SPSS Statistics

 View Only

Bitcoin's rise and fall in one chart

By Douglas Stauber posted Fri January 26, 2018 08:07 PM

  
In preparing for a future blog post, I discovered a chart visualization that summarizes nicely the difference between Bitcoin in 2017 versus so far in 2018.



This chart shows Bitcoin's volume vs the daily close (last transaction of the day) price in USD. Notice the strong positive slope between volume and price in 2017, but in 2018 this trend turned negative. This chart summarizes nicely the overall trend: in 2017, the bullish market meant as more people joined the crypto, the price shot up. In 2018, that trend reversed as higher volume was correlated with price decreases.

To better quantify this trend, I ran a Bayesian Pearson Correlation between Volume and Price in 2017 and 2018 using a uniform prior. In 2017, the correlation 90% credible interval was (0.767 to 0.829). Conversely in 2018, the correlation 90% credible interval was (-0.690 to -0.152). Quite a strong difference in only a few weeks.

The above analysis was quick to run. I used this dataset and SPSS Statistics Subscription to graph and run the Bayesian correlations. I encourage others to download the dataset and discover your own insights - there are plenty left to be discovered.








#bayesian
#bitcoin
#correlation
#datavisualization
#SPSSStatistics
#Usecases
1 comment
1 view

Permalink

Comments

Fri January 26, 2018 10:15 PM

very cool - and very telling.