or return to October 2000 Contents
READING ON THE ROAD
Editor,
I read your wonderful magazine cover to cover each time on airplanes and in hotels, and I never remember to acknowledge you and your team of editors and guest writers. Keep up the good work!
Jimmy Castro
Austin, TX
END POINT MOVING AVERAGE
Editor,
The end point moving average (EPMA) is an excellent least-squares curve-fit technique, per Patrick Lafferty's articles in the June 1999 and October 1995 issues of S&C. I use a 13-day EPMA for trend-determining indicators, and I am happy to have it. It should be a part of every trader's tool set.
Similarly, I have found several spreadsheet macros helpful. Would such Excel macros be of interest to other readers? For example: 1) a macro that converts points and 32nds to decimal points for bonds; 2) a formula for the date column that skips weekends; 3) a macro converting any variable over coordinate range A1-A2 to a different range AB1-AB2. This allows plotting/comparing variables together, on the same graph; 4) "perpetual lead contract" routine in the scratched part of a sheet that is used when contract rollover is under way. Volume proration is used to compute the lead contract O,H,L,CL,V,OI each day of the rollover period. This avoids step-function inputs or back-computing new contract data; and 5) a min and max Fibonacci move calculator.
Frank Suler, via E-mail
Key West, FLPerhaps our readers can let us know. Thanks for writing. -- Editor
TRADING PSYCHOLOGY ARTICLES
Editor,
I have not noticed a trading psychology article recently in STOCKS & COMMODITIES. Have you discontinued this section?
John P., via E-mail
We have not discontinued publishing articles on trading psychology; we recently published one by Adrienne Laris Toghraie in the June 2000 S&C. A trader's mental perspective while trading has a lot to do with his or her success in the markets. We try to mix things up from month to month in Technical Analysis of STOCKS & COMMODITIES, The Traders' Magazine. We also measure our readers' interest via survey responses on the advertiser product information card in the back of each issue. -- Jack Hutson, Publisher
CHAIKIN'S MONEY FLOW
Editor,
Christopher Narcouzi's article "Chaikin's Money Flow" in the August S&C was excellent! I use TradeStation, which contains an indicator called Chaikin's oscillator. Is this the same as Chaikin's money flow? In addition, what package did Narcouzi use to plot the charts? They were very well done.
Brett Nelson, via E-mailChristopher Narcouzi replies:
Glad you enjoyed my article. The Chaikin A/D oscillator is different than the money flow indicator that I discussed in my article. I use MetaStock, and here's the MetaStock formula for the Chaikin money flow:
Sum((((Close-Low)-(High-Close)) / (High-Low))*Volume,21) / Sum(Volume,21)
Good luck.
STRENGTH OF SUPPORT AND RESISTANCE
Editor,
I read Christopher Narcouzi's article "The Strength Of Support And Resistance: Measuring Overhead Supply" in the July 2000 STOCKS & COMMODITIES. It was very insightful. However, I have a question about measuring volume. For example, in Figure 1 on page 30, how do you conclude eight million shares at point C? Are there any rules?
Thanks for your reply.
Frank Chen, via E-mailChristopher Narcouzi replies:
You'll find the answer to your question in the first paragraph of the section under the heading "Support and resistance zone." In the second paragraph, I indicated that the zone was between 16.50 and 18. If you add up the volume between these two prices, then you will have eight million shares.
This is not an exact science. Each person will come up with a different number of shares for overhead supply. That's okay, unless you're way off. In addition, you might read the section on "Bottoms and tops" again.
Good luck.
BUYER BEWARE REVISITED
Editor,
Regarding the August Letter to S&C titled "Buyer Beware" and your reply, your recommendation to readers that they have some protection when purchasing a product with a credit card is generally not true when it comes to purchasing from a company outside the US. It is not widely known by credit card users that there is usually no recourse from a credit card company for charges made by a company that resides outside the US. Consumers should check with their charge card company for their specific policy.
I found this out years ago when I purchased a software program from a Canadian dealer. The program did not work. I immediately returned it, but the company refused to provide a refund. After calling my credit card company to dispute the charge, I found out that it did not take disputes or make chargebacks against purchases from a company located outside the US.
Readers should also know that many of the pirated programs that are sold by pirates in and outside the US are not executable programs. They are often upgrades that the pirates have downloaded from the software companies' Website. The upgrades are usually only a part of the whole program and are only executable if installed with the original program.
You get what you pay for. In most cases with pirated programs, you get nothing usable.
Robert Miner
Dynamic Traders Group, Inc.
Tucson, AZ
www.dynamictraders.comThank you for your informative letter that will surely be helpful to readers. -- Editor
FORMULAS
Editor,
I have been a reader of your fine magazine since 1990 and have all the back issues in my files. However, I finally have to ask you to address a problem that vexes me, and perhaps other readers.
I develop my own software. I use (gasp!) BASIC to write all my programs. I have used FORTRAN, ALGOL, PL-1, and others down through the years, but I got hooked on BASIC when I started using PCs back in the very early 1980s. (I'm in the midst of a cautious installation of Linux and C++ on a second hard drive, but that's another story.) I currently employ a comprehensive screening and technical analysis system to work on the entire list of US stocks, using a database that ranges up to 15 years of data. What I do has recently acquired the fancy handle of "data mining." (I used to call it "Let's see if we can figure out if this indicator works," or "Let's see if it works for me the way I invest/trade/speculate/burn money.")
