Assessing Anthropic: The company’s S-1 is expected soon. Ahead of its public filing, what should you look for in the prospectus? Our new report is your guide.

Bottoming out: Middle-market deal value fell to its lowest level since the 2022-2023 rate shock, but fundraising remains a bright spot. Read more in our Q2 US PE Middle Market Report.

Venture is having its biggest year ever: Exit value hit a record $2.1 trillion, dwarfing 2021’s $865 billion, public markets sit near highs, and AI mania is minting trillion-dollar companies. Venture was built to harvest exactly these cycles.

Yet the median VC-backed IPO has trailed the Morningstar Growth Index by 38.2% within 120 days of listing, and even SpaceX has traded erratically since its June debut despite its record-breaking $1.7 trillion debut. Public investors are hesitating to pay for venture’s elevated valuations.

This is a curse of venture’s own making. Deal value surged past its 2021 peak to $416 billion, with the top 10 deals absorbing 67.4%. OpenAI’s single round accounted for $122 billion. The widening gap between private pricing and what public investors are willing to pay has caused companies to wait. Of the 110 companies our model flagged as IPO-ready and backlogged since 2023, 86 have continued to raise venture capital.

Venture can no longer even rely on a few mega IPOs to reopen the shut window. Anthropic CEO Dario Amodei called for a slower pace of frontier AI development this month; soon after, Sam Altman ruled out an OpenAI listing this year. In the meantime, secondaries are filling this liquidity void, with a record $107.1 billion traded in direct secondaries over the past year.

Our Q3 2026 Quantitative Perspectives report unpacks venture’s liquidity paradox and what it means for underwriting discipline and portfolio construction.

It has always felt like there are too many acronyms to remember. With the rise of agentic AI, there are now way more.

The letters MCP and CLI, abbreviations for model context protocol and command-line interface, have appeared frequently in 2026 as more products are made accessible to AI agents.

Across retail, fintech, cybersecurity, and cloud infrastructure, companies have also introduced a wave of agentic payment protocols and rails. These include ACP, AP2, TAP, UCP, x402 and MPP, just to name a few.

The reason behind this growing alphabet soup of acronyms is the rise of the machine economy.

As AI agents become more capable and abundant, they are taking on a larger share of knowledge work. Many are now discovering and invoking various tools to complete their tasks, but this also means they need to pay for the data, APIs, and services they consume.

Scaled players and payment leaders want to be ready for this, since a meaningful share of payment volume could shift to agent-led transactions. However, finding investment opportunities in agentic payments is difficult, as adoption is still so early.

For the first eight months of 2026, just $491.8 million in disclosed venture capital went to startups focused on agentic payments.

But as with any new payment method, adoption will take time. Getting there requires systems that enable buyers to confidently transact through their agents and sellers to safely accept those payments.

Investing in payments for the machine economy therefore means investing in the solutions that build trust.

In our report Machine Economy Rising: How Payments Unlock New Agent Markets, we map over 180 startups emerging in this space and discuss which areas to focus on as agentic activity accelerates.