Which VC Sectors Generate the Best ROI? 2025 Data on IRR by Industry

Not all VC sectors deliver equal returns. Here’s what the 2025 data shows about which categories actually generate the best ROI.

Headline 2025 sector data

• AI — 65.4% of U.S. VC deal value, 39.4% of deal count (Pitchbook-NVCA Q4 2025).
• IT + Healthcare + Industrials — 85% of invested capital in U.S. VC index (Cambridge Associates H1 2025).
• Life Sciences + Biotech — consistent secondary share, growing with AI drug discovery.
• Climate tech — 4-6% of U.S. VC, up from under 2% pre-2020.
• Fintech — softer post-2022, recovering in 2025 led by AI-native fintech.

Historical sector return observations

Enterprise software (B2B SaaS)

• Why it wins: High gross margins (75–90%), predictable retention, recurring revenue.
• IRR pattern: Top-quartile funds in enterprise SaaS generate 20–30% net IRR historically.
• Examples: MongoDB, Datadog, Snowflake — fund-returners that drove top decile outcomes.

Consumer internet / marketplaces

• Why variable: Winner-take-all dynamics — power-law outcomes.
• IRR pattern: A single winner (Facebook, Airbnb, Uber) returns an entire fund.
• Risk: Most consumer bets write to zero; sector-level medians are weaker than enterprise SaaS.

Biotech / life sciences

• Why volatile: Binary clinical outcomes drive large wins and losses.
• IRR pattern: High-variance but potential for top-decile outcomes (Moderna, Loxo Oncology).
• Cycle dependency: Public market receptivity drives exit timing.

Fintech

• Why solid: Recurring revenue, sticky customers, regulatory moats.
• IRR pattern: Top-quartile strong but narrower dispersion than consumer.
• 2025 catalyst: AI-native fintech infrastructure is growing fast (Ramp’s trajectory).

AI foundation models

• Why uncertain: Huge capital requirements, long time to revenue, regulatory risk.
• IRR pattern: Too early to fully measure; 2022+ vintage foundation model bets are tracking well on paper but DPI lags.

Key insight from Cambridge Associates H1 2025

• Smaller funds generally outperform larger funds: The top-decile $100–300M fund often beats the $1B+ fund in IRR, because concentration of winners in a smaller base amplifies returns.
• 2021 vintage funds struggling: Median TVPI barely above 1.0x; 75th percentile IRR 5.9%.
• 2022–2023 vintages performing better: Entry valuations post-correction are healthier.

What actually wins in VC returns

1. Stage discipline: A fund that stays in its stage through cycles beats a firm that drifts.
2. Reserve discipline: Disciplined follow-on capital in winners multiplies fund-level DPI.
3. Sector specialization: Specialized funds often beat generalists in applied categories (health, fintech, cyber).
4. Portfolio construction: Power-law math requires betting small on many, large on winners.

NYC-specific ROI implications

1. B2B SaaS and fintech remain the most consistent NYC winners.
2. Consumer and marketplaces can produce outsized wins but require Bay Area-scale network effects.
3. Applied AI in finance and health is NYC’s highest-conviction 2026 category.

Practical takeaway

• Founders: Sector-specialist investors often deliver better post-money support than generalists.
• Investors: Durable returns come from sector discipline + reserve discipline, not deal count.
• LPs: Diversify across sectors AND stages; single-theme concentration is a top risk.

Sources

1. Pitchbook-NVCA Venture Monitor Q4 2025: https://nvca.org/wp-content/uploads/2026/01/q4-2025-pitchbook-nvca-venture-monitor.pdf
2. Cambridge Associates US PE/VC Benchmark H1 2025: https://www.cambridgeassociates.com/insight/us-pe-vc-benchmark-commentary-first-half-2025/
3. Carta Q4 2025 VC Fund Performance: https://carta.com/data/vc-fund-performance-q4-2025/