Tech Investing Comments . . . Despite the Tech Wreck in the stock market over the past three years, investors remain addicted to technology stocks. Old-timers warn that new bull markets are rarely led by the stocks that led the previous bull market, and that declined the most during the subsequent bear market.
The speculative bubble of the late 1990s was mostly dominated by grossly inflated tech stock prices. Nevertheless, while the S&P 500 Information Technology (IT) Index and the Philadelphia Stock Exchange Semiconductor Index (SOX) fell 83% and 84%, respectively, from their March 2000 record highs to their October 2002 lows, there were several huge rallies in these indexes along the way. And the latest stock market rally, whether measured from last year’s low or this year’s in March, was led by technology stocks.
I would love to join the Tech Cheerleaders’ Club. However, I am not convinced that the outlook is as rosy for tech investors as suggested by the high valuation multiples they seem so willing to pay to join the club. I do believe that the worst of the Tech Wreck is over. I was an early proponent of overweighting tech stocks back in the mid-1990s. I forgot to say “sell” at the top, but I was consistently bearish during 2001 and 2002. Now I am a tech opportunist. I do see some opportunities for profitable investments in the tech sector, especially in the computer hardware, computer software, storage systems, and wireless telecommunications industries. The outlook for the Internet survivors also looks to be very good, but those stocks are among the most expensive in the market.
During the past three years, there have been five significant rallies in the S&P 500 IT Index; the gains ranged between 28.0% and 52.9%, with an average of 40%. The losses during the subsequent five sell-offs averaged -43%, ranging between -19.0% and -55.2%. The swings in the SOX were even greater. This year, the S&P 500 IT Index and the SOX are up 23% and 29%, respectively, from their March lows. The previous rallies in the bear market were not supported by the underlying business and earnings fundamentals. The latest tech rally has more going for it. New orders for technology are no longer falling, and may be starting to recover. Industry analysts are no longer slashing their earnings forecasts for the current and coming years. In some cases, they are starting to nudge their numbers higher, if ever so slightly. However, the latest rally in tech stock prices once again places very high valuation multiples on these stocks. This is reasonable if earnings rebound sharply. I am not convinced that this is likely to happen across-the-board for the great majority of tech stocks that have participated in the latest rally. I think investors need to focus on groups within the tech sector, rather than the entire sector, selecting those with compelling growth stories.
The S&P 500 includes 82 technology companies. These companies are divided among 13 tech industry groups. The list includes many companies with sales around the world. Nevertheless, most of their business is in the United States. This explains why analysts’ forward earnings estimates for the S&P 500 tech stocks are so highly correlated with new orders for high tech. The orders data are compiled monthly by the U.S. Commerce Department and cover only “computers and electronic products.” Software, IT services, and internal company outlays are not included. The Commerce Department aggregate for IT orders includes “computers and related products,” communications equipment, and an “other” category. Semiconductor orders were dropped about a year ago because some of the major manufacturers refused to participate in the monthly survey. For the overall S&P Technology sector, both new orders and forward earnings rose sharply during the second half of the 1990s. Then both plunged together during 2000 and 2001. Both have subsequently stabilized and show tentative signs of recovering in recent months. Here are some highlights of recent developments in four major S&P 500 tech industries:
* The recovery signs are a bit stronger for the Computer Hardware industry group, where forward earnings are trending higher again and new orders are up 13% over the past year through May, though this time series is especially volatile.
* The new orders for nondefense Communications Equipment are also quite volatile from
month to month. However, they seem to have bottomed out since mid-2001 after declining by roughly two thirds from the peak of 2000. Forward earnings were negative during 2001 and 2002, but have recently turned back up just above zero (Figure 5).
* The forward earnings of the Semiconductor Capital Equipment group tend to be correlated with the industry’s book-to-bill ratio. Earnings estimates for the next 12 months have been rising since the start of the year despite downward revisions for this year because 2004 is expected to be well above 2003. Nevertheless, the book-to-bill ratio has weakened slightly in recent months.
