PureBytes Links
Trading Reference Links
|
I haven't looked closely at Paul and Rick's work yet but I thought it would
be
useful to establish what the more specific criteria are for winning.
What does it mean specifically to "replicate the weekly period DJIA 5,3,3
stochastic from daily data..." I understood, Paul, what
you meant when you said you had it "close" even if it wasn't exact. I had
the same problem. So let's agree on what is "close enough" before I have a
chance to look at it. I'm hoping that other list-members who are real
experts
on the use of the stochastic can help us to refine the criteria below.
I propose the following principles to define "close enough":
The synthetic weekly stochastic from daily data will be close enough if and
only if for the period 1994 to the present (to ensure that both bull and
bear markets are covered):
- the number of breakouts above 20 and breakdowns below 80 on the weekly
synthetic stochastic are equal to the number of the same on the actual
weekly stochastic;
- the number of stochastic direction reversals between 20 and 80 on the
weekly synthetic stochastic is equal to the number of the same on the actual
weekly stochastic;
- the lowest and highest values of the synthetic weekly stochastic
indicator at direction reversals are within x percentage points of the same
on the actual weekly stochastic; and
- the timing of 1) stochastic direction reversals between 20 and 80 and 2)
breakouts above 20 and breakdowns below 80 on the weekly synthetic
stochastic must be within 1 week (9 days) of the same on the actual weekly
stochastic.
Please understand that the purpose of setting out these principles is to
establish criteria for evaluating whether the MS 6.5 formula language can be
used to create a synthetic weekly stochastic from daily data even if it's
messy.
The purpose is NOT to try to deprive you of a prize. <G>
Do these principles need restating? Anyone like to propose additional
principles
or delete one of the principles listed above?
I especially want to hear from Paul and Rick since you two are the only
contestants.
Steven Buss
Walnut Creek, CA
sbuss@xxxxxxxxxxx
|