What is Anomalies? If Efficient Market Hypothesis holds, all securities should have the • same risk-adjusted returns. Lecture 2: Anomalies and Market o Therefore, observable stock characteristics such as size, PE ratio, or price-book value ratios will be useless in finding undervalued (i.e., Efficiency positive abnormal return) stocks or overvalued (i.e., negative abnormal return) stocks. Anomalies are empirical results that seem to be inconsistent with • maintained theories (e.g., CAPM) of asset-pricing behavior. o Fama (1970): Joint test of market efficiency and asset pricing theory o Jensen (1978): economic relevance After anomalies are documented and analyzed, they often seem to • disappear, reverse, or attenuate. Nattawut Jenwittayaroje, Ph.D, CFA Master of Arts program o Arbitrage strategy vs Statistical aberration in Applied Finance NIDA Business School 1 9.2 Market Anomalies: Size Effect Data Snooping Size Effect was first documented by Banz (1981) in the US • • The process of examining data and models affect the markets. likelihood of finding anomalies. Based on the sample period from 1926 to 1975, Banz (1981) • divided the stocks on the NYSE into quintiles based on their • Authors in search of an interesting research paper are market capitalization, and found that smaller firms earn higher likely to focus attention on ‘surprising’ results. returns than larger firms of equivalent risk. • Solution: Reinganum (1981) also documented that smaller firms had higher • o Test the anomaly on an independent sample: other average risk-adjusted returns than larger firms in the US stock countries, prior time periods, and/or subsequent data market prior to 1980. 9.3 4
Market Anomalies: Size Effect Market Anomalies: Size Effect • From Table 1 below, however, small-firm anomaly has Fama and French (1992) using data for the period 1963 to 1990, • group all stocks on the NYSE, AMEX, and NASDAQ into 10 disappeared since the initial publications of the papers that groups based on their market capitalization. discovered it, Banz (1981) and Reinganum (1981) Their results show that the average return of the smallest stock • group is 0.74% per month higher than the average return of the Size effects largest stock group. Dimson and Marsh (1986) and Chan et al. (1991) also find a small • firm premium for markets outside the U.S. (i.e., UK and Japan respectively). 6 www.dfafunds.com 5 Market Anomalies: Turn-of-the-year Market Anomalies: Turn-of-the-year (January) effect (January) effect Roll (1983) and Haugen and Lokonishok (1988) show strong • Average Return by Month of the Year differences in return behavior across the months of the year. o Specifically, they found that returns in January are significantly higher than returns in any other months of the year. Keim (1983) and Reinganum (1983) showed that much of the • abnormal return to small firms occurs during the first two weeks in January “turn of the year effect” or “January effect”. o Roll (1983): higher volatility of small stocks caused substantial short-term capital losses that investors might want to realize for income tax purposes before the end of the year. o Selling pressure in December and rebound in early January. 7 8
Market Anomalies: Turn-of-the-year effect January Effect: SET50 of the SET • Schwert (2003): from Table 2, it seems that the turn-of-the- year anomaly has NOT disappeared since it was originally documented. Whole period Sub-period 10 Market Anomalies: Weekend Effect Market Anomalies: Weekend Effect Weekend Effect – Gibbons and Hess (1981) study returns by days • • The weekend effect refers to the differences in returns of the week during 1962-1978. between Mondays and other days of the week, where the returns on Mondays are significantly negative, while the Average Daily Returns by Day of the Week: 1962 - 1978 returns on every day of the week are not. • The weekend effect is another return phenomenon that has persisted over long periods and over several international stock markets. • In the US market, French (1980) observed another calendar anomaly, by showing that the average return to the S&P portfolio was reliably negative over weekends (i.e., Friday to Monday) in the period 1953-1977. 11 12
Market Anomalies: Weekend Effect Market Anomalies: Weekend Effect • There is evidence of weekend effect in most major international markets Monday Rest of the Week Schwert (1990) shows estimates of the weekend effect from 1885 to 2002, • and shows that (in Table 3) there exists weekend effect in all periods, except the recent period (e.g., 1978-2002). Thus, like the size effect, the weekend effect seems to have disappeared • or at least substantially attenuated, since it was first documented in 1980. 13 14 Market Anomalies: “ Value” Effect Market Anomalies: Weekend Effect in Thailand PE Ratio- Studies by Basu (1977) and Basu (1983) investigate the • relationship between PE ratios and abnormal returns. Basu (1977, 1983) noted that firms with high earnings-to-price ratios • (i.e., low price-earnings ratios or “ value ” stocks) earn positive abnormal/excess returns (relative to the CAPM). PE Ratios and Stock Returns, 1952-2010 Annual Return Lowest PE Highest PE 16
Market Anomalies: “ Value” Effect Market Anomalies: “ Value” Effect There is also evidence of Value effect in other international • • Price-Book Value Ratios – Rosenberg et al. (1985), Fama and markets. French (1992) find that average returns on US stocks are positively related to the ratio of a firm’s book value to market Excess Returns on Low PE Ratio Stocks by Countries value. o In other words, low price-to-book stocks (i.e., “ value ” stocks) earn higher average returns than high price-to-book stocks. • Chan et al. (1991) find the same results in Japanese market. • Fama and French (1998) examine the “Value” (low P/B, low P/E) vs “Glamour/Growth” (high P/B, high P/E) stocks in international context. • “ Value ” outperformed “ Glamour/Growth ” in all markets under study. 17 18 Market Anomalies: “ Value” Effect Market Anomalies: “ Value” Effect Fama and French suggested that low price-book value ratios may operate as a measure of risk, since firms with prices well below book value are likely to be troubled firms. Therefore, investors require additional return for additional risk associated with such firms. o “Value” (low P/E and/or lowP/B ratios) characteristics represent a risk factor that are missing from the CAPM. Lakonishok et al (1994): Low P/E or low P/B ratio stocks are • generally characterized by low growth, large size, and stable business, all of which should work toward reducing their risk rather than increasing it. Therefore, the explanation for this finding is still not justified. • 19 20
Market Anomalies: Momentum Effect Market Anomalies: Momentum Effect • Jegadeesh and Titman (1993) provides an analysis of For a J/K strategy, at the beginning of each month t, securities are • relative strength trading strategies over 3 to 12 month ranked in ascending order on the basis of returns in the past J horizons. months. • They analyse NYSE and AMEX stocks over the period Ten equally weighted decile portfolios are formed. The top decile • 1965-1989. is the loser portfolio and the bottom decile is the winner portfolio. • Strategies considered select stocks based on their returns Strategy adopted in each month t is buy the winner portfolio and • over the past 1, 2, 3 and 4 quarters (i.e., formation periods) sell the loser portfolio, holding the position for K months. (J) • Holding periods considered vary from 1-4 quarters (K). • So a total of J x K = 4 x 4 = 16 strategies were empirically examined. 21 22 Market Anomalies: Momentum Effect Main Results: Returns of all zero-cost (i.e., • Rouwenhorst (1998) used data from12 European countries during • buy minus sell) portfolios a sample period of 1980-1995. are positive and nearly all are significant. Main Results: Most successful zero-cost • For each strategy, • strategy is to selects stocks past Winners based on their returns over outperformed past Losers by about 1% the previous 12 months and per month. then holds them for 3 months The returns range -> 1.31% per month. • from 0.64% to 1.35% per month, and all excess returns are statistically significant. 23 24
Recommend
More recommend
