CNADLESTICKS FORMULAS
Editor,I'm a long-time reader, first-time writer. The article "Candlestick Filtering" by Rudy Teseo in December 1999 inspired me to start writing my own formulas again. With the help of the MetaStock outline given in your Traders' Tips column also in the December 1999 issue, I have something to play with. Please keep up the good work.
Thanks for taking the time to write. Feedback from readers helps keep Technical Analysis of STOCKS & COMMODITIES, The Traders' Magazine, on track. Our goal is to publish articles each month that can help every reader trade profitably. Each person trades differently, so we try to publish a mix of material.Jack, via E-mailIf any reader develops something that could help other traders and would like to share the tip, please consider sending it to us for possible publication in Traders' Tips. -- Publisher
READERS' CHOICE AWARDS
Editor,Each year, your Readers' Choice Awards poll leaves me with the same puzzling reaction. CSI is without question the largest supplier of stock, fund, and index data through the Internet in the world. Millions of people access CSI data daily and we were not so much as mentioned in your Readers' Choice feature. Consider that we supply our timely and pristine data to the Websites of Yahoo, Microsoft Investor, Prodigy, Multex, AOL's Netscape, Go2Net, AltaVista, IMWorth, PeopleWeb, MoneyStar Communications, Stockpoint, Itochu Corp., and many more too numerous to mention, from whom we receive thousands of attributional referrals to our csidata.com Website daily, and you will get the proverbial picture. Why do you suppose we are the chosen source for nearly every major Internet financial site? The reason is obvious: Clean data is our hallmark.
There is no way that any data firm could have done better in spite of your ballot, your winner designation, and your method of information-gathering. I know it was not intentional, but just because the investor receives our financial information free is no reason to exclude us from recognition.
Bob Pelletier
CEO of CSI
Thank you for your letter. Actually, CSI won several awards. For those of you reading this who may not be familiar with our Readers' Choice Awards, each year in our annual Bonus Issue for subscribers, we present the results of a poll that asks subscribers by ballot to vote for their favorite products and services. The ballot included a list of products and services to select from, plus areas for write-in votes. We certainly apologize if we inadvertently left out CSI or any other product or service from our ballot in any category.CSI's products have won votes from readers for the past four years in our Readers' Choice feature. In the 1999 Readers' Choice Awards, which came out last month in our Bonus Issue, CSI data won Finalist in the futures data category; Honorable Mention in the options data category; and CSI's Quicktrieve and Unfair Advantage software each won Honorable Mentions in the category of data downloading software. -- Editor
ADX & CODING CANDELSTICKS
Editor,I have long enjoyed and profited from your magazine. The article "Adx Like The Pros" (STOCKS & COMMODITIES, October 1999) has really put my short-term trading into high gear. Thanks!
I would like to know if you have the candlestick code construction from page 44 of your November 1999 issue ("Coding Candlesticks" by Viktor Likhovidov) online so I can cut and paste it. I am a horrible typist and really appreciate the Traders' Tips area of your Website, from which I can cut and paste code.
David Jameson, via E-mail
We now offer long code and formulas from articles at our Website in a special subscriber area at https://technical.traders.com/sub/sublogin.asp. Login will require your subscriber number and last name. Your subscriber number can be found at the top of your magazine label before the dollar sign.In addition, the author of "Coding Candlesticks," Viktor Likhovidov, has provided equivalent code for use in TradeStation and SuperCharts. It is posted at https://www.forexclub.ru and at https://www.traders.com/Documentation/FEEDbk_docs/BackIssues.html under January 2000 Traders' Tips . -- Editor
HAMILTON LEWIS INTERVIEW
Editor,I'm not a subscriber to your magazine but I purchase it off the newsstand every month. I was very happy to see Hamilton Lewis interviewed in your November 1999 issue. You very rarely feature Africans or African-Americans, so this was a treat. It would be great to see more articles on people such as Joseph Jett and Jerry Favors. Keep up the good work.
We are always interested in interviewing persons with a demonstrated talent for analyzing or trading the markets and in publishing quality work from traders and analysts of all backgrounds. Color is not important; skill in making money in the markets is. By the way, we published five articles from Jerry Favors in 1992. -- EditorDoug Apperwhite, via E-mail
REVIEW OF OMEGA RESEARCH 2000i
Editor,Over the last year I have really enjoyed your magazine and the informative and educational articles and reviews. Based on the review your magazine did of ProSuite 2000i (October 1999), I decided to try the product/service for 30 days.
