Ill., Northwestern Univ., Diss.--Evanston, 2003. / Kopie, ersch. im Verl. UMI, Ann Arbor, Mich. - Enth. 4 Beitr.

Calif., Univ., Dep. of Economics, Diss.--Stanford, 2003. / Kopie, ersch. im Verl. UMI, Ann Arbor, Mich. - Enth. 3 Beitr.

Conn., Yale Univ., Diss.--New Haven, 2003. / Kopie, ersch. im Verl. UMI, Ann Arbor, Mich. - Enth. 3 Beitr.

Univ., Habil.-Schr.--Erfurt, 2004.

Zugleich: Diss. Regensburg, 2008. / Literaturverz.

Au cours de la dernière décennie, les chercheurs se sont intéressés au comportement de l'investisseur sur les marchés boursiers. Nombreuses théories en psychologie et sociologie sont mobilisées en finance comportementale pour dépasser les limites de l'hypothèse de l’efficience des marchés et expliquer la fragilité des marchés financiers. En analysant le comportement grégaire (mimétisme), les chercheurs tentent d'expliquer les anomalies et les grands mouvements du marché. Le comportement grégaire peut être décrit comme la tendance d'un investisseur ou d’un groupe d'investisseurs, à imiter les actions des autres acteurs du marché, ou à suivre la tendance du marché. Notre première étude examine la présence du comportement grégaire, en se concentrant sur le marché boursier français, au niveau du marché et au niveau des secteurs. Nous étudions le mimétisme selon différentes conditions macroéconomiques. Nous testons aussi l'existence du mimétisme au cours de la dernière période de crise financière et pendant les périodes caractérisées par des volumes (élevés ou faibles) de volatilité et de volume de transaction. En utilisant le modèle de CH 95, nous ne détectons pas de comportement grégaire, à la fois au niveau du marché qu’au niveau des secteurs, pendant les mouvements de marché extrêmes. Les modèles de CCK 2000 et Hwang et Salmon 2004 montrent des résultats mitigés. Même lorsque le comportement grégaire est détecté au niveau du marché, les secteurs se comportent différemment. La mesure que nous extrayons du state-space model montre différents niveaux de mimétisme dans les secteurs. La deuxième étude examine le mimétisme, sur les marchés boursiers américains et chinois, en introduisant une nouvelle dimension : l'asymétrie d'information. Nous utilisons plusieurs mesures de la disponibilité de l'information : la politique de dividende, le bid-ask spread, la taille de l'entreprise, la sophistication du marché, ainsi que l'état du marché (pré, post et pendant la période de crise). Cette étude nous permet d’analyser le mimétisme dans différents contextes selon la disponibilité de l'information. Elle permet de vérifier si le mimétisme est plus prononcé dans un contexte d'asymétrie d'information élevée. D’une part, les résultats du modèle de CH 95 ne montrent aucune preuve de mimétisme quel que soit le niveau d'asymétrie d'information entre les entreprises et les investisseurs, à la fois pour les marchés américain et chinois. D'autre part, le modèle CCK 2000 détecte des différences de niveau de mimétisme dans le marché boursier chinois en fonction du niveau d'asymétrie d'information. Les résultats suggèrent que les marchés émergents sont touchés par le mimétisme pendant la période de crise, quelle que soit la taille de l'entreprise. Enfin, le modèle de Hwang et Salmon 2004 montre différents niveaux de mimétisme dans les marchés américain et chinois, en fonction du niveau d'asymétrie d'information. Il est intéressant pour la recherche sur la modélisation des marchés boursiers d’examiner le comportement grégaire des investisseurs. De la même façon, les décideurs politiques pourraient être intéressés par les effets perturbateurs potentiels du mimétisme sur les marchés financiers. / Over the last decade, the academic research has highly focused on examining the investor’s behavior in stock markets. Many theories in psychology and sociology are used in the so called “Behavioral Finance” in order to explain the limits of the efficient market hypothesis and the financial market fragility. By analyzing the herding behavior, the researchers try to explain the market anomalies and the large market movements. Herding behavior can be described as the tendency of an investor, or a group of investors, to imitate the actions of other market participants, or to follow the market movement. Our first study examines the presence of herding behavior, focusing on the French stock market, at both market and sector levels. We investigate herding during different macroeconomic conditions. We also test the existence of herding during the last financial crisis period, and during the periods characterized by high or low volatility and transaction volumes. Using the CH 95 model, we do not observe herding behavior, both in market and sector levels, during extreme market movements. The CCK 2000 and Hwang and Salmon 2004 models show mixed results. Even when herding exists in the market level, various sectors behave differently. The measure we extract from the state-space model shows different patterns of herding at the sector level. The second study examines the investors’ incentives behind herding, in the US and Chinese stock markets, by introducing a new dimension, which is the information asymmetry. Using several proxies for information availability, such as dividend policy, bid ask spread, firm size and market sophistication along with considering the market condition (pre, post and during crisis period), this study allows us to investigate herding in different contexts of information availability, and to examine if herding is more pronounced in a high information asymmetry context. Findings of CH 95 model show no evidence of herding regardless of the level of information asymmetry between firms and investors in both the US and Chinese stock markets. On the other hand, the CCK 2000 model detects herding differences in the Chinese stock market depending on the information asymmetry level. The findings suggest that the emerging markets are affected by herding during the crisis period, regardless of the firm size. Finally, the Hwang and Salmon 2004 model shows different herding patterns in the US and Chinese stock markets depending on the information asymmetry level. Examining the herding behavior is interesting for the research in the market modeling field along with policymakers who may be interested in investigating the potential disturbing effects of herding on stock markets.

