HUB-FS Working Paper Series FS-2021-E-001 VAR-Based Turbulence Index: Change Point Detection for Upside Investments (original) (raw)

Kritzman and Li (2010) introduced the turbulence index (TI) based on the Mahalanobis distance for capturing the degree of multivariate asset price “unusualness” over time. We consider adding a sign to the unusualness indicator to detect change points toward both bull and bear markets. We find that unusualness is driven by government bonds when TI is extended to vector autoregression-based TI (VTI) to eliminate autocorrelation structure and introduce signed VTI (SVTI) based on this finding. Our simulation with simple dynamic asset allocation strategies using the TI, SVTI, and a static counterpart suggests that SVTI could enhance performance compared with the other portfolios.

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.