The finding of reversals in weekly returns has been attributed to a combination of microstructure issues and overreaction to information. I provide new evidence eliminating overreaction as a source of reversal. I show that well-known weekly contrarian profits are followed by a long run of momentum profits. In fact, these profits are strong enough to produce a significant momentum effect over the full year following portfolio formation. Thus, the market does not appear to view extreme weekly returns as excessive, as implied by an overreaction story. To the contrary, this return continuation is consistent with underreaction to the news driving extreme weekly returns. This is supported by cross-sectional tests in which I find this week's news is positively related to next week's returns. The evidence presented here is consistent with growing evidence that underreaction to firm-specific information is a pervasive feature of price formation. Therefore, if any short-run contrarian profits can be realized, they are better viewed as compensation for providing liquidity than as a reward for arbitrage.

While some prior studies document that investors overreact to information in sales growth as consistent with representativeness bias, other studies find no evidence of investor overreaction to either sales or earnings growth. Other recent studies also show that sales growth does not predict stock returns after controlling for changes in outstanding shares and asset growth. I reexamine the role of representativeness by investigating whether the effects of this bias are confounded by the presence of another effect that has been extensively documented - investors' underreaction to fundamentals. Adjusting for investor under-reaction to fundamentals, I document strong evidence that investors overreact to sales growth as predicted under representativeness despite adding accruals, asset growth, and equity issuance as additional controls. In cross-sectional regressions of future stock returns on predictive variables that control for fundamentals, changes in equity shares, accruals, and lagged 36 month returns, I find that the coefficient on sales growth is highly significant over both the full sample period 1970-2009 (t-stat -3.12). Furthermore, asset growth, equity issuance, and accruals lose much of their significance in favor of sales growth. I also provide evidence that rejects a theory based on fixation in favor of representativeness. These results document evidence of overreaction to past sales growth in firms where underreaction to fundamentals does not confound the overreaction due to representativeness bias.

The finding of reversals in weekly returns has been attributed to a combination of microstructure issues and overreaction to information. I provide new evidence eliminating overreaction as a source of reversal. I show that well-known weekly contrarian profits are followed by a long run of momentum profits. In fact, these profits are strong enough to produce a significant momentum effect over the full year following portfolio formation. Thus, the market does not appear to view extreme weekly returns as excessive, as implied by an overreaction story. To the contrary, this return continuation is consistent with underreaction to the news driving extreme weekly returns. This is supported by cross-sectional tests in which I find this week's news is positively related to next week's returns. The evidence presented here is consistent with growing evidence that underreaction to firm-specific information is a pervasive feature of price formation. Therefore, if any short-run contrarian profits can be realized, they are better viewed as compensation for providing liquidity than as a reward for arbitrage.

