Kirk, cyclicity: a poster on the Yahoo KLIC board has some
thoughts on cycles that are worth thinking about. Here is his message... If you prefer, here's a link
messages.yahoo.com
>by: crystalcharisma 1/25/00 10:41 pm
Msg: 6628 of 7160
Gottfried: Welcome aboard.
I appreciate you coming over to Yahoo to share your analysis.
Perhaps since I?m unfamiliar with your approach, although I
understand the derivation of the graphs, you need to explain the
message you?re trying to convey.
You must understand at the outset that a number of posters on this
thread have made a very strong case that the semi cycle is NOT ~2
years as is widely portrayed (and which appears to be the case in
your graphs) but is rather more likely on the order of 3-5 years. A
convincing argument against the two cycle, an argument that can be
applied to the apparent cyclicity of many stocks in the industry, is
the misinterpretation of the SOXX index. For example, it is often
claimed that the SOXX has shown an extremely consistent 2 year boom
and bust cycle since 1994. Two significant problems with this: 1)
the first ?bust? occurred in ~June, 1994....approximately 3-4 months
after the SOXX began. It impossible to state anything conclusive of
a statistical measure that had been around for such a short period.
Thus the 1994 ?bust? is bust and this is confirmed by the behavior
of many of the leading stocks in the sector - they paused but they
certainly didn?t dive during 1994.
There was a ?legitimate? dive in the industry in 1995-96,
approximately 4 years after the semiconductor rally began in 1992.
The second significant problem with the SOXX relates to the dive
that occurred from late 1997 to late 1998. While from the graphs it
appears this pullback falls perfectly in line with a two year cycle,
this is not the case. All of those familiar with the industry know
that 1998 (and 1999) would have witnessed the continuation of the
rally begun in early 1997...if it weren?t for the SEA crisis. Remove
SEA and there?s a very high likelihood the entire ?97-?98 dip would
not have occurred. And given this dip did occur, that it's
occurrence was brought about by events unrelated to the
semiconductor industry, and that the timing of the onset of the
crisis was completely arbitrary, the behavior of semiconductor
stocks during this period must be considered 'arbitrary' (i.e.
independent of the cycle) as well.
(Still more.....)
Gottfired-2
by: crystalcharisma 1/25/00 10:45 pm
Msg: 6629 of 7160
Thus no ?94 dip and no ?98 dip...and you?re left with a ~4 year
supply/demand-based semi cycle.
Thus, in my view, we?re left with two cycles to examine on your
graphs. In 1995, Klic?s stock price ran up pretty much in sync with
rising orders and then peaked as orders plateaued. Orders began to
pick up again but Klic?s share price plummeted until, eventually,
orders peaked. In 1998-99, Klic?s price also began to run up early
with the orders, but then stalled (it did NOT fall) as orders
continued to rise. This time, as orders picked up again, Klic?s
shared priced has soared. I?m not sure what the take home lesson is
here.
I would discuss this further - to compare Klic?s performance with
the other companies - but this is where I encounter great
difficulty. Since you?ve plotted the share price for each company on
an absolute scale (which differs significantly for each graph) while
the scale for the orders remains constant, it is virtually
impossible to compare these graphs coherently. The vertical axis for
the share price needs to be expressed on a percentage basis.
And finally, many of these companies are quite different now
compared to the previous cycle (no matter how you define ?cycle?)
and the timing and the degree to which they participate are likely
to be quite different with each cycle. For example, the most rapidly
expanding component of Klic?s business is its non-bonder units, a
fact VERY different than 4 years ago.
A fact VERY different than 2 years ago as well.
So, in the end, I?ll repeat: while I appreciate you bringing those
graphs to our attention, could you please explain their
significance? If you are implying some predictive value in these
graphs, could you provide your rationale...and whether you agree or
not about the 3-5 year cycle period.
I look forward to your input. |
