Many years ago, my first company raised its seed round from an institutional venture fund that managed about $250 million. One of its partners met us early, when we were unproven and the numbers that would later justify the investment did not yet exist. He wrote a $1.5 million check at a $5 million pre-money valuation and backed the conviction before the metrics arrived.
That investor is nearly extinct. A $250 million fund almost never writes $1.5 million seed checks in today's market, and funds of that size are increasingly not raised at all. The kind of firm that gave my first venture-backed company its start has been disappearing from the earliest stage of the market. The latest data shows how far the retreat has gone, and why it is unlikely to reverse in the form we knew it.
The second quarter of 2026 produced a set of venture records and a set of decade lows in the same dataset. Global deal value reached an all-time quarterly high, and half-year exit value exceeded any full year on record. Both figures are the product of a small number of very large, late-stage transactions concentrated in the United States and in artificial intelligence.
Beneath them, the measures that describe the early end of the market moved the other way. Fund formation fell to its lowest level in a decade, average fund size rose, and first-time managers were largely unable to raise. The data describes capital relocating up-market and concentrating there, and it supports a specific and narrower conclusion than the headlines suggest.
The institutional seed fund, the closed-end vehicle that has financed the earliest stage for a generation, is in terminal decline, and the mechanism that reversed past declines is being routed around rather than merely paused. Paradoxically, the capital that funds seed is not vanishing; it is leaving the tread-worn vehicles for others.
Therein lies the reformation of early-stage venture capital. And it carries generational implications for the emerging managers and founders who depended on it.
