"Can "Mac D" lead a Nasdaq rally?"
"A Fibonacci ghost blocks the bullish divergence's path "
By Tomi Kilgore, CBS.MarketWatch.com
Last Update: 12:00 AM ET Nov 27, 2000
NEW YORK (CBS.MW) - After hitting a new low for the year last
Wednesday, Nasdaq bulls were thankful for a 5.4 percent rally on Friday.
Although the retracement path is littered with resistance points, at least
one technical indicator has been hinting that a stealth rally may be in the
works.
But even if the moving average convergence/divergence (MACD)
indicator did suggest that a short-term rally might be in the works,
resistance should be waiting at around 3,030 and 3,210, backed by
centuries old Italian ratio.
George Appel developed the MACD lines using a
series of moving averages. Typically, a "trigger", or
a nine-day exponential moving average (EMA),
one that weights the recent closing prices more
heavily than older ones, is compared to the
difference between the 26-day EMA and the
12-day EMA. Some technical analysts use the
"Mac D" to follow a trend because it helps block
out a lot of the noise.
Pat Fitzgerald, Senior Technical Analyst at
Thomson IFR, pointed out that although the
Nasdaq Composite index ($COMPQ: news, msgs)
has been putting in lower lows since April, the
MACD has been putting in higher ones.
On Oct. 18, when the Comp hit a low 3,026.11,
the trigger stood at -160.42. And last Wednesday,
after the Comp fell to year-to-date low 2,754.14,
the trigger closed at 112.53. When the direction of
a charted instrument differs from that of a technical
tool, it is often referred to as a "technical
divergence."
In this case, it's thought to show that, although
bears have been the aggressor of late, bulls have been getting in some
good body shots. Although it's difficult to gauge the timing, the thinking is
that bears will eventually run out of gas and the Comp's price will
catch-up to where its guts have already gone.
But it shouldn't be a smooth ride by any means.
Resistance rears its head
Fitzgerald sees one of the first roadblocks coming
at roughly 3,030 for two reasons. First, it was the
Oct. 18 low (3,026.11). Previous lows, when
they are surpassed, tend to be resistances on
retracements considering initial buyers at that level
that have been burned may be very reluctant to try
again.
Second, it follows a "golden ratio" lovingly adopted by technical analysts
to calculate retracement levels.
Follow the 'golden' rule
An early 13th-century mathematician ghost named Fibonacci wrote about
a ratio - 0.618 - that he found to be prevalent in natural systems. Since
the market is perceived to be a living, albeit unforgiving, creature,
technicians have applied the ratio and its inverse - 0.382 - to their
analysis. As long as a retracement stays within 61.8 percent, it is still just a
part of the original move. If it passes that mark, the move takes on a mind
of its own.
If you calculate 38.2 percent of the difference
between the Nov. 6 high (3,480.01) and
Wednesday's low (2,754.14), you should get
277.28. Add that to 2,754.14, and you get the
first Fibonacci retracement level (3,031.42).
The fact that the 3,030 level was identified as an
impediment by using more than one method gives it added significance.
The 3,210 resistance area is also the junction of two different paths.
When looking at the drop from the Nov. 6 high, it's represents a 61.8
percent bounce. That's the most significant "Fibo," one that bears will have
every incentive to defend.
In addition, when the
Comp rebounded off
its Nov. 13 low
(2,859.39), it topped
out at 3,208.95 on
Nov. 15. If bears
managed to beat
back the bulls there
once, technicians feel
they may have the
confidence to take another crack at it.
A Fibo's ticking bomb?
Looking at where a
post-Thanksgiving
rally might stall may
be a bit pre-mature.
I applied the Fibo
technique to a
monthly chart,
subtracting 1,357.09
(November 1998
low) from 5,132.52
(March 2000 high), and multiplying the difference (3,775.43) by 61.8
percent (2,333.22). Take that off the March 2000 high and you get
2,799.30, a level surpassed briefly last Wednesday. Of course, that
would be close enough for horseshoes, hand grenades and some
technicians. But others may be running for cover.
If that scenario were to play out, I asked Fitzgerald where he saw support
in the Comp. If Wednesday's low were surpassed, he saw strong support
around 2,450.
Going to a
longer-term chart,
you will see two
weekly lows
(2,474.41 and
2,442.22) in that
area in the middle of
August 1999 (weeks
ending Aug. 6 and
Aug. 13). If
something works, you keep doing it until it stops working.
Below that, the 2,330-40 level stands out. That coincides with the low of
the week ended Apr. 23, 1999 (2,329.87) and the week ended May 28,
1999 (2,339.12).
cbs.marketwatch.com
Mang |
