Title: The short-run and long-run components of financial market volatility
Authors: Norbert Metiu - Deutsche Bundesbank (Germany) [presenting]
Giovanni Motta - Pontificia Universidad Catolica de Chile (Chile)
Abstract: An Evolutionary Latent Factor GARCH (ELF-GARCH) model is introduced which captures the volatility common to a large panel of financial variables by a small number of common volatilities. Each common volatility is decomposed into the product of two components which display different degrees of persistence. The first component reflects short-run dynamics, while the second component captures persistent movements in volatility. We develop a semi-parametric estimation theory, and the finite-sample performance of our estimators is investigated considering a variety of simulation scenarios. Our empirical illustration shows that the bulk of the variation in 157 monthly U.S. financial variables is captured by a single common factor. The short-run volatility of this factor gauges widely used proxies of risk and uncertainty in financial markets, such as credit spreads and the VIX index, while its long-run volatility reaches elevated levels around major financial crises. We embed the two volatility components into a standard macro-financial vector autoregression. An unexpected rise in short-run volatility is followed by dynamic responses of output and employment that closely resemble those typically associated with an adverse uncertainty shock. Long-run volatility Shocks trigger a significant decline in output and employment at the 2-5 year horizon, which highlights the role of long-run risks in shaping business cycle fluctuations.