International equity market integration: evidence from ETFs

We examine international equity market integration using VAR-based rolling cointegration analysis and coefficients of the error correction terms. Applying different sizes of window for the rolling cointegration analysis, we reveal that 500-day (or less) windows might not be long enough to study the cointegration analysis and are more likely to have biased results. We also find three sub-periods showing relatively strong cointegrating relationships, but results for the rest of the sample periods display weak integration among international equity markets. Our findings from the error correction model show that major markets with greater coefficients are more efficient than markets with relatively smaller coefficients.