What Is Venture Capital? Betting on Ten Companies to Find the One

A venture capital firm raises a fund, typically a few hundred million dollars, from limited partners, which are pension funds, university endowments, insurers, sovereign wealth funds and rich families, who commit the money for about ten years. The firm's partners then invest it in perhaps twenty to forty companies over the first three or four years, take seats on their boards, help them hire and raise further money, and wait; the return comes when a company is bought by a larger one or sells its shares to the public in an initial offering, at which point the fund sells its stake and passes the proceeds to the limited partners, keeping about twenty percent of the profit as carried interest and charging about two percent a year of the fund as a management fee along the way. The firm's own money is a small share of the fund; its value is in choosing.

The returns of a venture fund do not average; they concentrate. Of the companies a fund backs, roughly half fail outright, a third return something like the money invested, and a handful return many multiples, and the fund's whole performance depends on whether one of those multiples is large enough. A single investment that returns a hundred times pays for a hundred failures, and the industry's records are of that kind: the few hundred thousand dollars put into Google in 1998 returned billions, and the fund that backed WhatsApp turned 60 million dollars into three billion in five years. That arithmetic shapes everything the investors do:

• •They look for companies that could become enormous, since a solid business that will be worth ten times its investment is, to a venture fund, a disappointment
• •They prefer markets that are large or could be, technologies that scale without proportionate cost, and founders who intend to take over an industry
• •They invest in rounds, seed, then series A, B and C, each at a higher valuation if the company is growing, so that the risk is staged and the failures are cut off early
• •They take a large enough stake, ten to twenty-five percent, that a win matters, and rights that protect them if the company is sold cheaply
• •They accept that they will be wrong most of the time, and are judged on the size of their rights

The whaling voyages of nineteenth-century New England were financed the same way, by agents who raised money from investors for ships that mostly returned a loss and occasionally a fortune, and the modern industry began after the Second World War, with the American Research and Development Corporation in Boston in 1946, whose 70,000-dollar stake in Digital Equipment became 355 million. Silicon Valley made it a system: the firms that set up on Sand Hill Road in Menlo Park in the 1970s, Kleiner Perkins and Sequoia among them, backed Apple, Genentech, Atari and Oracle, and the concentration of investors, engineers, lawyers and universities around Stanford became a machine for turning ideas into companies that no other region has fully copied. The dot-com boom of 1999 and its crash showed the model's failure mode, in which money chases valuations rather than companies, and the boom of the 2010s, in which funds grew to billions and companies stayed private for a decade, showed it again in 2022.

For the company, venture money is expensive in a way that debt is not. The founders sell part of their company at each round and end, at a public offering, owning perhaps ten or twenty percent of what they started; they take on investors who can replace them, and a board that expects growth at a rate that a merely profitable business cannot deliver. The bargain suits companies that need capital they cannot borrow and that can grow fast enough to justify it, software above all, and it suits badly the many businesses that would be better off growing slowly on their own revenue and are pushed instead toward a scale they cannot reach. The industry's critics point out that its incentives reward the swing for the fences, that its returns after fees have not, on average, beaten the stock market since the 1990s, and that its money has gone overwhelmingly to a few cities and a narrow kind of founder.

Venture-backed companies are a tiny share of businesses and a large share of the ones that changed things: by one count they account for about half the value of all American public companies founded since 1975 and most of the spending on research among them. The model has spread to China, India, Israel, Britain and Europe, with mixed results, and into biotechnology, where the ten-year fund and the ten-year drug trial fit awkwardly, and into climate technology, where the capital required is far larger than software's. Whether its recent scale, with funds of ten billion dollars and companies valued at a hundred billion before they earn a profit, is the model working or the model breaking is the argument of the moment; the underlying trade, a stake in a company that will probably fail for a share of the one that will not, is the same one that sent the whaleships out.

Venture capital is investment in young, unproven companies in exchange for shares, made by funds that raise money from pension funds, endowments and the wealthy for about ten years, back twenty to forty companies in staged rounds, and expect most to fail while one or two return enough to pay for all of them, which is why they seek only companies that could become enormous. It financed the companies that define the modern economy, it is expensive for founders and poorly suited to slow businesses, and its cycles of boom and crash follow from betting on a power law.