Venture Capital and Silicon Valley: The Money That Made the Industry

Zusammenfassung

Venture capital — equity investment in early-stage companies with high risk and potentially unlimited return — was not invented for the technology industry, but it shaped the technology industry more profoundly than any other financial innovation. The particular form it took in Silicon Valley: staged funding, active board involvement, portfolio diversification across many bets expecting most to fail — emerged from a small number of investors in the 1960s and 1970s who were betting on semiconductor companies and their founders. The model they developed funded Intel, Apple, Genentech, Cisco, Google, and Facebook. It also created a selection mechanism that determined which technologies survived, which founders got backing, and which visions of computing became real. Understanding what venture capital rewards and what it ignores is understanding why the technology industry developed the shape it did.

Before Venture Capital: The Innovation Financing Problem

Technological innovation in the early 20th century was financed either by large corporations (Bell Labs, IBM Research, Xerox PARC) or by individual wealthy patrons. Both models had structural limitations: corporations funded research that served their existing business models; wealthy patrons were idiosyncratic in their interests and limited in number.

The research university provided a partial answer — MIT, Stanford, Caltech, and Berkeley generated technical knowledge that corporations could commercialize — but the gap between laboratory discovery and commercial product was the “Valley of Death” that neither universities nor corporations reliably bridged. A researcher with a working prototype and no path to market faced the same financing problem as any small business: conventional lenders required collateral and operating history that startups could not provide; equity sale to public markets required a track record that didn’t exist.

The solution required investors willing to accept illiquidity (they couldn’t sell their stakes immediately), high failure rates (most startups would fail), and patience (returns might come a decade later) — in exchange for the occasional extraordinary success.

Arthur Rock and the Semiconductor Investors

Arthur Rock is credited as the inventor of venture capital in its modern form. In 1957, he was a junior analyst at Hayden Stone & Company when William Shockley (inventor of the transistor, laureate of the 1956 Nobel Prize in Physics) called seeking investors for his new semiconductor company. Rock circulated Shockley’s letter among his network — and instead of investors for Shockley, he found eight engineers who wanted to leave Shockley Semiconductor and start their own company.

Rock helped the Traitorous Eight — including Robert Noyce and Gordon Moore — find funding from Fairchild Camera and Instrument Corporation, which took equity in the new Fairchild Semiconductor in exchange for startup capital. The structure was novel: an outside corporation investing equity (not debt) in a startup, taking ownership rather than a loan.

Fairchild Semiconductor’s success — it produced the planar integrated circuit and became the most innovative semiconductor company of the 1960s — validated the model. Rock moved to San Francisco in 1961, raised a $5 million investment fund, and became the first professional venture capitalist in Silicon Valley. His investments included Intel (1968, $2.5 million for 25% of the company, a stake worth billions a decade later) and Teledyne, Scientific Data Systems, and Apple (1978, $57,000 investment worth $14 million at the IPO).

Equity for Equity’s Sake

The defining feature of Rock’s model was that investors received equity — ownership — in exchange for their capital, not debt. This meant investors only made money if the company succeeded and its value grew. It aligned investor and founder incentives perfectly: both profited from the company’s success; both lost their investment if it failed. It also meant investors accepted that most of their portfolio companies would fail — the expected return came from the rare extraordinary success, not from average performance across the portfolio. Rock described his approach as “back the jockey, not the horse” — betting on the quality of the management team rather than the specific business plan, which would certainly change.

Kleiner Perkins and Sequoia: The Partnership Model

The institutional venture capital firm — a partnership raising capital from limited partners (pension funds, university endowments, wealthy individuals) and deploying it across a portfolio of startup investments — was pioneered in Silicon Valley in the early 1970s.

Kleiner, Perkins, Caufield & Byers (KPCB, 1972), founded by Eugene Kleiner (one of the Traitorous Eight) and Tom Perkins, raised a $8 million fund and invested in semiconductor companies, medical devices, and early software companies. KPCB’s most consequential investment was Genentech (1976), the first biotechnology company: $100,000 for a share of the company founded by Robert Swanson and Herbert Boyer to commercialize recombinant DNA technology. KPCB’s Genentech investment returned over 1,000 times the invested capital.

