Asymmetrical Volatility and Spillover Effects
This chapter takes a closer look at the volatility-specific component of beta. Some financial practitioners have advocated the use of relative volatility, standing alone, as a risk measure. This decision would eliminate the correlation component and its insights into the diversification value of financial portfolios. For its part, volatility is asymmetrical and prone to clustering. It tends to be greater on the downside of mean returns. Far from following a random walk, volatility is serially autocorrelated. Periods of high volatility tend to cluster together, as do periods of relative calm. Different approaches to the measurement of volatility, from time-series analysis to options-based implied measures and the Yilmaz-Diebold model of volatility transmission, reveal feedback and spillover effects.