Asymptotics for drift estimation of a Lévy-driven generalized Langevin equation sampled at high-frequency

Read full text Summarise with VIRAT
Authors: Felipe Sousa Quintino, Ary V. Medino, C. C. Y. Dorea
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Great interest has arisen in stochastic models with some memory effects, where Stochastic Integro-Differential Equations can play an important role. Considering that a Generalized Langevin Equation (GLE) generated our observed data, a generalization of drift estimation to a broader GLE class is proposed. The asymptotic behavior of the estimator is described for noise with finite activity and for processes observed at high-frequency time intervals. Although these limitations, we present a simulation study that extends beyond the established theory by exploring the estimator's applicability in settings with infinite activity and/or low-frequency observations. Computational time for estimation is compared with competitors' estimators, and a threshold parameter choice was proposed based on performance metrics for future predictions. To highlight the applicability of our framework, we present the modeling of two financial datasets.

Fields of Study & Subjects

Stochastic processes and financial applications Fractional Differential Equations Solutions Complex Systems and Time Series Analysis

Publication details

Journal / source
Figshare
Publication date
25 March 2026
Language
EN