Registered User Joined: 3/21/2006 Posts: 19
|
I reviewed the chart on TRB this weekend and observed the divergence in TSV and MS versus price in April vs late May. Could you provide a scan formula which would replicate these divergences?
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
What TSV settings are you using?
For MS, try these for a 1-month divergence:
Potentia1 1-month Money Stream Divergence (AVGC10<AVGC10.10 AND MS10>MS10.10) OR (MAXH10<MAXH10.10 AND MAX(MS1,10)>MAX(MS1.10,10)) OR (MINL10<MINL10.10 AND MIN(MS1,10)>MIN(MS1.10,10))
You should also try SORTING by the built in criterion MONEYSTREAM vs. PRICE 3-MONTH. Sort in descending order to find potential divergences over the last 3 months.
- Craig Here to Help!
|
|
Registered User Joined: 3/21/2006 Posts: 19
|
Craig: I am using a TSV 18 MOV AVG 10
|
|
Worden Trainer
Joined: 10/7/2004 Posts: 65,138
|
I'm sure Craig will provide a formula in line with your settings soon, but you may also wish to review the following in the meantime:
Visual Sorting by TSV Using Linear Regression Sorts to Help Spot Divergences Using Linear Regression vs Classical Peaks/Valleys for Divergence Analysis
-Bruce Personal Criteria Formulas TC2000 Support Articles
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
Try this:
Potential 1-month TSV18 divergence:
(AVG(TSV18,10)>AVG(TSV18.10,10) AND AVGC10<AVGC10.10) OR (MAXH10<MAXH10.10 AND MAX(TSV18,10)>MAX(TSV18.10,10)) OR (MINL10<MINL10.10 AND MIN(TSV18,10)>MIN(TSV18.10,10)) OR (MIN(TSV18,21)>0 AND AVGC5<AVGC5.5 AND AVGC10<AVGC10.10)
- Craig Here to Help!
|
|
Registered User Joined: 5/17/2006 Posts: 493
|
craig, when I plug in the tsv divergence it gives me a syntax error, is the formula right?
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
Fixed... try it now.
- Craig Here to Help!
|
|
Registered User Joined: 5/17/2006 Posts: 493
|
thanks, I think the MS formula is syntax as well.
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
Sure enough! Try it now.
- Craig Here to Help!
|
|
Registered User Joined: 5/17/2006 Posts: 493
|
also, how would you change it (both formulas) from 1 month to say 1 week or 2 weeks, what number would change?
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
Best to show by example:
Potentia1 2-week Money Stream Divergence (AVGC5<AVGC5.5 AND MS5>MS5.5) OR (MAXH5<MAXH5.5 AND MAX(MS1,5)>MAX(MS1.5,5)) OR (MINL5<MINL5.5 AND MIN(MS1,5)>MIN(MS1.5,5))
Potential 2-week TSV18 divergence:
(AVG(TSV18,5)>AVG(TSV18.5,5) AND AVGC5<AVGC5.5) OR (MAXH5<MAXH5.5 AND MAX(TSV18,5)>MAX(TSV18.5,5)) OR (MINL5<MINL5.5 AND MIN(TSV18,5)>MIN(TSV18.5,5)) OR (MIN(TSV18,10)>0 AND AVGC3<AVGC3.3 AND AVGC5<AVGC5.5)
- Craig Here to Help!
|
|
Registered User Joined: 5/17/2006 Posts: 493
|
sorry craig, could you show me for 1 week?
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
This is tougher as 5-days is not an even number. I will do it for 6-days.
Best to show by example:
Potentia1 6-day Money Stream Divergence (AVGC3<AVGC3.3 AND MS3>MS3.3) OR (MAXH3<MAXH3.3 AND MAX(MS1,3)>MAX(MS1.3,3)) OR (MINL3<MINL3.3 AND MIN(MS1,3)>MIN(MS1.3,3))
Potential 6-day TSV18 divergence:
(AVG(TSV18,3)>AVG(TSV18.5,3) AND AVGC3<AVGC3.3) OR (MAXH3<MAXH3.3 AND MAX(TSV18,3)>MAX(TSV18.3,3)) OR (MINL3<MINL3.3 AND MIN(TSV18,3)>MIN(TSV18.3,3)) OR (MIN(TSV18,6)>0 AND AVGC2<AVGC2.2 AND AVGC3<AVGC3.3)
- Craig Here to Help!
|
|
Registered User Joined: 5/17/2006 Posts: 493
|
Craig,
I need help with a 1 month tsv divergence as above but with the settings of tsv (28) and ma 19 (both exponential)
also can you provide me with a 1 month ms divergence with a moving avergae of 22 (also exponential)
and just to clarify when looking at these this will find the lowest low in a 1 month and a near term low shortly after with a divergence in tsv and ms, am I reading that right?