Title: Robust Monitoring Procedures for Dependent Data Author: Ondřej Chochola Department: Department of Probability and Mathematical Statistics Supervisor: Prof. RNDr. Marie Hušková, DrSc. Supervisor's e-mail address: huskova@karlin.mff.cuni.cz Abstract: In the thesis we focus on sequential monitoring procedures. We extend some known results towards more robust methods. The robustness of the procedures with respect to outliers and heavy-tailed observations is introduced via use of M-estimation instead of classical least squares estimation. Another extension is towards dependent and multivariate data. It is assumed that the observations are weakly dependent, more specifically they fulfil strong mixing condition. For several models, the appropriate test statistics are proposed and their asymptotic properties are studied both under the null hypothesis of no change as well as under the alternatives, in order to derive proper critical values and show consistency of the tests. We also introduce retrospective change-point procedures, that allow one to verify in a robust way the stability of the historical data, which is needed for the sequential monitoring. Finite sample properties of the tests need to be also examined. This is done in a simulation study and by application on some real data in the capital asset...

Submitted by Laura Simonsen Leal (arula@fgvmail.br) on 2016-06-22T12:59:34Z No. of bitstreams: 1 laura_tese14(final).pdf: 1036208 bytes, checksum: eac8007047195b00593f30884e72a3e2 (MD5) / Approved for entry into archive by GILSON ROCHA MIRANDA (gilson.miranda@fgv.br) on 2016-06-22T13:18:16Z (GMT) No. of bitstreams: 1 laura_tese14(final).pdf: 1036208 bytes, checksum: eac8007047195b00593f30884e72a3e2 (MD5) / Approved for entry into archive by Marcia Bacha (marcia.bacha@fgv.br) on 2016-06-29T13:38:50Z (GMT) No. of bitstreams: 1 laura_tese14(final).pdf: 1036208 bytes, checksum: eac8007047195b00593f30884e72a3e2 (MD5) / Made available in DSpace on 2016-06-29T13:39:56Z (GMT). No. of bitstreams: 1 laura_tese14(final).pdf: 1036208 bytes, checksum: eac8007047195b00593f30884e72a3e2 (MD5) Previous issue date: 2016-03-21 / The main purpose of this paper is to propose a methodology to obtain a hedge fund tail risk measure. Our measure builds on the methodologies proposed by Almeida and Garcia (2015) and Almeida, Ardison, Garcia, and Vicente (2016), which rely in solving dual minimization problems of Cressie Read discrepancy functions in spaces of probability measures. Due to the recently documented robustness of the Hellinger estimator (Kitamura et al., 2013), we adopt within the Cressie Read family, this specific discrepancy as loss function. From this choice, we derive a minimum Hellinger risk-neutral measure that correctly prices an observed panel of hedge fund returns. The estimated risk-neutral measure is used to construct our tail risk measure by pricing synthetic out-of-the-money put options on hedge fund returns of ten specific categories. We provide a detailed description of our methodology, extract the aggregate Tail risk hedge fund factor for Brazilian funds, and as a by product, a set of individual Tail risk factors for each specific hedge fund category.

