A Swedish Financial Stress Index

Abstract: In this thesis we develop a daily index that aims to reflect the functionality of the Swedish financial system. We extend previous research about systemic financial stress by introducing the Swedish Financial Stress Index (SFSI), a market data based measure consisting of 14 indicators aggregated into five subindices that reflect different parts of the financial system. Each component is chosen by its connection to key phenomena of financial stress as suggested by research consensus. The subindices are weighted with their time-varying cross-correlations to reflect the detrimental impact of systemic stress on the financial system. We validate our results and choose aggregation methods based on a comparison of the output of different specifications to the results of an academic expert survey. The SFSI peaks in periods classified as highly stressful by the academic experts. A threshold vector autoregression (TVAR) model is utilized to estimate the level of financial stress at which it starts to impact economic growth. The SFSI is constructed to become a macroprudential tool for the Riksbank.