September 2005 Letters To The Editor

or return to September 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


OFFLINE CHARTING SOFTWARE

Editor,

Is there any stock market graphing software I can use without going on the 'Net and working with just the closing prices? I track only around 10 stocks, and I use simple indicators with simple moving averages:

1MA, 2MA, 3MA, 10MA
Price oscillator (12MA-26MA)/26MA
Stochastic 19-1-1
Antoine Hebert
Borrington, NH

In these days of computers and Internet connectivity, most technical analysis software makes use of the Internet and most traders download their data via the Internet. Perhaps someone reading this can suggest an alternative program. However, you can always simply use a spreadsheet application such as Excel or a graphing application and input closing prices manually from the newspaper, then plot your charts, as was done in pre-Internet days.-Editor


PLOTTING THE TRUE STRENGTH INDEX

Editor,

I just finished reading Mark Phillips' article on the true strength index in the June 2005 S&C ("Less Is More"). I have been searching for a charting service that offers this indicator and have not yet found one. Could you please send me a list of charting services that offer this indicator?
Dave Traines
via email

Mark Phillips responds:
I have not encountered any charting service that provides the TSI (or any variant) as a "canned" indicator. However, there are some open-ended applications, such as TradeStation, for which members of the user community have programmed and then made available the functions necessary to have the TSI oscillator plotted on any chart, just as though it were one of the canned indicators. If you use TradeStation, you should be able to go to the TradeStationWorld website, enter the EasyLanguage discussion forum, and perform a search for TSI. One of the items that comes up is labeled "True Strength Index - TSI." That link will lead you to a downloadable .Eld file that will reproduce the TSI indicator on any of your charts in TradeStation.

Technical Writer David Penn adds:
One technical analysis program I have noticed that includes an implementation of the TSI is Prophet.net. See also the following letter and response.


CODE FOR THE TSI

Editor,

In his interesting article "Less is More," Mark Phillips analyzes the use of William Blau's true strength index. However, no computer code is referred to or indicated. Will MetaStock code for this be released or published? Where can I obtain this?
J.A. Basson
via email

Readers interested in learning more about the formulas for creating the true strength index (TSI) can visit the Online Store at our website, www.Traders.com, and download William Blau's original 1991 and 1992 articles published in S&C on that topic. Those articles included "The True Strength Index" from November 1991 and "Trading With the True Strength Index," May 1992. Both articles provided equations within the article and also a sidebar on calculating the TSI using Excel spreadsheet formulas.

These formulas could probably be recoded into MetaStock formula language. We do not offer the service of coding indicators for specific software, but some companies do, and code is also publicly available on the Internet. You might also try checking directly with Equis International, developer of MetaStock, or the MetaStock online community, for code.-Editor


TSI CHART CLARIFICATION

Editor,

I enjoyed Mark Phillips' article ("Less Is More," S&C, June 2005) and dug out past issues of S&C to get more details about the TSI formula.

In Figure 3 in the June article, the bottom chart shows two lines: the TSI (close,25,13) and another line. But which line is which, and how is the other line calculated? My guess is that the red line (fast line) is the true strength index (TSI) (close,25,13) and the slow line (blue line) is the exponential moving average (seven-day) of the TSI line. Can you clarify this for me?
Mike Caron
via email

Mark Phillips responds:
You got it absolutely correct. The red line is the raw TSI line, and the blue line represents the moving average of the TSI.


TSI AND ADX

Editor,

In his article "Less Is More" (S&C June 2005), Mark Phillips correctly points out that the stochastic is an oscillating indicator, useful during range-bound price action. But he then applies the indicator to a variety of stocks during time periods when they are trending, not range-trading. It's no surprise that the stochastic fired off erroneous signals during these time periods - it's the wrong indicator to use during trending action, when indicators like moving averages are more useful.

For example, Phillips applied the stochastic and TSI to RYLand Homes (RYL) during the July-December 2004 period. RYL was clearly trending higher from July through October, a period during which the stochastic issued four sell signals, the first three being early (and wrong) and only the last one being timely. During the same time period, the TSI issued its one timely, profitable sell signal. However, since RYL was trending, the stochastic was the wrong indicator to use under those conditions, so the incorrect sell signals were not unexpected.

So how does the trader know when to use oscillating indicators and when to use trending indicators? The answer to the question is the ADX.