* In the Semiconductor industry, forward earnings for the 16 S&P 500 components have been range-bound since the second half of 2001 following a major sinking spell during 2000 and the first half of 2001. This industry is among the most global of the tech sector. Forward
earnings are highly correlated with worldwide sales data compiled monthly by the Semiconductor Industry Association. These sales have been trending higher since January 2002, and are up 25% through May of this year. Still, May’s annual rate was $150 billion, well below the October 2000 record rate of $224 billion. Sales in both the Americas and Europe have been virtually flat for the past two years at the lowest levels since the first half of the 1990s. On the other hand, sales in Japan and Asia Pacific are recovering. Indeed, they are back at record highs in Asia Pacific.
During 2001 and 2002, industry analysts were forced to slash their earnings estimates for 2001, 2002, and 2003 almost every month, according to my “Earnings Squiggles” analysis. This year, the estimates for 2003 and 2004 have been very stable. Forward earnings are converging toward 2004 estimates, which were up 30% from 2003 estimates as of June. That’s a big increase, but that would raise next year’s S&P 500 Tech sector earnings back only to the 1997 level. Yet the forward P/E in June was 28.1 versus 20.1 during 1997, on average. P/Es of the 82 tech stocks in the S&P 500 using 2004 consensus earnings estimates are trading at multiples of 30 or higher. The S&P Tech sector’s P/E has been 50%-100% higher than the market’s valuation multiple since 1999. Before then, it tended to be about the same as the market’s P/E. Another way to see the same divergence is to compare the market-cap share of the S&P Technology sector, which was 16.3% in June, with the sector’s earnings share, which was 10.3% during the month. My hunch is that the two will converge again—though it may take a few more years to detox all the tech addicts out there. Certainly, industry analysts are still going through withdrawal as they continue to lower their expectations for long-term earnings growth. Perhaps the best indicator of irrational exuberance during the tech-stock bubble was the hallucination of industry analysts that technology earnings could grow by 28.7% per year. That was the consensus when the market peaked during March 2000. Now during June of this year, the estimate is down to “only” 14.6% per year.
Storage . . . Adaptec said its expectation for a strong business upturn in June 'failed to materialize' due to ongoing economic uncertainty. As a result, the company now sees revenue of about $107 million for the first quarter, below its prior projection for revenue ranging from $115 million to $120 million. Adaptec believes, however, that its bottom-line results for the first quarter will land in the middle of its previously disclosed range for earnings before items of 2 to 4 cents per share.
Network Equipment . . . MOT announced it is the first supplier to sign a contract with China Unicom for the Phase III expansion of the operator's CDMA 20001X network. The contract, worth more than $80MM, will bring advanced mobile communications services to customers in the cities of Nanjing, Suzhou, WuXi, Changzhou, Nantong and Zhenjiang in Jiangsu province, one of the most developed districts of China. This is confirmation that Phase III is on plan to begin this summer as expected and confirms view that the CDMA wireless infrastructure market remains a growing one, while the GSM outlook is decidedly more cautious. Expect to hear similar announcements from Lucent, Nortel and Ericsson.
Terayon Communication is saying it now expects a second-quarter loss of 18 to 20 cents per share on revenue of between $30 million and $31 million, better than its previous outlook for a loss of 20 to 24 cents per share on revenue ranging from $24 million to $28 million. Three analysts were looking for a loss of 22 cents per share in the period, on average. The Santa Clara, Calif., communications equipment firm attributed the better than expected performance to higher unit shipments of its Docsis 2.0 cable modem termination system in Asia and North America, a larger than anticipated order for its Multigate cable telephony system from a European customer and higher sales of its legacy S-CDMA product lines.
Semiconductor Equipment . . . KLA-Tencor has confirmed it will introduce its metal metrology product at Semicon West next week; this is seen as a negative for Rudolph Tech, since INTC has been the largest customer for RTEC on its similar system, but has been one of the primary beta testers for KLAC.
RobBlack.com MarketWrap:
robblack.com
Thanks for the table updates Don.
Happy Fourth Everyone!
And Michael...
You earned every dollar yourself!
Congrats on those gains.
RtS |
