A Midlife Crisis for Tech?

Don't bet on it. Despite recent wobbles, the sector's talent for making chips more productive assures its future

By Michael Wallace

Back in 1965, one of the founders of Intel (INTC ), Gordon Moore, posited that the number of components on a chip would double every 18 months. Moore's Law, as it has come to be known, is now something of a shorthand expression for the growth possibilities inherent in the technology sector as a whole. Well, Moore's Law turned 40 this week, and from the recent bleak performance of tech stocks it would seem the theory has lost some of its mojo.

But has the sector's expansion reached a tired middle age? Not really. Underlying strength in corporate investment in technology spending appears healthy and poised for rejuvenation.

You wouldn't know it from the recent performance of major tech indexes. From its post-bubble peak of 2,192 at the turn of the year, the technology-laden Nasdaq composite index has fallen back over 260 points, closing at 1,932 on Tuesday, Apr. 19 -- a decline of 12%. Likewise, the Philex SOX (semiconductor) index has plunged over 55 points from its 2005 peak in March to 393 -- a drop of 12%. These retreats suggest that equity investors are predicting a reversal in the heady corporate profitability levels reached in 2004 that will result in a comparable decline in business spending on productivity-enhancing technology.


  While we at Action Economics expect that U.S. fixed investment growth will moderate in the first quarter, it will still be advancing at a very respectable 8% "real" rate, i.e., as adjusted for inflation. Sluggish commercial-construction activity is the culprit restraining growth in fixed investment, not the equipment sector. The equipment and software component of U.S. gross domestic product (GDP) remains rock-solid, and we anticipate a healthy 14% real growth pace for it in the first quarter, even after sizable 18% gains in the prior two quarters.

And despite weakness in stock prices, the strength in profitability through the first quarter, regardless of some weakness in recent reports, suggests that earnings growth will continue to fuel rapid expansion in business investment through 2005. Year-over-year profit growth is closely correlated with investment growth, and the downside distortion on profits from last year's hurricanes should be unwound this third quarter, leaving healthy year-over-year profit gains through the summer months.

Still, the tech sector has seen some less positive signals. Exhibit A: On Apr. 19, the Semiconductor Equipment & Materials International group (SEMI) confirmed that the book-to-bill ratio of equipment orders to billings fell to 0.81 in March worldwide, for the seventh consecutive reading below the 1.0 waterline. This suggests that chipmakers are wringing out efficiencies from existing equipment rather than retooling, in an attempt to pare inventories and boost capacity use.

Also, the Semiconductor Industry Assn. (SIA) reported in April that sales of microchips globally had slipped marginally to $18.1 billion in February from $18.4 billion in January, consistent with the emerging first "soft patch" for the overall economy and seasonal weakness. Yet sales compared to the prior year ratcheted up 15.8% from $15.6 billion.


  The historically bullish SIA claimed that sales were surprisingly strong for this time of year and were led by individual consumer products that now make up more than 50% of end users of semiconductor goods. Sales of microprocessors for PC and cell-phone applications were up 11%, DRAMs (dynamic random access memory chips) gained 36%, and chips for wireless use surged 53%. Lean inventories also suggest that the sector will be quick to respond to any pickup in demand as 2005 unfolds.

Moreover, on the 40th anniversary of Moore's Law the SIA has expressed confidence that the Intel co-founder's semiconductor capacity yardstick will remain viable for the next 15 years to 20 years. As VLSI Research CEO Dan Hutcheson noted in a recent essay, in 1970 there were just a 1,000 bits of information on the on the first DRAM chip, compared to 32 billion bits on the modern 4-gigabyte version.

The resulting economies of scale have been a powerful deflationary force in the global economy, says Hutcheson. There's little doubt that the relentless downtrend in technology prices is a big force restraining overall U.S. price gains. Indeed, Moore's Law is the building block of the new productivity paradigm -- that economic output can be increased without undue inflationary pressure -- which remains alive and well. The cost per transistor (pre-chip) was $5.52 in 1954, but one-billionth of a cent per bit of DRAM in 2005.

Though the technology industry itself has proven to be one a beacon of this new paradigm, it's still subject to business cycles and the forces of supply and demand. By retaining its commitment to progressively enhancing the power of the integrated circuit, however, the industry will continue play a vital role in keeping the global economic engine firing on all cylinders. We can look forward to celebrating future anniversaries of Moore's Law with those trends intact.

Wallace is global market strategist for Action Economics

Before it's here, it's on the Bloomberg Terminal.