Cutting edge – multi-scale volatility in commodity markets

This paper deals with volatility estimation in commodity markets. Piotr Grzywacz and Krzysztof Wolyniec note that energy commodities have many time (volatility) scales, which has dramatic implications for mean-reversion and volatility estimation. They show that standard mean-reversion models dramatically overestimate cumulative volatility, especially over short- and intermediate time scales

We consider the problem of estimating cumulative volatility in energy commodity markets. We argue that the nature of volatility behaviour is very different from the standard stationary/non-stationary dichotomy popular in financial markets. We note that energy commodities have many time (that is, volatility) scales. This fact has dramatic implications for estimation of mean-reversion and cumulative volatility in those markets. More specifically, we show that standard mean-reversion models (such