Venture Capital (VC): The Complete Guide

Corporate Finance and Strategic Advisory · CGPH Banque d’affaires

Global venture funding hit $510 billion in the first half of 2026, already more than the whole of 2025, with over 70% of every quarterly dollar flowing into artificial intelligence companies.¹ To a founder raising a first institutional round, or an investor trying to make sense of headlines about $100 billion-plus valuations, the venture market can look irrationally exuberant and impossibly selective at the same time. Both are true. This guide explains what venture capital actually is, where it came from, how a raise is structured and priced stage by stage, what the data says about where the market stands today, which recent deals defined the current cycle, and why the entire industry is built around accepting that most individual bets will fail.

What is venture capital?

Venture capital is capital provided to early-stage, high-growth-potential companies in exchange for equity, by professional investors who pool money from limited partners — pension funds, endowments, sovereign wealth funds, family offices — into a fund with a fixed lifespan, typically eight to ten years. It occupies a specific point on the private capital spectrum:

• Angel investors — individuals investing their own money, usually at the earliest, pre-seed or seed stage, often the first check after founders and friends and family.
• Venture capital — institutional funds investing pooled capital in early- to growth-stage companies with high growth potential but unproven or nascent cash flows, taking minority equity stakes.
• Growth equity — investors backing already-proven, revenue-generating companies to fund expansion, typically with less risk, and lower target returns, than early-stage VC.
• Private equity — funds that typically acquire controlling stakes in mature, cash-generative companies, often using leverage (see our companion guide to private equity).
• Corporate venture capital — the strategic investment arms of operating companies, investing for access to innovation alongside financial return (see our dedicated article on this model).

What sets venture capital apart from every other form of private capital is its tolerance for a very high individual failure rate, in exchange for outsized returns from a small number of winners — a dynamic explored in detail below.

A brief history of venture capital

Modern venture capital traces its origin to a single institution: American Research and Development Corporation, founded in 1946 by Georges Doriot, a Harvard Business School professor often called “the father of venture capital.” Doriot’s innovation was to formalize what had previously been ad hoc — systematically investing in new, technology-driven companies through a professionally managed fund rather than personal wealth, and pairing capital with active mentorship. He also helped found INSEAD, Europe’s first MBA program, giving the discipline an early transatlantic footprint.²

From there, the industry moved through recognizable phases:

• 1950s–1960s — Formalization. Early firms following the ARDC model emerge on the US East Coast; the Small Business Investment Act of 1958 creates a licensing framework for VC-like funds.
• 1970s–1980s — Silicon Valley emerges. Kleiner Perkins and Sequoia Capital are founded, and the rise of semiconductor and personal-computing companies shifts the industry’s centre of gravity to Sand Hill Road.
• 1995–2001 — The dot-com boom and bust. Consumer internet companies attract unprecedented capital and public enthusiasm, followed by a severe correction that resets valuation discipline for most of the following decade.
• 2004–2013 — Web 2.0 and mobile. Social platforms, cloud infrastructure, and smartphones create a new generation of category-defining companies, and VC firms, at a fraction of the capital intensity of the dot-com era.
• 2020–2021 — The zero-rate mega-boom. Near-zero interest rates and abundant late-stage capital push valuations and round sizes to records, and “unicorn” creation accelerates sharply.
• 2022–2023 — The correction. Rising rates, down rounds, and a closed IPO window force a sharp re-pricing and a “flight to quality.”
• 2024–2026 — The AI supercycle. Foundational AI labs and AI-native applications absorb a majority of global venture dollars, reopening IPO and M&A exit markets and producing the largest individual funding rounds in the industry’s history — several of which are detailed below.

