The financial stress index: identification of systemic risk conditions

This paper develops a financial stress index for the United States, the Cleveland Financial Stress Index (CFSI), which provides a continuous signal of financial stress and broad coverage of the areas that could indicate it. The index is based on daily public-market data collected from four sectors of the fi nancial markets—the credit, foreign exchange, equity, and interbank markets. A dynamic weighting method is employed to capture changes in the relative importance of these four sectors as they occur. In addition, the design of the index allows the origin of the stress to be identified. We compare the CFSI to alternative indexes, using a detailed benchmarking methodology, and show how the CFSI can be applied to systemic stress monitoring and early warning system design. To that end, we investigate alternative stress-signaling thresholds and frequency regimes and then establish optimal frequencies for filtering out market noise and idiosyncratic episodes. Finally, we quantify a powerful CFSI-based rating system that assigns a probability of systemic stress to ranges of CFSI outcomes.