Optimum thresholding using mean and conditional mean square error

TL;DR

This paper develops an optimal threshold selection method for nonparametric estimation of integrated variance in jump-diffusion models, improving finite sample accuracy by minimizing mean square errors.

Contribution

It introduces a novel approach to select thresholds based on minimizing MSE or cMSE, with a practical implementation scheme and asymptotic characterization.

Findings

Optimal threshold proportional to Lévy's modulus of continuity

Proposed method outperforms existing thresholds in simulations

Asymptotic analysis confirms the efficiency of the approach

Abstract

We consider a univariate semimartingale model for (the logarithm of) an asset price, containing jumps having possibly infinite activity (IA). The nonparametric threshold estimator of the integrated variance IV proposed in Mancini 2009 is constructed using observations on a discrete time grid, and precisely it sums up the squared increments of the process when they are below a threshold, a deterministic function of the observation step and possibly of the coefficients of X. All the threshold functions satisfying given conditions allow asymptotically consistent estimates of IV, however the finite sample properties of the truncated realized variation can depend on the specific choice of the threshold. We aim here at optimally selecting the threshold by minimizing either the estimation mean square error (MSE) or the conditional mean square error (cMSE). The last criterion allows to reach a threshold which is optimal not in mean but for the specific volatility (and jumps paths) at hand. A parsimonious characterization of the optimum is established, which turns out to be asymptotically proportional to the L\'evy's modulus of continuity of the underlying Brownian motion. Moreover, minimizing the cMSE enables us to propose a novel implementation scheme for approximating the optimal threshold. Monte Carlo simulations illustrate the superior performance of the proposed method.