This study explores the implications of prompt corrective action and the Troubled Asset Relief Program on the behaviour of non-performing loans (NPLs) and real estate non-performing loans (RELs) in the United States over 1984?2015 using a Markov switching framework. We find that NPLs and RELs exhibit pronounced episodic behaviour switching between non-stationary and stationary regimes. Prompt corrective action and the Troubled Asset Relief Program have a significant impact on banking sector stability by influencing the probability of switching from non-stationary regimes to stationary regimes and by reducing the level of NPLs and RELs. These results are robust to various model specifications and have important implications for bank regulation as well as for the formulation of macro stress testing.
- banking stability
- Markov switching model
- prompt corrective action (PCA)