In my previous post I began the process of analyzing the use
of Bollinger Bands for trading bitcoins (BTC). I counted the number of
recommended trades and found that there would be 1-2 trades per week. But I
pointed out that the profitability of the trades would be more significant. I’ve
now had a look at this question and the results are disappointing.
I continued using the standard Bollinger defaults, last 20
days moving average and 2 standard deviations from this average to set the
limits. I did add the September prices which added 10 trades. Interestingly
they were all purchases. That’s not unreasonable because the price of BTC was
generally falling through the month. This seems characteristic of the Bollinger
trades, clumps of buys or sells. I started with a balance of $10,000, split
equally between cash and BTC. After 189 trades the combined balances were worth
$3,976. Not good.
There were periods when each balance became negative. I
decided not to tinker with these since you could use margin to overcome these. Next,
I’ll consider simulating trades if negative balances are not allowed. Follow
@billlanke on Twitter to know when I post this result.
No comments:
Post a Comment