Letters - February 2004

February 2004 Letters To The Editor

or return to February 2004 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

FOURIER ANALYSIS CODE

Editor,

I have been searching through your archives for articles on Fourier analysis and am looking for one that presents code or an algorithm that describes how to program the computer to calculate periods in price series. I am seeing several article titles from Volume 17 (April 1999), Volume 12 (September 1994), Volume 9 (March 1991), Volume 2 (April 1984), and Volume 1 (January/February 1983). I was hoping you could point me to one that shows some code on how to implement Fourier analysis on price series. I'd like to be able to extract periods within the price data from which to program other indicators.
--Bill Cooke, via email

Technical Analysis of STOCKS & COMMODITIES, The Traders' Magazine, has published more than a hundred articles that refer to Fourier. A few of those include computerized implementation as well as examples. Fourier is often used in engineering to design or tune data filters, like those that traders use as indicators.

Fourier analysis can be a great learning and filter design tool. Plot the output to visualize any underlying cycles in your input data. Cycle period length, at a cycle frequency, is equal to the number of input periods n divided by the cycle frequency (1 through n). Most technical analysis indicators use a period length, such as the number of days, to filter or smooth out insignificant, untradable market noise. In engineering, this type of filter is referred to as a finite impulse response (FIR) filter. Use the FFT on filtered data to see what cycles a filter is preserving or passing through.

The first Fourier article we published that included code was "Fast Fourier Transform," STOCKS & COMMODITIES, Vol. 1, No. 2 (pages 29?31). The following articles explain how to transform time domain series data into the frequency domain to produce cycle strength versus cycle frequency. Using this code can help visualize and optimize data filters. -Jack K. Hutson, Publisher

Warren, Anthony [1983]. "A MiniGuide To Fourier Spectrum Analysis," Technical Analysis of STOCKS & COMMODITIES, Volume 1: January/February.
Available from our Online Store at <https://store.traders.com/-v01-c02-minig-pdf.html>
Abstract: The basic idea behind Fourier analysis of time series is to decompose the data...

Warren, Anthony, and Jack K. Hutson [1983]. "Fast Fourier Transform (Fft) Basic language code," Technical Analysis of STOCKS & COMMODITIES, Volume 1: January/February.
Available from our Online Store at <https://store.traders.com/-v01-c02-fastf-pdf.html>

Hutson, Jack K. [1983]. "Using Fourier," Technical Analysis of STOCKS & COMMODITIES, Volume 1: January/February.
Available from our Online Store at <https://store.traders.com/-v01-c02-usfou-pdf.html>
Abstract: Joseph Fourier (1768-1830) was a French mathematician and inventor of Fourier analysis. His work helped show the way to expressing periodic curves...

DETECTING BREAKOUTS

Editor,

With regard to Markos Katsanos' articles on detecting breakouts in the April 2003 and September 2003 issues of S&C, as well as his follow-up letter that appeared in the November 2003 Letters to S&C column, I enjoyed reading the articles and would like to apply this technique. Please ask Mr. Katsanos if he would provide the two indicators (shown below) as ordinary line chart indicators rather than as a true/false indicator.
==================== period:=22;
pers:=40;
coef:=.1;
intra:=Log(H)-Log(L);
vintra:=Stdev(intra,period);
inter:=Log(Typical())-Log(Ref(Typical(),-1));
vinter:=Stdev(inter,period);
cutoff:=coef*(vinter+vintra)*C;
mf:=C-(H+L)/2+Typical()-Ref(Typical(),-1);
fve:=Sum(If(mf>cutoff,+V,If(mf<-cutoff,-V,0)),
period)/Mov(V,period,S)/period*100;
fve12:=Sum(If(mf>cutoff,+V,If(mf<-cutoff/100,-V,0)),12)/Mov(V,12,S)/12*100;
cbol:=1+((C-BBandBot(C,pers,S,2))/(BBandBot(C,pers,S,2)));
indbol:=1+((fve-BBandBot(fve,pers,S,2))/(BBandTop(fve,pers,S,2)-BBandBot(fve,pers,S,2)));
diverg1:=(indbol-cbol)/cbol*100;
diverg:=Mov(diverg1,3,E);
macdh:=MACD()-Mov(MACD(),9,E);
fve>-15  AND fve<20 AND diverg>30 AND LinRegSlope(fve,15)>.6 AND  LinRegSlope(fve12,15)>0.5 AND fve12>Ref(fve12,-1) AND
macdh>0
Linear regression April 2003, S&C
====================== pds:=22 ;
pds1:= 35  ;
mf:=C-(H+L)/2+Typical()-Ref(Typical(),-1);
fve:=Sum(If(mf>0.3*C/100,+V,  If(mf<-0.3*C/100,  -V,0)),pds)/Mov(V,pds,S)/pds*100;
bbfve:=(fve-BBandBot(fve,pds1,S,2))/(BBandTop(fve,pds1,S,2)-BBandBot(fve,pds1,S,2));
bbc:=(C-BBandBot(C,pds1,S,2))/(BBandTop(C,pds1,S,2)-BBandBot(C,pds1,S,2));
bbfve-bbc
Stuart, via email
Los Angeles, CA

