April 2001 Letters To The Editor

or return to April 2001 Contents

PRAISE FOR EHLERS

Editor,

I have subscribed to STOCKS & COMMODITIES for several years. It is no exaggeration to say I learn new and useful things from every issue. Many of the things I've learned are included in the consistently profitable trading method I have coalesced.

I first learned about the work of John Ehlers from his articles in S&C. Over the past year or so, I have studied and experimented with the various modifications and applications of the Hilbert transform/dominant cycle period as put forth by Ehlers. I have become as big a fan of him as I am of S&C.

I'm enjoying the Hilbert studies I purchased last week. I've selected a few favorites and converted two of them into show-me studies. They are a real-time marvel to behold. As an experiment after observing Ehlers's Hilbert stochastic/RSI/CCI, I built a Hilbert DMI, including the ADX. The power of accurately measuring the dominant cycle period is very apparent. They all generate signals within a heartbeat of each other and outperform the textbook period settings. That includes the market crossing the Hilbert trendline. The power of measuring the signal-to-noise ratio was the biggest surprise. The SNR outperformed even the Hilbert ADX and greatly reduced the number of whipsaws.

My thanks to S&C and Ehlers for the terrific work you do.

Lee Goldberg, via e-mail

We're a fan of John's, too. See his lead article in this issue for another innovative look at assessing prices. -- Editor
 

O'HIGGINS TRADING METHOD

Editor,

I have subscribed to STOCKS & COMMODITIES for many, many years but I cannot recall any space having been given to the O'Higgins trading method. If you have indeed published it, could you please advise where I should look in my back issues? If you have not published it, could you please give it some consideration?

Rudi Genis
Bryanston, South Africa

If any reader has an interest in this topic and would like to share his knowledge in the way of an article, please contact us for a copy of our writing guidelines. -- Editor

STATISTICAL SCRUTINY ON LUNAR SEQUENCE

Editor,

In the February 2001 Letters to S&C, Richard McKee noted an interesting four-day sequence in the moon data. The numbers, including those before and after the sequence, were: 49.2 42.9 52.3 56.3 59.3 54.7. He suggested that more study is needed.

I agree that the sequence is interesting, and that more data might yield more significant results. Possible, but I have my doubts about changing my conclusions [see Arthur Merrill's letter to S&C titled "Evidence Needed" back in the December 2000 S&C--Ed.]. The New York Public Library helped me obtain data on 60 moon cycles, a hefty sample!

My conclusion was that the data didn't give significant evidence of any influence on the Dow Jones Industrial Average near the full moon. I certainly stand by that conclusion.

I used the chi-squared test. To qualify for significance, statisticians like to have at least 80% confidence that a result isn't just pure chance. They prefer 90%, 95%, 99%, or 99.9%. The largest deviation in the figures above is 59.3% - 55.0% = 4.3%. Chi-square tells us that you can have only 50% confidence that this deviation isn't pure chance! The data may be accurate, but the deviations from expectation of any of the four figures in the sequence certainly wouldn't satisfy a statistician.

Arthur Merrill
Haverford, PA

INDEPENDENT ADVISOR

Editor,

I am interested in becoming an independent advisor to people trading stocks and options on stocks. I would like to charge a percentage of the profits as my fee. I am wondering if I could do this without a broker's license and what kind of regulations would apply if any. Could you advise?

Kyle Evans, via e-mail

You need to contact the Securities and Exchange Commission about becoming a registered investment advisor. Also, see "Retail Trading Myths" by Don Bright on page 64 of this issue. -- Editor

ENDPOINT MOVING AVERAGE

Editor,

Here is another word on the subject of moving averages and how to speed up their response with the endpoint moving average (EPMA). I doubt that it will be the last word.

As S&C readers know, and as Frank Suler pointed out in the October 2000 issue of S&C in his letter, the EPMA is an excellent least-squares curve-fit technique. Joe Sharp's article in the January 2000 issue, "More Responsive Moving Averages," gave a modification to the standard moving average formula that removes much of the inherent lag. It was based on an inventory model by Robert Brown in the 1950s. Tim O'Sullivan pointed out in the June 2000 issue that Brown's formula was identical to the EPMA described by Pat Lafferty in the October 1995 and June 1999 issues. The EPMA is the endpoint of the least-square straight line fit to the data. Don Kraska noted in the August 2000 issue that it is available in an Excel worksheet via the functions Trend and Linest.

What has not been said is that the slope and endpoint of the regression line fit to a set of security prices are each just the weighted sum of those prices. If, say, the security prices are numbered from p(0) for today's closing price, back to p(n) for the closing price n days ago, the end point of the regression line, representing the straight line estimate of today's closing price, is e:

e = w(0) p(0) + w(1) p(1) + . . . + w(n) p(n),

where the weights w(i) are (4n - 6i + 2)/((n+1)(n+2)), for i = 0, 1, 2, É , n.

