Assessing the Sustainability of China's Foreign Exchange Reserves: A Hybrid DEA-Text Mining Approach, 2020-2024

This study addresses the critical need for a comprehensive assessment of China's foreign exchange reserve sustainability by integrating Data Envelopment Analysis (DEA) and policy text mining.Traditional methods often overlook the dynamic interplay between economic indicators and policy directives.Our hybrid approach fills this gap by combining DEA models (CCR and BCC) to evaluate management efficiency with LDA-based text mining to extract policy themes from official documents.Key findings include an upward trend in technical efficiency (TE) from 0.85 in 2020 to 0.95 in 2024, pure technical efficiency (PTE) reaching 0.98, and scale efficiency (SE) at 0.97, indicating increasing returns to scale.Identified policy themes such as diversified investment and exchange rate stability align with sustainability goals.This integration of quantitative and qualitative analyses offers novel insights and practical guidance for policymakers, highlighting the importance of adaptive management strategies in a volatile global economy.