Therein lies my problem. As an example, the August 2000 "Stochastic Pops" article by Dave Steckler is a good article and makes sense. I have a lot of respect for Welles Wilder's work. However, it would be a lot quicker to set up in a database the 3,000 stocks in the US market with volume over two million per day, and then go against it with the program in question, test different values, smoothing periods, combinations of smoothing periods, trigger points, and so on, to see which stocks had good profits under which regimes (that is, mine the data).Could you lean on your authors to include the mathematical notation for the routines they present? Usually, there is an initial period, then a recursion that continues until the end; the mathematical notation is usually fairly compact once the program starts the recursion. That way, those of us who might use the routines for somewhat larger applications could develop them faster.
You probably don't get many requests for what appears to be more complexity. Please see what you can do. Thanks for a fine magazine.
Barry Harmon, via E-mailOn the contrary, you may have noticed recently in our Letters to S&C column the calls for more algebraic documentation. As I replied then, we're trying to include both the software code and algebraic notation whenever possible. We have some readers who want the full mathematical documentation so they can analyze the techniques themselves, and we have other readers who just want to plug in the code. We'll try to please both sets whenever we can. We also ask software vendors to contribute code each month to our Traders' Tips column. -- Editor
PRICE PATTERNS
Editor,
I had recently written you asking for sources for reading charts -- not classical patterns but more along the lines of support and resistance and so on. Well, lo and behold, my new issue of S&C arrived containing Martin Pring's article on price patterns ("Price Patterns, Part I," August 2000 S&C). Just the sort of thing I was after!
Don't suppose there's any way to get the follow-up article prior to publication of your next issue?
Geoffrey Levens, via E-mailWe're glad that issue of S&C provided just what you needed. And no, you can't! (You'll find Part II of Pring's article in this issue.) -- Editor
TESTING & OPTIMIZATION
Editor,
S&C has published many articles containing warnings against overoptimization and the importance of testing on out-of-sample data. Some articles suggest breaking data into thirds, optimizing on the middle third, and performing out-of-sample testing on the first and third segments. This method is one-time in nature and doesn't help when going live or on an ongoing basis.
S&C has also presented an excellent series of articles from Dennis Meyers using a technique called walk-forward testing. This method has the unique advantage of being adaptive in the same manner as adaptive moving averages are. The largest problem is that performing walk-forward testing in popular products such as TradeStation is extremely time-consuming, especially if working with many issues.
Are there any products that have walk-forward testing built in, or do you know of a third-party TradeStation add-on?
Brooks Rimes
Grand Island, NYWe don't know of any products that automate this process, but perhaps someone reading this can make a suggestion. -- Editor
NOVICE TRADERS' NOTEBOOK
Editor,
Great Website at Traders.com. Please keep up the good work on the Novice Traders' Notebook, as we beginners need this kind of information.
Al Purdy, via E-mail
PRICE EQUIVOLUME
Editor,
Regarding the letter from Brian Hershey in the August 2000 Letters to S&C ("Novice Traders' Notebook") in which he asks for resources for Equivolume charting, MetaStock also provides the Equivolume charting capability, besides the software products you listed.
Claud Baruch
Wayne, NJ
EUROPEAN STOCK EXCHANGE DATA
Editor,
In response to a letter to the editor in the June 2000 S&C in which Marino Campos inquired about obtaining European stock data, the Website Downloadquotes.com provides free world market and index data in different formats, including European data. As for the French stock markets, the Website Abcbourse.com provides free stock data. Go at the bottom of the main page and click on "historiques," then choose "format," and then click on "telechargez."
Traders can also download intraday data, but only for the RM (reglement mensuel). Daily data is available about three hours after the market close.
I hope this is helpful.
Patrick Rodot, via E-mail
France
PREFERRED TRADE CLARIFICATION
Editor,
In response to a letter in the August 2000 S&C, "Money Market Fund," in which your reader states that an account with Preferred Capital Market doesn't pay interest, I would like to correct this information.
In fact, Preferred Capital Markets pays interest of 4-1/2% on cash in accounts waiting investment. In the reader's letter, he indicated that his money market account at his brokerage pays 5.1%, so that on $100,000 in cash, he would earn $5,100 in interest a year. Likewise, an account at Preferred Capital Markets with an average daily balance of $100,000 would earn $4,500 in annual interest. In addition, we offer a money market fund into which customers can switch their money upon request.
Tim Taylor
Senior Managing Director
Preferred Capital Markets, Inc.
Member NYSE, SIPCThank you for your clarification. We published a review of Preferred Trade from Preferred Capital Markets in the June 2000 issue of S&C. -- Editor
ERRATA
Editor,
Thank you for the excellent article "Optimal Detrending" in the July 2000 STOCKS & COMMODITIES.
After sight-checking the detrend formulas (on page 26), I found the following discrepancy:
From the EasyLanguage code:
Value1 = Price+ 0.088 * Value1 [6]
From the algebraic language:
Value1 = (High + Low)/2 + 0.88 * Value1(t-6)
Which is correct: 0.088 or 0.88?Thanks for the detrended price technique, which I hope to apply as well.
Cláudio Araujo Castelo Branco
BrazilYou are correct -- it should have read 0.088. Thanks for letting us know. By the way, that value is the same as that shown in Figure 4, page 27. -- Editor