| A better factor model based on the investigation of anomalies in European stock market (Capture des anomalies par un modèle multifacteur amélioré : une preuve européenne) Xue, Ke - (2020-11-16) / Universite de Rennes 1 A better factor model based on the investigation of anomalies in European stock market
| |||
Langue : Anglais Directeur(s) de thèse: Lilti, Jean-Jacques Discipline : Sciences de gestion Ecole Doctorale : Sciences économiques et sciences de gestion Classification : Gestion et organisation de l'entreprise Mots-clés : Modèle multifactoriel, Risque idiosyncratique, Sentiment des investisseurs
| |||
Résumé : CAPM a été largement utilisé sur les marchés financiers, mais ses hypothèses sont trop strictes et irréalistes, ce qui entraîne l'échec de son application. En assouplissant les hypothèses de CAPM sur la diversification et la rationalité, cet article reconstruit le modèle par différentes manières pour améliorer les performances. Basé sur le Fama-French multifactoriel modèle , cet article mène une recherche empirique sur le marché boursier européen. Les résultats montrent que l'inclusion risque domestique, risque idiosyncratique et sentiment des investisseurs dans le modèle peut effectivement améliorer la performance du modèle, ce qui est très important pour le budget d'investissement, l'évaluation du portefeuille et l'analyse des risques. Abstract : The Capital Asset Pricing Model (CAPM) describes the relationship between the expected return of asset and relative systematic risk. It is widely used throughout finance. However, CAPM has serious limitations in real world, as most of the assumptions, are unrealistic. Sometimes, diversified investment is infeasible due to various restriction in practice. Besides, the participants of financial market are not always rational, which lead to the invalid of CAPM and lots of failure of application in practice. By relaxing its assumptions, this thesis explored three different method for improving the performance of asset pricing model in serving the financial practice. Firstly, this thesis explored the method that adding additional systematic risk factors. Based on the research of Fama and French(1993,2015), this thesis empirically investigate the local risk factor in Britain, French and Germany stock market. Moreover, we run the race between domestic model and European model. The second attempts is based on our investigation about the relationship between stock’s idiosyncratic risk and its return in European market. This thesis reconstruct the model by including idiosyncratic risk factor. At the third attempt, This thesis exceeded the rationality hypothesis and took the investor sentiment into consideration. This thesis construct a conditional multi-factor model, in which the Beta is dynamic. This thesis have three main finding. Firstly, the risk factor mimicking portfolio have different performance in UK, France and German. The domestic multi-factor model outperformed the European model in explaining the performance of local assets’ return. Secondly, the low risk effect exist in European stock market, which could not been explained by Fama and French five factor model. This gap could be filled by reconstructed the model by adding idiosyncratic risk factor. Thirdly, the European stock market is affected by investor sentiment. Bringing the investor sentiment information into the multi-factor model could effectively improve the model’s performance in explaining the performance of European stock’s return. The main contribution of this thesis is that constructing a better multi-factor asset pricing model according to three different method. This thesis empirically proved that incorporating the domestic risk factor, idiosyncratic risk factor and investment sentiment information into the multi-factor model could promote model’s performance in pricing financial assets. | |||