Inter-Asset-Class Volatility - A Forward Looking Measure Rooted in Investors' Realities

Sammanfattning: The aim of this study is to better understand the causes of repeated periods of extreme volatility in financial systems The study conceptualizes and quantifies Inter-Asset-Class Volatility (IAV) as a for-ward looking proxy for volatility in financial system to reframe the phenomenon in a reality oriented setting. IAV aims to capture the total revaluation potential arising from reallocations among all asset classes in the analyzed system. The cause for such volatility is related to investors' preferences best detectable in the objective functions of institutional investors that are implicitly and explicitly con-strained. Constraint induced deviations in portfolio weights from those of a purely mean-variance optimizing investor are argued to cause unexpected volatility by forcing investors to leave the market as soon as they reach their allocation limits. Consequently, the market shifts from a full-participation equilibrium to a limited-participation equilibrium, in which the price impact of trading rises due to a significantly reduced market depth. Building on Pastor and Stambaugh (2003), the resulting liquidity shortage in the primarily affected asset classes is related to strongly negative returns therein. Liquid-ity and relative valuation spill-overs to other asset classes create the system wide phenomenon IAV seeks to define. To detect evidence for this conjecture, the measures of Delta and Tau are introduced. Delta quantifies the constraint-induced deviations from the mean-variance efficient portfolio that are AuM weighted to arrive at the potential buying or selling pressure from an investor in a respective asset class termed Tau. A simulation study involving stocks, bonds, and cash forms the basis for a correlation analysis between Tau values aggregated from three modeled investor types with Pastor and Stambaugh's (2003) aggregate liquidity series. confirming the intuition. While correlations over the entire sample period remain insignificantly small, strong negative correlations are found in the 12-month windows around a set of 9 events characterized by extremely low liquidity levels and se-verely negative stock returns. The idea that build ups in aggregated Tau potentially forecast volatility regime shifts is, thereby, supported, pointing to investor preferences as a possible source of excess volatility in financial markets.