Another case point: suppose that Gruen had not formed a price-move termination triangle as it did. Are there any other criteria that could have been used to detect approach to the trading cycle peak? A much more generally applicable technique is needed, and is provided by the concept of non-real time envelopes.

The channel envelopes used so far are constracted in calendar or "real" time. But we know that cycle durations vary over a range. This variation coupled with inability to graphically separate components with accuracy sometimes leads to a situation where all of the information inherent in cycle presence is not displayed to us through channel construction. This effect can be minimized by forming channels from cychc highs and lows that are artiflciaUy placed equidistant from one another on a pseudo-real time scale. The detailed significance of time is thereby lost, but information is often gained that is very valuable. Heres how such channels are made:

• Go back over your chart and tabulate the successive prices of the stock at all intra-day highs and lows associated with any component that you wish to plot in non-real time.

• Use a separate sheet of graph paper on which to plot these.

• Plot the first low on the stock chart on the far left vertical of your new plot. On the very next vertical plot the next high. On the next vertical, plot the following low.

• Proceed in this manner until the data is exhausted.

• Form a smooth, constant, and minimum-width envelope through your data points. Adjust width until all points are as closely related to the resulting channel as possible.

Now what has been accomplished? Horizontal distances on your new chart still represent time, but equal distances between points no longer necessarily mean equal times have elapsed. In fact, component duration changes from any cause whatsoever are completely eliminated. This envelope will not differ very much from the envelope on your stock price chart-unless cycle duration change becomes significant But at that point, the non-real time envelope will give you a much clearer picture of what is happening than will the one plotted in real time.

Now try something else with your non-real time channel. On the lower channel bound and on the same vertical line as each "high" that is plotted, place a distinctive mark such as an asterisk. Do the same on the upper bound opposite each plotted low. Now, your last data point on the chart will be either a high or a low plot, depending upon whether the stock has last defined a cycUc high or low. Put an asterisk below or above this point, and spaced just the channel width from it. Now you can draw yoiu-non-real time channel, full width and right up to date. Note the slope and curvature (either up or down) of the channel at this point, and exteiKi both channel bounds two vertical lines on your chart into the future. The points of intersection of the extended channel bounds with the next two vertical chart lines represent a best estimate of the next high and low (or low and high) of the stock price at the times these should occur per cyclic criteria.

Now go back to your stock price chart. For the component in question, average sample durations and mark on your chart the predicted time locations of the next high and low. Use the sample variation from average to ow a time spread into which you expect these points to fall. Read the predicted stock prices for these times from your non-real time chart, add a reasonable plus and minus tolerance to these and plot the

result on your stock chart as two "boxes." To the best of your ability you have now located the price and time regions into which the stock will move next. ContmuaHy update this estimate by cyclic analysis as time passes and the stock price changes. You wiD be surprised at how often the stock behaves per prediction!

There is another advantage ui this type of graphical analysis. OccasionaUy the inevitable band of uncertainty connected with a real time envelope will mask the significance of a single days price action which actually portends a non-cyclic change in the situation. The non-real time envelope will often catch such a situation, making you aware of it in time to take appropriate action.

Needless to say, this technique may be combined with the sell signal criteria of previous paragraphs to sharpen up your selling (or short covering) transaction timing. In fact, lets see how it would have worked in the case of Gruen. Figure V-3 shows Gruen again up to and mcluding the day of termination triangle resolution. Our question on this day is how much further upside motion can we expect on this trading cycle move?

At this pomt in time, we do not know for sure whether or not point I on the chart is a low of the four-week cycle. However, assumption that it is will provide us with a conservative answer to our question, which is to be desired. We proceed to label each successive four-week high-low as A, B, C, D, E, F, G, H and our tentative I. The four-week cycle was chosen for this purpose because it is the component of next shorter duration than our trading cycle. Using the technique described above, the non-real time envelope for this case is drawn on the right, with corresponding peak and valley labeling.

Now note the effect of the low at I! To get to this point from the low corresponding to the high at H, the non-real tirae envelope must curve over sharply. We further note that if we are wrong about I being a four-week cycle low, the envelope would still be required to curve over unless this low were as high as 14! Since this is highly unlikely in view of the cyclic condition of things at this time, we are alerted that major components are indeed in the process of topping out, and we should be preparing to take profits. Note that this conclusion is identical with that derived from the triangle resolution analysis-but this time there is no dependency upon the triangle forming (in many cases it will not). We can go a little further. Continuing the turnover from I through the next two verticals on the non-real time chart establishes a conservative estimate of 12 3/4 for the next peak and 11 for the next low. With this information in hand the actual top-out at 13 and valid trend line sell signal at 12 come as no surprise, and lend credibility to your decision to take profits on the sell signal. It is interesting to note that the real time envelope to the same point in time provides no hint of a turnover for nearly another week-too late to act as the desired confirmation of the sell signal. This is typical of the use and value of the non-real tune envelope technique.

SELLING SHORT

In Chapter Four and so far in Chapter Five objective signals based upon the price-motion model have been developed for "wait," "buy," "hold," "sell," and "protect-Ioss" situations. Only two more are needed to round out our armory. These

Non-Real Time Envelopes