|
PureBytes Links
Trading Reference Links
|
Hello Jim Harmon,
<<As you can see, the DJIA April-July 1997 run-up (29.15% in 76 trading days,
or .38%/td) has exceeded that of 1987 (22.86% in 67 td, or 0.34%/td) in both
rate and extent. It has almost matched 1929 (29.93% in 68 td, or 0.44%/td).>>
Interesting post Jim. Certainly gets one thinking about the bigger picture.
A respectful comment or two. I wonder if you couldn't have said about the
same thing on or about 11/26/96. At that time the Dow had run up to about
6600 from the low on 7/16/96 of about 5200. Over that period July to
November 1996 the Dow rose 27% in 94 trading days 133 calendar days, and at a
rate of .34%/td (same as 1987). The similarities to 1987 continued with a
drop and then a retest of the previous high. As we all know we kept on going
up. The initial setup was "similar" to 1987, but the outcome was entirely
different.
I notice you use a linear or arithmetic scale as opposed to a log scale. My
experience has been that a log scale is a far better graphical representation
of how markets unfold. If you use a chart to look at the market going back to
1900 or 1960 on an arithmetic scale, you are basically saying there was no
inflation (one factor) from that period to today. It means 100 Dow points in
1960 carries the same weight, and implies the same ramifications as 100
points today. I'm not sure that's reasonable. What did the Dow drop in
1929? The same numerical drop today would hardly cause the same fuss. The
arithmetic chart may be interesting to look at, to see from where we've come,
but I think misleading for analysis. My Dow data only goes back to 1970, but
on a log scale 100 points in 1970 is equal to 1000 points today. Your 50%
retracement levels end up being about 38% on a log scale.
One last comment. Although I only have computer data back to 1970, on a
review of graphs going back to the early 1800s on a log scale, there seems
one obvious relationship between all the "major" correction. As time passes
each subsequent major correction is smaller than the previous one. I don't
see this mentioned ever, but there is a definite trend there. Doesn't mean
it can't change, but if it did, it would be counter to historical events
since the correction of the 1830s. The drop in 1929 appears less than 1835.
Nothing has come close to 1929 since on a log scale. 1974 was definitely
less than 1929. 1987 was about 25% less than 1974. I don't know if you would
consider 1990 or 1994 as "major", but they also reduced in size as a percent
on a log scale.
People like Prechter and others theorize about massive collapses, but the
evidence just isn't there yet. There is trend, and what's the expression?
Don't fight it?
Pattern recognition is interesting, but the same conditions never exist
economically during different time periods, to get a true comparison.
Russia, and China eventually, are going to be capitalist. Some third world
countries are bound to develop further. Who do you think these countries are
going to look to for experience on infrastructure, goods and the rest of
their needs? My guess is everyone listed in the Dow 30. Comparisons to past
market action must be approached carefully in my view.
You make some interesting points. Whether working with your 50% retracement,
or 38% on a log scale, the trick of course is determining from which point
you start to measure. If the argument is made that a correction, once
complete, has put the market back into a balanced state, then perhaps we
shouldn't be going back too far as a measurment starting point. Who knows?
Enjoyed your post. It also made me figure out what MIME files are <G>.
PJLaird
|