/ writing
Caveat Emptor
A point-in-time reassessment of my 2016 tech-market argument, focused on venture contraction, shared infrastructure, and the second-order effects of a funding reset.
Updated : Corrected copy errors, linked the funding figures that can be reproduced, and removed or qualified numerical claims that lacked a published model. The argument below remains a March 2023 perspective, not a current market forecast.
In 2016, I argued against the idea that technology was in a broad bubble, responding to warnings from Bill Gurley, Janet Yellen, Mark Suster, and others. My view then was that the technology economy was still expanding. By March 2023, I believed the conditions had changed enough to revisit that call.
Working inside the sector gave me a view of hiring plans, vendor choices, and the way venture-backed companies spend. My concern was not simply that startup valuations could fall. It was that a contraction in venture funding could travel through a shared operating stack.
The scale of the preceding expansion matters. Crunchbase projected that more than $1.5 trillion was invested in venture deals worldwide from 2010 through 2019. Its later year-end analysis estimated that global venture funding reached $681 billion in 2021, then fell 35% to $445 billion in 2022.
Private-market datasets use different definitions and are revised as deals are reported, so their totals do not line up perfectly. The direction is more useful than any single total: an extraordinary amount of capital entered venture-backed companies, followed by a sharp pullback.
That capital does not remain on a cap table. It pays for cloud infrastructure, software, hardware, offices, recruiting, and compensation. The PitchBook-NVCA Venture Monitor estimated that U.S. venture deal value reached $329.9 billion in 2021. My operating hypothesis in March 2023 was that when startups slowed hiring and controlled costs, the vendors and consumer businesses serving those companies and workers would lose part of their tailwind.
The shared technology stack makes second- and third-order effects plausible, not automatic. I previously attached a precise correlation to that relationship, but I do not have a linked, reproducible model that supports the figure, so I have removed it. Public data also does not establish how much venture funding ultimately became revenue for any individual platform company.
In my 2016 article I wrote, “I can tell you from my work, my co-workers’ work, that we haven’t missed a beat. We are all humming along just fine.” By March 2023, my view from inside the sector felt starkly different: the music had not merely slowed; it had stopped.
I cannot know the precise size or timing of a pullback from those observations alone. The defensible conclusion is narrower: a venture contraction can propagate beyond startup valuations into hiring, vendors, and consumer demand, and leverage can amplify the effects. That was the risk I believed too few people were pricing in at the time. Caveat emptor.