ALL INPredictions
Last updated Nov 29, 2025
Prediction
Chamath
← Back to hostEpisode
Chamath @ 01:14:45Inconclusive
ventureeconomy
Of the roughly $1 trillion invested in venture capital from 2018 through 2022, approximately $500 billion from those vintages—and around $600–700 billion including older vintages—will ultimately be lost (i.e., not returned to investors) as the cycle plays out.
what it basically tells you is about $500 billion of that trillion from 1819, 20, 21 and 22 is going to be destroyed.... we're talking about a 600 or $700 billion destruction of paid in capital.View on YouTube
Explanation

It’s too early in the life of the 2018–2022 venture-capital vintages to know whether roughly half of the ~$1T invested will ultimately be lost.

Key points:

  1. Fund lifecycles are ~10–12 years. Most VC funds are structured with a 10–12 year term, often with extensions, and exits tend to cluster in the latter half of that period. That means true outcomes (DPI—cash back to investors) are usually only clear a decade or more after the vintage year. (en.wikipedia.org)
    2018 vintage funds are only ~7 years into their life; 2019–2022 vintages are even younger.

  2. 2018–2022 vintages are still mid‑cycle, not fully realized. Cambridge Associates’ 2023–2024 benchmark data show that vintages 2014–2022 make up the bulk of VC net asset value and are still being actively invested and written up/down. Calendar‑year 2024 returns for major VC vintages ranged from 0.7% (2018) to 25.3% (2022), and 2016–22 vintages together account for about 80% of index NAV—indicating that much value is still unrealized and subject to future exits and revaluations, not final losses. (cambridgeassociates.com)

  3. Distributions have been weak, but that reflects timing, not final loss. Cambridge Associates notes that since 2022, capital calls have exceeded distributions and distribution yields have been low for several years, reflecting a poor exit environment (slow IPOs, fewer M&A deals) rather than a completed loss of capital. (cambridgeassociates.com) This supports the idea that the "cycle" Chamath refers to is still in progress.

  4. The $1T premise is broadly right, but we lack loss estimates. Contemporary analyses note that VCs deployed $600–675B in 2021 alone and over $400B in 2022, easily pushing total global VC deployment over ~$1T across the 2018–2022 period. (forbes.com) However, no credible, comprehensive study yet estimates ultimate loss of paid‑in capital for those vintages; we mostly see mark‑downs, down rounds, and anecdotal high‑profile write‑offs, not a bottom‑line figure like “$500–700B permanently destroyed.”

Because (a) VC funds from 2018–2022 are still within a normal 10+ year realization window, (b) a large share of their NAV has not yet been exited, and (c) there is no reliable aggregate estimate of eventual capital loss versus paid‑in capital, Chamath’s specific quantitative prediction (that ~$500B from 2018–22, and ~$600–700B including older vintages, will ultimately be lost) cannot yet be validated or falsified. The correct classification today is therefore **“inconclusive (too early).”