Research Article
Technological Change and Market Conditions: Evidence from Bitcoin Fork
Figure 1
Probability density function of log returns in the new and old forks. Following the Freedman–Diaconis rule [39], we plotted the histograms of both datasets with the optimal bin width of (Bitcoin) and (Bitcoin Cash). There are two strands of rules for determining the optimal bin width: the Scott rule and the Freedman–Diaconis rule. The former is best suited for data that are close to being normally distributed, whereas the latter is less sensitive to outliers and thus better suited for data with heavy tails. As Figure 1 indicates, the data do not match well with the normal distribution; thus, the Freedman–Diaconis rule is used. (a) Bitcoin. (b) Bitcoin Cash.
(a) |
(b) |