Darning the SOX: Semiconductor Stocks Look Ready to Consolidate

The Red Sox have been burdened with the Curse of the Bambino since Harry Frazee traded Babe Ruth to the Yankees. The Sox haven't won a World Series since 1918.

The White Sox were tarred by the infamous Black Sox scandal of 1919 and haven't won a World Series since 1917.

The Philadelphia Stock Exchange Semiconductor Index, or SOX, has not been cursed or rocked by scandal. But it has converted tremendous performances from first-ballot Hall of Famers into periods of great success.

Yet there are times when the action cools, and the odds of a correction are greater than a continuing advance. Like now.

Here's my take:

First, since the SOX began trading in 1994 -- my data show it first started trading on May 4 of that year -- the index has had three Hack Wilson-type RBI seasons.

Time period	Gain	Length of advance
May 4, 1994 - Sept. 12, 1995	+155%	16 months
July 16, 1996 - Aug. 21, 1997	+192%	13 months
Oct. 8, 1998 - Aug. 25, 1999	+193%	10+ months

After the first two of these Icarus-like flights, the SOX suffered serious slumps.

Time period	Decline	Length of decline
Sept. 12, 1995 - July 16, 1996	-54%	9 months
Aug. 21, 1997 - Oct. 8, 1998	-55%	13+ months

Second, by all momentum measures, semis are extremely overbought and their year-to-date returns make the following indices look like a combination of McGwire, Ramirez and Sosa. The Goldman Sachs Technology Industry Semiconductor Index is up 61%, the S&P 500 Electronic Semiconductor Index 54% and the SOX 49.5%. (Slightly different components are responsible for the performance differential, and the GSTI Semiconductor Index is the most inclusive of the three.) McGwire, Ramirez and Sosa indeed.

Now, I am the first guy to say that I don't use overbought signals as sell signals, because overbought is a sign of strength, and as a momentum player that's what I look for. Besides, strength in the semis has nothing to do with technicals and everything to do with fundamentals. By combining strong momentum and strong fundamentals, you'll get the stock equivalent of Ivan Rodriguez (who gets my vote as one of the top three players in the game).

So really, it's not the overbought signal that gets me concerned because I never want to punish a stock or group just because it is winning big. But there comes a time when I ask myself if I'm really that smart when it comes to gauging performance of this magnitude. The answer is usually no. What I mean is that it is very unrealistic to expect any index to power substantially higher from current levels given the rates of return (since the October 1998 low and on a year-to-date basis) without a consolidation or corrective phase.

Third, a look at the SOX roster reveals that all of the players are not on the field at the same time. For example, eight of the 16 stocks made their highs in July (Altera (ALTR Quote), Applied Materials (AMAT Quote), KLA-Tencor (KLAC Quote), Linear Technology (LLTC Quote), Motorola (MOT Quote), Novellus (NVLS Quote), Rambus (RMBS Quote) and Teradyne (TER Quote)), when the index closed at 530.08 on July 16. Since then, these stocks have been sideways to down, though the SOX has been essentially flat (-2%), because Intel (INTC Quote), LSI Logic (LSI Quote), Lattice Semiconductor (LSCC Quote), Micron Technology (MU Quote), National Semi (NSM Quote), Texas Instruments (TXN Quote) and Xilinx (XLNX Quote) have been strong.

Not taking anything away from the performance of INTC, et al., the split in the index suggests the SOX is human and investors should temper their expectations. And if you've got a July-high semi instead of an August -high semi, you haven't knocked in any runs in more than a month.

So what does all this mean?

• History, albeit thin, suggests the semis have already seen all or most of their advance.
• Gains since October 1998 and January 1999 are too heady to think this can continue without pause.
• Semis are not monolithic. Eight of the stocks in the index are lower than they were in mid-July.

I'm taking profits in these stocks for those reasons and because I don't think the market is prepared to work higher. And every semi analyst on the Street is either increasing earnings estimates, writing about higher utilization rates resulting in firming component prices, squawking about stronger cash flows or raving about increases in cap-ex budgets. I wonder if all of the good news is already priced into the stocks.

I mean, after all, a 193% gain in just over 10 months. I figure it's pretty safe to say these stocks, at least over the near term, have already discounted this sort of good news.