Cette étude examine l'impact de l'incertitude sur l'optimisme et l'excès de confiance des analystes financiers. Les données de l'étude, portant sur 1758 entreprises européennes et concernant la période 1997-2007, sont extraites de la base de données IBES. Étant donné que notre période d'étude comporte un krach boursier en 2000-2001, nos tests ont été conduits avec une comparaison pré- et post-krach. L'incertitude informationnelle est approximée par l'intensité technologique. Les firmes technologiques se caractérisent par une forte incertitude liée à l'aboutissement de leurs projets d'innovations, à la forte rapidité de l'évolution technologique et au traitement comptable spécifique des immatériels. La dispersion de prévision a été utilisée aussi comme deuxième mesure de l'incertitude pour tester la robustesse des résultats fondés sur le caractère technologique. A travers les études antérieures, nous avons montré que l'optimisme des analystes est, d'une part, rationnel, justifié par des incitations économiques et, d'autre part, comportemental expliqué par le phénomène de sur-réaction aux bonnes informations et la sous-réaction aux mauvaises informations. L'excès de confiance implique une sur-réaction aux informations privées et une sous-réaction aux informations publiques. Les résultats obtenus, en se basant essentiellement sur les méthodologies développées par Amir et Ganzach (1998), Easterwood et Nutt (1999) et Bessière et Kaestner (2008), montrent un optimisme plus fort pour les firmes technologiques qui s'atténue et même disparait lorsque l'horizon de prévision décroit. Le phénomène de sur-réaction aux bonnes informations et de sous-réaction aux mauvaises informations est plus fort pour les firmes de haute technologie lorsque l'horizon de prévision est lointain. Ce phénomène s'affaiblit plus pour ces firmes et même s'inverse lorsque la date d'annonce du bénéfice estimé s'approche. Les erreurs de prévisions deviennent moins optimistes pour ces firmes voir même pessimiste. L'analyse pré- et post-krach montre un optimisme plus fort pour les entreprises de haute incertitude avant le krach. Cet optimisme disparait après le krach essentiellement chez les firmes technologiques. L'étude montre aussi que les analystes sur-réagissent plus aux informations privées et sous-réagissent plus aux informations publiques relatives aux entreprises de haute technologie par rapport à celles de faible technologie. L'excès de confiance disparait progressivement à mesure que l'on s'éloigne de la date d'annonce de l'information publique. Il s'atténue significativement après le krach plus précisément pour les firmes de haute technologie. Les résultats trouvés avec la distinction haute/faible dispersion sont globalement similaires à ceux trouvés avec la décomposition haute/faible technologie. / This study examines the impact of uncertainty on optimism and overconfidence of the financial analysts. The data of the study, concerning 1758 European firms and covering the period 1997-2007, are extracted from the IBES database. Given that our period of study includes a stock market crash in 2000-2001, our tests were driven with a comparison pre- and post-crash. Informational uncertainty is approximated by technological intensity. The high-tech firms are characterized by a strong uncertainty linked to the culmination of their innovation projects, to the strong speed of technological evolution and to the specific accounting treatment of the intangible investments. The forecast dispersion was also used as second measure of uncertainty to test the robustness of results based on the technological characteristic.Across the previous studies, we showed that the optimism of the analysts is, on the one hand, rational justified by economic instigations and, on the other hand, behavioral explained by the phenomenon of over- reaction to good information and under-reaction to bad information. The overconfidence implies an over- reaction to private information and under-reaction to public information. Our results, principally based on the methodologies developed by Amir and Ganzach (1998), Easterwood and Nutt (1999) and Bessiere and Kaestner (2008), show a stronger optimism for the high-tech firms compared to low-tech firms. This optimism reduces and even disappears when the forecast horizon decreases. The phenomenon of over-reaction to good information and under-reaction to bad information is stronger for the high-tech firms when the forecast horizon is distant. This phenomenon decreases more for these firms and is even reversed when the expected earnings announcement date approaches. The forecast error becomes less optimistic for these firms and even pessimistic with the reduction of forecast horizon. The analysis pre- and post-crash shows a stronger optimism for the high-uncertainty firms before crash. This optimism disappears after crash principally to the high-tech firms.Study shows also that the analysts over-react more to the private information and under-react more to the public information related to the high-tech firms in comparison with low-tech firms. The over confidence disappears progressively as one moves away from the announcement date of public information. It decreases significantly after crash especially for the high-tech firms. The results found with the differentiation high / low dispersion are on the whole similar to those found with the decomposition high/low-tech.

Neste trabalho é proposto um novo modelo para análise empírica de sobre-reação e sub-reação no mercado de ações. O modelo proposto é baseado em uma técnica de classificação de padrão fuzzy, que permite estabelecer uma relação com as heurísticas de representatividade e ancoramento, oriundas da teoria de finanças comportamentais. O modelo é usado para classificar ações com base nos índices financeiros de companhias abertas. Resultados numéricos ilustram o procedimento de análise para ações do setor de petróleo/petroquímica e do setor têxtil do mercado brasileiro, com indicadores financeiros relativos ao período de 1994 a 2005. / In this work a new model for empirical analysis of stock market overreaction and underreaction is proposed. Such model is based on a fuzzy pattern classification technique, which is strongly connected to the representativeness and anchoring heuristics from behavioral finance. The proposed model is used for stock classification by exploring financial ratios of public companies. Numerical results illustrate the analysis procedure in the cases of the petroleum/petrochemical and textile stocks from the Brazilian market, with financial ratios ranging from 1994 to 2005.

學者研究如何去預測整個市場或是特定產業的走勢，而時間序列的使用在這個層面也被廣泛使用。實務者跟學者近年來也逐漸開始專注於技術指標的分析上，而近年來越來越多的理論架構是根據行為財務學而產生，像是過度反應以及反應不足。本篇研究的目的旨在透過實務與學術的交流，將兩者的優勢互相結合，首先在詳述其投資策略之方法論後，將其想法透過時間序列分析模型化，再則藉由專業的學術訓練修改其模型。簡言之，本篇論文結論如下:1.實務者的方法確實有其洞察力與預測上的價值。2.在將變數從動能訊號轉換成外資持有市值後，回歸的結果確實有進一步的改善。3.外國機構投資人在不同的產業所持有之市值比例確實在統計上顯著受到報酬率跟現金比率兩者的T值與係數之影響。 / Abstract Researchers have been exploring the subject of how to forecast the trends of overall market and certain of sectors from adequate information so far. The analysis and forecasting of time series are also extensively utilized in a variety of applications. Not only practitioners but also academics have been focus on technical indicators. And recently more and more theories based on behavior finance, such as overreaction and underreaction. This study is attempting to investigate the approach of an analyst according to time series analysis, and especially concerned about the momentum indicator, which is combined with overreaction and underreaction. Briefly, our conclusions are as follows： 1. The practitioner’s approach does really have its insight and predictive value. 2. After replacing variable from signals to holdings, the regression results have been improved. 3. We could indicate that the market value holding percentage of FINI in different sectors do really have significant influence toward T-statistics and coefficients of returnt-1 and cash ratiot-1. Key words：Momentum, Overreaction, Underreaction.

