Letters - July 2005

July 2005 Letters To The Editor
or return to July 2005 Contents
The editors of S&C invite readers to submit their opinions and information on subjects relating to technical analysis and this magazine. This column is our means of communication with our readers. Is there something you would like to know more (or less) about? Tell us about it. Without a source of new ideas and subjects coming from our readers, this magazine would not exist.
Address your correspondence to: Editor, STOCKS & COMMODITIES, 4757 California Ave. SW, Seattle, WA 98116-4499, or E-mail to editor@traders.com. All letters become the property of Technical Analysis, Inc. Letter-writers must include their full name and address for verification. Letters may be edited for length or clarity. The opinions expressed in this column do not necessarily represent those of the magazine. -Editor

S&C MARKET PROFILE ARTICLES
Editor,
I am conducting research on every article you have published on Market Profile. Can you please help me find all the links?
Alex Benjamin
via email
By performing a search at our website, www.Traders.com, you should be able to find all the references to Market Profile we've published. You could also search for Donald Jones, since he has written articles on that subject. Good luck.--Editor

INTRADAY END-OF-DAY DATA
Editor,
I am trying to find a vendor for intraday end-of-day data. Can you please email me a link or a list of the top 10?
Amir Sani
via email
A list of data service vendors is available on our website, www.Traders.com, under Traders' Resource. In addition, our November 2004 issue has a list of data vendors.--Editor

EXIT STRATEGY FOR MOVING AVERAGE PULLBACKS
Editor,
The article "Trading Moving Average Pullbacks" by Steve Palmquist [S&C, April 2005] does not include an exit strategy. Is it based on a target price or trailing stop, or a combination of both?
Ron Unzueta
via email
Steve Palmquist responds:
The exit strategy to "Hold for three periods" is given at the bottom of Figure 3. Positions are purchased when the setup and trigger conditions are met, as specified in the last paragraph of the first column. The positions are then held for three days again as specified in Figure 3, then closed.
The advantage of using a time-stop (sell after holding three days) rather than a price-stop is that the time-stop does not require the trader to monitor the market closely all day, only to wait three days and sell. Backtesting indicates that the time-stop exit strategy compares reasonably well to various price and pattern exit strategies, and is often simpler to use.
If readers have additional questions, they can contact me at maps@ daisydogger.com.

SEEKING TRADING SOFTWARE
Editor,
I am looking for a good, reliable trading software program and a source for real-time quotes that is compatible with the program. Can you help me?
Name withheld
via email
Deciding which trading software to purchase depends entirely on your trading style and experience level. Many trading programs are available, and each will have a list of datafeeds that are compatible with it.
Each issue of S&C contains thorough reviews of software products. I would recommend that you go through past issues to help narrow down the software that you think will fit your needs.--Editor

BEST STOCKS FOR BREAKOUT?
Editor,
I really enjoyed Markos Katsanos' May 2005 article ["Measuring Flags And Pennants"]. In fact, I was so impressed I bought all his previous articles on the S&C website (I'm not a subscriber) and even upgraded my old version of MetaStock.
I noticed in his article that the lower-priced small-cap stocks did very well in his test of 13 stocks. I'm currently trying to decide which stocks to put in my database, and am wondering if it would be best to focus on the lower-priced small caps. I also thought about including IBD 100 stocks. Any thoughts on what types of stocks perform the best for the pennant/flag breakout system?
Chris Russell
via email
Markos Katsanos responds:
When I started researching for the article, I screened more than 2,000 stock charts trying to identify 100 flags and pennants to include in my sample list for further statistical analysis. When I finally identified 100 flags, my stock list included only six large-cap stocks (A, BHP, WPI, NXTL, MBT, NVLS) and a few mid-caps, with the rest small-cap stocks.
The average price before the breakout was $14 with a maximum of $73 (MBT) and a minimum of $0.60. This is because it is much harder for a stock like MSFT with a float of more than 10 billion shares to move by more than 5% in 10 days than a stock like Gap with a float of less than 40 million shares. Performance is also positively correlated with historical volatility, as this was depicted in the correlation matrix in Figure 7 of the article.
Considering that you can't screen for market capitalization with most technical analysis software, you can only limit the price to less than, say, $50. If the list is too large, you can reduce it further by screening for volatility and including only volatile stocks. I haven't tested the Ibd stocks, but I suppose that strong momentum stocks are good candidates for your list.

