PYXIDA Institutional Repository
and Digital Library
Collections :

Title :1. Handling data from Thompson’s DataStream 2. The Cross-Section of European Returns: an application of asset pricing models, anomalies and predictability in a Pan-European framework
Creator :Landis-Conrad, Felix-Michel
Contributor :Skouras, Spyros (Επιβλέπων καθηγητής)
Athens University of Economics and Business, Department of International European Economic Studies (Degree granting institution)
Type :Text
Extent :153p.
Language :en
Abstract :This paper refers on Handling data from Thompson’s DataStream (TDS). Although DataStream is a rich data source for approximately 50.000 equities from more than 65 markets around the world with more than 25 years of coverage, its data should be handle with care. An extensive research on that subject is already published by Ince and Porter (2006-“Individual Equity Return Data from Thompson DataStream: Handle With Care”). So our paper comes as a supplement to it, providing, though more information especially in the common stock identification processes for European markets and proposing extra screens for unreported errors. At the same time it, jointly describes the methodology we applied to obtain the European sample we used in our tests throughout the paper 2 .
This paper concentrates on the performance of asset pricing models in the Euro-area for a universe of European stocks. The Euro area has faced a high number of monetary and policy changes in the recent past as the consequence of the European integration process, and naturally, these developments have important implications for portfolio diversification and asset pricing. So we apply Pan-European versions of asset pricing models: CAPM, Fama - French 3 factor model. In addition we exploit for predictability in European returns and also test for a number of well known asset pricing anomalies such as reversal, momentum, small firm, value and turn of the year (January) effects. Finally we apply newly reported, in the New York stock exchange (NYSE), portfolio optimization processes that enhance return through forming optimal mean variance efficient portfolios with huge ex-ante sharpe ratios.Our analysis clearly rejects CAPM, a European market factor alone is unable to capture most of the variance of average European returns, in a Pan European context. We document very strong value and small firm effects, especially during the period 1999 to 2007, strong Momentum and Reversal effects as long as evidence that European returns are predictable in long horizons. A Pan European Fama-French 3 factor model can capture most of the priced risk of average European stock returns, connected with these attributes, except Momentum. Finally, we apply conditional mean variance optimization processes to a set of Fixed attribute portfolios producing “optimal” portfolios with a huge level of annualized Ex-Ante Sharpe Ratio.
Subject :Handling data
Thompson’s DataStream
Stock identification processes
European markets
Pricing models
Fama–French three-factor model
Date :2008
Licence :

File: Landis_Conrad_Master_Thesis.doc

Type: application/pdf