Markos Katsanos replies:

The MetaStock code that I provided in the November 2003 Letters to S&C column ("Errata: Detecting Breakouts In Intraday Charts And Divergence Methods") was appropriate for MetaStock explorations only and was not meant to generate a chart. The code for the finite volume elements indicator (FVE) is available in my article in the September 2003 issue of S&C, or from S&C's website at www.traders.com. [Editor's note: S&C subscribers can log into the Subscribers' Area of www.Traders.com to find the FVE code, and nonsubscribers can purchase the individual article through the Online Store at www.Traders.com.]

Here is the code for the Bollinger Band (BB) divergence indicator that will generate a chart:

Copyright © 2003 Markos Katsanos
D1:= Input("PERIOD FOR BB",5,300,40);
PERIOD:= Input("PERIOD FOR FVE",5,80,22);
COEF:=.1;
INTRA:=Log(H)-Log(L);
VINTRA:=Stdev(INTRA,PERIOD);
INTER:=Log(Typical())-Log(Ref(Typical(),-1));
VINTER:=Stdev(INTER,PERIOD);
CUTOFF:=COEF*(VINTER+VINTRA)*C;
MF:=C-(H+L)/2+Typical()-Ref(Typical(),-1);
FVE:=Sum(If(MF>CUTOFF, +V, If(MF <-CUTOFF, -V,0)),PERIOD)/Mov(V,PERIOD,S)/PERIOD*100;
IND:=FVE;
CBOL:= 1+((C- BBandBot(C ,D1 ,S ,2 ))/( BBandTop(C ,D1 ,S ,2) -BBandBot(C ,D1 ,S ,2 )));
INDBOL:=1+((IND- BBandBot(IND ,D1 ,S ,2 ))/( BBandTop(IND ,D1 ,S ,2) -BBandBot(IND ,D1 ,S ,2 )));
DIVERG:=(INDBOL-CBOL)/CBOL*100;
Mov(DIVERG,3,E)
Go to the indicator builder, click New, name it "BB divergence," and copy the above code. Values >0 indicate positive divergence, while values >30% indicate significant positive divergence.

I have found from experience that this indicator has a positive bias where penny stocks are concerned. Do your due diligence (that is, check that the stock won't get delisted from the Nasdaq soon or the company won't go bankrupt soon) before buying a penny stock.

If you want to scan the market for stocks with positive FVE divergence (that is, BB divergence >30%), click on the exploration (binocular) icon, click on New, and name the exploration "BB divergence" and copy the code that was provided in the November 2003 Letters to S&C column. [Editor's note: Readers can find this code at our website, Traders.com, in the back-issue archive area, under the issue of November 2003.]

You can then run the exploration to scan the market for stocks with divergent FVE.

FOREX TRADING

Editor,

I enjoyed your December 2003 issue, which was dedicated to forex. However, some of the opinions expressed by author Matt Blackman and interview subject Philip Gotthelf are misleading.

Regarding the account size recommended for beginners, the impression was given that $15,000 or even $500,000 is needed to trade successfully. I do not see how that is safer than starting with $1,000, as you can still use the 2% capital-per-trade rule with 0.5% margins now available. With proper stop placement, it is virtually impossible to get "hacked to pieces."

What is dangerous is huge overnight gaps in the futures market. This market, recommended by Gotthelf, with its gaps, slippage, and commissions, is definitely less desirable for a small retail trader. As far as daily versus intraday time frames, a longer time frame is not preferable in forex trading, as intraday charts often capture opposite trends when dailies get whipsawed or stay flat. Daily time frames are riskier and less consistent, and in the long run less profitable. Anybody who actually trades actively using technical analysis can see that.

