Anyone who tries to evoke the insanity of the technology bubble today -- remember Webvan? -- might launch into superlatives. The words "unreal," "extreme" and "titanic" come up -- even "unhinged." How else to describe the phenomenon of the world's smartest people throwing gobs of money at imaginative leaps of faith in the Internet and telecommunications sectors? A mass delusion of epic proportions? Investors surely were guilty all around of aiding and abetting the craziness in some fashion.

"There were no innocents," one venture capitalist allows. By the time the stock markets collapsed in spring 2000, the venture capital industry had been engulfed by the biggest bubble in its 30-year history.

If the tech boom was a test of venture capitalists' investment discipline, its collapse, combined with the impact of Sept. 11, left the industry in tatters. VCs took some time to recoup as the effects of the downturn sunk in. The next several years turned Darwinian as capital commitments dried up. Just when the VC industry was returning to some semblance of normalcy, with signs of a rebound in exits through initial public offerings in 2007, the world changed yet again.

Call it a protracted hangover. Until the tech bust, VCs had never had it so good. Things haven't been the same since. Returns have been mediocre, at best. To be sure, what transpired during the tech boom was an extraordinarily potent combination of transformational information and communications technologies and unprecedented liquidity channeled into it. Something like that comes maybe once in a lifetime.

Everyone agrees that for the industry to regain its shine changes need to happen, but exactly what changes ought to occur aren't clear. While the industry itself continues to shrink, and that's not necessarily bad, size isn't the only issue. The path to liquidity has been a challenge for VCs and for limited partners that pin their hopes on the asset class for above-average returns. And so the debate rages: Has the industry right-sized enough? Are VCs building enough transformative companies? Which is more important?

"An era of lost promises and high expectations" is how Kevin Delbridge, the senior managing director of Boston investment advisory firm HarbourVest Partners LLC, speaks of the past decade. "Looking back, there's been truly a tremendous amount of promise, but it's also been a decade when returns have not matched expectations. There's been a mismatch industrywide of putting portfolio companies on the books and then getting liquidity out of VC portfolios."

LPs may be unhappy today over the industry's woes but there was a time in the mid- to late '90s, when VC returns outperformed all asset classes. Globally, vintage 1995 funds' median internal rate of return, as reported by the sample funds Dec. 31, 2008, was 51.7%, according to London industry research firm Preqin Ltd. The 1996 median was 23.6%, and it was 35.9% for 1997. (The sample funds are few, but they serve as a reasonable proxy for the industry as a whole.)

In the run-up to the peak, VCs kept rewarding investors with better and better returns. One vintage '98 fund posted an IRR of more than 1,025%. Investors tend to move lemming-like historically. By 1999, new capital commitments to venture capital had quintupled, to $56 billion, from about $10 billion in 1995, according to Thomson Reuters. That number peaked in 2000, when VC funds collected more than $100 billion.

VCs appeared to be putting as much money to work as they received in commitments from LPs in a single year. Venture investors deployed roughly $52 billion in 1999, and $102 billion in the following year.

To say that it was a time unlike any others doesn't quite do it justice. "A venture capitalist would put $10 million into a company at a post-money valuation of $20 [million] or $30 million. Six months later, they'd be selling the company at a $300 [million] to $400 million valuation, and in the aftermarket, it would be $800 million," recalls Sanford Robertson, founder of technology bank Robertson Stephens & Co. LLC and a founding partner of San Francisco buyout firm Francisco Partners. "It was bizarre."

Robertson says as early as spring 1999 he had warned the industry was getting overheated: "I said the stock market would go down to 10% of where it was. Then markets got frothier still."

In 1999, there were 269 venture-backed IPOs in the U.S., raising more than $21 billion, according to Thomson Reuters. In 2000, although IPOs were fewer, the takings were greater.

For a poster child, look no further than Webvan Group Inc., then the second-largest online retailer after Amazon.com Inc., which venture firms led by Sequoia Capital and Benchmark Capital took public in November 1999. Goldman, Sachs & Co. led an army of investment banks, including Robertson Stephens, to tout a company with projected annual revenue of $11.9 million and a loss of $73.8 million. Coordinated by George Shaheen, former chief of Accenture Ltd., the online grocer raised more than $1 billion from a phalanx of VCs, then spent it on a highly ambitious distribution system with warehouses and trucks, only to realize that it couldn't quite figure out how to make money out of the scheme.

