| Welcome Guest, please sign in to participate in a discussion. | Search | Active Topics | |
Registered User
Joined: 3/12/2005
Posts: 2
|
Hi, I am looking for a PCF for a divergence of [Price] versus [RSI14], for a period of one to two weeks. Thanks
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
I will assume your averaging period is 1. I will also assume that you are looking for price to be trending one way in the last 2 weeks and RSI the other.
Here is a PCF to try:
(AVG(RSI14.1,5)-AVG(RSI14.1.5,5))*(AVGC5-AVGC5.5)
The lower the number it returnsm the more opposed the trends are between your RSI and price. Put a 10-bar LR on both price and your RSI so you can clearly see the trend of each then sort by this PCF in ascending order so you can get a feel for the results.
- Craig
Here to Help!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 4,308
|
I tried a more basic approach.
RSI14.1 must be up over 5 days. (RSI14/RSI14.1.5)*100
Price Must be down over 5 days. Use Price Percent Change 5-Day
Make sure your PCF is updated, then put both conditions in an EasyScan.
For the first condition, use the range selector to find values above 100...in fact, quite a bit above 100 will show a more marked divergence.
For the second condition, use the range selector to find values below zero, the further below, the more marked the divergence.
Call the scan Positive RSI Divergence 5 day
This will find stocks where RSI went up over 5 days, while price declined over the same period...a positive divergence. Of course, you could adjust the formula to any time period. If you wanted to go 10 days, you would have to write a PCF for the second condition. It would be (C-C10)/C10)*100.
The nice thing about this scan is that you can also use it to find negative divergences as well. Open up the negative divergence scan ddialog.
For the first condition, use the range selector to find values well below 100...the farther below, the more striking the divergence.
For the second condition, use thee range selector to find values greater than zero...the higher above, the more striking the divergence.
To save this second scan, click the Save As button, and call it Positive RSI Divergence 5 Day.
Now you'll have both scans.
- Doug
Teaching Online!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
...low values on my ONE PCF find both negative and postive divergences at the same time!
*throws glove at Doug's feet*
- Craig
Here to Help!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 4,308
|
*bows down to Craig*
Clearly, Craig's solution is more elegant, and more versatile.
Actually, Craig's PCF finds more kinds of divergences than my method does. Mine only finds the kind where price moved in one direction while RSI moved in the other. Divergences can also occur when both move in the same direction, but perhaps between two points, price moves up to hit a new high, while RSI moves up but falls well short of a new high. Craig's PCF finds these divergences as well.
The only advantage I can see to my method is that with Craig's PCF, the positive and negative divergences are all mixed together...because low values simply mean a more marked divergence...positive or negative. With my method, one scan finds positive divergences, the other scan finds negative divergences.
Ever so humbly offering another solution...
- Doug
Teaching Online!
|
|
Worden Trainer
Joined: 10/1/2004
Posts: 18,819
|
The ability to split the potential divergences into different scans is a huge advantage of Doug's approach. It really depends on what you like!
Here are two other videos you may enjoy...
Learn to quickly identify when price and volume are in or out of step
Using Linear Regression Sorts to Help Spot Divergences
- Craig
Here to Help!
|
|
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
|
Here's another way to look at it.
This PCF will yield +2 for a positive divergence, and -2 for a negative divergence (zero otherwise).
It's a model of the slope of 7-day linear regression slopes of price vs RSI (7 days is median of 1-2 wks).
Sgn( 3*RSI14.1.0 +2*RSI14.1.1 +1*RSI14.1.2 -1*RSI14.1.4 -2*RSI14.1.5 -3*RSI14.1.6 ) -
Sgn( 3*C0 +2*C1 +1*C2 -1*C4 -2*C5 -3*C6 )
To see it visually, plot a 7-day LinReg child on price, and on the RSI14.
Jim Dean
|
|
Registered User
Joined: 10/7/2004
Posts: 799
Location: Duluth, GA
|
P.S. please note that PCF is NOT based on Wilder's Smoothing, but rather TC's native RSI formula
|
|
Registered User
Joined: 3/12/2005
Posts: 2
|
Many thanks for a quick response. I will try to experiment these great ideas. It is very important to me, Nick
|
|
|
Guest-1 |