A Neural Network with Shared Dynamics for Multi-Step Prediction of Value-at-Risk and Volatility

37 Pages Posted: 3 Jul 2021

See all articles by Nalan Basturk

Nalan Basturk

Maastricht University - Department of Quantitative Economics

Peter C. Schotman

Maastricht University - Department of Finance

Hugo Schyns

Maastricht University - School of Business & Economics - Department of Finance

Date Written: June 17, 2021

Abstract

We develop a LSTM neural network for the joint prediction of volatility, realized volatility and Value-at-Risk. Regularization by means of pooling the dynamic structure for the different outputs of the models is shown to be a powerful method for improving forecasts and smoothing VaR estimates. The method is applied to daily and high-frequency returns of the S&P500 index over a period of 25 years.

Keywords: Neural Network, Value-at-Risk, Volatility Models, Equity Returns, Risk Management

JEL Classification: G17,C32

Suggested Citation

Basturk, Nalan and Schotman, Peter C. and Schyns, Hugo, A Neural Network with Shared Dynamics for Multi-Step Prediction of Value-at-Risk and Volatility (June 17, 2021). Available at SSRN: https://ssrn.com/abstract=3871096 or http://dx.doi.org/10.2139/ssrn.3871096

Nalan Basturk

Maastricht University - Department of Quantitative Economics ( email )

P.O. Box 616
Maastricht, 6200 MD
Netherlands

Peter C. Schotman

Maastricht University - Department of Finance ( email )

P.O. Box 616
Maastricht, 6200 MD
Netherlands
+31 43 388 3862 (Phone)
+31 43 388 4875 (Fax)

Hugo Schyns (Contact Author)

Maastricht University - School of Business & Economics - Department of Finance ( email )

Netherlands

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
50
Abstract Views
299
PlumX Metrics