I confess that I have no idea what's going on with Chandler's formula, or Wolfgang's. BNS1 is basically CCI rescaled so it can be used to look for crossovers with BNS2, right? BNS2 is a price oscillator, rescaled to fit the rscaled CCI.
BNS1: (100-(100/(1+((((((H+L+C)/3)-Mov(C,34,S))/(0.015*Std(C,34)))+2500)/2500)))) [A rescaled version of CCI. That standard deviation junk in the divisor is part of the original formula for CCI -- we've still got to work in Bill's mean deviation formula:
Std(((h+l+c)/3),28) Divisor problems or devisor problems? Or was it me, the reviser?]
BNS2: (100-(100/(1+((( OscP(13,55,E,% ))+2500)/2500)))) (What? A rescaled price oscilator?)
Those are the two elements of the ultimate combination, BNS for WOW:
Fml( "BNS1" ) - Fml( "BNS2" )
But what happened to the second part of Bill's New System (or BradCCI), the mean deviation, which was expressed like this:
Std(((h+l+c)/3),28)
Now it appears that the price oscillator has been substituted for the mean deviation formula, but to what purpose? And what about the original basis for BNS, CCI and mean deviation working together to get a "less noisy" CCI. And then looking at these in conjunction with the RedLine indicator?
BNS2, with its rescaling, reduces the price oscillator to essentially a trend line at 50, so it can fit within the range of BNS1. But does this faithfully follow Bill's system of eyeballing CCI and mean deviation for concurrent moves, crossovers, etc. (I'm still not sure exactly what we're supposed to be looking for), along with the price oscillator and the RedLine indicator?
And was this the intent of the rescaling of BNS2, to get what amounts to a trend line? It appears that the old problem with the rescaling of the two elements of the BradCCI formula has resurfaced, but just 50 pints higher. The standard deviation line used to be reduced to a zero line when plunked down in the same window with CCI. Now the same thing is happening with the price oscillator line when plotted with CCI, a CCI that's been rescaled, too. And why use price oscillator as a proxy for the old standard deviation?
I'm hopelessly confused. I thought the formulation had to work in, at least: CCI, mean deviation and the RedLine indicator (or the proxy, slow stochastics).
I don't think the one fuse for all these indicators has been found yet, but whaddo I know? |