ADX measures the strength, but not the direction of a trend. As a rule of thumb, when the ADX rises above 30 and continues rising, the stock is entering a trending period. When the ADX falls below 25 and continues falling, the trend is ending and the stock is entering a range-bound period. In practice, ignore stochastic signals when the ADX is above 30 and rising; act upon stochastic signals when the ADX is below 25 and falling.

The buy signal in late July was valid because the ADX was very low - below 15 - and the stochastic was oversold. In late August when Mr. Phillips shows the first stochastic sell signal on RYL, the ADX was around 40 and rising. Clearly, RYL was trending so this sell signal should have been ignored. The subsequent stochastic sell signals in September should also have been ignored because the ADX was as high as 51 that month before falling.

With the proper use of stochastics, RYL would have been purchased around $37 (split-adjusted) in late July 2004 and held until $65 in March 2005, when it slipped below its 50-day moving average. This is a much better return than using TSI, which also signaled a purchase around $37 in July, but sold it in late September around $47.
David R. Steckler
Member, American Association of
Professional Technical Analysts
via email

Mark Phillips responds:
While it is true that the ADX indicator can be useful for telling the difference between trending and range-bound price action, I think you may have missed the main point of my article. That point is that stochastics cannot be used alone, due to the wide variance in many different market conditions' behavior. If you're using stochastics as a primary trade indicator, you're absolutely correct in noting that other indicators are necessary to clarify the signals given. As noted in the opening paragraph of the article, I was looking for technical indicators that can be interpreted in a black and white manner, not shades of grey. I think we can all agree that we cannot make such a claim about stochastics - other tools are required to qualify the signals given by that oscillator.

ADX is a great tool to provide that qualification, but it requires interpretation as well. What do we do if ADX is 28 and trending up or 27 and trending down? Those situations require some interpretation as they are close to the trend/no-trend line. Do we use the standard 14-period setting or is there a better one? If we use a different setting, is the threshold of 25/30 still correct, or should it be higher or lower? By introducing the ADX in addition to stochastics, we now have two indicators to interpret, both of which are sensitive to the parameters chosen. I won't debate that using the two together can be a viable approach, but not for someone who prefers a more black and white approach. Recall that the TSI is relatively impervious to the settings chosen - we can't just tweak the settings to make the indicator look the way we want. Also there are no overbought/oversold levels to deal with (or change), meaning that TSI does not lose its effectiveness at the far end of the range as stochastics does. We only need deal with the crossover buy and sell signals, producing clean signals without the need for interpretation.

Certainly, even an indicator as simple as the TSI is not the holy grail of technical indicators, but it is a good starting point. As I noted near the end of the article, "...the best indicators exhibit a high degree of robustness, performing well in most market conditions and minimizing false signals the rest of the time." Stochastics does not pass muster on that criterion, but TSI certainly does.

Of course, no indicator should be used in a vacuum and that also holds true for TSI. It should be used in conjunction with other tools to make up an overall trading methodology. The primary point is that TSI provides a better starting point due to its higher degree of robustness and clarity of interpretation.

When we try to make real-time trade decisions, starting with an easily understood indicator makes it simpler to incorporate other tools, without getting mired in too many shades of gray. The net result is that less really is more.


COLOR-CODING NEW BIDS

Editor,

I enjoyed Thomas Carr's article, "The Tape Reading Edge" (S&C, July 2005). But the author didn't sufficiently explain or define what "new bids" are, or explain which piece of information on my Time & Sales screen is a "new bid," which is something I need to know to use his color-coding system. How can I find out? Thank you very much!
Denton Anderson
via email

Thomas Carr responds:
A "new bid" is a new order initiated by someone who wants to buy the stock in question. If the new bid raises the current "inside bid" (the price at which the stock is currently bid), then the new bid becomes the new "best bid." In most charting programs, you can color-code new bids, and I prefer to code them yellow. That way, whenever a yellow bar flashes in my Time & Sales window, I am alerted to a new bid on the stock. As I stated in the article, when you see the number of new bids increasing, and the size of those bids increasing, then you have a bullish condition and the price is likely to rise over the next few minutes. Hope that helps!


AVERAGING PORTFOLIO RATES WITH HARMONIC MEANS

Editor,

Since I began investing in 1957, I have been using geometric means to average portfolio growth rates, as it was understood that that was the correct average, confirmed by Technical Market Indicators Analysis & Performance by husband and wife team, Drs. Bauer and Dahlquist (1999). However, in rereading Facts From Figures by J.M. Moroney (1960), I find the correct average for speed - and presumably growth - rates is the harmonic mean.