Submitted by Eduardo Toledo (dudutol@hotmail.com) on 2016-09-15T14:15:58Z No. of bitstreams: 1 Dissertação Eduardo Rietmann Toledo - Final.pdf: 4612894 bytes, checksum: ab143894eec8b23502f2222a0b9a5a27 (MD5) / Approved for entry into archive by GILSON ROCHA MIRANDA (gilson.miranda@fgv.br) on 2016-09-21T14:12:18Z (GMT) No. of bitstreams: 1 Dissertação Eduardo Rietmann Toledo - Final.pdf: 4612894 bytes, checksum: ab143894eec8b23502f2222a0b9a5a27 (MD5) / Approved for entry into archive by Marcia Bacha (marcia.bacha@fgv.br) on 2016-09-23T12:48:52Z (GMT) No. of bitstreams: 1 Dissertação Eduardo Rietmann Toledo - Final.pdf: 4612894 bytes, checksum: ab143894eec8b23502f2222a0b9a5a27 (MD5) / Made available in DSpace on 2016-09-23T12:49:08Z (GMT). No. of bitstreams: 1 Dissertação Eduardo Rietmann Toledo - Final.pdf: 4612894 bytes, checksum: ab143894eec8b23502f2222a0b9a5a27 (MD5) Previous issue date: 2016-05-31 / The aim of this study is to assess the impact of credit risk in the Brazilian stock cross-section return, and evaluate if a strategy based on this feature is able to generate positive and significant alpha. To measure credit risk, credit ratings were used assigned to Brazilian companies by the american agency Standard and Poors in the period between 2009 and 2014. First we divided ali firms in portfolios according to their credit risk and after we analyzed monthly returns of their stocks. Empirically, we find an evidence that the credit risk has a positive correlation in stock returns. However, as the strategy to operate based on credit risk was not able to generate significant alpha, it was not possible to statistically validate this effect on return. / O objetivo deste estudo é verificar o impacto do risco de crédito no retorno cross-section acionário brasileiro e avaliar se uma estratégia baseada nesta característica é capaz de gerar alfa positivo e significativo. Para mensurar o risco de crédito, foram utilizadas as notas de crédito atribuídas às empresas brasileiras pela agência americana Standard and Poors no período entre 2009 e 2014. Dividimos as empresas em portfolios de acordo com seu risco de crédito e analisamos os retornos mensais de suas ações. Empiricamente, encontramos indícios que o risco de crédito possui uma correlação positiva com o retorno acionário. No entanto, como a estratégia para operar baseada no risco de crédito não foi capaz de gerar alfa significativo, não foi possível validar estatisticamente este efeito sobre o retorno.

Submitted by Carlos Roberto Simões Chaves (carloschaves_88@hotmail.com) on 2017-07-20T14:14:35Z No. of bitstreams: 1 Dissertação_Carlos Chaves_final_entrega.pdf: 1016182 bytes, checksum: e5eeabf21bc225b6b1308e739fd8bf80 (MD5) / Approved for entry into archive by GILSON ROCHA MIRANDA (gilson.miranda@fgv.br) on 2017-08-29T15:27:45Z (GMT) No. of bitstreams: 1 Dissertação_Carlos Chaves_final_entrega.pdf: 1016182 bytes, checksum: e5eeabf21bc225b6b1308e739fd8bf80 (MD5) / Made available in DSpace on 2017-09-06T19:49:12Z (GMT). No. of bitstreams: 1 Dissertação_Carlos Chaves_final_entrega.pdf: 1016182 bytes, checksum: e5eeabf21bc225b6b1308e739fd8bf80 (MD5) Previous issue date: 2017-05-19 / This paper examines the impact of expected inflation on stock returns and earnings per share projections for the next 12 months. We used the Ibovespa's weekly real returns and FOCUS survey for IPCA and Industrial Production growth. A one-percentage point increase in projected inflation over the next 12 months is associated with a decline of 0.56 percentage points in the weekly Ibovespa real change to a significance level of 1%. No statistically significant relationships were found between the expected inflation and the projections for Ibovespa's profits. It was verified that the Ibovespa's weekly returns also react negatively to the 5-year CDS oscillations and the VIX index. / Este trabalho examina o impacto da inflação esperada sobre os retornos das ações e as projeções de lucros por ação para os próximos 12 meses. Utilizamos os retornos reais semanais do Ibovespa e as expectativas da pesquisa FOCUS para o IPCA e crescimento da Produção Industrial. Um aumento de 1 ponto percentual na inflação projetada paras os próximos 12 meses está associado a um declínio de 0.56 pontos percentuais na variação real semanal do Ibovespa para um nível de significância de 1%. Não foram encontradas relações estatisticamente significativas entre a inflação esperada e as projeções para os lucros das empresas do Ibovespa. Verificou-se que os retornos semanais do Ibovespa também reagem negativamente às oscilações do CDS de 5 anos e o índice VIX.