Venture Capital: Should Intel Stick to Its Day Job?

When Intel Corp. (INTC ) announces third-quarter earnings on Oct. 16, losses on its venture-capital investments are likely to take a big bite from its bottom line. Over the past 10 years, Intel has pocketed net gains of $4 billion from the portfolio run by its Intel Capital unit, which invested in the likes of BlackBerry pager maker Research In Motion (RIMM ), Red Hat Software (RHAT ), and incubator CMGI (CMGI ). But shortly before the September 11 terrorist attack, Intel warned that VC losses were mounting and would exceed interest income in the third quarter. The losses, the first since the chipmaker began disclosing VC results in the fall of 1999, could well drain $200 million from the parent's pretax operating income.

Even for the tech giant, that's a serious hit. It's about as much as all the operating income Intel reported under generally accepted accounting principles for the second quarter and about one-fifth of such income calculated under the pro forma accounting the company prefers. And it's a reversal from the $716 million profits the unit contributed during the same quarter a year ago, as well as a huge comedown from the heyday of the tech bubble in 2000, when the unit had a record $2.1 billion windfall.

Worse may come. Intel's VC portfolio is valued on its balance sheet at $2.7 billion, as of June 30. Just a small portion of that, $726 million, is in quoted securities, which may be revalued at market prices each quarter. The other $2 billion is investments in private companies, mostly still carried at their original cost. That private pool could turn into an ocean of red ink if sales by Intel or others set a lower reference price for the companies.

Faced with such prospects, most companies would quit VC. Already, many have: VC investments by corporations fell 90% in the first half of 2001, to $353 million, says PricewaterhouseCoopers.

Not Intel. Leslie L. Vadasz, president of Intel Capital, says Intel is trying to keep up the pace of its venture investments. But try as he might, he's not finding opportunities to invest even half the amount he did last year--$1.3 billion in 300 deals. The big problem: VC firms are offering few new deals to Intel. They're too busy with problems among their portfolio companies.

STRATEGIC MOTIVE. Intel can plow on regardless because it's not in VC primarily to make money. When it began to invest a decade ago, it was looking for companies that were developing products that could help it with design and manufacturing. By the late 1990s, Vadasz was putting money into everything from PC and server manufacturers to the Internet portals--companies that could fan demand for products and services using Intel chips. "This was, is, and continues to be a strategic program," says Vadasz.

Intel's own assessments of its investments differ from conventional VCs. Consider the now-defunct eToys. Although the e-tailer collapsed, Intel cites it as a successful venture because its dynamic Web site encouraged others to improve their sites and increase demand for computing. Intel nearly tripled its $4 million investment in eToys by selling part of its holdings before the company failed. Profits from the deal were a fraction of the $1 billion it made from Micron Technology Inc. (MU ) But to Intel's thinking, Micron was a strategic flop because it didn't advance the production of random access memory as Intel had wanted.

With the tech sector depressed, Intel may now have to curb its appetite for VC investments. Although the stock market might accept a few down quarters in VC results, its patience is unlikely to be unlimited.

By David Henry in New York

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