Welcome Guest, please sign in to participate in a discussion. | Search | Active Topics | |
Registered User Joined: 10/7/2004 Posts: 43
|
A cumulative indicator for price/volume relationships....
Up Condition: c-c1>0 and v>v1*1.05 Down Condition c-c1 and v>v1*.9 Value to Add/Subtract abs(v-v1)
I use a smoothing of 1 and apply a 5 and a 13 bar MA.. But you can use what ever you like, it’s just to see which side of the trend your on...
The indicator: moves up if today’s close is up and the volume is at least 5% greater than the prior day.. moves down if today’s close is down and the volume is at least 90% of the prior day.. moves sideways if (today’s close is up and the volume is up by less than 5%) and (today’s close is down and the volume is less than 90% of the prior day)..
In other words it will go flat or move down on an up trend with decreasing volume and go flat or move up on a down trend with decreasing volume.. Give a warning of failing trends..
However, be aware, it can give some false indications around holidays (any abnormal volume day)..
You can find some interesting charts by doing a visual sort on the indicator..
Tisync
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 4,308
|
Thanks for sharing this idea. I'm gonna give it a try!
Looks like in your down condition, you may have meant C-C1<0.
- Doug Teaching Online!
|
|
Registered User Joined: 10/7/2004 Posts: 43
|
QUOTE (Doug_H) Thanks for sharing this idea. I'm gonna give it a try!
Looks like in your down condition, you may have meant C-C1<0.
Yes, you are correct... sometimes I type faster than I think..
|
|
Registered User Joined: 10/7/2004 Posts: 22
|
I am trying to use something similar to this to simply count recent big volume up days. My starting formula (tune later) to test are:
up condition: c>c1+.5 and v>avg(v,50)*1.5
down condition: c<c1 and v>avg(v,50)* 3
value to add/subtract 1
1) When I use this I get an indicator, but I can't tell how far back in time the indicator starts - it seems to change depending on what is displayed - can you tell me what date is used to start the accumulation?
2) I would like to use this indicator as a sort. I tried it with actual value sort but it does not seem to work for me. I only get values of +- 0, 1, or 2 in the sort field display of the watchlist window, instead of the values displayed on the chart scale of the indicator. What am I doing wrong?
Thanks!
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
The starting plot varies. You will want to use VISUAL VALUE for your sorting.
- Craig Here to Help!
|
|
Registered User Joined: 10/7/2004 Posts: 22
|
Thanks! but.... Using visual sort with those formulas results in most of the list at 100% - all the lines trend up. I'm trying to sort by the VALUE of the indicator as displayed to the right of the chart to see which has the highest actual count of up days, but the sort value displayed does not seem to track the chart scale values. Is that possible?
Perhaps there is a better way to accomplish what I want? Count the number of days up recently with over $.5 move and volume over 150% of 50 day average. (an approximation of the IBD "Stocks on the Move" tables.)
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
GrahamSanJose, You didn't specify a timeframe and it may not be what you want, but you might want to try this (it's based on 1-Year):
Select Chart Temlplate | Add Indicator | Indicator. -Visible: Checked -Center Zero Line: Unchecked -Plot Using Price Scale: Unchecked -Smoothing Average: 252 -Average Type: Simple -Indicator Formula:
((C<C1 AND V>AVGV50*3) - (C>C1+.5 and V>AVGV50*1.5))*252
It should include results from the last 252 trading days in its count. You can adjust the Smoothing Average and the value at the end of the formula to adjust the number of trading days included. The difference is that older data is also removed from the indicator instead of only newer data being added.
It will look quite a bit different than your Cumulative Indicator, but it is looking at the same thing. You can demonstrate this for yourself by setting the Smoothing Average to 1. The up spikes should correspond exactly to points where your Cumulative Indicator is going up and the down spikes should line up with the downward blips.
If you want something that looks more like the Cumulative Indicator (albeit in a curvy sort of way), you might want to try changing the Average Type to Exponential (OK, maybe it doesn't look like the Cumulative Indicator, but you might want to try it anyway).
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 10/7/2004 Posts: 22
|
Thanks Bruce.
That does it.
I find I can do an indicator sort on the Actual Value of your formula to get what I'm looking for.
The formula needs an extra set of parens I believe: ((C<C1 AND V>AVGV50*3) - (C>C1+.5 and V>AVGV50*1.5))*252
Now to play with the values and see if this is useful....
ps. I still don't understand why the "Actual Value" sort on the cumulative indicator I started with does not sort on the value as displayed. It seems to actually only sort on the past 2 or 3 days of data, not the chart as displayed..... ah well, irrelevant for now.
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
You are correct about the formula. I had copied the formula from the window where I was typing the post and never pasted it back after testing it. I've added them in the original post.
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Registered User Joined: 1/1/2005 Posts: 2,645
|
GrahamSanJose,
Another method to do the sort is to do a SortBy on your Custom Cumulative Indicator using:
Sort Method: Net Change Period: 252
The sort order and the values obtained are identical to that obtained by Bruce's method. The advantage is that you do not have to change the PCF when you change the Period and you do not need the additional indicator. The disadvantage is that the sort takes about 3.27 as long (on my machine with the S&P 500 Components WatchList). My guess is that the time disadvantage becomes worse as the Period increases.
Thanks, Jim Murphy
|
|
Registered User Joined: 10/13/2005 Posts: 13
|
Is there a pfc for the Williams accumulation/distribution with horizontal lines at 95 and 5.
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 4,308
|
I have replied to your question posted in duplicate at the following link:
Williams a/d
- Doug Teaching Online!
|
|
Guest-1 |