Time series (re)sampling using Generative Adversarial Networks

Abstract

We propose a novel bootstrap procedure for time series data based on Generative Adversarial networks (GANs). We show that the dynamics of common stationary time series processes can be learned by GANs and demonstrate that GANs trained on a single sample path can be used to generate additional samples from the process. We find that temporal convolutional neural networks provide a suitable design for the generator and discriminator, and that convincing samples can be generated on the basis of a vector drawn from a normal distribution with zero mean and an identity variance-covariance matrix. We demonstrate the finite sample properties of GAN sampling and the suggested bootstrap using simulations where we compare the performance to circular block bootstrapping in the case of resampling an AR(1) time series processes. We find that resampling using the GAN can outperform circular block bootstrapping in terms of empirical coverage. Finally, we provide an empirical application to the Sharpe ratio.

Keywords: Bootstrapping; Dependent processes; Generative adversarial nets.