|
Are You Missing the Moves Because of
'Default' Charts Settings?
How to Scale Your Charts
The very first aspect of charts we need
to clarify is something that new chart readers miss. Indeed, it
often takes considerable time before realizing that something isn't
quite right. Most software applications for chart reading have two methods
for scaling. Scaling is the way we see the prices on the side,
and the dates at the bottom of a chart.
The settings can have several
variations of names, depending on the software. They are Arithmetic
(also known as Linear, Logarithmic , and
possibly others), and Logarithmic (which has also been
called Log, Automatic, and Semi-Log). Yes, it gets confusing
right from the start, especially considering both chart styles have been
called Logarithmic. Most software defaults to Logarithmic
/ Semi-Log. What
this means is that, as the price of a stock rises, the scale is
averaged out to give more weight to the lower prices. This is useful
for comparing the percentage of change between a low
priced stock and a high priced stock, or the percentage of change within the same
stock over a period of time.
Be sure to find out exactly how your
charting software determines scaling. Go to the help and look up
each of the words above. Read their definitions so you are sure how
they are determining scale.
When looking at the chart, a small dip
a year ago would look the same size as a larger dip today, even though the stock has
risen substantially within that year. But the percentage of the
change would be the same.
Why would you want to do this? Because
if a $50 dollar stock is gaining 2% day, and a $150 stock is gaining
2% day, it may not be as apparent by looking at the charts. The Logarithmic
scale
would allow you to see the two stocks side by side as a pattern,
exclusive of price.
But if you are drawing trendlines, you
could end up with a distorted picture.
Lets look at the Nasdaq. This chart is
in Logarithmic scaling from 1975 through 2000. See the nice
clean uptrend? Also notice that the Crash of 1987 and the Crash of
April 2000 are about the same size. And note that from the bottom of
the chart scale up to 1000 is almost 2/3 of the chart. From 1000 to 5000
takes up the remainder of the chart. This is obviously giving you a
distorted picture of the actual price action.

Now view the same chart in Arithmetic
format. What a difference! When we hear media announcers tell
us that we are in the midst of the biggest bull market in history, the chart
above gives us a very different picture than the chart below. The real bull market
started at the end of the gulf war in the early 1990's, and then
picked up steam as the masses gained access to the Internet. And see
how the Crash of 2000 had a more devastating affect, dollar wise,
than did the Crash of 1987, even though the percentage of the plunge
was roughly the same. Even adjusted for inflation, these two charts
offer vastly different inferences. Note the key difference in the
chart scaling, where each 1000 point increment gets equal spacing.

In my opinion, the Arithmetic chart
offers a much more accurate view of what is happening. As we shall
see, this precision is very important in the anticipation of future
price.
Let's look now at a weekly chart of the
Nasdaq, to determine when price falls to the trendline. The chart
here is in Logarithmic mode.
Here we see that the trendline from the
1998 low moves up to the April 2000 low and the is confirmed by the
low in September -- a very predictable stopping place that shows the
trend doing just fine, thank you. Chances are, there is nowhere
to go now, but up.

But wait. Not so fast. Now look at the Arithmetic
chart. Oops! We haven't even reached the trendline yet. Instead of
bouncing, we may have more downside -- maybe all the way to about the 3500 level
before things get any better. Optimism suddenly turns to caution.
Keep in mind that we are only looking at the trendline here -- there
may be other factors, like Support, for example, that could
influence where the actual Bottom will be.

We know that the Arithmetic chart is
the more accurate reading, yet the Logarithmic is very compelling
evidence that someone, it seems, is buying and selling based
on the wrong chart!
The problem we face is that various
books and websites may use one or the other method of scaling. This
makes things more difficult, and certainly muddies the water. I have
come to the conclusion that the safest bet when viewing any stock I
am truly interested in is to examine the chart with both
scaling methods. That way, I know what both camps are thinking,
which will better enable me to plan a strategy accordingly.
|