"Valuations were insane," recollects Dixon Doll, co-founder and general partner at DCM of Menlo Park, Calif. As his firm was poised to take telecom equipment maker Foundry Networks Inc. and connectivity provider Internap Network Services Corp. public in September 1999, a banker bet him that their stocks, which priced at $20 apiece, would trade at $100. "I thought I couldn't lose, but I did," he says of his own Webvan moment.

Investment banks can fairly take part of the blame. In 1999 Robertson Stephens was a bookrunner on 44 IPOs in the U.S. and globally, and 27 IPOs in 2000, according to research firm Dealogic. Before parent FleetBoston Financial Corp. unceremoniously dispatched it into oblivion in 2002, the bank was among at least a dozen institutions that paid steep fines stemming from class actions relating to the allocation of about 300 tech issues.

When valuations unraveled in 2000 and companies began failing, things got ugly, and uglier still after Sept. 11. Of the 11,686 companies first funded between 1991 and 2000, about 18% failed, while a further 35% faded away quietly, according to Thomson Reuters.

Unlike the buyout boom, where the speculative momentum was fueled by leveraged finance, the tech boom was an equity bubble with the value placed very much on the equity itself, and stock investors may have longer memories than debt investors, suggests Charles van Horne, managing director at investment advisory firm Abbott Capital Management LLC. That might explain in part why the industry has taken awhile to recover.

By now everyone accepts what many had suspected all along: Funds that invested quickly in 1999 and early 2000 will be the industry's worst performers. "Judging from the returns, the managers during that period were penalized by the speed with which they deployed the capital," says Brien Smith, managing director at Neuberger Berman LLC.

In North America, vintage '99 funds are reporting a median IRR of minus 8.6%, or a return of 91 cents for every dollar invested, according to Preqin's analysis as of Dec. 31. Vintage 2000 funds are posting a minus 1% IRR, though that vintage should really be divided between those that started deploying capital before March and those that started investing after October, LPs say. The two groups will have totally different portfolio valuations, Smith says.

The true measure for both vintages, the LPs say, is getting capital back. "That's the only thing that counts, and there'll be few people who can stand up and raise their hands on that one," Smith says.

For most VCs the post-bubble years were traumatic. Sept. 11 was a tragic episode in U.S. history that shook everyone's confidence. VCs went from denial to a deep funk. Firms eventually purged portfolios of businesses with unsustainable models, but the process didn't happen overnight. "We just went through a cataclysmic change, and it took us awhile to recognize the realities," says Stephen Holmes, general partner at InterWest Partners LLC, a Menlo Park-based early-stage investor. VCs struggled in many cases whether to walk away or to stay with companies.

By 2002, the so-called denominator effect of the stock market collapse on LPs' asset allocations had also kicked in. Firms raising money that year scraped bottom. By contrast, fewer funds in 1995 raised more capital.

The average fund size was cut by half as the pace of investing slowed dramatically. Members of the "billion-dollar club" -- about 35 firms that raised $1 billion or more between 1999 and 2001 -- were pressured by their LPs to return capital. Atlas Venture of Waltham, Mass., pared back, as did Mohr Davidow Ventures and Kleiner Perkins Caufield & Byers, both of Menlo Park. Accel Partners of Palo Alto, Calif., initially attempted to split its $1.4 billion, mid-2000 fund into two, retaining uncalled commitments to use for a later period. But investors preferred not to be locked in, and it later acquiesced. New Enterprise Associates Inc. (see box) and VantagePoint Venture Partners were among the exceptions, though funds raised immediately after were smaller.

What followed from 2002 to 2004 was a "reasonably steady correction," says Clinton Harris, managing partner at investment advisory firm Grove Street Advisors LLC of Wellesley, Mass., whose firm first began advising pension funds such as California Public Employees' Retirement System in 1999. Those vintages "looked to be relatively good years" before the downturn, he adds.

But as markets slowly recovered, liquidity once again flowed like the Mississippi, this time coming from not just the U.S. or Europe.

The trouble is, while the supply was nothing like the levels seen during the tech bubble, there was still a lot of it. "We came out of the '99-2000 vintages thinking that the industry would shrink in the numbers of managers," says one LP. "But the funds that people felt were going concerns 10 years ago have continued to thrive."

Part of it is a function of the name brands' ability to perpetuate their franchise on the basis of their names, LPs say. An investor new to the asset class would no doubt be less discriminating than an existing investor. Generational shifts that should be occurring have also moved glacially. So a new crop of younger VCs are setting out on their own, hoping to reinvigorate early-stage investing. Look for the likes of Tugboat Ventures and Retro Venture Partners among the newer names.

