A New Investment Method with Autoencoder: Applications to Cryptocurrencies

Abstract

This paper proposes a novel approach to the portfolio management using an AutoEncoder. In particular, the features learned by an AutoEncoder with ReLU are directly exploited to the portfolio construction. Since the AutoEncoder extracts the characteristics of the data through the non-linear activation function ReLU, its realization is generally difficult due to the non-linear transformation procedure. In the current paper, we solve this problem by taking full advantage of the similarity of the ReLU and the option payoff. Especially, this paper shows that the features are successfully replicated by applying so-called the dynamic delta hedging strategy. An out of sample simulation with crypto currency dataset shows the effectiveness of our proposed strategy. Furthermore, we investigate the background of our proposed methodology, which suggests that the first principal component is quite important.