Abstract
The formation, ampli2ication and transmission of systemic risk in Indian banks are examined by integrating dynamic panel GMM estimation, VAR, CoVaR, network analysis and Markov-switching models. The analysis uses annual banklevel and macro-2inancial risk variables covering 2015 to 2022 and a network sample of 24 Indian banks (11 publicsector and 13 private-sector), excluding foreign banks and institutions not represented in the analytical network. Market-based analysis over the same study period indicates that systemic risk is procyclical, time-varying and concentrated among a small group of highly connected institutions. Non-performing assets, leverage and rapid credit growth emerge as important vulnerability channels, whereas stronger capital and liquidity buffers are associated with greater resilience. VAR analysis and Granger-causality tests indicate feedback between credit growth and asset-quality deterioration. CoVaR and network measures show that systemic importance is associated not only with balance-sheet size but also with centrality and bridging positions in the banking network. Markov-switching estimates indicate stronger risk transmission during stressed regimes. The integrated framework therefore supports macroprudential surveillance, stress testing, capital and liquidity planning and early-warning assessment in the Indian banking system.
Keywords: CoVaR, Financial Stability, Indian Banking Sector, Network Analysis, Systemic Risk.