CÍMLAP
Kocsis Zalán
Global, regional and country-specific components of financial market indicators: an extraction method and applications
CONTENTS, ABSTRACT
Contents
Abstract
1 Introduction
2 Methodology and data set
2.1 Factor analysis with Procrustes rotation
2.2 Data set
2.3 Discussion of prior selection
3 General factor analysis results
3.1 The global factors
3.2 Regional factors
4 Applications
4.1 Variance shares of factors
4.2 Online monitoring of components of the Hungarian CDS spread
4.3 Extension to multiple markets
4.4 Channels of risk propagation from the eurozone periphery
5 Conclusions
References
Appendix A Additional tables and figures
Abstract
This paper presents a variance decomposition method - factor analysis with Procrustes rotation - that is capable of separating the global, regional and idiosyncratic components of various financial market indicators. The method is applied to indicators of five key financial markets: sovereign CDS spreads, stock indices, exchange rates, EMBI Global bond spreads and 10-year reference yields of domestic government bond markets. The results support the finding of the literature of a significant global component in most markets, but also point out the importance of regional correlations. Based on the method two practical applications are proposed: one, which is useful in the daily monitoring of financial markets to identify magnitudes of risk premium shocks of global, regional and country-specific origins; and another one, which gauges channels of risk propagation from the eurozone periphery.