A STUDY OF EXCESS VOLATILITY OF GOLD AND SILVER

In this study we discuss the case of strong path dependency in asset prices from the theoretical and empirical standpoints. We examine the volatility of gold and silver using the extreme value estimator RS proposed by Rogers and Satchell (1991) and the VRatio proposed by Maheswaran et al. (2011). We demonstrate persistence of excess volatility in the gold spot price data that engenders excessive path dependence, whereas we find that it is not the same with silver.  The multiple-days’ time horizons used to analyse the 16-years’ time series data (January 2001 to December 2016) allow examination of volatility across different time frames and provide better approximation of Brownian motion in the data. Bootstrap simulations are employed  to compute standard error and also to check for the statistical validity of our analysis. Unlike the initial observations, there is excess volatility in gold prices when compared to that of silver over multiple-days’ time horizons. We capture the excess volatility in the gold data using the Binomial Markov Random Walk model.  We also utilise the expected lifetime shortfall (ELS) ratio, as a measure of risk to test for the presence of mean reversion in asset prices. Using this ratio, we observe that the strong mean-reverting characteristic in gold makes it a better investment choice than silver, in general, in the medium term. The mean-reverting characteristic is not significant in silver, that is, if prices move up then they are more likely to stay up for a relatively long time. Similarly, if prices for silver go down then they are more likely to stay low for a longer time than the prices of gold. Gold is less volatile when held over the medium term because of a negative correlation between successive price changes. Hence, from an investor perspective, the portfolio should consist of gold rather than silver.