Scenario-free analysis of financial stability with interacting contagion channels
提出一种基于特征值的无场景方法，分析多个交互传染渠道对金融系统稳定性的影响，发现交互渠道的不稳定性远超各渠道单独作用之和，并推导出大机构数量下的稳定性阈值公式。
Financial stress tests that capture multiple interactions between contagion channels are conditional on specific, subjectively-imposed stress scenarios. Eigenvalue-based approaches, in contrast, provide a scenario-independent measure of systemic stability, but so far only handle a single contagion mechanism. We develop an eigenvalue-based approach that brings the best of both worlds, enabling the analysis of multiple interacting contagion channels without the need to impose a subjective stress scenario. Our model captures the solvency-liquidity nexus, which allows us to demonstrate that the instability due to interacting channels can far exceed that of the sum of the individual channels acting in isolation. The framework we develop is flexible and allows for calibration to the microstructure and contagion channels of real financial systems. Building on this framework, we derive an analytic stability criterion in the limit of a large number of institutions that gives the instability threshold as a function of the relative size and intensity of contagion channels. This analytical formula requires comparatively little data to elucidate the mechanisms that drive instability in real financial systems and thus complements the insights gained from traditional stress tests.