WHAT IS GOING ON IN VENTURE 2025?
Part I: The State of Venture Liquidity
THESIS: The venture market in 2025 is defined by liquidity — or the lack of it. While headline investment totals look healthy, deal counts remain depressed, fundraising takes longer, and exits are scarce. IPOs and M&A are finally showing signs of life, but they are nowhere near the levels needed to unclog the system and recycle capital. A backlog of unicorns and cautious LPs means that, despite rising valuations for a select few winners, the broader venture ecosystem is still stuck in a slow-motion liquidity crunch. Despite all of this, there is still opportunity.
A month ago, Refinery Ventures held its annual meeting, where I presented our state of the market report. Every year, I end up saying something along the lines of “well, you know, it’s a pretty wonky market for XYZ reason”. And the interesting thing, is that every year, the “XYZ” on the surface seems to be pretty different from year to year, but at the root is almost always just about one thing: Liquidity.
That wasn’t any different for this this past year.
The market is currently wonky because we have a K-Shaped economy in venture. If you are building a certain kind of startup and are at a certain stage doing so, you are in for a windfall of money and a superb valuation. And if you aren’t doing that, you are definitely on the outside looking in, asking where all of the money has gone. As far as bubbles go, it’s pretty brutal out there.
As a part of my presentation, I shared several charts that illustrate this point. The first is just the general purpose deal count & value aggregator from each quarter going back to 2020:
First, if we look at the actual deal flow metrics (deal count and deal value), it would appear at first blush that the last 9 months, ending in June 2025, have been pretty good to venture. Total amount invested into the asset class has gone up pretty significantly, almost close to 2021 levels. But even if you only look at deal count, you will realize that the decline we saw post-ZIRP is still on a steady trend downward. It’s really only aggregate deal value that looks like it is bouncing back. Meaning that there are fewer names getting funded, but with more capital.
RATE OF DEALS IS STILL LOW
And if you peel that back one more layer, you see that the number of deals really has declined. Just not as many deals are being done. Volume has really fallen off a cliff, especially this year.
MULTIPLES APPEAR BACK ON TRACK
And valuations are doing pretty well, but really only if you are one of the top performers. According to data from Jamin Ball’s Clouded Judgement, the highest growth companies are the ones that are garnering elevated valuations, compared to the last three years. Mid and Low growth companies are pretty flat to slightly up. And while this data is just public cloud companies, it’s a decent proxy for the private markets as well.
So if it feels like valuations are up, that is somewhat true, but mostly just for top performers (or those with an AI story to tell). This is true for the public AND private markets.
Interestingly, where the data really skews upward for private markets is at two disparate rounds: Pre-Seed & Series C. They represent the largest jumps in valuations since the year prior on a percentage basis, and appear to be climbing.
Pre-seed is likely doing so well because most companies being started today are “AI-native”, building on the bleeding edge of LLMs and Agents (or at least appear to be). It’s a hot space, and the limited supply has cause a euphoric and competitive market.
On the flip side, the uptick experienced by Series C companies is likely because companies at this stage that are raising were either AI companies well positioned to capitalize on the current wave of AI post-2022, OR, they were the companies that made it through the slog of 2023 and 2024 and are actually accelerating. Otherwise, they wouldn’t be raising. However, if that were true, I would be curious why Series B isn’t seeing as much of an uptick as well.
ON AVERAGE IT STILL TAKES A WHILE TO FUNDRAISE
And despite these growth rates in valuations, it’s still taking a lot longer to raise your next round. It appears that actually raising a Series B is one of the hardest rounds to raise, which could contribute to the lack of valuation increases as well. Regardless of the reasons, the fact remains: it is taking a lot longer for companies to raise rounds. In frothy markets, that is not a typical occurrence. You would expect the opposite to be true.
It is entirely possible that companies are just seeking less capital in general. More on that in Part II.
STARTUP EXITS REMAIN ELUSIVE
We just don’t have enough liquidity in the venture ecosystem to help spin the capital flywheel. Yes, IPOs have started to tick back up, with the most notable trial balloons being Figma and Klarna. Figma had a huge pop (up almost 250% on first day of trading from listing price), but has since come back down to Earth. It is still, however, priced well above listing. Klarna just went public last week, and it didn’t really have the pop that Figma experienced, but it was still what you would consider a successful IPO (although, how does one really define successful IPO these days, other than just being able to go public at all).
