US: More Early-Stage VC Funds are Targeting Fewer Startups

The number of new early-stage VC funds with a focus on US-based pre-seed, seed, and early-stage startups grew from 163 (vintage 2014) to 490 (vintage 2019) [1]. As I wrote two weeks ago in “Rapid Decline of New U.S.-based IT Startups”, the number of newly founded US startups in information technology declined from 5,856 (2014) to 1,293 (2019) [2].

In a nutshell: More early-stage VC funds are targeting fewer US early-stage startups.

The right graph with the sum of final close amounts by vintage might be misleading. We see an increase in the number of smaller, sub-$20 million funds; and an increase in fund size at some firms due to different reserve strategies; and some early-stage firms where their next fund is more of a multi-stage fund. I’m not sure what that graph means yet. Fewer early startup deaths? More startup deaths if follow-on funds are not growing at the same rate? Larger early-stage rounds (as I wrote about here)? Higher early-stage valuations? More down-rounds at Series B? More angry LPs in five years?

So What?

My worry as a fund manager: A perfect storm and critical mass of disenchanted LPs if a market correction should come in the next two to three years. Who knows what sort of regulations the VC market could see.

My worry for entrepreneurs: You will have to have the very best firms as investors, with the most flexible way of financing possible. Pricing discipline is hard if everyone around you seems to raise insane amounts of money for stupid ideas and with a mediocre team.

Update [September 2019]

I came across a long-form article — a 25-minute read! — at https://www.toptal.com/finance/venture-capital-consultants/state-of-venture-capital-industry-2019 with lots of additional data, such as:

Footnotes

[1] as of July 2019, per Prequin; vintage might be different from year of final close (See https://en.wikipedia.org/wiki/Vintage_year); final close amounts do not include amounts of funds that had a close but are still open (should be obvious, but I thought I point it out).

[2] as of July 2019, per PitchBook