Registered User
Joined: 4/18/2005
Posts: 4,090
|
curious if you could shed a littlelight on the VPCI indicator you gve me in realtion to the Compqx. This is a negative divergence going back as early as may. What exactly is it telling me?
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Scott, keep in mind that just because the indicator is on the decline does not mean that there is a divergence.
It is when the indicator contradicts price that there is a divergence. On the Nasdaq you will see that there has been a divergence since July 3rd. (Price is rising as the indicator is declining).
A good example of a possitive divergence would be XMSR. Scroll back to 5/17/07 (zoom 7) see how since early May the price is on the decline as the indicator is rising. You need to also understand what is causing the divergence. The indicator analyzes the relationship between the daily close vs volume. (remember Martha says that if volume is week that a price rally cannot be sustained) think of it in those terms...
Good luck.
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
so you wouldnt say that the High mark in VPCI and the high mark on COMPQX price in JAN vs. April vs. NOW wouldn't constitute a longer term divergence?
Because, acording to the theory of reading TSV which is also plotted arround a centerline that would constitute a longer term divergence. Yes I clearly see the short term deeper one.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Scott, to be honest with you I am not sure if this indicator should be used in that way. In the magazine article they mentioned that this indicator is for short term trading only. Perhaps Diceman will chime in with the answer to that. In the meantime I will be looking at past charts to see how long term divergences played out.
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
That is an interestig concept if you think about it....Time decay on the past information of an indicator. Don Worden states tht charts should be read from the the left to the right.... and to many traders put all their emphasis on the right edge. But if a short term indicator looses validity in readings that are on screen just a few months ago Then i wonder what that means for other indicators that are set for short term settings?
Perhaps this may have aplications in reading stochastics for example at the left of the screen. I feeling is that the equation is still plugging along and plotting what it plots... and that the anomalies would still have value... but if not that is worth understanding more about.
|
|
 Gold Customer
Joined: 11/11/2006
Posts: 359
|
Tom: I agree with you on this one. The article defines VPCI as being a snapshot of the influence of volume upon price trend over a specified time.
I feel that divergences diminish in predictive value in relation to the number of price fluctuations that occure between indicator peaks, or valleys. Like gaps, they hold less value to me if they happened 300 bars ago than last week.
Tom, the article refers the VPCI crossing its "smoothed, volume weighted average of VPCI"
(Page 26)
Do we not use a standard 30 bar MA with VPCI?
Does it make a difference?
Mammon
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
"I feel that divergences diminish in predictive value in relation to the number of price fluctuations that occure between indicator peaks, or valleys. Like gaps, they hold less value to me if they happened 300 bars ago than last week."
Inteeresting becasue Elder and Worden both indicate that divergences can build and magnify over time. I personaly find that there is a capitulation point with price... IN TSV divergences I consider it when TSV crosses zero... or in come cases it's MA. Other indicators are a bit more tricky but you can kinda guage a point where the decent of price and the ascent of the indicator reach a critical level... then price will show the affects of the divergence, if it's going to. sometimes it's with fireworks and other times it's more akin to those toy guns with the cork in the end .
Some divergences are longer and shallower and IMO these are the ones that it's important to look for the capitulation point. Others are deep and very stark... these short term ones can be very powerful as well and IMO worth consideration as an entry. I have more experience with the longer sort but have had some good experiences with the shorter period ones as well.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Mammon, I do a lot of freelancing. I just took it upon myself to add a 30 day moving average. I always back-test my choices and this appears to work well in most cases.
Scott, I fear that my choice to use the term "Divergence" in reference to the VPCI was a bad one. No where in the article do they use that term. They refer to Volume vs Price contradictions. You are applying existing thoeries in reference to other indicators and I am just not sure that this indicator is built that way. I am convinced that Diceman will read this and put our minds to rest. But in the meantime I would not give much credit to any longterm implications involving the VPCI..
