|
PureBytes Links
Trading Reference Links
|
Rudolf,
I couldn't agree more. I use a combination of LinReg, SMI, & TSI
(Kaleidoscope) for the most of my studies. As for the CMO and cocoa: one
look at the "perfectly" structured downtrend and one can discern that the
parameters (variable periods and trigger levels) would not produce the same
results if cocoa starts to trade above it's current channel. I have a more
utilitarian formula for CMO studies. I call it the FibCMO (for obvious
reasons):
(CMO(C,3)+CMO(C,5)+CMO(C,8))/3
All one really has to do to take advantage of the cocoa downtrend (or any
trend), is to sell against the 13 day moving average put a stop in at the
34 day moving average. It would have resulted in a successful trade the
last 20+ times in cocoa.
Neither the CMO nor the moving averages would show stunning results, in
cocoa, if the nature of the market changes. I don't buy into the
suggestions that one must be able to back test a system for 5 years with
various samplings to validate an approach. It's been fairly obvious that
cocoa has been in a well defined trend for quite sometime. For over six
months, I have posted the signals for the CMO (and taken each signal) and
fully expect it to continue to produce profits, providing cocoa doesn't
break out of it's current channel.
Maybe it's just a coincidence that cocoa posted it's strongest "up day" in
four weeks on Friday. After enjoying the profits from the last short
position at 13.31 (2/19/99), it was sure fun (and profitable) to cover the
position on the opening yesterday (10.74), see the market make a low, one
tick below entry (10.73), and then carry a 24 point profit home for the
weekend.
The markets are dynamic. To think that a universally applied approach
(mechanical) would consistently work on a commodity or many commodities is
not a very good premise to trade by. If that was the case, I would be
"flashing" CMO results in everything that moves. The post was a simple
demonstration of how a system can be applied to a specific situation.
In 1979, while working for Paine Webber, I traded cocoa 19 times in a row
for a profit (I thought I was the second coming). At the time it was
chopping sideways and I had a little moving average oscillator that was in
tune with the market's nuances. Unbelievable? When the market stopped
chopping around, it locked the limit, for three days, against my hefty
position. I got crushed (lost tens of thousands).
Nothing last forever, but in the meantime, until cocoa demonstrates that it
wants to trade out of it's channel, I will continue to take the signals
that this little secular system provides.
Happy trades,
Steve Karnish
CCT
----------
> From: rudolf stricker <rst@xxxxxxxxxxx>
> To: metastock@xxxxxxxxxxxxx
> Subject: Re: CMO & Cocoa
> Date: Saturday, April 10, 1999 12:53 AM
>
> On Fri, 9 Apr 1999 09:28:03 -0700, you wrote:
>
> >There is no substitute for good old TA grunt work, but for the "TA
> >impaired", here's the latest "auto-pilot" trade using Chnde's Momentum
> >Oscillator on Cocoa.
>
> When I did similar things as a basis of my "Know-how Recycling" work
> on DAX options, I found the result very sensitive to the number of
> days used in the CMO indicator. Seemingly, there has to be a
> "resonance" between the "frequency" of the CMO and the "main
> frequency" of the price to make things working well.
>
> To cover a range of years (with bullish & bearish periods and very
> different price "frequencies"), I was not successful using _one_ "CMO
> frequency". Even using several (up to 5) "CMO frequencies" in parallel
> was not satisfying (and could not be handled easily in MetaStock
> because of its stupid restrictions).
>
> The way out I found, is a "variable frequency CMO" I constructed by
> interpolation. It seems to work reasonably, but I use this type of
> indicator only as a back-up at this time, because I found a more
> powerful and more flexible one, combining simple ROC and
> LinearRegressionSlope.
>
> mfg rudolf stricker
> | Disclaimer: The views of this user are strictly his own.
|