The purpose of this paper is to explore the relationship between the returns of momentum strategies and macroeconomic factors. The empirical results indicate that the phenomenon of underreaction is found in Taiwan stock market in the short term, and adoption of momentum strategies can slightly result in significant positive abnormal returns, while no phenomenon of overreaction is found in the long term, and no significant positive abnormal returns are gained if the contrarian strategies are applied. After dividing the market status into bull market and bear market, we find that the underreaction phenomenon appears in the bull market in the short term, and significant positive returns may be gained if the momentum strategies are used; on the other hand, the overreaction phenomenon appears in the bear market in the long term, and the adoption of contrarian strategies may offer significantly positive returns. In addition, either positive or negative excess returns of momentum strategies are found in the bull and bear markets. The value of (alpha) of the returns mostly exceeds zero after the adjustment of Fama and French three-factor model. Finally, the predictive value of macroeconomic analysis and the analysis of returns of momentum strategies reveal that the rate of return of momentum strategies is higher when the expected rate of return of macro economy in the bull market is getting lower, and the rate of return of momentum strategies is lower while the expected rate of return of macro economy in the bear market is high. We conclude that macroeconomic factors are unrelated to the returns of momentum strategies, regardless of bull market or bear market.

Neste trabalho é proposto um novo modelo para análise empírica de sobre-reação e sub-reação no mercado de ações. O modelo proposto é baseado em uma técnica de classificação de padrão fuzzy, que permite estabelecer uma relação com as heurísticas de representatividade e ancoramento, oriundas da teoria de finanças comportamentais. O modelo é usado para classificar ações com base nos índices financeiros de companhias abertas. Resultados numéricos ilustram o procedimento de análise para ações do setor de petróleo/petroquímica e do setor têxtil do mercado brasileiro, com indicadores financeiros relativos ao período de 1994 a 2005. / In this work a new model for empirical analysis of stock market overreaction and underreaction is proposed. Such model is based on a fuzzy pattern classification technique, which is strongly connected to the representativeness and anchoring heuristics from behavioral finance. The proposed model is used for stock classification by exploring financial ratios of public companies. Numerical results illustrate the analysis procedure in the cases of the petroleum/petrochemical and textile stocks from the Brazilian market, with financial ratios ranging from 1994 to 2005.

No description available.

This dissertation consists of two essays. In the first essay, I examine the source of momentum in stock returns. The reversal of momentum returns has been interpreted as evidence that momentum results from delayed overreaction to information. I examine momentum and reversals conditional on firms’ share issuance (net of repurchases) during the momentum holding period and show that (1) among losers, the momentum returns are statistically significant, but the reversals are non-existent, for both issuers and non-issuers; (2) among winners, momentum and reversals are restricted to issuers, but are non-existent among non-issuers. After further conditioning on firm size, I find that winner reversals are restricted to small, equity issuing firms. After excluding these small issuers from the sample, the remaining firms have strong momentum profits with no accompanying reversals. The evidence suggests that the return reversals are a manifestation of the poor performance of equity issuing firms. Further, while investor overreaction potentially contributes to the momentum among winners, a large fraction of firms do not earn any significant abnormal returns following initial price continuation, suggesting that underreaction, and not delayed overreaction to information, is the dominant source of momentum in stock returns. In the second essay, I examine alternative explanations of reversals in stock returns. George and Hwang (2007) find that long-term reversals in stock returns are driven by investors’ incentive to defer payment of taxes on locked-in capital gains rather than by overreaction to information. I show that return reversals are instead attributable to the negative relationship between firms’ composite share issuance and future stock returns documented in Daniel and Titman (2006). The ability of locked-in capital gains measures to forecast stock returns is largely subsumed by the composite share issuance measure. My results do not support the hypothesis that capital gains taxes drive long-term return reversals. / Ph. D.