Download software Tutorial videos
Subscription & data-feed pricing Class schedule


New account application Trading resources
Margin rates Stock & option commissions

Attention: Discussion forums are read-only for extended maintenance until further notice.
Welcome Guest, please sign in to participate in a discussion. Search | Active Topics |

Profile: mohtc
About
User Name: mohtc
Groups: Gold User, Member, TeleChart
Rank: Registered User
Real Name:
Location
Occupation:
Interests:
Gender: Unsure
Statistics
Joined: Sunday, March 20, 2005
Last Visit: Tuesday, July 28, 2009 12:33:54 PM
Number of Posts: 78
[0.03% of all post / 0.01 posts per day]
Avatar
Last 10 Posts
Topic: Min Max PCFs
Posted: Tuesday, July 28, 2009 11:57:20 AM

Okay, Bruce.

If PCF's had variables it really would speed up a 21 bar MIN or MAX MACD Exp. Hist. calculation, by eliminating dozens of duplicated calculations...  But if wishes were horses I would need a lot of hay...

I guess I will stick with manual visual sorts and flags (Sort by one thing and flag everything that fits your desired parameters, then sort by something else and unflag everything that does not fit your desired parameters for the second item, etc...).  Sometimes it is faster and easier to do things the "hard way"

Thanks again for your posting of the templates; I may find them useful for other, less complex PCF's.

Topic: Min Max PCFs
Posted: Monday, July 27, 2009 9:32:27 PM
QUOTE (mohtc)
Bruce, I copied the 21 day set for Min & Max, and used a Wordpad file to find and replace all of the various PCF occurances in your template, and tested the Max finder on MOV after updating on July 24, 2009.  It worked correctly, and produced the correct value, but it took 1:35 to execute.  If I were to add this Max finder and the Min finder, that would add 3:10 to the calculation time for every chart.  With 7264 charts, it would take nearly 16 DAYS to calculate my charts!!!!  Is there any way to make TC2007 cough up the value that it displays on the screen already*, to avoid having to wait two weeks for an update?  How does the program find the minimum an maximum values to display on the chart so quickly?  Is there some way to mathematically simplify the formula?  If you wish, I could post it here, but it is very long...

* Change the scale of the windw with MACD in it to the MACD scale, and either the bottom value will exactly match the lowest visible low, or the top value will exactly match the highest visible high.  I can write down the number, and use it outside of TC2007, but cannot use it within any TC2007 formula...
Topic: Min Max PCFs
Posted: Monday, July 27, 2009 9:20:10 PM
Bruce, I copied the 21 day set for Min & Max, and used a Wordpad file to find and replace all of the various PCF occurances in your template, and tested the Max finder on MOV after updating on July 24, 2009.  It worked correctly, and produced the correct value, but it took 1:35 to execute.  If I were to add this Max finder and the Min finder, that would add 3:10 to the calculation time for every chart.  With 7264 charts, it would take nearly 16 DAYS to calculate my charts!!!!  Is there any way to make TC2007 cough up the value that it displays on the screen already, to avoid having to wait two weeks for an update?  How does the program find the minimum an maximum values to display on the chart so quickly?  Is there some way to mathematically simplify the formula?  If you wish, I could post it here, but it is very long...
Topic: Min Max PCFs
Posted: Sunday, July 26, 2009 1:20:55 AM

After reading through the MACD video posts and this topic, my brain is pretty well fried for the day.

Is it possible to apply this method to find the highs and lows of the formulae for exponential histogram MACD 5, 20, 30 ??

I am looking for a way set up two PCF's:  

1.  MACD within 10% of the highest recent value.
2.  MACD within 10% of the lowest recent value.

Topic: Sort by Visual Value to find new highs and lows in price and indicators
Posted: Sunday, July 26, 2009 1:06:31 AM
This is indeed a very powerful too, and I use it a lot.  For example, I set the price scale to Tight (pre-set to 5% per horizontal line), then sort by the visual slope of a linear regression of price, to find out which stocks are climbing at the fastest rate (hit the home key), sliding sideways or have changed direction recently (scroll down to item 250 of 500), and the stocks that are in the steepest decline (500 of 500).

I have recently watched the MACD video and added the exponential histogram version of MACD 5,20,30, and found out that I cannot compare the MACD of one stock to that of another except by using visual sort, because they have different scales.

How does TC2007 determine the minimum and maximum values so it can place the max at the top of the window, and the min at the bottom?