The industry did shrink some post-tech bubble, and has shrunk again post-2007, though perhaps not as much as the industry might hope. If anything, despite the initial drop in numbers post-bubble, the industry got larger between 1998 and 2008. In 1998, there were 624 VC firms, according to Thomson Reuters. Last year, there were 882. Funds have risen numerically, to 1,366, from 1,085, with the average fund size now at $104.4 million, from $60.6 million in 1998.

Meanwhile larger, more established firms have branched out into "multiproducts," such as growth equity funds, clean technology funds, India and China funds. Groups like Sequoia and Kleiner Perkins say these are growth areas. Their LPs say it's also a way to boost the firms' management fee streams, depending on how the firms scale up, and create more carried interest pools so that losses in one fund will not adversely affect the carry pool in another. "The troubling thing for us is that, because of the lack of exits, fund managers are more than ever addicted to management fee streams," one LP bemoans.

A recent study by the Ewing Marion Kauffman Foundation looked at the VC industry's size and how it affects performance. Its conclusion: "A five-fold increase in venture capital commitments by limited partners led to a collapse in performance from which the sector has never recovered." The study suggests that LPs are expected to shrink their annual allocation to VC to about half the current $25 billion rate or higher.

Performance isn't just a function of the industry's size, however. It boils down to the lack of exits, investors say. And after the 1999 to early 2000 era, there hasn't been a robust market for IPOs, which have historically offered the multiple pop for venture investors.

For a time it seems, VCs and entrepreneurs were backing companies that simply filled out a product void that large public enterprises would likely acquire. Over a four-year period up to 2006, only five VC-backed IT companies achieved billion-dollar exit valuations: Google Inc. ($24.6 billion), Semiconductor Manufacturing International Corp. ($5 billion), Alibaba.com Corp. ($4 billion), Salesforce.com Inc. ($2 billion) and Skype Technologies SA ($2.6 billion).

At the same time, markets became risk-averse. People got burned badly, and memories are long. The multiples of companies such as Microsoft Corp., which serve as proxies for the tech industry, have fallen substantially. The global financial crisis hasn't helped.

Many blamed the Sarbanes-Oxley Act, passed in 2002 to overhaul the financial reporting requirements for public companies, saying it added layers of costs smaller companies find hard to absorb. Many venture-backed companies that went public prematurely have struggled with the higher costs of corporate governance, and their population is much smaller.

On the other hand, Neuberger Berman's Smith contends, "I'd venture to say that companies created post-SOX -- where those considerations are built in from day one -- are better prepared for it than others."

"In some years we may have had 50 to 60 venture-backed IPOs, but when the industry is funding 3,000 companies a year, that's not nearly enough liquidity for the industry to meet investors' expectations," says HarbourVest's Delbridge. "M&A transactions have helped, but the industry needs a vibrant IPO market."

Perhaps VCs are just as much to blame for not providing adequate foundations for sustainable businesses. "It's not about just getting your portfolio company public. Time and time again, these IPO'd companies have been poor after-market performers," says Jeff Cavins, CEO of venture-backed Internet telephony company CallWave Inc., which recently decided to take itself private again. "We really weren't enjoying the benefits of being a public company."

There are folks like Alan Salzman, CEO and managing partner at VantagePoint Venture Partners of San Bruno, Calif., who believe the real challenge for the industry is how to return to "creating more Googles and Ciscos." Venture groups should go back to their roots of finding and funding innovation that can transform industry sectors and society as a whole, much as the PC and biotech transformed the computer and pharmaceuticals industries, he argues. That transformation has run its course, he asserts, followed by the changes in data communications in the '90s. "But in the post-transformative period, there's been nothing truly exciting," Salzman adds.

Drawing from the lessons of previous tech cycles, Salzman's firm decided in 2002 to emphasize clean technology. "We define cleantech as innovation seeking to address the challenges of limited resources," he says. Vantage Point now boasts some 18 investment professionals focused on cleantech, its largest category. Admittedly, cleantech is in its early stages, despite increased capital available.

All this doesn't make the LPs' job of finding the next VC winners any easier these days. VCs are still trying to figure out how to extract the kind of returns that their LPs are looking for. In venture capital, there's an old saw that fewer than 20% of firms have provided the lion's share of returns historically. That hasn't been the case in recent years, because not many firms have shown consistency of returns.

"The real question is, can VCs make a case to LPs that they'll be rewarded going forward?" Salzman says. "If you can't make a case for the industry as to why we're going to generate horrific returns, then it's kind of hard to expect the LPs to make a case for you."