Dad..
Thanks for compliments. SOX call was based on <<FIBONACCI NUMBERS AND RETRACEMENT>>. Of course I was looking at SOX very closely and jumped in at the right time. I am still a novice and have a lot to learn. Thanks a lot for providing all three of us the opportunity not only to learn swimming but to encourage us to venture at the deep end.
Rehan
Now see what Jim Jubak has to say about LSI TXN.
The calendar and the charts say this is the time to buy chip stocks.
In my opinion, the argument is overwhelming. In fact, to my mind the only issue is whether you want to pay up for super-growth stocks in the sector or look for some of the bargains that combine more moderate, but still hefty, growth with very modest prices. Your choice will depend in part on how you read the character of this market going forward.
Summer is almost always cruel to stocks of the companies that make semiconductors. And this year has been no exception. From June 20 through Aug. 10, the Philadelphia Semiconductor Sector fell 25%.
But in most years, the dog days of August are the time to load up on chip stocks. In three out of the last four years, buying the Philadelphia Semiconductor index on Aug. 15 would have resulted in a better than 30% gain by Dec. 31. Score that 36% in 1999, 32% in 1998 and 43% in 1996. Only in 1997 would that strategy have backfired. That year, the semiconductor sector peaked in late summer and early fall, then declined through the end of the year, driving the Philadelphia Semiconductor index down 31% for the period. (1997 is also the only year in the last five -- back to 1995 -- in which the S&P 500 index outperformed the Nasdaq Composite.)
This year is starting to look like it will run the recent record to four out of five. After hitting a low of 921 on Aug. 4 -- not quite matching the May 23 low of 870, but coming close enough to scare another group of investors out of the sector -- the index moved back above its 200-day moving average.
Investors can find the same evidence of a likely turnaround in the charts of individual stocks. Atmel hit 26 3/16 on Aug. 11, then moved up to 32 1/2 by the close on Aug. 15. Intel (INTC:Nasdaq - news) traded at 62 on Aug. 10, but closed at 67 7/8 on Aug. 15. Conexant Systems (CNXT:Nasdaq - news) dropped to 28 1/16 on Aug. 10 and then rebounded to 33 11/16 on Aug. 15. I could go on -- but I think you get the idea.
Behind the Sector's Ups and Downs
What's behind this seasonal pattern?
It was easier to explain when companies making semiconductors exclusively for personal computers dominated the chip industry. PC sales typically peak in the fourth quarter of the year and hit a trough in the summer months. That leads to the kind of pattern that stands out in Intel's financials year after year. In 1999, for example, sequential revenue growth (that's growth from the immediately preceding quarter) fell by 5% in the June quarter before picking up by 9% in the September quarter and 12% in the December quarter. Operating income dropped by almost 13% from the March to June quarters before climbing by 34% from the September to December quarters.
Today, the chip industry is a lot more diverse than in the days when PC makers constituted the dominant market. But the end markets remain heavily seasonal. Electronics makers -- which consume millions of chips in DVD players, digital cameras and game players -- aim their product introductions and marketing dollars toward the November-December retail spending frenzy. Car makers -- another major consumer of chips -- launch new models with big ad budgets in the fall, and gear their chip orders to build up inventories for new models during that period. Wireless-phone makers plan to launch the bulk of new phone models in the fall and see a measurable uptick in sales during that period.
OK, so if you buy into the logic of this seasonal pattern, then what do you want to buy now?
Dropping Names
I'd divide chip stocks into two groups, each with different fundamental characteristics and very different risk/reward profiles. I'd call the first group super-growth stocks and the second, bargain growth stocks.
I think we can all tick off some members of the super-growth group. The names became familiar to investors during the momentum-driven market that stretched through 1999 and into 2000. Broadcom (BRCM:Nasdaq - news), PMC-Sierra (PMCS:Nasdaq - news), Applied Micro Circuits (AMCC:Nasdaq - news) -- these are stocks with stunning estimated earnings-growth rates. Analysts project that Broadcom will grow earnings by 97% in 2000, PMC-Sierra by 122% and Applied Micro Circuits by 133% in the fiscal year that ends next March. And they have stunning price-to-earnings ratios to match. Broadcom sports a trailing 12-month multiple of 359, and Applied Micro Circuits trades at 411 times trailing earnings. PMC-Sierra seems a bargain at a P/E multiple of just 296.
It's not surprising that stocks in this group also turned in astonishing returns in 1999 -- Broadcom was up 373% for the year, PMC-Sierra, 378% and Applied Micro Circuits, 608%. What has surprised me this year is how well these astonishingly expensive stocks have held up in the technology selloff that started in March. From March 10 through Aug. 15, PMC-Sierra is down just 13%, Broadcom is down just 1% and Applied Micro Circuits is actually up 10%. It's not that these stocks didn't fall during the technology bear market -- PMC-Sierra, for example, traded as low as 125 9/16 on April 13 and 133 7/16 on May 23. It's just that they've recovered faster than most of the stocks in the sector.
They've recovered much faster than the stocks in the second group, certainly. At the close on Aug. 15, for example, LSI Logic (LSI:NYSE - news) is still down 59% from its March 10 price. (Atmel (ATML:Nasdaq - news) is down 46%, Conexant is down 68% and Texas Instruments (TXN:NYSE - news) is down 30%.)
