Is there still room for a mid-sized fund

Three arguments, weeks apart, same conclusion

During 2026 three independent analyses arrived at close to the same claim: venture is bifurcating into small funds and very large ones, and the middle is being squeezed out. They come from different vantages and use different evidence, which is what makes the pattern worth taking seriously rather than treating as one commentator's take.

The claim matters here because the middle, in fund-size terms, is where the Cadence stage lives.

The arithmetic argument

Value Add VC works the problem from the return math. To return 3x net, a $500 million fund needs roughly $1.5 billion in proceeds. Getting there takes something like nine positions each producing $200 million of proceeds, which means backing nine companies that each reach around $2 billion and holding meaningful ownership all the way through dilution.

Their conclusion is the sharp part: that is not an outlier strategy, it is a batting-average strategy, and venture has never been good at batting average. The whole asset class is built on a small number of extreme outcomes carrying everything else. A fund size that requires nine large wins rather than one or two enormous ones has quietly stopped relying on the distribution that makes venture work.

Hence their framing of the squeeze from both sides: $500 million is too big to return a multiple from seed-stage outcomes, and too small to win the growth rounds that multi-billion-dollar funds are taking.

The market-structure argument

TechCon Global comes at it from where the capital is actually going. In February 2026 alone, 83 percent of the $189 billion invested globally went to three companies: OpenAI, Waymo, and Anthropic.

The implication they draw is that the market has developed two workable positions and very little in between. At one end, entering early enough that ownership is cheap and a single outcome can carry a fund. At the other, being large enough to write into the rounds where the dollars are actually going. A mid-sized fund is too late for the first and too small for the second.

The performance argument

VC Stack assembles the returns case. Drawing on PitchBook data analysed by Santé Ventures, smaller funds show an average cumulative IRR of 17.4 percent against roughly 9.7 percent for large funds, with large defined above $750 million. A separate Carta cut of 2017 to 2021 US vintages points the same way: a median IRR of 13.8 percent for funds between $1 million and $10 million against 9.8 percent for funds above $100 million.

Two independent datasets agreeing is a materially stronger claim than either alone. Their reading of why is structural rather than about skill: funds above $1 billion have become asset managers whose economics work on fees, while funds below $100 million still live on carry and therefore have to chase concentrated multiples. Different incentives produce different behaviour, and the middle inherits neither the fee base of the first nor the return math of the second.

Where the argument is weaker than it sounds

Three caveats, none of which dissolves the case but all of which should be held alongside it.

The IRR comparison is not obviously like-for-like. Small funds are more numerous, more varied, and more prone to survivorship effects in reported data, and a small fund can post a high IRR on a quantity of capital that would not move an institutional allocation. IRR also flatters shorter holding periods, which small funds are more likely to have. None of that makes the gap fake; it does mean 17.4 against 9.7 is not the clean like-for-like it reads as.

The concentration figure describes an extraordinary moment. A month in which three companies absorbed 83 percent of global venture dollars is not a normal month, and building a structural claim about fund sizes on the tail of an AI cycle risks generalizing from the most unusual period in recent memory.

And the arithmetic argument is a statement about a specific fund size rather than about the middle as a region. The nine-outcomes problem gets easier at $250 million and harder at $750 million; the analysis picks the number where it bites and names the whole band after it.

How this maps onto the three stages

Read against the lifecycle framing, the barbell argument is close to a claim that Conviction and Continuity are viable and Cadenceis not. That is a real challenge to the model's framing and it deserves a direct answer.

A firm can hold real operating rhythm, multiple funds, a working team, and formal process at $200 million, which sits inside the Cadence band and well clear of the size the arithmetic argument attacks. Equally, a firm can raise $600 million while still selling nothing but the founders' judgment, which is a Conviction firm carrying a Cadence-sized fund and is arguably the configuration the barbell argument is really describing.

Which suggests the sharper version of the claim is not that the middle stage is unviable. It is that a middle-sized fund has to be genuinely run as a Cadence firm, with the process, team scale, and diversified construction the stage implies, because it can no longer win on concentration alone and cannot yet win on scale. The firms getting squeezed are the ones occupying the size without the structure.

What the model does with this

Nothing directly, and that is deliberate. The Colibrí Architecture model has no view on what a fund should be worth, does not recommend a fund size, and does not read a firm more favourably for sitting at either end of the barbell.

What it reads is coherence at whatever size the firm actually is. The capital math reading asks whether the cheque this fund can write buys a position at the stage it enters, which is exactly the mechanism the arithmetic argument describes, applied to one firm rather than to a band. The lifecycle fit reading asks whether the configuration matches the stage declared, which is where a Conviction firm carrying a Cadence-sized fund shows up.

Separately, the transition question, why getting from a first fund to a second is structurally hard regardless of size, is a different problem with a different cause and is covered on the Conviction page.

Sources

• The Death of the $500M Fund: Why Mid-Sized VC Is Getting Squeezed From Both SidesValue Add VC, June 2026The return arithmetic for a mid-sized fund, worked through to the number of large exits it implies.
• Why The Future Of Venture Is Either Pre-Seed Or Mega-Fund, With Nothing In BetweenTechCon Global, 2026The deal-level concentration figure for February 2026, and the two-extremes framing.
• The Barbell of Venture Fund SizesVC StackPerformance by fund size, drawing on PitchBook data analysed by Santé Ventures and a separate Carta cut of 2017 to 2021 vintages.