Stages of venture capital funding

A company typically raises capital across several discrete stages, each with its own risk profile, instrument, and expected dilution:³

• Pre-seed and seed. Backed by angels and micro-VCs, usually through SAFEs (Simple Agreements for Future Equity) or convertible notes rather than a priced round. The gating milestone is a working product with early users; founders typically retain a median of roughly 56% of the company after this stage.
• Series A. The first institutional, priced round, typically $5–15 million at a median pre-money valuation in the tens of millions, and the point at which a “repeatable” revenue model starts to matter. The bar has moved quickly: where $1 million of annual recurring revenue growing 3x year-on-year was once considered Series A-ready, institutional investors in 2026 more commonly look for $2–3 million or more, net revenue retention comfortably above 100% (treated as a baseline rather than a target), and a burn multiple below 2x.⁸ Founder ownership typically falls to roughly 36% post-round, partly from option-pool top-ups negotiated alongside the new capital.
• Series B. Growth-stage investors back companies with proven unit economics and a clear payback period; governance shifts meaningfully as investor board representation increases. Founder ownership typically falls to roughly 23%.
• Series C and beyond. Late-stage and crossover funds, which also invest in public equities, back category leaders at scale. The central question shifts from execution risk to “outcome risk” — whether the company can achieve the exit scale needed to justify its valuation.

Preferred shareholders sit ahead of common shareholders — founders and employees — in the liquidation waterfall, and later investors typically sit ahead of earlier ones. It is one of several reasons why the headline valuation of a later round is not the same thing as what earlier shareholders will actually realize in a downside scenario.

How a venture round actually gets done

However large or small the round, the process tends to follow a recognizable sequence:

1. Preparation. A data room, a financial model, and a clear narrative of traction and market size — the same discipline that underpins any capital raise, private or public.
2. Sourcing and outreach. Direct introductions dramatically outperform cold outreach; this is where an experienced advisor’s investor network shortens the timeline materially.
3. Pitch and initial diligence. Partners assess the team, market, product, and early traction data before committing partnership time.
4. Term sheet. A non-binding document setting out valuation — pre-money and post-money — the amount raised, liquidation preference, board composition, protective provisions, and pro-rata rights: terms that matter every bit as much as the headline valuation.
5. Due diligence and definitive documentation. Legal, financial, and increasingly technical diligence, followed by the definitive investment agreements.
6. Closing and post-close governance. Funds are wired, board seats and information rights take effect, and — critically — the relationship with existing and new investors is managed for the follow-on rounds still to come.

Where the market stands in 2026

Venture capital in 2026 is defined by concentration — of capital, of company, and of geography:^{1 4}

• Record volume. Global venture funding reached $510 billion in H1 2026, already surpassing all of 2025’s $440 billion; Q2 alone accounted for $205 billion across more than 5,000 companies, the second-largest quarter on record after Q1’s $305 billion.
• AI dominance. Over 70% of Q2 global venture capital went to AI companies, up from roughly 50% a year earlier. OpenAI and Anthropic alone captured $217 billion between them — 43% of all H1 2026 global venture funding.
• Mega-rounds are the market. Sixteen companies raised billion-dollar rounds in Q2 2026 alone, totaling $108.6 billion, or 53% of the quarter’s funding, including several China-based frontier AI labs alongside defense, robotics, and healthcare startups.
• The exit window has reopened. 32 companies went public above $1 billion valuations in Q2 2026, and 24 acquisitions exceeded $1 billion, totaling $113 billion — the highest quarterly M&A total for venture-backed companies on record. IPO volumes and proceeds grew 20% and 84% respectively over the past twelve months, and secondary markets — an increasingly normalized liquidity route for LPs, founders, and employees — were projected to exceed $210 billion in 2025 alone.⁵
• The US still dominates, but slightly less than before. US companies captured two-thirds of Q2 2026 global venture capital, down from 83% in Q1 — still highly concentrated, but a signal that capital is beginning to find opportunities elsewhere, including in Europe.

For founders outside the small circle of frontier AI labs, the practical read-through is what the Harvard Law School Forum on Corporate Governance calls a “flight to quality”: capital is increasingly concentrated in companies with the strongest competitive positions, and selectivity, not the general availability of capital, is now the defining constraint on most fundraises.⁵

Landmark venture deals of recent years

A handful of transactions from the past eighteen months capture both the scale of the current cycle and its full lifecycle, from mega-round to IPO to acquisition:

Two things are worth noting for a European audience in particular. First, the SpaceX sequence — an IPO, then a $60 billion acquisition days later — shows how quickly a well-capitalized venture-backed company can move from private mega-round to public listing to using its own stock as acquisition currency, compressing into weeks a cycle that used to take years. Second, Mistral AI’s trajectory is a rare European counterweight to an AI funding landscape otherwise dominated by two US labs and a cluster of Chinese frontier players — useful context for European founders and investors judging what a credible, fundable AI strategy looks like outside Silicon Valley.

