Statistical Analysis of the (In)efficiency of Bitcoin : A comparison of the weak form efficient market hypothesis in the US and Venezuela

Abstract

Bitcoin is a phenomenon that has received a lot of attentionduringthe last years. Although the literature on the subject has expanded, there is still need for more research. This paper replicates the work of Urquhart (2016) and examines whether there isevidence ofweak form efficiency in the Bitcoin market. He found that the Bitcoin market in the US showed signs of moving towards weak form efficiency. We contribute to the existing literature by adding recentdata and comparing two different markets;the USand Venezuela. To obtain robustness of the results, the analysis is conducted by performing six different statistical tests,to detect whether Bitcoin returns can be viewed as realizations of independently identically distributed random variables. Despite that the findings are somewhat contradictory for the US, our conclusion is that the market is considered as inefficient. However, we found evidence of weak form efficiency in our last subsample period. This is the period that includes more recentdata, hence our findings support Urquhart ́s statement. For Venezuela, the findings are more coherent, and we conclude that there are no signs of weak formefficiency in the market. Keywords: Bitcoin, cryptocurrency, efficient market hypothesis,i.i.d.,long memoryrandomness, statistical analysis