Identifying systemic risk of assets during international financial crises using Value at Risk elasticities

Abstract

Since their emergence in the late 1980s, the Value at Risk models have quickly become the global standard in risk management and forecasting. However, their role during global financial crises has raised the concern about the effectiveness of such tools in managing risk during extreme events while some researchers even attribute global financial contagion to Value at Risk practices. We propose a new measure called Value at Risk elasticity to measure the responsiveness of a portfolio's Value at Risk to a percentage change in the weight of each investment asset. Using BIS quarterly data on consolidated, international claims, we test our new measure with five historical financial crises. Specifically, we estimate Value at Risk elasticities for the portfolios of major banking centers prior to each crisis as an indicator for an increased risk factor in assets and contrast our predictions of high-risk assets with the actual losses in the aftermath of the crises using stock market data. The results show that our proposed measure performs well in the prediction of asset losses during several global crises. Based on our findings, we recommend including Value at Risk elasticity as an index to better manage risks in portfolios.