Venture investors keep selling their winners too soon

Doug Leone, the longtime investor at Sequoia Capital, the Silicon Valley venture firm, was explaining why failures never surprise him. He plans for them. What surprises him are the winners, which keep climbing long after he expects them to stop. Then he put it bluntly: “Every venture investor has made the mistake of selling too early their winners.”

Leone argues that great companies keep compounding for 20 years after they sell shares to the public. He points to Nvidia, Google, and Meta as examples. If a company has what he calls an unbounded market, a market that can keep growing for years, he believes an investor should never sell a share, because a $5 billion gain can compound into a $50 billion gain.

He learned it from his own sales. Sequoia once owned about a quarter of Cisco Systems, the company that makes the equipment that connects computer networks, and sold it for a total gain of $90 million. At its height, he said, Cisco might have been worth half a trillion dollars. He said he also pushed to distribute shares of ServiceNow, the software company, and left what became a 10-fold gain on the table.

Sequoia has held pieces of giants. Leone said the firm owned about 20 percent of Nvidia, 10 percent of Google, was the first investor in Apple, and owned a quarter of Cisco. Yet he said even Sequoia sold too soon.

Leone adds one limit to his own rule. Most companies held too long disappear, he argues, because technology changes. But if an investor can identify the five or 10 companies in a cycle that can keep growing and hold them for 20 or 25 years, he said, “there’s really nothing else you have to do in life.”