The record VC funding year that's actually a drought

Every headline this month is saying the same thing: 2026 is on track to be the biggest year in venture history.

The number is real. US startups raised $412.7 billion in the first half of the year, per the PitchBook-NVCA Venture Monitor. That's already 29% more than all of 2025, which closed at $319.2 billion. We're on pace to break the 2021 record with two full quarters to spare.

Here's what that headline hides. If you're starting a company right now, almost none of that money is available to you. US startups raised a record $412.7 billion in the first half of 2026, but 87.5% of it went to megadeals of $100 million or more, which means the early-stage market where new companies actually get funded is the tightest it has been in a decade.

I've sat on both sides of this. I've been the founder raising at the small end of the market, and I've worked inside a venture firm watching where the checks actually go. And I keep landing on the same conclusion: venture doesn't have a capital problem. It has a supply problem. The mid-year data didn't soften that take. It sharpened it.

Where did the record $412.7 billion actually go?

Start with the shape of the year. Q1 came in at $268.9 billion. Q2 came in at $143.9 billion. That's not a plateau. Q2 nearly halved Q1.

The report frames Q2 as the second-highest quarter in a decade, which is true. It's also the wrong frame. The trend inside the record year is down, and it's down fast.

Now look at where the money actually went. Megadeals of $100 million or more took 87.5% of all dollars in the first half, per the PitchBook-NVCA data. Everything under $100 million, which is the entire market most founders live in, drew just $51.4 billion.

And that slice has been collapsing for three straight years.

Read that again. The part of venture that funds normal new companies went from nearly half the market to an eighth of it in two years. During what's on track to be the biggest fundraising year ever recorded.

A record and a drought at the same time. Both are true. Only one will make the headlines.

It's an AI story wearing a venture costume

If you want to know where the record actually lives, follow the concentration. AI took 86% of all US venture dollars in the first half (PitchBook). Q2 alone saw seven rounds of $1 billion or more, totaling $87.2 billion. Five of the seven were AI companies.

The biggest one tells the whole story by itself. Anthropic raised $65 billion at a 157% pre-money step-up, landing at a $965 billion post-money valuation. One round.

So let's be honest about what "the biggest venture year in history" means. If you're building a frontier model, this is the best fundraising market that has ever existed. If you're building anything else, the record year is happening to someone else.

Even the exits are a mirage

The exit story is getting the same celebratory treatment as the fundraising record. It deserves the same skepticism.

Yes, SpaceX went public at $1.7 trillion, the largest IPO ever. And yes, Anthropic and OpenAI have both confidentially filed for IPOs. Two more trillion-dollar exits are probably coming. The window is open. Right? ... Right? 

As Venture Monitor said themselves, "a few good exits do not make a great liquidity market." IPOs are still running well below the rates the report's own model predicts, and most investors and their LPs won't see distributions from them. The value is landing with a small circle of funds. Kingmaker exits are not a thaw. They're headlines.

This is not a capital problem. It's a supply problem.

Capital has never been more abundant, yet fundable new companies have never been scarcer.

Think about what happens when those two things are true at once. Too many dollars chase too few real opportunities. Entry prices go up. A 157% step-up stops being an outlier and starts being the market clearing price for anything that looks like a winner. Returns get harder at exactly the moment the checks get bigger. And investors respond the only way they can: by circling the same handful of names, bidding against each other for allocation in companies everyone already knows about.

That's the market the Q2 data describes. Not a boom. A shortage, with a bidding war on top of it.

The scarce asset in venture right now is not capital. It's a new company worth funding.

How do you win a concentrated venture market?

If the scarce thing in venture is fundable companies, the move is obvious and most of the market won't make it: stop bidding for companies and start manufacturing them.

That's the entire logic behind venture building, and it's why I think this cycle favors builders over scouts. A company created through corporate venture building starts life with things no amount of capital can buy at the seed stage: a real asset to build on, a first customer with an actual budget, distribution that took someone else decades to earn. It's purpose-built for a validated problem instead of discovered after the fact and marked up at auction.

You don't have to outbid anyone for a company that didn't exist until you built it. You set the entry price. You choose the problem. You start with an advantaged startup instead of paying a premium for the possibility of one.

In a market where 87.5% of dollars flow to the largest deals and everyone else fights over the remainder, that's not a nice-to-have. It's the only reliable way to get exposure to early-stage venture at a sane price.

The power law is steepening

Every venture cycle concentrates returns. This one is concentrating inputs too: the dollars, the deals, the attention, all stacking up behind a few dozen companies while the base of the pyramid thins out underneath them.

So when you're reading about record venture funding, take it for what it is — a story about the top of the market in pre-IPO companies.

The winners in this cycle won't be the investors bidding up the same handful of companies that already exist. They'll be the ones creating the companies that didn't exist six months ago.

So consider this a call to arms. There has never been a better time to build a company. The answer to a supply crisis is supply. We need more new companies, not more capital circling the ones that already exist. That's why I want to see venture building happen at scale. Build more of them, on purpose, with advantages baked in from day one.

The record is real. So is the drought. Build accordingly.