Some Bitcoin volatility statistics

[Daal]

Here is some interesting information about the volatility of BTC that doesn't get talked about much:

BTC has had 144 "large up days", days where it went up by 10% or more (52.8% was the largest up day). It had 104 large down days, days where it went down by 10% or more (-38.8% was the largest down day). The kurtosis of the large up days is 5.94 whereas of down days is 3.27.

That is, the tails are twice as fat to the upside versus to the downside. So when people look at the volatility and think its risky, yeah, its risky. But its twice as risky to the upside as it is to the downside, and who doesnt like upside risk? Something for you skeptics to think about

 

[longshort]

Characteristics of a log-normal distribution? You lose 38.8%, you make 52.8%. Now you're down 6.5% from start.

Once you lose 90% (which BTC did several times, albeit not in a day) you need to make 900% just to break even. Both changes are equally likely and represent the same deviation:

-1 * log(1 - 90%) = log(1 + 900%)

 

[Daal]

Characteristics of a log-normal distribution? You lose 38.8%, you make 52.8%. Now you're down 6.5% from start.

Once you lose 90% (which BTC did several times, albeit not in a day) you need to make 900% just to break even. Both changes are equally likely and represent the same deviation:

-1 * log(1 - 90%) = log(1 + 900%)

Yes you get hurt more by the drops than rises but the rises are more frequent and larger. Up days are 54.55% of the sample, down days are 45.45%. In terms of large up days its 58% vs 42% large down. The overall result is that the surprises tend to be more of the good kind than of the bad