The slope of the line is s:
s = v(0) p(0) + v(1) p(1) + . . . + v(n) p(n),

where the weights v(i) are (6n - 12i)/(n(n+1)(n+2)), for i = 0, 1, 2, É , n.

For example, for n = 5, we get:

i        w(i)      v(i)
0      0.5238    0.1429
1      0.3810    0.0857
2      0.2381    0.0286
3      0.0952   -0.0286
4     -0.0476   -0.0857
5     -0.1905   -0.1429

This is easily put into a spreadsheet. Let's say that 30 days of price data are in the first column, labeled A1 through A30. We want to fit a running six-point line (n=5) to these data, starting, say, on day 9. The last six prices are A9, A8, A7, A6, A5, A4, which become p(0), p(1), p(2), p(3), p(4), p(5), in that order. They will be multiplied by w(0), w(1), w(2), w(3), w(4), w(5) to get the endpoint, and by v(0), v(1), v(2), v(3), v(4), v(5) to get the slope. Let's put the endpoints in column B, right next to the corresponding datapoints in column A. In cell B9, type the expression

= 0.5238*A9 + 0.3810*A8 + 0.2381*A7 + 0.0952*A6 - 0.0476*A5 - 0.1905*A4

and hit Enter. Put the slopes in column C, right next to the corresponding end points. In cell C9, type the expression

= 0.1429*A9 + 0.0857*A8 + 0.0286*A7 - 0.0286*A6 - 0.0857*A5 - 0.1429*A4

and hit Enter. These formulas say, in effect, that the end point and slope on day 9 are each the weighted sum of the prices from day 9 back to day 4. Now highlight cells B9 through C30 and click Edit|Fill|Down. The columns B and C are filled with the end points and slopes of the data in column A from day 9 to day 30.

Finally, I would note that the formulas for w(i) and v(i) may be made recursive, as Joe Sharp has shown. But considering the speed of current machines, I wonder if the reduced computation time justifies the added complexity in programming. In fact, if you are working in a spreadsheet, the TREND and LINEST functions are the simplest way to get the endpoint and slope. The formulas above are most valuable in custom-designed programs.

John F. Bellantoni, via e-mail
Washington, DC

TC 2000 SCAN

Editor,

I enjoyed Gary B. Smith's interview in the October 2000 issue. Inspired by it, I bought TC2000 and had the same problem that John Wagner wrote about in the January 2001 Letters to S&C. I could not get it to work. I e-mailed the Personal Criteria Formula to TC2000 and they said it looked okay to them; they could not identify my problem. Then I found the answer. Push the Update All Criteria button after every download. Now excuse me while I try to find a Gartley setup (very interesting)!

Mac McCann, via e-mail


Sounds like a good follow-up article. See also the response to the next letter. -- Editor

STOCK SCREENING

Editor,

With all the hoopla about new real-time programs, does any one of these programs offer an alert system? This question remains untouched so far or is even avoided. By alert system I mean a system in real-time programs by which one can find a stock turning up or down in regard to an indicator by a graphical signal in a list of stocks, not by furiously clicking chart after chart in a 12,000-plus stock universe.

Chris Pratsch
Houston, TX

There are several real-time programs that offer these features -- TradeStation 2000, MetaStock Professional, and CyBerCorp, for example. -- Editor

MONTE CARLO SIMULATION

Editor,

I read your review of @Risk in the February issue with interest. @Risk is a versatile risk analysis program. However, your comparison of @Risk to my Monte Carlo-based EasyLanguage function for fixed fractional position sizing is a bit of an apples-to-oranges comparison. I wrote Monte Carlo for the very specific problem of fixed fractional trading, whereas @Risk is general-purpose.

It is certainly true that @Risk could be used for fixed fractional risk analysis, but to do so would require setting up an Excel spreadsheet to simulate fixed fractional trading, including the calculations for rate of return and peak-to-valley drawdown. It would also be necessary to model the distribution of trades, as you demonstrated in your review. My experience with @Risk suggests that many traders would experience a fairly steep learning curve with this software. Nonetheless, I believe that given the importance of risk analysis to trading, @Risk is worth a look.

For traders who want a simpler approach to the problem of quantifying the risk of a fixed fractional trading approach, I invite them to consider the enhanced version of the Monte Carlo user function offered for purchase on my website, www.BreakoutFutures.com.

Michael R. Bryant, via e-mail

MONITORING STOCK BASKETS

Editor,

Do you know of any Internet-based software program or a combination of stand-alone charting/Internet-based real-time datafeed that would allow me to create and monitor my own stock baskets, updated and charted in real time? While many programs will monitor the individual stocks within the basket, I would like to monitor the basket as a whole (as if it were a mutual fund). Thank you.

