Working Paper: NBER ID: w9510
Authors: Geert Bekaert; Campbell R. Harvey; Angela Ng
Abstract: Contagion is usually defined as correlation between markets in excess of what would be implied by economic fundamentals; however, there is considerable disagreement regarding the definitions of the fundamentals, how the fundamentals might differ across countries, and the mechanisms that link the fundamentals to asset returns. Our research takes, as a starting point, a two-factor model with time-varying betas that accommodates various degrees of market integration between different markets. We apply this model to stock returns in three different regions: Europe, South-East Asia, and Latin America. In addition to providing new insights on contagion during crisis periods, we document patterns through time in world and regional market integration and measure the proportion of volatility driven by global, regional, and local factors.
Keywords: No keywords provided
JEL Codes: G15; G12; F3; F36; F15
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| factor volatility (G17) | market correlation (C10) |
| increased volatility in U.S. or regional markets (N21) | higher return covariance between these markets and others (G15) |
| increased trade integration (F15) | higher correlations (C10) |
| crises (H12) | increased correlations among markets (G15) |
| higher return covariance between these markets and others (G15) | rejection of the null hypothesis of no contagion (F65) |