Which factors can explain the returns of Bitcoin and other cryptocurrencies?

Abstract

This paper aims to determine the factors that affect the returns of a majority of 44 of the largest cryptocurrencies, which already existed on October 30th, 2017. Prior studies used factors like trade volume, gold, oil, the stock market, and Google searches to explain the price movements of Bitcoin. To examine the effects of those factors on multiple cryptocurrencies, I used three different types of models (normal time series, Distributed-Lag and Vector-Autoregressive) in combination with daily and weekly data. To check whether Bitcoin was the dominant factor for the other cryptocurrencies, a variable with the Bitcoin returns was added to those models. The regressions show that trade volume and the S&P500 have a significant effect on the returns of Bitcoin and a majority of the altcoins, with Bitcoin having a significant effect on all altcoins as well. The Vector-Autoregressive model with weekly data proves to have the best goodness-of-fit for almost every cryptocurrency, indicating that the addition of a lagged variable of the returns of the cryptocurrencies increases the explainability of the model.