Deep Learning, Jumps, and Volatility Bursts

25 Pages Posted: 20 Sep 2019 Last revised: 14 Mar 2020

Date Written: September 12, 2019

Abstract

We develop a new method that detects jumps nonparametrically in financial time series and significantly outperforms the current benchmark on simulated data. We use a long short- term memory (LSTM) neural network that is trained on labelled data generated by a process that experiences both jumps and volatility bursts. As a result, the network learns how to disentangle the two. Then it is applied to out-of-sample simulated data and delivers results that considerably differ from the benchmark: we obtain fewer spurious detection and identify a larger number of true jumps. When applied to real data, our approach for jump screening allows to extract a more precise signal about future volatility.

Keywords: Jumps, Volatility Burst, High-Frequency Data, Deep Learning, LSTM

JEL Classification: C14, C32, C45, C58, G17

Suggested Citation

Bashchenko, Oksana and Marchal, Alexis, Deep Learning, Jumps, and Volatility Bursts (September 12, 2019). Swiss Finance Institute Research Paper No. 20-10, Available at SSRN: https://ssrn.com/abstract=3452933 or http://dx.doi.org/10.2139/ssrn.3452933

Oksana Bashchenko

Swiss Finance Institute - HEC Lausanne ( email )

Chavannes-près-Renens
Switzerland

Alexis Marchal (Contact Author)

EPFL ( email )

Station 5
Odyssea 1.04
1015 Lausanne, CH-1015
Switzerland
‭+41 21 693 09 23‬ (Phone)

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
541
Abstract Views
2,396
Rank
94,413
PlumX Metrics