Sequoia Capital (1972), founded by Don Valentine, also emerged from the Fairchild network. Valentine had been a sales manager at Fairchild and then national sales director at National Semiconductor; his understanding of technology markets was commercial rather than engineering. His 1977 investment in Apple Computer — $150,000 for a stake in Jobs and Wozniak’s company — returned over $200 million. Sequoia later invested in Oracle, Cisco, Yahoo!, Google, and YouTube.

These two firms established the institutional venture capital model that dominated Silicon Valley for decades: a small partnership of experienced technologists-turned-investors, raising closed-end funds with 10-year lives, taking board seats in portfolio companies and providing strategic guidance alongside capital, and making a concentrated number of large bets expecting to generate most returns from one or two exceptional outcomes.

The Returns Structure and Its Consequences

Venture capital economics are driven by a power law: in a typical fund, half the companies fail entirely, a quarter return roughly the invested capital, and the remaining quarter generate all the returns — concentrated heavily in one or two exceptional outcomes. A fund that invests in 30 companies might return 90% of its capital from two companies.

This structure has profound consequences for what gets funded:

Big markets are necessary. A company that could become worth $500 million is uninteresting to a large venture fund — the return, even if the investment succeeds, is insufficient to matter at the portfolio level. Investors systematically prefer companies targeting markets worth billions. This creates systematic underfunding of companies solving important problems in small markets.

Growth over profitability. A company that achieves immediate profitability at small scale is less interesting than a company that loses money quickly while growing revenue 3× per year. Venture economics reward the latter because the expected value of “explosive growth” is higher than “steady profitability.” The preference for growth over near-term profit was a structural feature of the model long before the zero-interest-rate environment of the 2010s amplified it.

Founder founder-friendly terms evolved over time. Early venture investments (1970s–80s) often gave investors extensive control: preference liquidation stacks, board majorities, anti-dilution provisions. The competitive market for deals in the 1990s and 2000s shifted leverage toward founders, culminating in the dual-class share structures at Google (2004), Facebook (2012), and Snap (2017), which gave founders permanent voting control over public companies.

Andreessen Horowitz and the Institutionalization

Andreessen Horowitz (a16z, 2009), founded by Marc Andreessen (creator of Mosaic, co-founder of Netscape) and Ben Horowitz, represented a deliberate challenge to the existing venture model. Where traditional VCs were investors who offered strategic advice, a16z built a platform of full-time operational specialists — a recruiting firm, a marketing agency, a business development team, an executive network — that portfolio companies could access. The argument was that in a competitive market for deals, the value of a venture investor was the platform, not the capital alone.

Andreessen’s 2011 essay “Software Is Eating the World” articulated the firm’s investment thesis: every major industry would be disrupted by software companies, and the defensible position was to fund the disruption rather than the disrupted. a16z invested in Facebook, Twitter, Airbnb, Lyft, GitHub (acquired by Microsoft for $7.5 billion), Coinbase, and Stripe — a portfolio that spanned social media, transportation, developer tools, cryptocurrency, and financial infrastructure.

By the early 2020s, a16z had raised over $35 billion in committed capital across multiple funds, including crypto-focused, bio-focused, and American dynamism funds. Its scale had shifted it from a venture firm toward a more general asset manager, a shift that attracted criticism from investors who saw the GP-LP relationship as incompatible with management of multiple distinct asset classes simultaneously.

Dead End: The Bubble and the Selection Filter

Venture capital creates selection pressure on which technologies survive to scale. This pressure has been consequential in ways both constructive and problematic.

The constructive case: venture-backed companies are forced to demonstrate market traction, build operational discipline, and compete for talent against well-funded rivals. The resulting companies — those that survive and IPO — are genuinely capable organizations.

The problematic case: venture returns depend on a small number of companies achieving massive scale, which means the selection filter systematically favors winner-take-all market structures, network-effects businesses that exclude competition once established, and growth models that externalizes costs to labor, regulation, or public infrastructure. The social media companies and gig-economy platforms that venture capital disproportionately funded created externalities — privacy erosion, labor precarity, algorithmic manipulation — that the investment model had no mechanism to price.

The dot-com crash (2000–2002) wiped out approximately $5 trillion in market capitalization and destroyed hundreds of venture-backed companies that had been funded on metrics — page views, “eyeballs” — that were not connected to sustainable business models. The surviving companies (Amazon, Google, Salesforce) were those that had developed genuine unit economics underneath the growth. The crash functioned as a harsh filter that killed companies whose business models were theoretical; it did not change the structural incentives that produced them.