While the arithmetic average is the sum of items divided by the number of items, the geometric mean is the more complicated nth root of the product of n items. The harmonic mean is the even more complicated n divided by the sum of 1, divided by each item.

To prove this, I took data for the period of the 1987 crash, which included two years' gains, as well as two years' losses. Thus, starting 1986 with £100,000, a profit of 39.18% was made; starting 1987 with £139,180, a profit of 19.15% was made; starting 1988 with £165,832.97, a loss if 15.47% was made; starting 1989 with £140,178.61, a loss of 38.08% was made. The year ended with £86,798.60.

Thus, the total loss was £13,201.40, or £3,300.35, or 3.30% annually.

The arithmetic average is a gain of 1.2%, the geometric average a loss of 3.48%, and the harmonic average a loss of 3.33% (coded by adding 200 and dividing by 100).

I hope this enlightens other readers misled by Drs. Bauer and Dahlquist.
Dieter Babel-Thomson
Fellow of the Royal Statistical Society
Dublin, Ireland


EXCEL CODE FOR UNIVERSAL CYCLE INDEX

Editor,

In the May 2005 S&C, I was very interested in the universal cycle index (UCI) that was presented in "Cycles In Time And Money" by Stuart Belknap.

Unfortunately, the formulas provided with the article are for MetaStock users, and your Traders' Tips section that issue offered little help to Excel users, and I am not familiar with the other code languages presented. In addition, some developers appeared to have used some modifications to Belknap's code (forward/backward-looking references).

Can you provide the rationale or construction steps for the UCI indicator, making it more accessible?

Thank you for a great magazine!
Marco Alves
via email

Unfortunately, we do not develop code in-house for any of the articles. You might try visiting our message boards area at our website, www.Traders.com, and ask other readers if they have coded the UCI in Excel.-Editor


TECHNICAL ANALYSIS SOFTWARE

Editor,

I used to use TradeStation 2000i for technical analysis, but the company practically no longer supports this product. Do you have any magazine articles I could purchase that offers comparisons of different technical analysis software?
Peter
via email

We generally don't compare products side-by-side in our publications, but we do provide a list of software products and their features in the Traders' Resource area of our website, www.Traders.com. We also publish reviews of technical analysis software in every issue. Also helpful is the Readers' Choice Awards feature in our annual Bonus issue (available only to subscribers). Here, you will find a list of award-winning products and you'll be able to see what other products belonging in the same category as TradeStation are rated highly by our readers.

You could also seek out other traders' opinions of different software via the message board at our website, or through the many other trading forums that are widely available on the Internet. Hope this helps.-Editor


CODE FOR FRACTAL DIMENSION INDEX

Editor,

In the May 2003 issue of Stocks & Commodities, you said the MetaStock code for the fractal dimension index and the Hurst exponent were available in the article ("Making Sense Of Fractals"), but I couldn't find them anywhere. I would like to have the MetaStock code for this stuff.
Don Mateja
via email

The basic formulas for the fractal dimension index were provided in a sidebar to the "Making Sense Of Fractals" article on page 52 of that issue. The author, Erik Long, did not work in MetaStock. These basic formulas could probably be coded into MetaStock formula language.

If you are referring to the line in our Traders' Tips column in that issue under the MetaStock heading that said code for "Time And Money Charts" was provided in the article, this was a different article. Unfortunately, MetaStock did not submit code for the fractal dimension index in that issue.-Editor


ONLINE STOCK-SCREENING SERVICES

Editor,

I wonder if you could help me out with online stock screeners. In the past, I have used marketscreen.com, which your magazine reviewed back in 2003. Do you happen to have any more recent suggestions for online stock-screening services?
Greg
via email

There are several online stock screeners available. A simple Google search will bring up many results. Of course, you still have to determine which ones suit your specific needs.-Editor


ERRATA: TRACK 'N TRADE REVIEW

At the end of our Track 'n Trade review in the July 2005 issue of S&C, the author refers to the Worden Brothers company and TC2000 as "a team of intrepid designers ... out of Florida." That should have read "out of North Carolina." S&C regrets the error.


Back to September 2005 Contents