Expected risk and excess returns predictability in emerging bond markets

Chih Ling Lin, Ming Chieh Wang, Yin Feng Gau

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

This study examines the predictability of expected excess returns from eight emerging bond markets within an international asset pricing framework. Two sets of instruments are used, which include both world and local factors, to forecast emerging bond returns. Besides investigating the influence of the macroeconomic factors in specific countries on bond returns in those countries, this study also divides local factors into macroeconomic and financial factors. Unlike previous studies, we apply macroeconomic instruments that contain more information on excess returns as a proxy for local risk factors via principal component analysis methodology. The information variable approach enables the prediction of excess bond returns based on world and local factors and facilitating understanding of the degree of integration between emerging bond markets and developed bond markets. The results indicate that the bond market in emerging world is partially integrated to that in the developed world and the predictability of local factors that include both financial and macroeconomic information variables can forecast around 25-66% of the returns of emerging bonds. Incorporating the macroeconomic variables increases the explanatory power of the model. Both world and country-specific local instruments can forecast excess bond returns, but local instruments appear to be better predictors of such returns, particularly the local credit spread to US. Additionally, this study finds that investor risk aversion is significant among most of sample countries.

Original languageEnglish
Pages (from-to)1511-1529
Number of pages19
JournalApplied Economics
Volume39
Issue number12
DOIs
StatePublished - Jul 2007

Fingerprint

Dive into the research topics of 'Expected risk and excess returns predictability in emerging bond markets'. Together they form a unique fingerprint.

Cite this