Leveraging the users graph and trustful transactions for the analysis of Bitcoin price

Cryptocurrencies are notorious for their exchange rate high volatility, and are often tools of wild speculation rather than decentralised value exchange. This is especially true for Bitcoin, still, nowadays, the most popular cryptocurrency. This paper presents an analysis to detect the influence of a set of topological properties of the Bitcoin Users Graph on Bitcoin's exchange rate. We consider, besides classical properties, a novel notion of Trustful Transaction Graph introduced to describe partial Users Graphs derived by chains of 0-confirmation transactions. We present a temporal analysis of the evolution of a set of features with a single day granularity. Afterwards, we applied autoregressive distributed-lag linear regression to assess whether and with which strength and duration a change in the considered features is likely to influence the exchange rate up to a prespecified number of days (fifteen) in the future. The results show that some of the considered features significantly influence the exchange rate up to several days, and that such relationships are likely not to be spurious, since we found that those features contribute significantly to decrease the error in predicting the exchange rate.