SAN FRANCISCO (Business 2.0 Magazine) - To everything there is a season - including venture capital.

In the venture-capital world, fundraising runs in three-year cycles, when VC firms hit up pension funds, corporations, and other investors to provide them the cash they dole out to startups. The current season, which began late in 2003, has been a return to sanity after the 1999-2001 cycle, which saw VCs salt away a staggering $200 billion - almost three times their normal take for a fundraising cycle. The industry then watched in horror as the cash-rich startups funded from those monies spent the money like drunken sailors.

During this cycle most venture-capital firms have cut the size of the fund they're raising in half. Top firms like Sequoia Capital, Kleiner Perkins Caufield & Byers, and New Enterprise Associates have all raised considerably smaller funds. The current season is expected to end this year with a modest $70 billion raised.

The big exception among top-tier VC firms, however, is Norwest Venture Partners, which has just raised a $650 million fund - that's two-thirds bigger than the $400 million fund it raised in 2001, and its largest fund ever.

Norwest typically invests in young startups, so the size of its fund reflects a growing optimism among investors that opportunities to invest in fast-growing companies are getting better. But Norwest plans to make more investments, not larger ones - a strategy that, if the firm sticks to it, should prevent a repeat of the bubble in which the valuations of startups grew out of control, hindering VC's ability to make money.

When the current season began in 2003, the watchword for venture capitalists burned by the bubble was "discipline" - and they practiced it almost to a fault. Though managers of pension funds and university endowments were asking to invest more, top-tier VCs wouldn't expand the size of their funds to let them in. It was prudent behavior, to be sure, but it also belied an uncertainty about the future of startups, and about the economic environment in which these young companies were setting off to do business.

"The reason we could rationalize a larger fund [now] is that the general environment is much better now," says Promod Haque, managing partner of Norwest, which is based in Palo Alto, Calif. Norwest's sole investor is Wells Fargo, the San Francisco-based financial services giant, which evidently shares Haque's optimism.

But does Norwest's move presage a new wave of billion-dollar VC funds that could crash like the last bubble?

Mark Heesen, who's seen several fundraising cycles in his 15 years as president of the National Venture Capital Association, doubts that a bubble's in the offing. For one thing, he notes that Norwest has always been a disciplined firm when it comes to fundraising. In his view, Norwest's decision has more to do with its ambition to move up a "weight class" among its peers. The size of a fund ultimately dictates the size and number of investments its partners can pursue; a larger fund can invest in more startups. (Norwest added five partners to its ranks in conjunction with the fundraising.)

"This is a firm that thinks it has the experience, the people and the mettle to invest a larger fund," Heesen says. "If we had 60 firms suddenly going out and raising a lot more than they have in the past I might be worried, but not with Norwest."

Norwest has a good track record of backing winners like Cerent, a networking-equipment company which Cisco (Research) bought for $6.9 billion in 1999 - an investment Haque led. Other startups the firm has backed include PeopleSoft, bought by Oracle (Research), and Tivoli, bought by IBM (Research).

Haque sees a huge opportunity in software-as-a-service companies like Salesforce.com, especially as corporations adopt software delivered over the Web.

"It's a huge shift from the traditional enterprise application approach," Haque says. "We saw a similar change when client-server architecture came on, back in the early '90s." With its new fund Norwest will also expand its investments in India, Israel, and China.

But none of these investments will be large - they can't be if Norwest hopes to make any money. That's because today, it's far more common for startups to get acquired than to go public. It's a reliable way to make money on a startup, although it leads to a less lucrative outcome for investors and entrepreneurs. So it's less likely that venture capitalists will make a killing on a multibillion-dollar IPO like Google's (Research) - the sort of financial windfall that can subsidize other failed investments.

Startups typically sell for a mere $100 million to $200 million, so venture capitalists must spend much less than that on them in order to return money to their investors, Haque says.

"If you are a software company, we expect to invest $20 million to get to cash-flow breakeven, $30 million if you are a hardware company," he says. "No more."

Even at the most optimistic of venture-capital firms today, discipline still reigns.