|
PureBytes Links
Trading Reference Links
|
There is a clear spreadsheet presentation of a CMO-based Vidya in TASC
v.13 p.424. Comparing the MS formula (p.461 in v.6.5)to that, it is
clear that in the smoothing constant there is a misplaced decimal: 0.78
and 0.078 should both read 0.078. This figure seems to come from a TASC
v.10 article with a 25-period (weekly) Vidya. The TASC v.13 article
uses a smoothing constant determined by the lookback period of Vidya.
This is different from the book.
Smoothing constant: Smo = 2/(period + 1)
Volatility Index: VI = Abs(CMO)
Vidya = (Smo * VI * Close) + ((1 - (Smo * VI)) * Yesterdays Vidya)
The first day only needs to be seeded with the previous days Close.
Perhaps a more experienced programmer would offer the definitive version
that we can all copy into our Indicator Builder. Are you listening
Equis?
Harvey Pearce, Victoria, B.C., Canada.
-------------------
Rick Mortellra wrote:
>
> Hi Max,
>
> The MS interpretation of Chande's VIDYA indicator can be found under the
> built-in VMA moving average. HOWEVER, EQUIS BOTCHED THE WHOLE CONCEPT AND
> IT IS COMPLETELY SCREWED UP!
>
> To see what I mean simply plot a SMA, EMA, and the VMA of any security
> using the same "period" input for each. Logically, the VMA should be the
> most sensitive, almost hugging the price bars. But as you'll notice MS's
> VMA is the least sensitive. In fact, it's downright dead!
>
> Further, the fact that MS even asks for a "period" input for the VMA is
> dumb, since the "period" by definition is supposed to be VARIABLE anyway.
> The only inputs required under Chande's formula definitions are for the
> "volatility index" and a scaling factor. MS says it has fixed these as the
> 9-period CMO for the index with a scaling factor .78, so no inputs should
> even be required. Finally, a quick glance at the built-in formula given in
> the manual shows Equis didn't even get the scaling factor part right.
>
> Fortunately, the VIDYA is pretty easy to build and in MS 6.5 the new
> "input" function lets you play around with the scaling factor without
> rewriting the formual each time. Changing the period of the volatility
> index doesn't affect the average significantly. Nor does using different
> variables for the index. Afterall, it only varies from 0 to 1 anyway.
>
> I've discussed this with Equis tech support when 6.0 first came out, and
> they've recluctantly agreed the VMA might be implemented wrong but
> suggested I contact Chande myself to check. Yeah, right! Too bad, because
> the VIDYA concept minimizes the time lag problems of a crossover system,
> and lets you grab more of a price move.
>
> Hope this helps,
> Rick Mortellra
> Tokyo Japan
>
>
> > ----------
> > > From: Max Turner <m.turner@xxxxxxxxxxxx>
> > > To: metastock-list@xxxxxxxxxxxxx
> > > Subject: Vidya and other Chande/Kroll indicators
> > > Date: Sunday, September 28, 1997 8:07 AM
> > >
> > > List'rs
> > >
> > > Has anyone programmed the indicators from Chande and Kroll's book (The
> > New
> > > Technical Trader) into Metastock formulae? In particular I am
> interested
> > in
> > > VIDYA and the Stops indicator at the end of their book.
> > >
> > > Nowadays many authors give the formulae but in this interesting book
> they
> > > give the general formulae only.
> > >
> > > Would appreciate any replies.
> > >
> > > Thanks in advance.
> > >
> > > Max
> > >
> > >
> > > K
> > >
|