Deep Neural Networks and Fan Tokens’ Pricing in Football Clubs

Abstract

Technological evolution is causing the development of alternative financial assets that are gradually attracting the attention of investors. The appearance of crypto assets has created expectations for new economic and social tools for the future of companies. Fan Tokens have represented an opportunity for different types of companies, especially sports. It is a crypto asset that brings fans closer to their clubs, generating synergies and greater involvement. The Fan Token also represents a business opportunity for clubs; the issuance of these assets allows them to generate income and retain fans. Determining the price of fan tokens is essential to make them profitable. In this study, five models of Deep Learning Neural Networks have been used to study the price, the impact, and the most relevant variables. A Quantum Neural Network is applied to compare estimations. The Convolutional Neural Networks (CNN) model has provided the best indicators, even so, both Gated Recurrent Unit-Convolutional Neural Networks (GRU-CNN) and Quantum Neural Networks, have made predictions with very optimal values. The importance of the Accumulation/distribution indicator (A/D), the Momentum (MM), or the Szymansky Ranking (SR) have been highlighted and it has also been possible to rule out certain variables that are not significant.