KATSANOS: LOG SCALE VS. Y-AXIS
Editor,
I read Markos Katsanos' article in STOCKS & COMMODITIES about measuring flags and pennants, and I found it very helpful. I am not a specialist in technical analysis but I really enjoy reading such articles. I have a few questions: What is the advantage of using semilog scale of Y axis? Can I use it? Can I use the formula for symmetrical triangles and what is the accuracy? How about the bearish pennants - how can I use the formula? I get negative numbers and the formula does not work. I look forward to your reply!
Theodore Kousathanas
Mykonos, Greece
Markos Katsanos responds:
I recommended using the log scale, as it is easier to compare percentage changes in the pole and breakout segment of the flag. For example, if a stock rises from 10 to 20 in the pole and then, after forming a flag, breaks out from 20 to 30 in the arithmetic scale, both moves would be depicted equally, whereas in the log scale, the pole would be double the breakout.
Concerning your second question, I understand that research on bear flags would have been more useful under current market conditions. However, when I was researching for the article (in the summer of 2004), the market was still bullish. Considering that a lot of investors are not comfortable selling stocks short, I decided to limit my research to bullish flags. However, if you want to use the formula for bearish flags, you can take the absolute value of the pole, so there wouldn't be any negative numbers. My research did not include symmetrical triangles either.

KATSANOS: FLAGS & PENNANTS & CODE
Editor,
I appreciate Markos Katsanos sharing his excellent ideas with traders. I have been using MetaStock Professional 9.0 for a year and a half, and I have studied his articles in S&C. In his May 2005 article, "Detecting Breakouts From Flags & Pennants," he shared a MetaStock exploration for a flag and pennant pattern-recognition system. After reading this article, I have the following questions.
When I enter the information into Explorer, I get the following error message for Target: "This is not a recognized name, constant, or operator." How can I correct this?
In the same article, Katsanos also shared MetaStock code: "System Test, Flags & Pennants." When I enter the code, follow the instructions, and run the simulation for 209 stocks, it completes the test with no successful trades. Under status,"Completed. Had errors" is displayed for each of the stocks. Do you have any suggestions on how I can get this to work?
From the System Test, I entered the "Enter Long" and "Sell Order" code into a new Expert, but received the following error message: "An Msx Dll is reporting an internal error. Referenced ‘Simulation.Currentpositionage' Dll Message: The System Tester encountered an error inspecting the ongoing simulation. This inspector was unavail." Do you have any suggestions on how I can get this to work?
Jim Mann
via email
Regarding the first question, please see the errata section at the end of this section for the correct code.
Regarding the second question, Markos Katsanos responds: Concerning the system test error, I used MetaStock 8.0 and ran the test numerous times with no problems. Try loading more data and if you still get no trades, you will have to contact MetaStock technical support.
Concerning the error in Expert Advisor, simulation functions only work in the system test and should not be used in Expert Advisor. The following is from the MetaStock Help section concerning simulation functions:
These functions let you use values that occur in a system test simulation to determine rules for buying and selling. For example, if you want the simulation to place a Sell Order when the equity dips below 5,000, you could write:
 if(Simulation.AccountCash < 5000,1,0)
Note that these functions only work in the Buy Order, Sell Order, Sell Short Order and Buy to Cover Order rules in the Enhanced System Tester Dialog. Also, the CurrentPosition simulation functions only work in the Sell Order and Buy to Cover rules.--Editor

KATSANOS: DISAPPEARING ENTRY POINTS?
Editor,
I have been a stockbroker and MetaStock user for over 10 years. Having copied the filter in its entirety into a new indicator file and dragged it over a chart, I found two entry points, which failed to appear when I run the exploration on the indicated days. How can this be? The entry points on the chart are pretty good, but any exploration would have failed to pick these up. The stock is Medical Marketing (LondonSE:Mmg) and the flag appeared at the end of February, beginning of March. Any ideas?
Bill Rook
Henley on Thames, England
Markos Katsanos responds:
Because of the dynamic nature of the zig function, which can change at a later date, the system code published in the magazine was meant to be used only with explorations and not as an indicator.
The flag you mentioned was detected by the MetaStock explorer on three occasions: 2/23/05, 2/25/05, and 2/28/05. If you still can't verify these, you should load at least 300 records (because of the VFI calculation) and run the exploration again.