Well, I had terrible difficulties with it. Granted, I have only been trying to get this program up and running for the last three days, but I think you should have talked more about:
1. How difficult it is to load basic symbols -- it took three calls to tech support, four wasted hours of looking in the manuals, searching the Website, and so on, just to find out that sometimes, when you enter a stock's symbol, it is aborted for no apparent reason and you have to enter the company name.I only mention this because if Omega Research expects customers to pay $5,000 for what is supposed to be a top product, they should have a top-notch support system.
2. The inaccurate/incomplete instruction manuals.
3. Terrible customer support -- after different calls, the shortest being eight minutes' wait and the longest being 24 minutes, you get on the line with someone who makes you feel as though you are on the clock and don't want to spend too much time with you. (They should take a lesson from Dell Computer.)
4. The lack of a toll-free phone number or at least 24-hour/seven-day per week support.I have reread your product review and can only surmise one of three things:
1. I am the problem -- which wouldn't be the first time.
2. Your staff really didn't try this out as a first-time customer would -- you probably had someone from the company showing you how to use the program, or the writer was experienced with the previous versions of TradeStation, or didn't even try it out at all and instead took the press package along with the ad contract.
3. The ad revenue is more important then your reputation as an objective magazine. Sorry, but I will be very reluctant to ever believe one of your product reviews again.Sounds like you enjoyed the same kind of frustrating experience we recently had upon installing Microsoft Office 2000 on two of the magazine's computers. While the installation went great on most machines, these two machines both required an hour or two on the telephone with Microsoft technical support. Lucky for us that calling Microsoft is a local call. One of the machines still starts with an odd error message that we have learned to just ignore.K.M. JunkinsWe try to be evenhanded about reviewing products in Technical Analysis of STOCKS & COMMODITIES, The Traders' Magazine. When reviewing products, we require that the software developer send us exactly what he would a customer, including manuals and packet material. We speak to product users for their input, and we do not generally review demo or beta products.
A 30-day free trial of the $4,800 ProSuite 2000i from Omega Research is still a good deal. Twenty years ago, we spent $7,500 for software (CompuTrac) and an Apple ][+ computer with no data that did a lot less than the ProSuite 2000i TradeStation module does for $2,400 with data, less the computer. -- Publisher
52-WEEK HIGHS/LOWS
Editor,Do you know of any Websites that list stocks at their 52-week highs and stocks at their 52-weeks lows?
Frank Brucato
Bronx, NY
Try investertech.com and stockcharts.com. Investertech.com lets you screen stocks based on yearly highs and lows. -- Editor
BACKTEST RESULTS
Editor,I backtested the system Jay Kaeppel describes in his July 1999 article "A System For Trading Fidelity Select Funds," and I received different results, I believe because I started the backtest on a different date. I started my backtest on March 5, 1993, because my database on Fidelity funds did not extend beyond this. Using Kaeppel's system for buying and selling funds, I received 11 buy signals on approximately the following year of data. When these were closed out, the result was four winners and seven losers, with an average gain per trade of 1.1%. This was during a period when the Dow Jones Industrial Average went up about 7%.
Kaeppel's results for this period were much better because he started in 1988. If I were to use the simple crossover of 50-day and 200-day moving averages as a system for buying and selling these same Fidelity funds, I would get much better results. For example, for the same one-year period, you would have eight trades, six winners, and two losers, with an average gain per trade of 50.3%, and over the last six years, 63% winners and average gain per trade of 26%.
There are no easy answers on determining how many years of data is appropriate for a backtest. Naturally, using different periods for a backtest will yield different results, so how does a trader determine the best parameter values to use for future trades based on backtesting?Dwight Cook, via E-mailKaeppel states in his article that he uses 10 years of data on which to backtest because of that market's particular trading history, "although many system developers believe that 10 years is not really all that much history from which to draw definite conclusions."
Developing a system requires careful backtesting on out-of-sample data to determine reasonable parameters that aren't simply curvefitted to the data. Further, selecting the length of time on which to backtest a system may depend on the market, the holding period for trades, and some statistical principles having to do with the number of trades needed for good statistical results. For more on system testing, see some of Dennis Meyers's past articles in STOCKS & COMMODITIES.
While it is generally agreed that too-short periods of historical data don't allow for the most useful backtests, and longer periods present more variety in the data, even so, there is never any guarantee of future results. There is only probability. -- Editor
HISTORICAL INTRADAY S&P 500 CHARTS ON WEB
Editor,Do you know of any Website that would let me pull up historical intraday charts for the S&P 500 (or DJIA) going back a couple of years? I'm interested in doing some research into the patterns the indices make at the open and close, and correlating this to how the market did that day and is expected to do the next day. Have you ever published any articles on this?
