Market Turbulence and Financial Contagion in Emerging Economies: A Meta-Analysis

Abstract

Quantitative syntheses of the financial contagion literature are scarce, with frontier markets at the edge of the emerging-market category particularly under-represented. We meta-analyse 312 effect sizes from 87 studies (2002–2024) spanning 26 economies. A three-level random-effects model with cluster-robust variance estimation and Hartung–Knapp–Sidik–Jonkman adjustment returns a pooled standardised contagion coefficient of 0.187 (95% CI: 0.156–0.218; I-squared = 76.4%; tau-squared = 0.022). Multilevel meta-regression attributes 28% of between-study variance to methodological choice: DCC-GARCH and TVP-VAR connectedness measures exceed the Forbes–Rigobon adjustment by 35–45%, consistent with heteroscedasticity-induced bias. Regional moderators diverge sharply: the BRICS bloc registers 0.241, against 0.124 for Central Asia and Mongolia (k = 14), the lowest sub-sample mean. Selective-reporting diagnostics—Egger regression, trim-and-fill, PEESE, p-curve, p-uniform-star and a Vevea–Hedges selection model—indicate moderate but bounded distortion. Leave-one-out and Cook-distance diagnostics confirm that no single study drives the result. We draw three conclusions: methodological choice has first-order consequences for measured contagion; small open frontier markets exhibit attenuated direct co-movement but pronounced commodity-mediated transmission; and Central Asia and Mongolia warrant a focused research agenda calibrated to commodity- and currency-channel measurement.