The Concentration Problem Behind Israel's Record Exit Year
The headline number says 2025 was the best year ever. The composition underneath says the ecosystem is narrowing — fewer companies, fewer sectors, bigger bets.
A record is not the same as a recovery, and 2025 is the year that distinction matters most. Israeli tech posted 74.3 billion dollars in exits and 15.6 billion in private funding — figures that read like a triumphant return to form. But the composition beneath those numbers tells a more uncomfortable story: an ecosystem concentrating into fewer companies, fewer sectors and fewer, larger bets. The headline says broad health. The internals say narrowing.
The exits were a handful of deals
Start with the M&A figure. The 74.3-billion-dollar total — compiled by Startup Nation Central — was dominated by two transactions: Google's 32-billion-dollar Wiz acquisition and Palo Alto Networks' 25-billion-dollar CyberArk purchase. By one accounting, nine deals accounted for roughly 53 billion of the total. A number that looks like a thriving exit market is, on inspection, a small number of extraordinary trophy sales sitting on top of a much more ordinary base. Strip the giants out and the picture is solid but unremarkable — M&A excluding the top two deals rose about 12 percent.
The funding went to fewer rooms
The private-funding side shows the same compression. Deal volume fell to 717 rounds — the lowest in a decade — even as total dollars rose, which means the money concentrated into a shrinking number of companies. The median deal size jumped roughly 67 percent to about 10 million dollars, and mega-rounds above 100 million captured roughly half of all capital raised. Investors are making fewer decisions and writing much larger checks. That is healthy for the companies that get funded and worrying for the ones that now never will.
And into fewer sectors
The concentration is sectoral too. Cybersecurity and generative AI together absorbed around 70 percent of all capital — roughly 40 percent of funding rounds in just those two categories. Business software and cyber led the value tables; almost everything else competed for what was left. An ecosystem that once spread its bets across a wide range of categories is increasingly a two-theme economy, dependent on the continued global appetite for Israeli cyber and AI specifically.
Why concentration is fragility
This is where the record becomes a risk. An ecosystem that depends on a few mega-deals, a few funded companies and two hot sectors is an ecosystem with less margin for error. If global appetite for cybersecurity assets cools, or the AI funding cycle turns, or a single trophy acquisition falls through, a disproportionate share of the headline disappears with it. The breadth that makes an innovation economy durable — many companies, many sectors, many buyers — is precisely what 2025's numbers show thinning. Even the early-stage data, where funding recovered but the number of rounds kept falling, points to a narrower funnel feeding the whole system.
The long-term cost is more corrosive than any single bad year. Concentration this severe means fewer funded founders, which over time means fewer companies reaching the stage where they can become anything at all. It means a narrower innovation pipeline, as capital clusters around proven categories and known winners rather than the unproven bets that produce the next category. And it means declining ecosystem diversity — the quality that let Israel pivot from chips to security to AI over three decades. An economy that funds only its current champions is an economy slowly optimizing away its ability to produce the next ones. That is the warning underneath the record: not that 2025 was weak, but that its very strength was distributed in a way that makes the future thinner.
The achievement is real; the Wiz and CyberArk deals prove Israeli technology can reach genuine global scale and strategic weight. But a record year built on concentration is a record that depends on everything continuing to go right in a small number of places — and a model already built to sell those winners abroad (see The Exit Is the Business Model). The number to watch in 2026 is not the headline total. It is the deal count — and whether the funnel widens again, or keeps narrowing toward a handful of bets.