About George Soros "having his technicals right," Soros is a pure fundamentalist. Yes, after his comments in late May, EUR/USD continued its uptrend, but only for a few days. It nosedived right after, from 1.19 to 1.09, for a huge loss for Soros. A competent chart reader would have caught both moves easily. Technical analysis is the way to go in forex, and your magazine's efforts are highly appreciated.

Ivo Toman
Tahoe City, NV

FOREX BROKERAGES

Editor,

With the recent increase in retail forex trading, it would be very useful if you wrote a feature comparing the different forex brokers - for example, FXCM, GFTForex, CMS-Forex, CMCForex, etc. As a trader, I have been confused with the variety of spread margins, software features, rollover interest criteria, interest paid on balance, and so on. While the various brokerages' websites detail the advantages of opening an account with them, there are some hidden elements such as liquidity, execution, and so on, which the customer cannot gauge.

Venkat Dharan

Thank you for your feedback. We can certainly try to publish an article on the "hidden" elements, but not one that compares various forex brokers.-Editor

BULL AND BEAR POWER

Editor,

In "Bull And Bear Balance Indicator" by Vadim Gimelfarb in the October 2003 S&C, Figure 4 displays a histogram of the bull and bear balance (BBB) indicator. At the end of June 2000, the market almost gapped down about 15 points. However, the graph still shows bull power is more than bear power. Am I missing something?

Gomu Vetrivel
Chennai, India

No, you are not missing anything. The chart displays the indicator correctly.-Editor

SYSTEM TRADING

Editor,

In your interview with Robert Deel in the October 2003 S&C, he comments that trading systems are not as functional as the human brain and a pair of trained eyes. Is it not subjective if I am trying to change my risk management based on my trained eyes? Can it be done by everyone? If it is subjective, how can I backtest the system and get confidence in trading that system?

Gomu Vetrivel
Chennai, India

Yes, it does make it subjective, and the point that Robert Deel is trying to make is that trading systems should be individualized. In other words, when you see that the market dynamics are changing, you need to make adjustments to your trading systems.-Editor

HURST EXPONENT

Editor,
"Finding Out About Fractals" by Clem Chambers (November 2003) was a great read, but I'm puzzled about how to use the Hurst exponent. I did some research on the Internet, and found an extensive article:

<https://www.bearcave.com/misl/misl_tech/wavelets/hurst/>

which seems to indicate that it is not easy to estimate and use the Hurst exponent for financial series. Is this something suitable only for those with deep pockets and a Ph.D.? Could you suggest some tools I can apply?
Ian Johnson, via email

The Hurst exponent is not easy to estimate. I would recommend Chaos And Order In The Capital Markets by Edgar E. Peters, although I should warn you that it still didn't help me in calculating the Hurst exponent.-Editor

THE TRENDILATOR

Editor,
Author Matt Blackman in the article "In Search Of The Trendilator" (STOCKS & COMMODITIES, October 2003) very clearly demonstrated how to build an indicator combining a trend and an oscillator. He further substantiated the indicator by testing over a wide group of 50 stocks. The performance of the trendilator was quite impressive.

Xavier Maria Raj
Chennai, India

READERS' CHOICE AWARDS

Editor,

For the past several years, your company has distributed awards for certain categories of products and services, be they data, software, or online brokers.

I've been trawling the Internet and your website trying to find a list of awards given out by you over the past five years. Would you be so kind as to either point to the correct place at your website or email me the list of the lucky winners of your awards, as I cannot find any reference to them? The only visible reference seems to occur when various vendors claim (in their marketing) that they've received such awards.
Dorian Porecki, via email

The annual Readers' Choice Awards are based on a subscriber poll we take every year. The awards are based on tallies of reader votes for various products and services related to technical analysis and trading. The awards are not editors' picks.

Sorry, these awards are not posted at our website. The results of this annual subscriber poll are available only in our Bonus Issue, which is a premium magazine sent to subscribers as a thanks for subscribing. The only way to obtain a copy of the Bonus Issue is to subscribe. You can contact our subscription department for more details at 800 832-4642 or circ@traders.com.-Editor

ERRATA: Divergence Bias

Editor,

In the caption under Figure 1 in my December 2003 article "Divergence Bias," there is an error. The proposition, "There is a similar hold for fracMACD and MACD," is inaccurate. The statement should have read: "There is a similar hold between fracMACD and %MACD."

Giorgos Siligardos

Back to February 2004 Contents