[One note: not sure I really consider Klarna a “tech” company. My problem with fintech as a category is that it is really just financial services with a good tech wrapper. Klarna may be an exception to this rule, but I don’t know it well enough. Similarly, Stubhub went public yesterday (9/17) and had a mediocre debut. Again, somewhat cautious to call that a tech company PLUS the metrics for StubHub are nowhere near as good as Figma’s. It’s losing money & not growing all that much.]
In general, the market has seen 153 IPOs in 2025, compared to the 150 we saw all of 2024, well ahead of pace, but way behind 397 in 2021 (peak ZIRP) and 230 in 2019. On top of that, S-1 filings, the first step in going public, are up 29% year over year. So the market is feeling more confident than it did last year regarding tech public listings. “IPO pops” (first-day gains) are near 10-year highs, with tech leading that surge. Compared to 2021, we are still well below those marks. Which is probably a good thing – 2021 was mania. But compared to 2019, IPOs are still below trend. Today, we have had 25-30 IPOs in tech in 2025 and in 2019 we had 70-80. The market could still catch up to that pace, but we would need an extremely active back half of the year (not out of the question).
It's important to note that Figma (and in some ways Klarna) is the golden child of the VC-backed tech company set. Its metrics were all incredible – profitable, growing quickly, big market, etc. So, if that IPO didn’t go well, we would have been in a lot of trouble. But that’s not what happened. It was one of the surest trial balloons we could have had.
However, it’s not like Figma is the only company with strong fundamentals. There are a lot of companies that are allegedly doing just as well as they are, but it’s hard to know since we are talking about private companies who don’t report with the robustness that public companies do, if at all. There are a lot of companies that could possibly go public and continue to test the waters. Some notable ones are Stripe, Databricks, Discord, Skydio, and Plaid. They have been well-known hyperscalers that are well funded and check a lot of the boxes that excite public investors. But for a variety of reasons, they have still held off.
But if the Figma trial balloon was so successful, wouldn’t it stand to measure that more companies would come forward to want to go public? Well, there are several companies that have filed S-1s since Figma’s IPO. I am not saying that these companies waited to see how Figma’s IPO went and then decided to go public as a result, but I am certain that the bankers working with these companies paid very close attention to the results and updated their models accordingly.
IT’S NOT JUST ABOUT IPOS
IPOs are incredibly important, but most venture-backed companies don’t go public. IPOs do help price expectations because new technologies are able to provide adequate comps. Before Salesforce went public, how could you know how to value a cloud-based SaaS business in a private transaction? But these comps are only partially responsible for price-setting. Especially in strategic acquisitions, which care less about multiples and are more financially incentivized.
Beyond IPOs, M&A activity has accelerated in 2025. Deal counts are up sharply compared to 2024 and 2023, which were already somewhat elevated versus pre-pandemic years, though still well below the 2021 peak. A significant share of these transactions are driven by liquidity pressure (startups running out of cash or preferring a sale over a down-round) unlike 2019–2020 when fundraising was abundant. And while the number of deals is rising, the picture is mixed: many are smaller, distressed tuck-ins, but there are also true strategic megadeals (Google–Wiz, HPE–Juniper) that lift aggregate deal value.
UNICORN BACKLOG
Ultimately, the biggest question is what are we going to do with all of these unicorns? Yes, 2021 created a lot of them, because of ZIRP era financing activity & prices, so a question remains how many of them are actually worth >$1 billion. But that doesn’t change the fact that a lot of money is tied up in these companies by investors who are going to continue to look for a palatable exit. Many of them are good companies, just trying to grow into their valuations, and many of them could go public in a more stable economic environment / IPO window. This lack of liquidity here is a problem – a lot of potential VCs & LPs who would be deploying capital into the market in 2025 are sitting around waiting for some cash (or a better DPI story) to do so.












Great read Peter! Looking forward to Part 2🙌🏼