Title: Influencer detection between cryptocurrency sectors via sparse network analysis
Authors: Kexin Zhang - City University of Hong Kong (Hong Kong) [presenting]
Simon Trimborn - University of Amsterdam (Netherlands)
Abstract: The rapidly changing cryptocurrency market evolved from a market focused on payment systems into various sectors operating on blockchains but dedicated towards entirely different goals. The emergence of sectors within the cryptocurrency market raises the question of the interdependence between asset classes belonging to different groups. We introduce a 3-layer sparse network AutoRegressive estimator to identify the influential cryptocurrencies and sectors. We study the asymptotic properties of the estimator and validate its performance in extensive synthetic data experiments. We study 55 cryptocurrency sectors and highly capitalized cryptocurrencies during 2015-2020 for their influence on each other. During the bear market of 2015-2016, the lead-lag effect is not significant throughout the years. However, during the bull market of 2017-2019, including the 2017 market frenzy, Monero and Bitcoin frequently determine the market direction and influence sectors' performance. From 2020 onwards, following a period of strong market growth, various sectors and Bitcoin influenced each other's performance. This suggests that the market entered a new stage since its behavior suggests interdependencies between previously unrelated sectors. Further, it suggests that Bitcoin still plays an important role despite the market entering a new stage.