Data From: Analysing social media forums to discover potential causes of phasic shifts in cryptocurrency price series

The recent extreme volatility in cryptocurrency prices occurred in the
setting of popular social media forums devoted to the discussion of
cryptocurrencies. We develop a framework that discovers potential
causes of phasic shifts in the price movement captured by social media
discussions. This draws on principles developed in healthcare
epidemiology where, similarly, only observational data are
available. Such causes may have a major, one-off effect or
recurring effects on the trend in the price series. We find a
one-off effect of regulatory bans on bitcoin, the repeated effects of
rival innovations on ether and the influence of technical
traders, captured through discussion of market price, on both
cryptocurrencies. The results for Bitcoin differ from Ethereum,
which is consistent with the observed differences in the timing of the
highest price and the price phases. This framework could
be applied to a wide range of cryptocurrency price series where there
exists a relevant social media text source. Identified causes
with a recurring effect may have value in predictive
modelling, whilst one-off causes may provide insight into
unpredictable black swan events that can have a major impact on a system.