hi everybody
I did some statistical analysis. It might be interesting for some of you.
Data:
btc/usd price of bitstamp
year 2013 [31.12.2012 - 31.12.2013]
Statistical test: ANOVA
Trading Volume hourly
http://i59.tinypic.com/14tgm1.png Time of day (=time_category) has a significant effect on the volume of trade (p=0.000). 2.1% of the variance of volume can be explained by time of day.
http://i60.tinypic.com/nyh5x0.png Trading Volume daily
http://i59.tinypic.com/1zbvsdx.pngdays have significant effect on the volume of trade (p=0.013). 4.4% of the variance of volume can be explained by weekday_category.
http://i60.tinypic.com/kcgn61.png Profit or Loss, daily close-priceshttp://i62.tinypic.com/34e81i1.png days hav no significant effect on change_previous_day (p=0.455). Hard to believe that there are no statistically signifcant differences, when you're looking at the chart.
http://i58.tinypic.com/2n8zujm.pngProfit or Loss, hourly
http://i62.tinypic.com/x423nt.png time of day (=time_categories) have a significant effect on the profit/loss (p=0.042). 0.4% of the variance of profit/loss can be explained by time_categories.
http://i59.tinypic.com/v5b8sm.png Regards