Article ID Journal Published Year Pages File Type
5058187 Economics Letters 2016 4 Pages PDF
Abstract

•We propose test statistics for testing Brownian motion against fractional Brownian motion.•Our test framework is robust to finite large jumps.•We extend bi-power variation to the inference of Hurst index.

Fractional Brownian motion embeds Brownian motion as a special case and offers more flexible diffusion component for pricing models. We propose test statistics based on bi-power variation for testing Brownian motion against fractional Brownian motion alternatives. To filter out the prevalent existence of finite large jumps, a truncation method based on Hurst index estimator is proposed. Simulation results confirm the consistency of jump truncation framework with desirable empirical size and viable empirical power for our tests.

Related Topics
Social Sciences and Humanities Economics, Econometrics and Finance Economics and Econometrics
Authors
, , ,