Regional, global, and hybrid asset-pricing models in heterogeneously integrated regions: evidence from local industries

This study evaluates the performance of global, regional, and hybrid asset pricing models across 67 countries, employing the most refined regional classification to date, which captures both geographic and developmental distinctions. The results indicate that hybrid and regional models consistently outperform global models, which generate the largest pricing errors. Sub-period analyses reveal no evidence of increasing dominance of global factors, showing that full financial integration has not yet been achieved. The best-performing individual model varies by region. Regional models are more effective in segmented markets, while hybrid models perform better in partially integrated ones, indicating that model effectiveness depends on the degree of regional integration. Factor-spanning tests indicate that the market factor is the most globalized, followed by momentum, while investment remains largely regional. A novel globalization measure highlights high integration in North America, Western Europe, and developed Asia-Pacific, contrasted with persistent segmentation in MENA, Latin America, and Eastern Europe.