The departures of Intel Capital executives John Miner and Claude Leglise signaled to some that the chipmaker’s new CEO plans to make significant changes in the company’s venture capital unit. Add to that the announcement that Arvind Sodhani, Intel’s longtime treasurer, replaced Miner as president of Intel Capital, and gossip hounds had further evidence that changes were afoot.
But the official line from Intel is that it’s business as usual-even though Intel Capital’s portfolio has lost about 30% of its value in the past four years. Miner told VCJ that he doesn’t expect his successor to make any big changes. “Our portfolio performance has improved and our strategic performance has improved,” he says. “I’m very pleased with both of those. Our initiatives and strategy should be stable.”
The consensus from interviews with about 20 VCs who work or compete with Intel Capital is that the venture unit is too important and has been too successful for its parent to follow the current trend among technology companies to shutter or cut back their VC efforts.
Asked directly if Sodhani’s elevation means a new focus on the bottom line and a change in investing for the group, one Intel Capital insider says: “Its business as usual-so far.” Another says he has no fear that Sodhani will lead from the bottom line. Sodhani is well liked within the organization, and “he’s no bean counter,” says the source, an Intel Capital investment professional who spoke on condition of anonymity. “He’ll spend an hour telling you that he’s not an accountant.”
Others find it hard to believe that Intel won’t take a closer look at the bottom line. “You have to realize that corporate venturing has undergone enormous restructuring at all companies,” says the managing partner of a venture firm that has co-invested with Intel Capital. “Anyone who is a new executive, like [new Intel CEO Paul] Otelleni, will look at this activity hard.”
The VC points to the large and growing list of corporations that have sold off or shut down their VC arms recently. The list includes Boeing, Infineon, Applied materials, and Nokia. “Dell sold most of its portfolio assets to Lake Street Partners, Deutsche Telekom sold part of its portfolio to Cipio Partners, and Daimler Chrysler sold its portfolio to Cipio Partners,” the VC says. “Even Intel sold off part of its portfolio to Lake Street.” Given that context, it isn’t unreasonable to expect Intel to tighten its focus or spending.
Paul Otellini, who officially takes over as Intel’s CEO from Craig Barrett this month, is faced with shareholders who want to see a bump in the company’s lackluster stock price. Shareholders are counting on him to get Intel’s processor strategy back on track, make up for the lost years and billions of dollars expended on the Itanium, and to fend off the successful technology challenge from rival AMD. Intel’s venture unit may be a convenient place for the new CEO to make some quick changes.
Intel’s annual earnings statements show that the total value of Intel Capital’s portfolio has declined by about 30% over the past four years-from just over $1.7 billion in 2001 to $1.17 billion in 2004 (see chart, page 16). On one hand, the value of the shares Intel Capital holds in portfolio companies that have gone public has increased. On the other hand, the value of its so-called “non-marketable equity securities” (stakes in privately held companies) has fallen sharply-just as portfolio valuations across the venture industry plummeted after the dot-com bust.
An Intel Capital spokeswoman says the venture unit isn’t concerned about the decline. “The carrying value is not a measure of your potential exit or cash return,” says Laura Anderson. “As you know, valuations fluctuate as the overall economy does and impacts all VC firms across the board. However, because we’re a publicly traded company with a large portfolio and have an obligation to our shareholders, we report the carrying value of the portfolio and cash gains/losses for the organization. LLPs do not do this. If they did, you’d find that the valuation for their portfolios fluctuates along the same lines that ours does.”
Intel Capital has had a good run of exits in the past year. Three of its portfolio companies went public: CSR plc, which makes wireless chips; SiRF Technology Holdings Inc., which produces GPS semiconductor and software solutions, and Jamdat Mobile, a publisher of wireless gaming and other applications.
In addition, another eight of its portfolio companies have been acquired in the past 12 months, including Iridigm Display Corp., a MEMS-based display maker acquired by Qualcomm Inc. for about $170 million in cash; Musicmatch Inc., a personalized music software company bought by Yahoo for about $160 million in cash; OSA Technologies, a maker of firmware for embedded devices that was bought by Avocent Corp. for about $100 million in cash and stock; and E2O Communications Inc., an optical transceiver developer acquired by JDS Uniphase for about $60 million in cash.
