But in December that changed. There appeared to have been a rapid deceleration in the growth of the US economy. John Chambers, president and chief executive of Cisco Systems, explains: "Is the economy slowing? Absolutely. It is slowing faster than people realise. In December some people in business were saying it was like they had hit a brick wall, or the switch had been turned off. This is a US phenomenon."

That deceleration has hit even the once fast-growing information technology sector. The downturn in the growth of demand was sudden, argues David Peterschmidt, chairman and chief executive of Inktomi. He had expected his company to achieve record profits for the December quarter. But, as his finance team totted up year-end revenues, it became apparent that something at the end of December had gone very wrong for the Californian-based internet software supplier.

"Our customers were all set to sign. The contracts had gone through legal and were sitting on the chief executives' desks, and then they either didn't sign or significantly scaled back the order," says Mr Peterschmidt.

"They really didn't know what to do. There was consternation among many executive teams at the end of the last quarter. It was sudden . . . a collective loss of confidence."

Nor was Inktomi's experience: "It's clear market confidence wavered [during the fourth quarter]," admits Michael Capellas, Compaq's chairman and chief executive officer. Carly Fiorino, chief executive of Hewlett-Packard, adds: "Our earlier assumptions were based on the expectation of a soft landing for the economy. Uncertain economic conditions and the slowdown in IT spending have forced us to revise our guidance."

The result was that before the earnings season began in mid-January companies across the industry were warning of lower-than-expected earnings. Many of those that actually made their fourth-quarter targets then spoilt investor confidence by predicting declining rates of revenue growth and lower earnings for the coming year.

If the last quarter was disappointing, immediate prospects also look bleak. During 2000, US IT budgets, which include staff, hardware, software and services, increased 11 per cent, according to a study by Merrill Lynch, the investment bank. But the report estimates that spending growth this year will decelerate to just 5 per cent. That is bad enough, but the report's authors warn that 30 per cent of those polled had not adjusted their budgets for the economic slowdown.

Not surprisingly, given the slew of profits warnings and general slowdown, Nasdaq continued its steep decline. By mid-January, the index was down more than 50 per cent, a crash echoing that of Japan's Nikkei 225 average in the early 1990s.

The destruction of value in the valley has been astonishing. Many dotcoms ceased trading as internal cashflows remained negative and external sources of funds dried up. Even the few internet companies that were profitable suffered. Yahoo!, for example, was down nearly 90 per cent from its peak: its market capitalisation, which reached $139bn, collapsed to just $14.4bn.

"We've had some sectors fall more than the market did during the Great Depression," concludes Brad England, head of origination, equity capital markets, at Merrill Lynch.

The effect on Silicon Valley's economy has been almost immediate. Art galleries, European car showrooms and upscale architects have all reported a marked slowdown in orders. House prices and rentals - some higher than Manhattan - appear to have halted their relentless rise, and increases in the prices for commercial property seem to be moderating.

Not least, the personal wealth of many valley inhabitants has been hit badly. A huge number own shares and share options. A poll in the last quarter of 2000 by Gallup showed that one in three households in Santa Clara county owned options, 23 per cent in San Mateo, 21 per cent in San Francisco, and 15 per cent in Contra Costa. For many, the shares that would once have bought a house will now buy a piece of furniture - a nice one, it is true, but a piece of furniture nonetheless.

The broader ramifications are equally worrying. Mr Peterschmidt warns that the deterioration in demand is across the board: companies reining in spending are not just cash-strapped dotcoms but, more worryingly, large, well-established groups.

If he is right that corporate America is scaling back, postponing or even cancelling information technology investment, then the implications for the US economy would be serious indeed, for the triumph of the US economy during the second half of the 1990s was based largely on capital spending on technology. Not only did this directly boost economic growth, but it also increased productivity - in theory by a considerable amount. This in turn reduced inflationary pressures and allowed the Federal Reserve to keep interest rates low, encouraging economic growth.

Part of the problem may be that although productivity growth was boosted by technology investment during the late 1990s, in all the enthusiasm for IT there was over-investment in projects that would not generate a return above the cost of capital.

"There's no doubt corporate clients have been disappointed by the results of information technology investments," admitted Larry Ellison, chairman and chief executive of Oracle, the software group, at the Las Vegas Comdex trade show.

One reason for the disappointment is the complexity of contemporary computer systems which makes them unreliable, requiring constant maintenance from hugely expensive IT professionals.

The problem, according to Mr Ellison, is that companies have been buying equipment and software from different vendors whose systems will not work properly together. "Everything is custom-made. That strikes me as insane. When everything is custom-built it cannot be reliable or cost-effective. It's like asking for a Mercedes, but with a BMW engine, a Toyota fuel injection and a Cadillac's air conditioning, and complaining that the vehicle won't work properly."

"You then have to call in garage mechanics - sorry, systems integrators like Andersen, IBM or EDS to fix the whole thing with glue guns. Worse, you end up with your own IT people spending all their time maintaining the system, trying to keep it running," says Mr Ellison. "No wonder there's a shortage of IT people"

Mr Ellison's answer to this, naturally enough, is for people to buy software from one vendor - notably his own company, Oracle.

Not all agree with Mr Ellison's diagnosis, let alone his prescription. Dick Brown, chairman and chief executive of EDS, ripostes: "Larry's a salesman. You have to deal with legacy systems and you can't meet every customers' needs with one answer. You have to tailor the solution to the problem."

But even Mr Brown admits there is growing frustration among users with IT systems' reliability and high cost.

Such frustrations and doubts about the efficiency of IT investment have been around for some time. But there is a real danger that the economic slowdown could crystallise those concerns, convincing US corporations, rightly or wrongly, that they have overspent on IT.

Nonetheless, if IT spending actually begins to fall it will not affect all parts of the sector equally. "Enterprises will not cut their IT spending across the board," says Christopher Mortenson, technology analyst at Deutsche Bank Alex. Brown. He predicts that companies will still invest in areas that offer a high and quick return on capital, notably e-business software and storage.

The question now is how deep and long the downturn will be. What is certain is that the US will not repeat the mistakes of Japan during its lost decade of the 1990s. Japan has been stuck in sub-optimal growth because of structural barriers that prevent the reallocation of resources within the economy. In contrast, the US economy is relatively efficient at recycling capital, labour, land and technology.