In our previous post we looked at 9 different combinations
of parameters for Bollinger Band simulations. The default parameters had only
average results. The variety of resulting net worth’s leads to the possibility
that there are combinations that might be profitable. So, I expanded the limits
of the parameters. The number of days ranged from 4 to 25. The standard
deviation multipliers went from 0.3 to 2.5. The resulting net worth’s ranged
from $795 to $8,126. Following are the
top 10 results.
|
Net Worth
|
Days Desired
|
Std Dev Multiplier
|
#Trades
|
# Buys
|
# Sells
|
|
$8,126.41
|
15
|
1.7
|
195
|
101
|
94
|
|
$8,103.39
|
14
|
1.7
|
203
|
105
|
98
|
|
$8,023.37
|
13
|
1.7
|
210
|
108
|
102
|
|
$7,977.60
|
12
|
1.7
|
219
|
113
|
106
|
|
$7,974.25
|
7
|
2.1
|
218
|
113
|
105
|
|
$7,939.24
|
8
|
2.1
|
198
|
102
|
96
|
|
$7,904.43
|
14
|
1.9
|
173
|
89
|
84
|
|
$7,807.13
|
9
|
2.3
|
154
|
80
|
74
|
|
$7,799.28
|
16
|
1.7
|
190
|
97
|
93
|
|
$7,571.09
|
16
|
1.9
|
165
|
85
|
80
|
A few observations. None of the results were positive
(remember we started with $10,000). Both parameters were smaller than the
defaults. None clearly stood out. This tells me that I should not pick a set yet,
but pick a narrower set of ranges to test in future simulations.
I have lots of questions I want answers to. The 1st
is how are the parameters impacted by the general market direction. The period
we’ve covered is 2018 through the first 3 quarters of 2019. Interestingly 2018
started with a BTC price of $13,850 and ended at $3,747. It dropped rapidly
early and was generally had a downward move. In 2019 it started at $3,747 and
was at $8,327 on September 30th. If we break the simulations into
separate years, we might learn something.
We’ll address the results in my next post. See @billlanke on
Twitter to know when this occurs.
No comments:
Post a Comment