Methods

Data for this section includes digital assets historical monthly returns (Coincheckup.com), on-chain metrics (Coinmetrics.com), and off-chain web and social analytics (EconomyMonitor.com and click-stream data providers). The period of study is August 2017 to January 2018.

The characterization of flows

When a blockchain split event occurs, a race (competition) for attention begins. Demand stars flowing from search engines, price trackers, faucets, wallets, educational sites, and the many services that support a crypto economy. One such event occurred on August 1st 2017, when the Bitcoin blockchain forked, creating two competing digital assets, BTC and BCH \cite{cash}. We obtained monthly data for 177 of those web services, specifically, the share of usage of each service towards two of the official communities in both networks; this off-chain activity acts as an inferential sensor, an indication of interest and trust. The sources were the largest contributors in the six-month period, their share is weighted by contribution. The resulting arrays (Figures 5 and Figure 6) depict the time period in the vertical axis, and the source variables in the horizontal axis; the target variable (returns) are included in the last column. Red tones encode incomplete data.
In the case of Bitcoin (digital asset ticker BTC), we observe an active economy, with creative destruction (services that come online or stop contributing, as time progresses).