The persistence of these huge drops has helped to give the stocks in this group a very different price/growth profile than the stocks in group one. They all had lower -- but still very respectable -- earnings-growth rates going into this technology bear market. And now they have much, much lower P/E ratios as well.
Here's how some of the stocks in this group shaped up on these measures as of Aug. 15.
LSI Logic had a trailing P/E ratio of 27, with analysts projecting 120% earnings growth in 2000 and 45% in 2001. Texas Instruments showed a trailing multiple of 43 and estimated earnings growth of 39% for 2000 and 29% in 2001. Atmel's had a P/E ratio of 42 and estimated earnings growth of 162% in 2000 and 62% in 2001. If you back out some big one-time charges, Conexant's P/E ratio is 39, and estimated earnings growth is 27% in the fiscal year that ends in September 2001.
Stocks That Could Go Bang
I think both groups of stocks will do well in the next few months, but I don't think there's any doubt that the second group now gives an investor much greater potential reward "bang" for the potential risk "bust."
Take a look at GlobeSpan (GSPN:Nasdaq - news), one of my favorite stocks in group one. UBS Warburg initiated a buy recommendation on the stock this week with a 12-month price target of 175. That would be a rather juicy 50% gain from the stock's Aug. 15 closing price of 116 1/2. The investment firm said it expects 30% revenue growth in 2000 and 57% in 2001 from this specialist in DSL (digital subscriber line) chips for the speedy delivery of broadband Internet signals over ordinary copper telephone wire. Analysts are projecting 269% earnings growth in 2000 and 75% earnings growth in 2001.
That means the stock is now trading at a price-earnings multiple of 364 times projected 2000 earnings per share. Using analyst estimates of the 2001 earnings-growth rate, GlobeSpan trades at a forward PEG ratio (that is, the 2000 P/E divided by the 2001 growth rate) of 4.9. Any investor putting money into this stock is betting that the market will continue to reward high-growth stocks with extraordinary multiples. That's essentially a bet that the market for the rest of 2000 and into 2001 will look like the momentum market of 1999 -- the one that made holding Broadcom, Applied Micro Circuits and PMC-Sierra so rewarding.
Compare that to the reward/risk ratio of LSI Logic, Texas Instruments or Atmel -- all Jubak's Picks. LSI Logic, for example, at the Aug. 15 price of 36.94, was trading at just 30 times estimated 2000 earnings per share of 1.22. That gives the stock a forward PEG ratio -- using analyst estimates of 45% earnings growth in 2001 -- of just 0.67. Atmel has a similar forward PEG ratio of just 0.51. Texas Instruments is higher at 1.82, but still a relative bargain for a company that dominates the DSP (digital signal processing) sector.
You can find similar bargain-priced growth stocks in the sector outside Jubak's Picks, too. STMicroelectronics (STM:NYSE ADR - news), for example, has a forward multiple of just 42 and a forward PEG ratio of 0.93. My favorite growth bargain, however, is Cypress Semiconductor (CY:NYSE - news), which has a forward P/E ratio of just 20 and a forward PEG of 0.78. And I think analysts could be underestimating earnings for 2001. The company has been aggressively marketing its SRAM memory products to communications infrastructure firms. The biggest win on that front to date is a $200 million- to $250 million-, one-year deal with Cisco Systems (CSCO:Nasdaq - news). Analysts speculate that Cisco's move to Cypress' product could lead to other networking contracts. Be that as it may, the company's growing networking business means higher margins for Cypress.
There's Life in Momentum
At the very beginning of this column, I noted that your decision on how many semiconductor stocks to buy from the super-growth stocks of group one, and how many from the bargain growth stocks of group two, would depend on your opinion on what investing style was likely to be most successful in the market going forward. I have to admit to being agnostic on the issue, currently.
I can see evidence that the momentum style that was so successful in 1999 -- and that favored super-growth stocks -- has more life in it. I certainly understand the logic of those who say that investors are turning toward growth stocks with reasonable prices. And the arguments for out-of-favor value stocks make sense to me as well. So my semiconductor selections in Jubak's Picks include both super-growth and bargain growth stocks. I own super-growth stocks like PMC-Sierra, and recommend buying the stock here. (Or even better, on one last, end-of-summer dip -- that's why I'm taking the mental stop-loss limit off this stock. I'd be a buyer at 190 rather than a seller.) I own bargain-growth stocks such as LSI Logic and Atmel.
I'm also just about equally interested in super-growth GlobeSpan and bargain-growth Cypress Semiconductor. Just about -- but not quite.
After the technology bear market of the summer, opting for Cypress feels like a better decision. That's a personal reaction, of course, and it's always dangerous to extrapolate from your own emotions to those of the market. But it does make me wonder if other investors feel the same way. And if they do, shouldn't I be thinking about shifting the style of my portfolio?
Like most of you, I've read a lot in the last few months about how investing in value will be the winning strategy for the rest of 2000, or how buying growth stocks at a reasonable price is the way to go, or about how momentum will bounce back. In my next column, I'm going to take a look at the wisdom of shifting styles, and whether this is the time to consider a big move in one direction or another. |
