CMStatistics 2022: Start Registration
View Submission - CFE
Title: On empirical challenges in forecasting market betas in crypto markets Authors:  Jan Sila - UTIA AV CR, v.v.i. (Czech Republic) [presenting]
Michael Mark - Ecole Polytechnique Federale Lausanne (Switzerland)
Ladislav Kristoufek - Czech Academy of Sciences (Czech Republic)
Abstract: The predictability of market betas for crypto assets is investigated. The market beta is the optimal weight of a short position in a simple two-asset portfolio hedging the market risk. Investors are, therefore, keen to forecast the market beta accurately. Estimating the market beta is a fundamental financial problem, and we document pervasive empirical issues that arise in the emerging market of crypto assets. Although recent empirical results about US stocks suggest predictability of the future realized betas is about 55\%, predictability for the universe of crypto assets is at most 20\%. Our results suggest that the crypto market betas are highly sensitive not only to the beta estimation method but also to the selection of the market index. Thus, we also contribute to the discussion on the appropriate market representation.