KATSANOS: THE DISCUSSION CONTINUES
Editor,
I've been running the pennant/flag scans for a couple of days now (on all stocks, just to get a feel for it). I've been impressed with its ability to identify the formation, but I do have a couple more questions that came up from the scans.
1. I noticed that it will identify a flag that is only in its third or fourth day. Is there a recommended minimum length for a flag?
2. I also noticed that a stock may show up on consecutive days. Case in point is Stmp. It showed on my scan for Friday and Monday.
However, it recalculates the target price each time so there was a different target price for each scan. What should be the valid target price-the day or first signal?
Chris Russell
via email
Markos Katsanos replies:
According to the statistics in Figure 1 of the article, 90% of the flags lasted from four to 21 days with a median (most likely) of nine days. If you are not comfortable entering a trade too early, you can either change the code to exclude flag durations of less than five days by changing the 16th line from X1>2 to X1>4 or you can put the stock in your watch list and wait for a breakout.
The target price calculation in column F of the exploration is based on the current price, so it's normal to get slightly different targets for two consecutive days. My statistical results and the formula, however, were based on the price before breaking out from the flag formation, so you should use the lowest of the two. In the case of Stmp, this is 26.26, derived from the exploration on 5/13/05.

TSI OR MACD?
Editor,
In regard to Mark Phillips' "Less Is More" article in the June 2005 issue: What is the difference between the TSI and the MACD? They look exactly the same when plotted next to each other. Thanks!
Brent Lindaman
Omaha, NE
The TSI and MACD are similar indicators, although there are subtle differences in their calculations. Please refer to the S&C Traders' Glossary at our website, www.Traders.com, or on page 75 for the formula of TSI.--Editor

ERRATA: FLAGS & PENNANTS FORMULA
Editor,
I read with interest your article "Measuring Flags And Pennants" in the May 2005 S&C. However, when I tried copying the MetaStock formula you provided to MetaStock 9.0 for the exploration, columns E and F weren't usable because the word "Target" was not recognized. Any suggestions on correcting this so I can add that to the exploration?
Eric Simrod
Marietta, GA
Markos Katsanos replies:
First, the word "target" should be in the column heading, not in the formula space. Second, I apologize that there was an error in the code. You should delete the word "pole" from the code in columns E and F, so the final code should read:
Column E
Name: Target %
Formula:
 zz:=Zig(C,17,%);
X:=BarsSince(ZZ < Ref(ZZ,-1)AND Ref(ZZ,-1) > Ref(ZZ,-2));
X1:=LastValue(X)+1;
c2:=Ref(HHV(C,(2+X1)),-1);
c3:=Ref(LLV(C,26),-X1);
POLE:=(C2-C3)/C3*100;
TARGET:=1.94* Power(pole ,.724);TARGET
Column F
Name: Target Price
Formula:
 zz:=Zig(C,17,%);
X:=BarsSince(ZZ < Ref(ZZ,-1)AND Ref(ZZ,-1) > Ref(ZZ,-2));
X1:=LastValue(X)+1;
c2:=Ref(HHV(C,(2+X1)),-1);
c3:=Ref(LLV(C,26),-X1);
POLE:=(C2-C3)/C3*100;
TARGET:=1.94* Power(pole ,.724); (1+TARGET/100)*C
If you still get an error, send me the formula for columns E and F and I will try to see where the error is.

ERRATA: BELKNAP FORMULAS
This note is for the readers who have sent me messages about my May 2005 S&C article, "Cycles In Time And Money." Thanks for your compliments and for taking the time to convey your difficulties with the table that appeared on page 35. You are correct; there is a typographical error that makes the MetaStock formulas in the table dysfunctional.
The easiest way to correct the error is to replace the formula name "Volatility (Sigom%)" in your own indicator builder with the name "_sac-sigom." (In an alternative version of the table, all the indicator names started with the underscore so that they would group together in the indicator builder list.) It may help to detail the necessary steps.
With any MetaStock Smart or Windows chart displayed, right-click your Indicator QuickList and select Indicator Builder. Highlight the existing (previously typed) formula name "Volatility (Sigom%)," press Edit, and retype the name "_sac-sigom." In the absence of a previously typed formula, at the Indicator Builder, press New, type "_sac-sigom," and in the formula window, type the six lines of text shown in the table on page 35 below "Volatility (Sigom%)."
As an alternative correction method, for any indicator of interest such as the minor cycle index, replace the statement:
sigom:=FmlVar("_sac-sigom", "sigom")
with the statement:
sigom:=FmlVar("Volatility (Sigom%)","sigom")
This, of course, requires eight times as much typing to correct all indicators.
The objective here is to use a MetaStock formula language function to reference variables within custom indicators (see any MetaStock user manual -- for example, page 252 of Pro Version 8.0 or 8.01). This avoids retyping the six lines of text eight more times, which would expand the table well beyond one page.
Thanks again for pointing out this very instructive error. Please feel free to address any additional questions or comments to stu@seanet.com.
Stuart Belknap

Back to July 2005 Contents