Swing Strength Comparison

BarroMetrics Views: Swing Strength Comparison

I received a few e-mails asking why I keep saying that the 13-week swing direction is statistically overbought.

I have two reasons:

  1. The LOG percentage changes and
  2. Swing Strength

In Figure 1, I have shown the LOG percentage change of the larger 13-week impulse swings. Notice that even among the sample of 10, the current move is 5th. If you include all the swings since 1952 (inception of the S&P), you’ll appreciate that the current swing lies in the 3rd standard deviation. In addition, we are not comparing apples with apples.

I have used swings in a strong trending market rather than swings in a sideways market (where impulse moves are smaller).  Had I used the DJIA and compared the 1966 to 1982 and 1937 to 1942 with the current swing, you’ll find that the current swing is among the top 5% using log percentage change.

The second reason is the swing strength formula I use to measure the strength between swings. In Figure 2, I compare the 2002 rally to the current one.  The current swing is the 2nd strongest in this structure and 2.5 the strength of the rally from the 766 low to the 1560 high.

So the above are the reasons why I view the 13-week rally off the March low as being overbought. This does mean I would be seeking to short it. It does mean I would not go long without a 13-week correction unless there are compelling reasons to go long.

(The swing strength formula:The difference in the prices squared divided by the difference in the number of Bars)

2009-10-19-blog-13w-stats.jpg

FIGURE 1 LOG % CHANGE

2009-10-19-blog-13w-stats-2.jpg

FIGURE 2 SWING STRENGTH

2 thoughts on “Swing Strength Comparison”

  1. Hi Ray, I am curious as to how you compare the strength between swings, when you mean strength do you mean the rate of change (i.e. % chg / time)? The reason of my asking is because the volatility environment of the recent period seems to be higher than that of 2002, hence the ‘mean’ rate of change to be expected should be higher as well?

  2. Hi TK

    The formula is in the article. Basically it is range squared/time of two points.

    If you compare the current vol with 2002, it is not that different. It was a different story in the last quarter of last year and then I’d have agreed with you.

Leave a Reply

Your email address will not be published. Required fields are marked *