing” n De. Bondt Thaler: “Over Reaction” n Buy past losers; Sell past winners n Value? Investing; Contrarian Investing n Jegadeesh Titman: “Price Momentum” n Possibly “Earnings Momentum” Behavioral Finance Predictability Apr 11, 2017
Data in Fama and French n 1962 -1989 data n Book Value (leverage and price/earnings) at previous year end n Returns starting on July 1 of the following year (also use the market n n n equity as of July 1 for size, but use market equity at previous year end for B/M calculation) Calculate monthly returns Each month the cross-section of returns is regressed on explanatory variables. Prior research used “portfolio betas”; F-F use individual stocks Sort stocks into “size deciles” n n Sort each size decile into 10 portfolios based on beta Calculate equal weighted monthly returns on the portfolios for the next 12 months (from July to June). Behavioral Finance Predictability Apr 11, 2017
Results on Beta n Portfolios in size deciles (without breaking them into 10 beta portfolios) show a relationship between beta and return n Large size means lower beta and lower returns n When size deciles are subdivided into beta ranked decile portfolios Larger size firms have lower returns n “no relation between average return and beta” n Behavioral Finance Predictability Apr 11, 2017
Results on Book/Market n What is book to market Book is firm net worth reported on 10 -Ks n Market is: shares outstanding times price Book/market is positively related to returns n Size still matters but B/M is much more important n B/M swamps leverage and E/P Leverage: book or market leverage? January “slopes” twice slopes of other months Overall largest decile book to market beats smallest decile book to market by 1. 53 % per month n n n Behavioral Finance Predictability Apr 11, 2017
De. Bondt-Thaler 1984 n “Over-Reaction” Hypothesis n Suggests that: n After a period of “over-reaction, ” markets “revert” back and go the other way. n n n Behavioral Finance Stocks that have done well in the past, do poorly in the future Stocks that done poorly in the past, do well in the future Their article is designed to test whether or not “mean reversion” is true. Predictability Apr 11, 2017
Data n NYSE data n Jan 1926 through December 1982 n Monthly return data n Begin with three year lookback in Dec 1932 n Monthly data from Jan 1930 through Dec 1932 n 36 months or three years data n Form portfolios of L(osers) and W(inners) n Then see how they do for the next three years Behavioral Finance Predictability Apr 11, 2017
De. Bondt and Thaler: “Does the Stock Market Overreact” (1985) n L – three year loses n W – three year winners n Question: How do the W’s do in the next three years? How do the L’s do in the next three years? n Other things worth noting Almost all of the impact is in January n When the W portfolios are formed, they have very high P/E ratios, the L portfolios have low P/E ratios at the time of formation n Behavioral Finance Predictability Apr 11, 2017
De. Bondt-Thaler conclusions n Definite evidence of mean reversion (a form of serial correlation): n L portfolios consistently outperform W portfolios n n 19. 6 % better than the market after end of 3 years W portfolios consistently underperform the market n Behavioral Finance 5 % less than the market after end of 3 years Predictability Apr 11, 2017
Interesting facts n Most of the excess returns are in January n Loser effect more pronounced: Losers earned 19. 6 % more than the market n Winners earn 5. 0 % less than the market n Loser portfolio minus Winner portfolio return = 24. 6 %!!!!! n Most of the return difference is during 2 nd and 3 rd year n Larger loses become larger winners; larger winners become larger losers n Behavioral Finance Predictability Apr 11, 2017
Ball & Brown 1986 n Market “underreacts” to earnings surprises n Article generally ignored until Jagdeesh- Titman n Time span suggests that Ball-Brown effect may be the same thing as Jagdeesh-Titman Behavioral Finance Predictability Apr 11, 2017
Jegadeesh and Titman (1993) n Relative strength strategies, sometimes called n n n “earnings momentum” strategies Find past winners and past losers (using 3 to 12 month holding periods) generate gains (winners gain; losers lose) Construct W portfolio and L portfolio W-L (using 6 month periods) earns more than 12 % better than market portfolio Longer term portfolios do best in next 12 months Interpretation in “event time” Doesn’t work in January Behavioral Finance Predictability Apr 11, 2017
Chan, Jegadeesh, Lakonishok 1996 n Is it earnings? Is it price? n They 7. 7 percent six month gap between winner portfolios and loser portfolios using price momentum. n Conclusion (page 1709): “ In general, the price momentum effect tends to be stronger and longer-lived than the earnings momentum effect. ” Behavioral Finance Predictability Apr 11, 2017
Chordia-Shivakumar, 2006 n Is it “pricing momentum” or “earnings momentum” that drives the “under-reaction” phenomenon? n Conclude the earnings momentum is the key factor. n Price momentum variables are a “noisy proxy” for earnings momentum Behavioral Finance Predictability Apr 11, 2017
Hong, Lee & Swaminathan 2003 n Earnings Momentum is the real driver of price momentum n Systematic relationship between earnings momentum and future GDP growth – hence a “risk factor” n This matters, because if there is a risk factor, then momentum might be consistent with EMH (which price momentum generally is not) Behavioral Finance Predictability Apr 11, 2017
The End Behavioral Finance Predictability Apr 11, 2017
- Slides: 16