[EquisMetaStock Group] adjusted moving averages & zero lag oscillators

["Vasanth Mohan G Buddaan" <vgbudawn@xxxxxxx>]

Talking of zero lag moving averages, there is an 'adjusted' oscillator on which there was a good amount of discussion earlier - RMO - which is also good at tagging onto the flow of the prices like the TEMA

 

Hope my below given chart gets posted on the yahoo (otherwise it would be meaningless)...

 

 

Have taken 2 weeks of 5min chart of nifty (Indian Index) in which period there is trending in both up & down sides, flat periods and a couple of gaps which were followed by contra market movement by which I mean that the prices moved opposite to the gap on both occasions - once after violating the low and once without violating it on 15th & 20th respectively. When the price violated the low formed on the gap many (trend following) systems would have given a 'sell' which would have got whipsawed much later in a normal MA cross over system (like the one shown at the top in green colour - which is a difference of a long & short period moving averages).

 

Whereas the RMO given at the bottom in dark blue colour immediately turned up in the direction of the recent price behaviour ignoring the 'gap down' - the area marked between the yellow - dotted & plain lines. On 20th, it was so swift the whipsawing would have been minimum as compared to the normal MA crossover system given at top of the chart.

 

Among the Moving Averages - TEMA (red line) at the bottom and EMA (yellow) at the top - both of 10d - obviously TEMA has the minimum lag and hence a faster recovery after the gap. To me it looks like RMO is largely a moving average cross over system as may be seen by the white & black lines giving intermediate tops & bottoms in RMO which mostly coincides with the price extremes or does so with a negligible lag. There is quite possibly a component of a conventional oscillator built into it. The main problem is that the indicator does not come with the flexibility of choosing our own parameter. It is so good at capturing the trends but does not have much of the predictive ability of an oscillator resuting in suffering the 'gaps' - may not be 'after' but 'prior' to them.

 

Just a post to throw up an idea at altering a MA Cross Over system...

 

 

gv

----- Original Message -----

From: Lionel Issen

To: equismetastock@xxxxxxxxxxxxxxx

Sent: Tuesday, February 17, 2009 11:22 PM

Subject: RE: [EquisMetaStock Group] Re: adjusted moving avs

Do you need a zero-lag moving average?

 

From: equismetastock@xxxxxxxxxxxxxxx [mailto:equismetastock@xxxxxxxxxxxxxxx] On Behalf Of pjrbutler
Sent: Tuesday, February 17, 2009 3:44 AM
To: equismetastock@xxxxxxxxxxxxxxx
Subject: [EquisMetaStock Group] Re: adjusted moving avs

 

Thx for your replies
I had a look at Jurik's. It's a pity he charges so much , but it does
look a very good mov av. Tillson comes out second best. Has anyone
got his address ? I googled him. There were 186,000 pages of
Tillson's in Colorado !

Thx
Pat

--- In equismetastock@xxxxxxxxxxxxxxx, Code 2 <Code2@xxx> wrote:
>
> Mark Jurik's JMA is a very nice low-lag moving average with little
> overshoot. See http://www.jurikres.com/catalog/ms_ama.htm#top
>
>
>
> From: pumrysh <no_reply@xxxxxxxxxxxxxxx>
> To: equismetastock@xxxxxxxxxxxxxxx
> Date: Sunday, February 15, 2009, 10:20:18 AM
> Subject: [EquisMetaStock Group] Re: adjusted moving avs
>
> Hi Pat,
>
> The problem with any moving average is the lag that is introduced
when
> you begin manipulating them. So the question then is are you really
> improving them? There is a formula out there that was discussed
several
> years ago at:
> http://finance.groups.yahoo.com/group/equismetastock/message/23694
>
> I'm not aware of any that restrict the advance/decline by a
percentage
> or points...seems that would defeat their purpose.
>
> Another thought is an adaptive moving average which is set to a
small
> lookback period at the beginning of a trend then adjust to longer
> lookbacks as the trend progresses based on an indicator value.
There
> are DLL's in the files section that will help you with this task.
>
> Preston
>
>
>
>
> --- In equismetastock@xxxxxxxxxxxxxxx, "Patrick Butler" <pat494@>
> wrote:
> >
> > Hi,
> > Our old friends moving averages do a good job and are generally
> useful. However to improve them and their forecasting ability, is
it
> possible to negate their larger than normal swings up and down ? A
> spike of more than X points or a percentage perhaps ? Perhaps there
is
> a formula already out there somewhere ? Jurik may have done
something
> along these lines.
> > Thanks
> > Pat

__._,_.___

Your email settings: Individual Email|Traditional
Change settings via the Web (Yahoo! ID required)
Change settings via email: Switch delivery to Daily Digest | Switch to Fully Featured
Visit Your Group | Yahoo! Groups Terms of Use | Unsubscribe

__,_._,___