Venture Liquidity, AI Bubbles, and the IPO Reality Check

Every month, it feels like we could save ourselves time by copy-pasting “is anyone else seeing this?!” under every new Bog AI fundraising announcement. However, the themes dominating venture capital right now—post-bubble liquidity fallout and the brewing AI bubble—are too significant, too consequential, and frankly too complex to ignore.

Last month, we teased these ideas. This month, the evidence is piling up, and even some of the big data shops are starting to poke holes in the “bad market conditions” excuse that’s been thrown around for years. Let’s unpack what’s actually happening.

IPOs: Participation Trophies for VCs

Pitchbook’s 2025 IPO report painted a fascinating picture of the venture-backed IPO market:

1. Public appetite is real. Of the 18 VC-backed IPOs so far this year, 10 were north of $1 billion, averaging $7.3 billion. Add in 17 M&A exits over $1 billion (averaging $4.25 billion), and it’s clear the public market still wants mature tech.

2. Late-stage VCs are bleeding. Investors who came in during the final one to three rounds have lost money on all but two IPOs. Even those who entered slightly earlier rarely made more than 3x.

3. Losses at scale. Chime’s June IPO at $9.8 billion was objectively a success. Yet investors from its 2021 $25 billion private round took a staggering 60% hit.

   
   

Here’s the irony: demand for good companies is strong, but the number of listings remains tiny. The “bad market” narrative doesn’t hold up when U.S. indexes are at all-time highs and the S&P’s P/E ratio is hovering around 30x—a level seen only during the dot-com peak or recovery years.

The real questions aren’t “when will conditions improve?” but rather: how much better can it get—and how likely is it to get worse? At extremes like these, the risk of correction looms larger than any promise of easier exits.

The AI Bubble: Same Game, Higher Stakes

Just as late-stage investors lick their wounds from Bubble 1.0, Bubble 2.0 is already inflating. This time, the center of gravity is AI, and the strategy is even riskier.

Four years ago, mega-funds tried a high-stakes “spray and pray” with over 1,000 deals in a single year. Today, they’ve swung to the opposite extreme: hyper-concentrated bets on a handful of AI companies at nosebleed valuations.

As Euclid Ventures points out, the herd mentality is unmistakable. Pedigreed founder? YC batch darling? A mega-fund planting its flag? Everyone piles in. Demand dictates quality; quality dictates price. It’s groupthink at its most dangerous, and it mirrors the flawed logic that fueled the last bubble.

Are we really betting the endgame is $3 trillion IPOs for companies burning billions annually? Or is this just self-deception dressed up as strategy?

As one commentator put it, “This time it will be different” is the last bastion of both poor logic and bold showmanship.

Final Thought

The first loss is often the best loss. For late-stage managers who are underwater on 2021 vintages, holding out for valuations that aren’t forthcoming is a dangerous game. And as the AI bubble inflates, the déjà vu is impossible to ignore.

We’ve gone long here, but this is only scratching the surface. For a deeper dive into venture liquidity, the AI bubble, and what it all means for founders and investors, read the full September newsletter here.