Finally you should know by now that I am not convinced of the power of divergence on the same level as your self and Hohandy. Money Stream is the only indicator that I really pay attention to a divergence, it seams to have a better track record of success or accuracy.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
When looking for toping formations and Bottoming formations I also give some credit to divergences in Stochastics. Only because I have found success doing so.
|
|
Gold Customer
Joined: 11/11/2006
Posts: 359
|
Thanks, Tom. That was my feeling also, but I knew you were more through in backtesting than I, but I needed to ask the question.
Thanks,
Mammon
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
"But in the meantime I would not give much credit to any longterm implications involving the VPCI.."
OK 
"Finally you should know by now that I am not convinced of the power of divergence on the same level as your self and Hohandy. Money Stream is the only indicator that I really pay attention to a divergence, it seams to have a better track record of success or accuracy."
Yea.. Don't seek out only divergences and as I said TSV divergences don't mean much to me unless they cross zero... which is just healthy action for TSV anyway. MS divergences are worth noting... but even there I have seen long term MS divergences lasting as many as 4 years... so caution is still advised. but I don't discout them.
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
The core calculation of VPCI is VPC.
VPC= (LT VWMA) - (LT SMA)
LT VWMA= Long-term volume weighted moving average.
SMA = Simple moving average
------------------------------------------------------------
In an 4 period SMA each day is 25% of the value:
1
2
3
4
SMA= (1+2+3+4) / 4 = 2.5
-------------------------------------------------------------
In a VWMA if volume is equal each day the
value will be the same as the SMA
(If volume is 1 each day)
1*1
2*1
3*1
4*1
equals:
(1+2+3+4) / 4 = 2.5
VPC= 2.5-2.5=0
----------------------------------------------
When there is higher volume that close will have more weight.
1*1
2*1
3*1
4*3
In this example volume on the 4 close was 3 times the
other days.
VWMA= (1+2+3+12) / 6 = 3
VPC= 3 - 2.5 = .5
VPC is now positive because the last day has a
50% weight (3/6) and the other days have a 16.67%
weight (1/6) in the final equation.
------------------------------------------------------------------------
So we have an equation where VPC should rise in strong
rallies. (higher price and volume) and fall in sell-offs
(lower prices and higher volume)
All depending on volume
-------------------------------------------------------------
Because volume is a factor in the calculations. Holidays
will be a factor. VPCI does not know if its a holiday or
a "weak" day.
Because VPCI is calculated from price. Its values will
not be adjusted for price.
(In the case of COMPQX closing at about 2640. A value
of 4 in the VPCI would be about +.15%.
Equivalent to .015 on a 10 dollar stock)
So its possible that a better version of VPCI would
be:
(VPCI / C)*100
Hope that helps
Thanks
diceman
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (diceman) Because volume is a factor in the calculations. Holidays will be a factor. VPCI does not know if its a holiday or a "weak" day.
I gotta agree with Diceman here. There are times where volume is or should be very indicative of market dynamics, but there are also times where volume is heavily influenced by non-market factors such as a holiday week where everyone is away or even because of shortened trading hours. So in evaluating this indicator, I would be careful before deciding on hard and fast rules based upon a short period of time that includes a holiday period. I think a much truer test of this indicator will come as we leave the July 4th holiday further behind.
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Diceman, You have confirmed that my interpretations are correct and That is appreciated, but in another thread I explained to Hohandy that after looking over the volume patterns for the major Holidays for the last three years there was no standard pattern. Some years the volume increased on a given Holiday and the next year the volume decreased. (Same probabilities as every single market day)
It is all relative after all we are just discussing indicator/Stock market interpretation. No need for a crusade... 
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
I would put it in the category of something to be aware
of.
It all depends on the price and volume action around the
holiday.
If price and the indicator are falling. It will probably fall
less on holiday volume. (but still may not be a noticeable
change)
In the case of COMPQX. We had a rally into July 9.
Volume dropped to about 50% of its normal weight on about
July 3.
This is a good recipe for the "under performance" of the indicator.
Rally days on lower volume had less weight compared to the
simple average.
So on this one. I would put more weight on holiday volume.
(I put the individual values into a spreadsheet and the VWMA
started to really under perform on July 3.)
Thanks
diceman
|
|
|
Guest-1 |