In 2022, auditors reportedly told venture GPs to compare notes to figure out how to value their portfolios. That’s the state of independent oversight, “…I don’t know, ask your neighbor?” There is a fish market paper that sheds some light on why this is a structural problem; here’s me trying to unpack the argument for making startup valuations public.

This is Networked Conviction 009, my weekly investing journal: one idea I’m still working out, one startup I’d want to see built, and a handful of things from my notes worth resurfacing. The first section is free for everyone; the rest is for paid subscribers.

Under Construction: Make Startup Valuations Public

Around the time of the valuation reset in early 2022, Dan Gray surfaced this from Beezer Clarkson: “I heard stories that last year, end of 2022, when the auditors were getting involved, they were telling GPs to talk to each other to try to figure out how to value things.”

When external oversight arrived, the answer pushed people towards consensus. Mark your portfolio at whatever everyone else marks theirs.

The idea I’ve been sitting on since January 2025 is rooted in a paper about fish markets that the Ramp team flagged for me when we collaborated on a piece back then. The paper? Robert Jensen’s The Digital Provide. The finding was that the more information a market has, the less price variance it shows. Between 1997 and 2001, mobile phone service rolled out across Kerala, India, and once fishermen and wholesalers had phones, price dispersion across coastal markets dropped dramatically, wasted catch disappeared entirely, and prices converged on the Law of One Price.

The uncomfortable extension for private markets is that we built an entire asset class on the assumption that price opacity is not just acceptable but structural. A startup valuation is set by one term sheet; you only need a single willing counterparty. Microsoft needs thousands of market participants to agree it’s worth $3 trillion. So what happens when that single counterparty is working from comparables that are also private?

You get marks-by-committee. This week Eric Fitzgerald described the late-stage venture market as “marks, not distributions.” The AI megadeal markups that drove late-stage returns in 2025 were largely on paper, not necessarily in cash. Those marks were set in the same information vacuum. The deeper question is who does that kind of opacity actually protect?

My working theory is that it protects everyone on a micro level but hurts the broader market (the macro level). It protects founders from downward pressure on their pricing, it protects GPs from marked-to-market accountability, and it protects LPs from the quarterly heartburn of watching the number go down. What it doesn’t protect is accurate price discovery, which is impactful for everyone’s long-term interest.

The question I keep coming back to: what breaks if startup valuations become public? Clearly no one is really incentivized to push for that. You get a taste of it with secondary markets where there’s a bit more liquidity but it almost never reflects actual price, just slight leading indicators of demand.

That’s something I’m still puzzling on.

Below, you can check out this week’s Request for Startup (knowing what you own), ten things from my wiki (five I added this week and five from the historical stacks), and a look back at the Unholy Trinity.