Focus

By Saeed Shah

By Saeed Shah

18 October 2000

When Jonathan Joseph, a California-based investment analyst with Salomon Smith Barney, reduced his stand from "outperform" to "neutral" on the semi-conductor industry in July, he was said to have received death threats. It is not as if he even said "sell", but he did point to signs of declining chip shipments, falling prices, a slowing of mobile phone growth and a peaking of capital expenditure by chip-makers.

In Silicon Valley, calling the top of the silicon cycle is no laughing matter. Jobs and investment depend as nowhere else in the world on the microprocessor industry, and in the suburbs of San Francisco, belief in its potential for unlimited growth is unshakeable.

It is easy to see why the dot.com bubble burst in the spring. Many of these companies had no discernible revenue model, let alone a route to profit. Sooner or later, even brokerages ramping these stocks had to own up -these emperors had no clothes.

But the technology industry is real, making real products, with many of its biggest companies enjoying huge margins and pulling in billions of dollars a year in revenues. If anyone in business has a future, it surely must be producers of microchips, telecoms equipment, computers and mobile devices.

The semi-conductor industry has had an average yearly growth rate of 18 per cent for the past 30 years, driven by what, all those years ago, would have seemed incredible advances in the processing power of chips. Much personal and corporate wealth in America is tied to the fortunes of the likes of Cisco and Intel.

Silicon chips are the vital, most basic technology product, used not just in mobile phones, laptops, and servers, but in virtually all electronic products of any complexity. They are the cement between the building blocks of the New Economy. But they are not a driver in themselves. Semi-conductor manufacturers rely on the health of their customers, who make the end-product. For instance, two-thirds of the world's mobile phones use chips made by Texas Instruments. Any downturn in the mobile phone market would be catastrophic for the company.

So is the industry heading for a downturn? Commentators fret, with good cause, about the ability of fixed-line telecoms companies to fund the cost of laying fast new networks to carry internet traffic. They also question the ability of mobile phone service providers to pay for the huge cost of "third generation" licences, which will allow them to run the networks of the future.

More tangibly, in recent weeks, we have seen a spate of profit warnings from many of those seen as hi-tech superstars, including Lucent, the telecoms equipment company, Dell and Apple, the computer manufacturers, and Motorola, the mobile phone and chip-maker. But most significant of all was a profits warning from Intel, the microchip behemoth whose processors are used in more than 80 per cent of the world's personal computers.

Intel said on 21 September that weaker than expected demand for PCs in Europe meant third-quarter sales would fall short of forecasts. The statement resonated round the world's markets and wiped $91bn off Intel's value the same day.

Technology stocks, from Cisco to Microsoft, suffered with it. Mr Joseph's startling market call appeared to be coming true even faster than he had anticipated. With Nasdaq now off about one-third since the euphoria of March, some are starting to believe we are staring at the abyss.

But the prophets of doom do not have it all their own way. The semiconductor industry is extremely cyclical and it is supposed to be in the middle of a boom. From 1996 to 1998, the sector was in a slump, caused by oversupply and sluggish demand, compounded by the Asian financial crisis. In 1999, the market picked up.

Gartner Dataquest, the technology research firm, believes sales in the semi-conductor industry will grow this year by a staggering 37 per cent, to $232bn. Gartner does not forecast a downturn until late 2002.

And Andrew Griffen, an analyst at Merrill Lynch, said the pattern of industry's cycles did not point to a significant slowdown soon. "Three years of above-trend growth caused the last downturn," he said. "We have had very rapid growth this year, but it is the first year in five with above-trend growth. It's a catch-up year. We are not in a crazy over-supply situation."

Some companies support this, including Sony and Ericsson who have not been able to get hold of sufficient supplies of chips. In the UK, Pace Microsystems said this summer it was "extremely frustrated" by a shortage of the flash-memory chips it needs to make its digital TV set-top decoder boxes.

The source of these complaints points to another reason to be optimistic about the state of the microchip industry - there are now so many more outlets for its product. From Sony's PlayStation 2, to the Palm Pilot, to set-top boxes, demand has mushroomed. Keith Woolcock, an analyst at Nomura, said: "The difference this time is that, in 1995, the chip industry was the PC industry. This time, the industry is much more heterodox."

And although a few technology companies are not performing up to the demanding standards they set for themselves, competitors in the same markets are booming. That makes for a confusing picture.

For instance, Advanced Micro Devices (AMD), Intel's biggest rival, last week beat forecasts with the news that third-quarter sales had jumped 82 per cent. The company's chief executive could not resist rubbing salt into Intel's wounds. "We are at the beginning of a growth cycle," he said. "It turns out our business in Europe looked just fine. Ourcompetitor, Intel, never had any real competition. For the first time, we are offering a better solution."

Intel has had well-documented problems over the past year, with shortages of product and delays in the launch of new chips. And it is true there has been a marked slowing in the PC market, especially the business segment Intel, Apple and Dell have thrived in.

AMD had the good fortune of grabbing a product lead on Intel with its Athlone chip. And AMD is focused on the home PC market, which has not slowed to such an extent as the corporate and education markets.

To some extent, Intel has simply found itself in the wrong market. PCs are becoming yesterday's news. The juicy pickings are now in the high-growth markets of communications, servers and networking. Here, it must face other dominant players, such as Sun Microsystems. In the hugely promising area of mobile devices, where the emphasis is on minimal electricity consumption and little heat generation, the forthcoming Crusoe chip from Transmeta must cause Intel's management sleepless nights.

There appear to be company-specific factors at work at Lucent too. The group's warning of a fall in demand for optical network equipment is in stark contrast to some competitors, who have reported sales of this product up 50 per cent this year. Andrew Norwood, senior analyst at Gartner Dataquest, said: "Intel has lost the plot somewhat. Its problems, and those at Motorola, do not spell gloom and doom. The problems are to do with their product mix. These are company-specific, internal difficulties."

But Mr Norwood would not agree that the chip industry has broken out of its cycle. "In 1995, they said it was the end of the silicon cycle and look what happened," he said. "This industry never learns. Everyone wants 20 per cent market share. When you have 20 companies chasing that, you get overcapacity."

Just like dull Old Economy sectors, such as paper and steel manufacture, chip-makers invest too heavily in new plant in the good times. Each new factory, which costs $2bn and takes about 18 months to complete, adds enormously to supply. When all these facilities come on-stream, there is a glut, which leads to a collapse in prices. That is the silicon cycle, but we are nowhere near overcapacity yet. For now, apart from some difficulties in the PC market, supply-side worries may be overdone. But, the other side of the coin is, as ever, demand. Here the semi-conductor industry could be hit by macro- economic factors well beyond its control - the high oil price, the reaction of the US consumer to the drop on Nasdaq, and inflation.

Merrill Lynch's Andrew Griffen said: "The question is, what will the global economy do in 2001? Will there be a soft landing? That's where the danger for the industry comes from. If there's a big economic slowdown, people may not change their cell phone so often.

"I'm a silicon industry analyst. People ring me to ask where we are in the silicon cycle and if there's going to be a downturn. But they're asking the wrong person. I tell them they must ask an economist."