thanks
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
TSV 28,19
(AVG(TSV28,10)>AVG(TSV18.10,10) AND AVGC10<AVGC10.10) OR (MAXH10<MAXH10.10 AND MAX(TSV28,10)>MAX(TSV28.10,10)) OR (MINL10<MINL10.10 AND MIN(TSV28,10)>MIN(TSV28.10,10)) OR (MIN(TSV28,21)>0 AND AVGC5<AVGC5.5 AND AVGC10<AVGC10.10)
Potentia1 1-month Money Stream Divergence (AVGC10<AVGC10.10 AND MS10>MS10.10) OR (MAXH10<MAXH10.10 AND MAX(MS1,10)>MAX(MS1.10,10)) OR (MINL10<MINL10.10 AND MIN(MS1,10)>MIN(MS1.10,10))
These look for either:
Higher recent low in the indicator with a lower recent low in price OR Higher recent high in the indicator with a lower recent high in price OR TSV greater than 0 with a downtrending price
- Craig Here to Help!
|
|
Registered User Joined: 5/17/2006 Posts: 493
|
Craig,
what is usually the entry, the next bar? Do these look for short positions as well. I'm trying to divide it up so I have both ms and tsv divergence for buys and shorts. How would I break these up? It looks like the tsv looks for buys, is that correct?
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
I cannot begin to answer these types of questions. I cannot provide anything close to trading advice.
These PCFs are designed to help you find positive divergence between the indicators and price. You have to decide what to do (or not do) with any stocks these PCFs find.
- Craig Here to Help!
|
|
Registered User Joined: 1/1/2005 Posts: 2,645
|
QUOTE (Craig_S) Potentia1 1-month Money Stream Divergence
(AVGC10>AVGC10.10 AND MS10<MS10.10) OR (MAXH10<MAXH10.10 AND MAX(MS1,10)>MAX(MS1.10,10)) OR (MINL10<MINL10.10 AND MIN(MS1,10)>MIN(MS1.10,10))
Craig,
Here is your original:
Potentia1 1-month Money Stream Divergence (AVGC10>AVGC10.10 AND MS10<MS10.10) OR (MAXH10<MAXH10.10 AND MAX(MS1,10)>MAX(MS1.10,10)) OR (MINL10<MINL10.10 AND MIN(MS1,10)>MIN(MS1.10,10))
I do not see any difference.
Why is the way you are looking for negative divergences so completely different from the way you are looking for positive divergences? Are you suppose to be looking for positive divergences only and have the first two inequalities backwards?
If one month is 21 days, I can give reasons for using the PCF:
(AVGC10<AVGC10.11 AND MS10<MS10.11)
for positive divergences. For a 20 day period, this becomes:
(AVGC10<AVGC10.10 AND MS10>MS10.10)
They are the consequences appropriate least-mean-square filtering of the 21 or 20 data points, respectively.
I know of no logical reason for the rest of the PCF. Maybe you can help.
I'll bet you wish that I would go to bed and stay there?
Thanks, Jim Murphy
|
|
Registered User Joined: 1/1/2005 Posts: 2,645
|
Errata:
(AVGC10<AVGC10.11 AND MS10<MS10.11)
should read
(AVGC10<AVGC10.11 AND MS10>MS10.11)
Sorry, Jim Murphy
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
This PCF:
(AVGC10>AVGC10.10 AND MS10<MS10.10) OR (MAXH10<MAXH10.10 AND MAX(MS1,10)>MAX(MS1.10,10)) OR (MINL10<MINL10.10 AND MIN(MS1,10)>MIN(MS1.10,10))
Should be
(AVGC10<AVGC10.10 AND MS10>MS10.10) OR (MAXH10<MAXH10.10 AND MAX(MS1,10)>MAX(MS1.10,10)) OR (MINL10<MINL10.10 AND MIN(MS1,10)>MIN(MS1.10,10))
(I will fix it in the above posts)
Anyone can make the periods used anything they like. If you want to do 10 and 11 for 21 versus 10 and 10 for my 20 you are welcome to.
As for your sleeping patterns, the inability to sleep and stay asleep appears to be one thing we share.
- Craig Here to Help!
|
|
Registered User Joined: 1/1/2005 Posts: 2,645
|
QUOTE (Craig_S) Anyone can make the periods used anything they like. If you want to do 10 and 11 for 21 versus 10 and 10 for my 20 you are welcome to.
Craig,
Please notice that in following your approach the PCF I suggested for a 21 day period:
(AVGC10<AVGC10.11 AND MS10>MS10.11)
uses only SMAs of period 10, not 10 and 11.
I know of no logical justification for using SMAs of periods 10 and 11. You may have misread the PCF as:
(AVGC10<AVGC11.10 AND MS10>MS11.10)
Thanks, Jim Murphy
|
|
Worden Trainer
Joined: 10/1/2004 Posts: 18,819
|
I understood your suggested PCF for 21 days. When I said "period" I meant the period of time looked at on the chart. Your suggestion covers a 21 day period by "skipping" the day 10 days ago.
- Craig Here to Help!
|
|
Guest-1 |