Is there a formula to calculate the visual value of indicators, to use in easyscans?

I wanted to set up PCF to detect when MACD (exponential histogram) 5,20,30 is within the top 10% visually, and another to detect when MACD is within the bottom 10% visually.  

Then I could use the MACD PCF's in easyscans that I have set up to look at Stoc 5,3,2 (highest 10%; lowest 10%) AND price above $1 AND avg volume above the bottom 10% AND etc...

This will avoid having to UNFLAG ALL IN SYSTEM, visual sort all stocks by MACD, Flag the top 10%, Flag the bottom 10%, switch to ALL FLAGGED SYMBOLS, and run my easyscans on this watchlist...
Topic: Understanding MACD
Posted: Saturday, July 25, 2009 9:36:58 PM
This is all great stuff, thanks guys!  : )
I was going to ask how to find the minimum and maximum values of the Exponential MACD Histogram, but you answered that already; thanks.

However I have a question that I did not see answered above.

Is it possible to use the High price instead of the Close?  And the Low price?

Is there some way to use the High price when the stock price is above average, and use the Low price when the stock price is below average (I use linear regression to 'eyeball' above and below average; 15 bars for all of my charts, and one month, two month, etc... linear regression lines in some of my charts).  In other words, I am trying to make the Exponential MACD fit a little better with the stock's actual realtime action but do it on a daily chart.  The stock may reach a new high on a given day, but have a lower close than the prior day, which shifts the MACD that uses Close instead of High.  In some cases it makes the highest bar of the exponential MACD sit under a different daily price bar than the one with the highest high.  (Have I muddied the waters enough, or should I try to explain myself further by getting into the dips in price where the lowest low is not always on the day with the lowest close?)
Topic: Updating Old Watchlists
Posted: Monday, May 16, 2005 9:23:14 PM
Bruce, I accidently found out that if I add a 'Watchlist
Tracking Price' column to a watchlist tab, the price is
zero for a 'weed' chart that has replaced a retired one.

I don't yet know if this is always the case; I was
playing around with sorting by 'Symbol' first and by
'Watchlist Tracking Date' second, and got off onto a
sidetrack of finding that sometimes the chart Symbols
get randomized by the second sort. See my Watchlist
Sort... bug report, earlier tonight. Anyhow, I ended up
seeing only one case of 'WTP' being zero, because I got
distracted by the neat bug I saw, and forgot that I was
trying to pull weeds.
Topic: Updating Old Watchlists
Posted: Sunday, May 15, 2005 4:57:42 PM
Thanks, Bruce

That will help for tracking dates as far back as about
January 2000. I'll figure out something else for the
really old ones (2/97-12/99). I probably shouldn't care
about those anymore, anyhow, but I occasionally like to
review old "what if's" and look for signs that I should
have seen just before moves that I missed, or drops that
I did not miss (Ouch!). Over the years I have added a
lot of stocks to a lot of lists.

MOHTC
Topic: Half-Point+ Breadth Indicators skewed by high priced stocks
Posted: Saturday, May 14, 2005 9:41:29 PM
Thanks, Bruce. I don't usually see the bottom of the
Worden Reports; I stop scrolling at the trend indicators.
Topic: Updating Old Watchlists
Posted: Saturday, May 14, 2005 9:29:22 PM
Thanks for the quick response, Bruce.

I don't understand the custom cumulative indicator that
you proposed, but the idea to "Add a Watchlist Column:
Tracking Days In List" is a good one. I had no idea that
I could do that. That will help in the weeding process,
because I could then sort the watchlist by that column,
descending and zoom out to include the tracking date.
Scrolling through the watchlist will reveal the weeds by
the sudden changing of the zoom, or the absence of the
watchlist tracking pointer. Those can be flagged for
later deletion (with 'delete all flagged'), so I could
continue to scroll through, watching for the orderly
movement of the tracking pointer from left to right.

I am going to log off from the internet now, and play
with it a bit. But I also had another question that I
will 'Watch This Topic' for:

What does your custom cumulative indicator summarize?
Is it an attempt to count days in which the price went
up, or down? If so, what about days in which the price
was unchanged?

What I need is a way to compare the watchlist tracking
date against the date of the oldest price data for any
given chart. If the watchlist tracking date is less than
the date of the oldest price data, then the chart is a
weed and needs to be pulled from the watchlist.

Is there any easy way to find the first date with price
information for a given chart?
"All Stocks"