Why most VC bets fail — the power law

Venture capital’s economics only make sense once one accepts a counterintuitive premise: most individual investments are expected to fail. Long-run research puts the failure rate for venture-backed startups at roughly 75%, measured as failing to reach an exit that returns capital to all equity holders, with some estimates running higher still.⁶

The industry’s return model is not an average across a portfolio; it is a power law — a small number of outsized winners generate the overwhelming majority of a fund’s returns, while most portfolio companies return little or nothing.⁷ The practical implications of that dynamic run in several directions at once:

• For fund managers, it means underwriting every investment for the possibility of an outsized outcome rather than a merely “good” one. A marginal, moderately successful company does not move the needle on fund returns the way one breakout winner does, which is also why VCs typically decline to run distressed workout processes for struggling portfolio companies, preferring to let them fail quickly and redeploy attention to potential winners.
• For founders, it explains why VCs will sometimes prefer a larger addressable market and a more ambitious, higher-risk strategy over a safer, smaller path to profitability: the fund’s math requires swings, not singles.
• For companies raising capital outside that model — the profitable, well-run mid-market businesses CGPH Banque d’affaires advises most often — this is precisely why venture capital is not always the right tool. Founders seeking a partner for a steady, profitable growth trajectory are frequently better served by private equity, private debt, or a strategic transaction than by an investor whose model depends on a small probability of a very large outcome.

The role of a fundraising advisor

Raising venture or growth capital well is a full-time exercise in project management, positioning, and negotiation leverage — one most founding teams are running for the first time while also running the business. An experienced advisor typically adds value by:

• Structuring the narrative and the data room before the first investor conversation, rather than after the first round of questions exposes the gaps.
• Running a genuine process — approaching multiple investors in parallel on a defined timeline, rather than a single, sequential conversation that erodes negotiating leverage.
• Negotiating the terms that matter beyond valuation — liquidation preference, board composition, protective provisions, and pro-rata rights, all of which shape founder and early-investor outcomes regardless of the headline number.
• Sequencing the capital strategy, including recognizing when venture capital is not the right instrument at all, and a growth-equity, private-debt, or strategic-partner path better fits the company’s actual growth profile and risk tolerance.

Sources

1. Crunchbase News, Global Startup Investment Hit Record $510B in H1 2026 As AI Boom Accelerates Funding and Exits — news.crunchbase.com
2. Harvard Business School, Working Knowledge, The Founder of Modern Venture Capital — library.hbs.edu
3. Qubit Capital, Stages of VC Funding: Cap Table & Dilution at Each Round — qubit.capital
4. Crunchbase News, Sector Snapshot: Venture Funding to Foundational AI Startups in Q1 Was Double All of 2025 — news.crunchbase.com
5. Harvard Law School Forum on Corporate Governance, Venture Capital Outlook for 2026: 5 Key Trends — corpgov.law.harvard.edu
6. Harvard Law School Forum on Corporate Governance, Startup Failure — corpgov.law.harvard.edu
7. The VC Factory, Understanding the Power Law: Do Venture Capitalists Take Enough Risks? — thevcfactory.com
8. CRV, Series A Metrics VCs Expect in 2026; Crunchbase News reporting quoting Uncork Capital on the shift away from the “$1M ARR” Series A benchmark.

Additional data points on individual transactions (Anthropic, OpenAI, Databricks, Anduril, xAI, SpaceX, Mistral AI) were sourced from contemporaneous reporting by CNBC, TechCrunch, Forbes, Al Jazeera, and company newsroom announcements, cited by name in the text above.

This article is provided for general information purposes and does not constitute investment advice. Valuations and funding figures for private companies are as reported publicly at the time of writing and may change; figures for in-progress or reported-but-unconfirmed rounds (e.g., Anduril’s July 2026 talks) are noted as such.