The Scale Effect on Chart Work

[zdo 20]

This post is for the small group that might both ‘need it’ and ‘get it’ that they ‘need it’.

btw the following was written when almost everyone still drew charts by hand or received them in the mail… in computer charting it is easy to make adjustments, but our brains are basically at the same evolutionary level as they were when it was written.

Some people get it and automatically, almost without thinking make the proper adjustments. Others don’t quite so unconsciously see the need to make scaling compensations or replot for analysis purposes…

 

Anyway if you ‘chart’ various stocks, or multiple commodity groups, or multiple fx pairs, or various treasury instruments (or even multiple trading indexes on the same time frame) – making the proper scaling adjustments can be a tiny change that brings significant transformation in your perceptions.

 

DEALING WITH THE SCALE EFFECT

 

This is another highly significant factor when dealing with charts. The kernel of the problem is that no chart is unique. The time-price pairs that constitute the data on which the chart is based are unique. Once established, they appear in your Wall Street Journal just exactly as they appear in “Friend Jack’s Journal,” 2000 miles away.

 

However, when you construct a chart of this data and Jack charts the same data you will sometimes have difficulty believing that the two charts describe the same stock. In fact the significance of time-price relationships in a stock can be suppressed and /or lost if care is not use in the selection of scale factors!

 

Now, the time scale factor that you choose is simply the space that you allot on you chart to represent a particular unit of time (day ,week, etc.). Similarly, the price scale factor is the spaced you allot on your chart to represent a specific unit of price (one-eighth, on-point, etc.). Thus by choice of scale factor, a chart can minimize time effects while emphasizing price motion, or vice versa.

 

If you fix your time scale factor, you can accomplish the same results by varying the price scale factor.. Or you may select both so that you must stand yards away in order to see relationships without receiving a surrealistic effect. Through compression of both scales, you can convey a sense of unimportance and insignificance for both time and price motions – hence of the stock itself!

 

The problem is thus seen as one in which you must first know what it is you’re looking for on a chart. Then you must choose scale factors which optimize the ease of perceptivity of the desired information.

 

You will to want to experiment on you own in this area, but a few guidelines can be drawn:

 

…

4. In general, arrange the scale factors sot that the price motion over the time period of interest forms a nearly square chart.

5. To suppress the effect of short components while emphasizing the longer ones, plot the price motion in less space while retaining the same scale factor for time.

6. To suppress the effect of long components while emphasizing the short ones, plot the price motion in more space while retaining the same scale factor for time

7. If a pattern you’re analyzing (a triangle, for example) seems insignificant to you, enlarge both price and time scale factors.

8. If you attention seems riveted on what you know are insignificant phenomena, reduce both price and time scale factors.

 

Chart services present a different version of the same problem. In such charts, the space available is fixed, as is the time span covered. This means that the time scale factor never varies from chart to chart and issue to issue.

 

At the same time, the price motion of all stocks charted by the service must be displayed in a fixed amount of chart space fro the total time covered.

 

Thus a highly volatile issue will have a compressed priced scale (with all that this implies), and a sluggish issue will have an expanded price scale (with undue emphasis on short duration fluctuations). …

 

All of this means that the same types of… phenomena occurring in two different stocks charted in this manner require a shift in your thinking in order to effect the required same interpretation. Either this, or one of the other of the two charts must be replotted.

 

… practice negating the effect of scale factor change from chart to chart…

 

 

From The Profit Magic of Stock Transaction Timing J.M. Hurst pg 165

[BlowFish 10]

Very worthy of consideration. By using fixed scaling you can immediately get a better feel for momentum and volatility. You can see price action more clearly when a chart isn't re scaling every few ticks. You get a much better feel of whether price is up or down and by how much. Also geometry is far more apparent. Many moons ago I spent a bit of time corresponding with Howard at ensign and the fixed scale option they have is the result. Sadly I no longer use ensign.

 

Another thing that is nice to have is a chart start at the left edge at session open then fill to the right closing on the right edge. This has a similar effect in so far as you know in exactly where you are in the day.

 

I asked the guys at MC to implement this, guess I should ask the Ninja guys too but its probably a bit late for version 7.