Abstract:
This article applies a bootstrap rolling-window causality test to assess the causal relationship
between economic policy uncertainty (EPU) and stock returns in China and India. Empirical literature
examining causality between two time series may suffer from inaccurate results when the underlying
full-sample time series have structural changes. However, the bootstrap rolling-window approach enables
us to identify possible time-varying causalities between time series based on sub-sample data. Using a
twenty-four-months rolling window over the period 1995:02 to 2013:02 in China and 2003:02–2013:02
in India, we do find that there are bidirectional causal relationships between EPU and stock returns in
several sub-periods rather than in the whole sample period. However, the association between EPU and
stock returns is, in general, weak for these two emerging countries. Our findings have important implications
for policy makers and investors.