Miner told VCJ that even though Intel Capital made a move three years ago to “change [its] performance measurement from the number of deals done to the quality of deals done,” that move hasn’t resulted in its number of deals falling significantly. Intel Capital did 110 deals last year, compared to 120 in 2003. Also, the amount it invested last year ($130 million) was on par with the $127 million it invested in 2003-when you factor out two large investments it did in 2003 that totaled $573 million.
Still, the venture unit isn’t investing at the crazed pace it set during the bubble years, when it pumped out $1 billion in 1999 and 2000. As recently as 2001, it put $350 million into 175 companies (see “Inside Intel Capital,” VCJ, April 2002).
To understand how Intel Capital might change, it’s useful to remember what the group has done to date. Since its founding in 1991, Intel Capital has invested about $4 billion in more than 1,000 companies.
Compared to traditional venture firms, Intel Capital has an army of investors. Among the 400 staffers it has scattered around the world are 200 investment professionals. “It’s an institution,” Leglise told VCJ. “It’s been around for 14 years, has over 200 investors and we’ve collectively built it into the preeminent corporate investor.”
That doesn’t mean its track record is spotless. In the tapering of the Internet bubble the Wall Street Journal reported on one quarter’s $350 million loss in Intel’s portfolio and the group’s projection of a further $280 million loss in the succeeding quarter. Despite those results, Les Vadasz, president of Intel Capital at the time, told the WSJ that the firm would continue its venture program as a way to keep pushing demand for advanced semiconductors, to keep abreast of new technologies and to improve operations such as chip manufacturing and design.
Vadasz told the WSJ that even some of the worst investments financially-for example, into online toy seller eToys and sports adventure site Quokka Sports-were strategic-because the idea was to help create rich Internet content that would spur demand for higher-end chips. “The losses hurt your ego, but strategic [results] have always been more important [than financial results].”
One VC says that while there were quarters when Intel lost money after the bubble burst, cumulatively the group has probably returned billions of dollars to its parent over its life. “People forget that before the bubble burst Intel Capital was returning so much money to the company that for a while they had to break the group’s earnings out in the annual reports because the numbers were so big.” Indeed, one of the often-repeated statements about the group is that “we’ve made enough money for Intel to pay for a new fab,” suggesting that the group has returned in the neighborhood of $2 billion to $3 billion.
More recently, Intel Capital has been perceived as being more disciplined. “In the past four years, they have been taking a much more goal-oriented approach and investing large sums into technologies that will further Intel sales,” says a longtime semiconductor investor. “Particular examples are 802.11 products and services and WiMax. I am not sure if the fund is profitable, but the strategic results for Intel appear to be strong.”
And, like the large corporate venture groups run by companies such as IBM, Cisco and Motorola, Intel Capital has had genuine global reach, recently investing as much as 60% of its annual total outside of the United States. In 2003, for example, Intel invested $600 million in Asia alone-not including investments in fabs-plus another $100 million to $200 million in Japanese startups. It invested in more than 100 companies across the region and in a wide variety of technology segments, including software, power, display controllers, IC design, 802.x and storage companies.
Despite Intel Capital’s perceived success as a VC arm, one investment banking analyst says, “Intel Capitaldid not do much of what should havebeen its primary role: tofund the next generation of semiconductor companies.” Those sorts of investments would have been a source of R&D and entrepreneurial talent for Intel in the way that Cisco has used acquisitions of Silicon Valley startups to supplementits R&D and management.
“The attitude of Intel Capital was that all other semiconductor companies were competitors and that it had nothing to learn from them,” the analyst says. The group’s thinking was that “it could put them out of business any time it wanted,so why should it help competitors? This reflected the insular attitude of Intel itself.”
A top-quartile venture firm told VCJ, “Intel Capital as a whole has had mixed success. Their mission is not the same as ours, which is to make money for our investors and create companies. Their reason for existence is to generate more demand for their products and also to monitor what may endanger their market position.”
Intel Capital at a Glance