Heavy-tailed features and dependence in limit order book volume profiles in futures markets

This paper investigates fundamental stochastic attributes of the random structures of the volume profiles of the limit order book. We find statistical evidence that heavy-tailed sub-exponential volume profiles occur on the limit order book and these features are best captured via the generalized Pareto distribution MLE method. In futures exchanges, the heavy tail features are not asset class dependent and occur on ultra or mid-range high frequency. Volume forecasting models should account for heavy tails, time varying parameters and long memory. In application, utilizing the generalized Pareto distribution to model volume profiles allows one to avoid over-estimating the round trip cost of trading.