Emanuel Schroeter, via e-mail

Try MetaStock Pro's Composites feature. -- Editor

TRADING SYSTEMS

Editor,

I have subscribed to S&C for 10 years. I love this magazine with all of my heart! In response to the February 2001 Opening Position, I have also noticed the extinction of trading systems articles month by month. I think it is related to the fact that buy on dips so far was better than anything in the brains of the majority of investors. Developing a trading system requires great effort. After all, why bother buying databases, software, books, and a PC when you can just get the money in the Nasdaq? Well, not anymore -- or at least not so easily now.

I believe technicians always shine when there is a bear market because people want to know where the bottom is. The same thing is true for systems. Then there's the Internet, not only to sell as you correctly pointed out, but also to publish. So there is less and less material waiting to be published in S&C.

Bull markets make investors lazy, and so does technology. Now there is so much computer power, software, and data with which to analyze, compared to my Apple II in the late 1980s. I also believe people with ideas -- good ones -- prefer to trade rather than to make public their approach, so we see still turtles and volatility breakout systems after all this time.

Check out the book market for investors. After the crash in tech stocks, there were no more IPOs and no more tech books, but many educational, discipline, and psychology books -- a market for all seasons. I think there is a cycle in this phenomenon. Maybe chaos will be back as neural networks. Maybe there is no more to say on trading systems, but I don't think so. Maybe this is simply the calm before the storm, where the storm is a major breakthrough in trading systems development or in time series analysis. Time will tell. In the meantime, I am thinking about submitting an article to repay in a modest way all the good things I received over the years reading S&C.

Riccardo Ronco
London, UK

Still no offerings! -- Editor

TRADING ROOMS IN LAS VEGAS

Editor,

I am thinking of moving to Las Vegas and am tired of the isolation of my home office. I am looking for a trading room there that I might check out on my next trip. I would like a place with a bit of a social atmosphere, where I could lease equipment or set up my own. Could you suggest such a facility? Perhaps someone on your staff could refer me to a website that lists these for the whole country? I like to trade both futures and equities. Any information you can provide would be most appreciated.

I have been subscribing to S&C for about eight months now and it is excellent! Keep up the good work!

Henry Fiddler, via e-mail

One of the more popular trading rooms in Las Vegas is Bright Trading, LLC. You can find a list of daytrading outfits at https://www.daytradingstocks.com/training.html. -- Editor

OPTIONS EXPIRING WORTHLESS

Editor,

I enjoyed Larry McMillan's article in the January issue regarding the "truism" of 90% of options expiring worthless. As to why this often-quoted "fact" stays alive, McMillan may have overlooked one possibility, a typographical error in the original study: 90% of options don't expire worthless, 90% of options expire worth less.

Keep up the interesting work.

Viktor Brandtneris, via e-mail

GETTING STARTED

Editor,

I just received a complimentary issue of STOCKS & COMMODITIES. Most of it was way over my head. Please keep this magazine simpler. For technical articles, maybe you could tell readers what they should know before they get into the article, or give more articles, books, and other references. I have several questions.

Do any of your advertisers offer a limited version of their product on a CD?

Thank you for considering my letter.

Ronald Brongiel, via e-mail
Shawnee, KS

You might try our companion publication, Working Money, which offers educational content geared more for the novice investor. Meanwhile, a list of index stocks can be found at https://www.amex.com under Index Shares. Most quote vendors, such as TC2000, PCQuote, and eSignal, provide the service of monitoring an unlimited number of stocks, as well as being able to sort them. Vendors such as TC2000 and PCQuote also offer their products on CD.

ERRATA: READERS' CHOICE AWARDS

Editor,

In the 2001 Bonus Issue of STOCKS & COMMODITIES, we incorrectly listed the recipient of a Readers' Choice Award in the category of Technical Analysis Websites. The Honorable Mention recipient, as determined by reader ballot, should have read DynamicTraders.com from Dynamic Traders Group, not DynamicTrader.com from DynamicTrader Online. We sincerely regret the error. Dynamic Traders Group can be reached at 520 797-3668, fax 520 797-2045, dt@dynamictraders.com, or www.dynamictraders.com.

ERRATA: BUFF AVERAGES

Editor,

First, I want to say that your magazine has been instrumental in teaching technical analysis. I look forward to reading each and every issue. Keep up the great work.

In the February 2001 Traders' Tips, the MetaStock formula for Buff averages was incorrect. Thank you for including the algebra in your articles; it makes it easier to fix mistakes.

Dan Graham
Cobourg, Ontario

Cheryl Abram of Equis International replies:

Here is the corrected formula. It can be written one of two ways:

1
X := Input( "Time Periods",1, 500, 25 );
Sum( V * C, X ) / Sum( V, X )

2
X := Input( "Time Periods",1, 500, 25 );
Sum( V * C, X ) / (Cum( V ) - Ref( Cum(V), -X ))
 

Back to April 2001 Contents