Try QCharts (https://qcharts.quote.com/qcharts). As for articles, you may enjoy some of the discussion in the following S&C articles. Abstracts for these and other S&C articles are available at our Website, www.traders.com.Michael Barrow, via E-mailChande, Tushar S., Ph.D. [1992]. "Forecasting Tomorrow's Trading Day," Technical Analysis of STOCKS & COMMODITIES, Volume 10: May.
Hartle, Thom [1996]. "Living The Dream: Trader Gary Smith," interview, Technical Analysis of STOCKS & COMMODITIES, Volume 14: May.
Merrill, Arthur A. [1992]. " Closing Tick," Technical Analysis of STOCKS & COMMODITIES, Volume 10: February.
_____ [1991]. "Merrill Directional Patterns," Technical Analysis of STOCKS & COMMODITIES, Volume 9: April.Widner, Mel, Ph.D. [1998]. "Automated Support And Resistance," Technical Analysis of STOCKS & COMMODITIES, Volume 16: May.
--Editor
MASS INDEX
Editor,The formula and explanation for determining the mass index in the sidebar was so poor, can anyone provide a better definition of this index? It does approach timing from a different viewpoint and I would like to program it into my computer, but as it stands now, I would be creating a new formula, not entering the mass index.
Vernon B. Brunelle, via E-mail
The sidebar on the mass index provided with the article "Detecting New Trends Early" was not clear that the two moving averages of the close are to be used in conjunction with the mass index. They determine trend. In Donald Dorsey's original formulation (which can be found in Dorsey's article "The Mass Index," Technical Analysis of STOCKS & COMMODITIES, Volume 10: June 1992), once the mass index indicated a possible trend reversal, you would trade against the direction of the two moving averages. That is, you would buy the market if the first moving average were below the second, and you'd sell if the first moving average were above the second. We regret the lack of clarity. -- Editor
VOLATILITY STOP
Editor,I read Mark Vakkur's article "Volatility Stop System" in the October 1999 STOCKS & COMMODITIES with interest. I believe it is quite an effective system, especially in a trending and volatile market. I knew about this system after I learned about it from Chuck LeBeau's Website more than a year ago. Vakkur's system is no different from that of LeBeau's. As Vakkur's article did not give any credits to LeBeau, it seems that he must have discovered it independently, by coincidence. You might pass this along to Vakkur.
Tean Tan, via E-mail
Mark Vakkur replies:I can't comment on the specifics of Lebeau's ideas or his Website, but I doubt he could have invented the volatility stop, since that claim most probably goes to J. Welles Wilder. The volatility stop system is not presented as an original system in my article per se, but as a variation on a theme. I have (as I'm sure you have) encountered many variants of this system; the most recent inspiration was Van K. Tharp's book, Trade Your Way To Financial Freedom. (In it, he reviews the idea of volatility stop on pages 256-258.)
I believe the idea of a trailing-volatility stop was first used by J. Welles Wilder. He also found a 2.7 to 3.4 range to be most profitable, although he used daily charts. The reader is under the impression that I am claiming to have "invented" the volatility stop. I hope nothing I wrote implied that claim, which would be as ludicrous as saying I invented the bar chart or the exponentially weighted moving average, both of which were also used in the article. The point of my articles is not to present entirely new systems per se, but to build on classic, simple building blocks and illustrate how to pull it all together, test across different assets and iterations of the inputs, and so on. The only unique aspects of the system I presented was that I used the last bar's high as an entry (rather than a trailing volatility buy stop) and used weekly rather than daily charts.
As far as giving credit is concerned in articles, I generally leave that to the editors, recognizing that if I were to cite every contributor who helped give us the tools we use, the citations would be longer than the article! We stand on the shoulders of giants. ... I will give citations in the future if you like, but as you know it's often hard to find out who first developed a particular indicator, anyway. I most certainly do not want to give the impression that I am reinventing sliced bread (to mix metaphors). Thanks for the feedback.
ERRATA: CANDLESTICKS
Editor,I just received my November 1999 STOCKS & COMMODITIES and scanning through noticed Figure 5 on page 77, which may be confusing for some readers, as while it shows a hammer bar, it also shows dark cloud cover patterns that are marked but not labeled, nor mentioned in the text, I don't think. Perhaps the other marked patterns could be commented on briefly for readers in a future issue. Thanks for a great magazine!
Nick Lindridge, via E-mail
Good point. The author provided these but we missed labeling them in the editing job. We regret the error. -- Editor