.png)
The venture capital industry is operating on flawed data, and it's time to call it out.
If you've been reading headlines about the "decline in fundraising" and a "quiet year" for venture capital, you've been misled. The numbers being cited, primarily from PitchBook, are logically impossible and create one of the largest industry misinformation scenarios we've seen in years.
Let's break down why.
The Math Doesn't Work
Consider what we know happened in 2025:
On the startup side:
- OpenAI raised $40 billion at a $300 billion valuation, the largest private tech raise ever recorded
- Anthropic raised $13 billion at a $183 billion valuation
- xAI raised $20 billion at a $230 billion valuation
- AI startups raised a record $150 billion in 2025, surpassing the previous record of $92 billion set in 2021
- North American startups raised $280 billion in 2025, up 46% from 2024
On the fund side:
- Andreessen Horowitz raised $15 billion, representing 18% of all U.S. venture capital dollars allocated in 2025
- Lightspeed Venture Partners raised $9 billion across six funds, the largest in their 25-year history
- Founders Fund raised $4.6 billion for its third growth fund
- Global venture capital reached $512 billion in deal value in 2025, nearly matching the 2022 record high
So where is all this capital coming from if fundraising is supposedly declining?
The answer: It's not declining. The data is wrong.
The PitchBook Problem
PitchBook relies on a survey-based approach to collect venture capital data. The fundamental challenge? Many fund managers don't report to them.
When your data collection methodology depends on voluntary survey responses from an industry where reporting is inconsistent, you're working with an incomplete picture. And that incomplete picture is being reported as fact by major publications, shaping narratives, influencing LP allocation decisions, and creating a misleading view of the venture capital landscape.
Why does this matter?
Inaccurate data doesn't just misinform, it discourages participation. When headlines paint a picture of a contracting industry, it deters:
- Founders from pursuing venture-backed paths
- New fund managers from launching funds
- LPs from allocating to venture capital
At Decile Group, we see the opposite of contraction. We believe the expansion of venture capital is critical because more venture funds mean more support for founders solving humanity's biggest problems. Emerging managers who don't fit the traditional mold deserve to know how healthy the market really is.
What We're Actually Seeing
From our vantage point at Decile Group, where we work with hundreds of emerging fund managers every year, the reality looks very different:
2025 was larger than 2024 for VC fundraising.
2026 is looking to be larger than 2025.
Our December 2025 numbers were extraordinary. We closed multiple funds in a single month. LPs were wiring capital on New Year's Day because they didn't want to miss allocation opportunities.
This is not what a "quiet year" or "declining fundraising environment" looks like.
The Exit Market Is Reopening
After a slow period for IPOs, the window is opening again. In 2024 and 2025, we saw meaningful public market debuts:
- Rubrik went public on NASDAQ
- Netskope raised over $908 million in its NASDAQ debut at an $8.6 billion valuation
- Navan raised $932 million on NASDAQ at a $6.2 billion valuation (Alternatives Watch)
- Discord is reportedly preparing to file
The backlog of venture-backed companies ready to exit is substantial, and the returns from top-performing funds demonstrate that venture capital continues to deliver exceptional outcomes.
Bloomberg's recent profile of Index Ventures illustrates this point:
In 2025 alone, Index generated approximately $9 billion in realized gains and unsold shares from six exits: Wiz (sold to Google), Figma (IPO), Scale AI (partial sale to Meta), Dream Games, Nexthink, and Wealthfront (Bloomberg).
Their 2012 fund delivered a DPI of 11x, meaning $11 returned for every dollar invested. The industry average for funds from that vintage? Just 1.5x.
These are not the results of a struggling industry. These are exceptional returns being generated by firms that know how to pick winners and support them to exit.
The Real Story of Venture Capital in 2025 and 2026
Here's what's actually happening in venture capital:
AI is creating a new wave of massive outcomes. The companies being built today have the potential to be the largest in history. OpenAI, Anthropic, and others are raising at valuations that would have been unthinkable five years ago.
Emerging managers are thriving. First-time fund managers with differentiated strategies and strong networks are finding LP interest. The "emerging manager premium" that institutional LPs seek is real, and capital is flowing to smaller, more nimble funds.
LP appetite for venture remains strong. Despite what the flawed data suggests, institutional LPs, family offices, and fund-of-funds continue allocating to venture capital. They understand that the best returns come from this asset class, and they're not sitting on the sidelines.
A Call to Action
Venture capital fuels innovation. It supports founders building solutions to climate change, healthcare, education, and countless other challenges. When the narrative suggests the industry is struggling, it has real consequences:
- Talented people don't pursue careers in VC
- Founders don't start companies
- LPs allocate elsewhere
- Fewer problems get solved
When you see posts or articles citing data that paints an incomplete picture of the venture capital industry, ask questions. Look at what's actually happening: the mega-rounds, the fund closes, the IPO pipeline.
The venture capital industry deserves accurate data. LPs making allocation decisions deserve accurate data. Emerging managers trying to raise their first funds deserve to operate in an environment where the narrative matches reality.
Pontificating on incomplete data is a giant waste of time. Let's start demanding better.
The reality: 2025 was a strong year for venture capital. 2026 is shaping up to be even stronger. Don't let incomplete data tell you otherwise.