|
PureBytes Links
Trading Reference Links
|
Mike:
This is a most informative posting. It would be great to have something
like the Trendsetter program for plotting trendlines. If you come across
something, please let me know about it.
Thanks much for your posting
Lionel
At 12:04 AM 6/1/97 -0500, Michael L. Robb wrote:
>Michael Arnoldi;
>
>Hurst's 1970 book on the Profit Magic of Stock Transaction Timing (Prentice
>Hall) introduced the concept of "valid" trend lines. Valid meant trend
>lines that crossed over important cyclic bottoms (under,tops). As an
>example; for daily data, he demonstrated a "model" of 80D, 40D, 20D, and 10D
>cycles.
>You pinpoint cycles from visual inspection of the data. You pace them off by
>first and second finger measurements. You draw downward trend lines over
>taops of the smaller cycles.
>
>When the trend line passes over the expected place of a major, longer term,
>low (40D, 80D, 20W, 40W, 80W), dthen the subsequent breakout above
>dastsrendline is a "valid" breakout and may be followed.
>
>There is another aspect to his work, which makes a "system", and that is the
>"half-span" moving average. This is simply the MA of 1/2 the period of the
>dominant cycle. The "dominant" cycle is the one having the greatest
>magnitude, determined via visual inspection/.
>
>Now that we have MS Fourier analysis tsechniques built is as an indicator,
>we can analyse price movement for specific cycle lengths. If you believe MS
>Fourier is accurate, you will use it to determine your dominant cycle.
>
>There are fine points of sub-dominant cycle analysis that can be included in
>this approach; where ,i.e. smaller waves come in threes instead of twos,
>etc; but that is not the princicpal thought in defining the valid trendline.
>
>Only when important longer term cyclic lows are expected (and subsequently
>made) are trend lines to be drawn, that can be considered valid.
>
>The half-span MA technique is the use of MA's of 1/2 the period of each
>identifyable cycle, lagged by half their span. When you see a layered
>cascade of these MA's above your valid trend line, you can calculate your
>upside price objectives by multiplying distance from low to break of MA by
>2; fosr each MA in turn.
>
>A very intriguing system. Only Pardo from Chicago has produced software for
>this entire system, but it is older and may no longer be available. You can
>use the MS Fourier analysis instead to achieve virtually the same results.
>
>Another use of trendline is one developed by Jeff Borowitz of Trendsetter
>(LA, CA); proprietary, and available only for the Mac. It uses an
>iterative process to produce trendlines, on the basis that all important
>moves result in the breaking of some long trendline. What software does is
>produce a set of short,medium, and long term trend lines based on actual
>prices, but without using "hindsight". Buy and sells are calculated at most
>profitable distance over, or under, trendlines. Results in excellent entry
>prices when right.
>
>Trendlining the momentum indicators has also been used successfully by
>commodity traders in the Wilder- Patterson millieu to find entry and exit
>points.
>
>In addition there are many interesting (but complicated) formulae available
>at the Market Research Association website.
>
>Regards,
>
>Mike
>
>
>
|