Quantile convolutional neural networks for Value at Risk forecasting

Abstract

This article presents a new method for forecasting Value at Risk. Convolutional neural networks can do time series forecasting, since they can learn local patterns in time. A simple modification enables them to forecast not the mean, but arbitrary quantiles of the distribution, and thus allows them to be applied to VaR-forecasting. The proposed model can learn from the price history of different assets, and it seems to produce fairly accurate forecasts.

Keywords

• Quantile regression
• Convolutional neural network
• Value at Risk
• Forecasting