Venture Capital
Venture Capital (VC) is equity investment in early-stage privately-held companies. Structured similarly to PE — fund vehicle with multi-year drawdown — but with longer hold periods (10–12 years) and far more skewed return distribution. Top-decile VC funds deliver 20%+ net IRRs; median funds deliver mediocre returns; a meaningful fraction lose principal. Stage-graded into Seed, Series A, B, C, D corresponding to company-development milestones.
VC return distribution is the defining feature. A typical VC portfolio of 30 investments expects: 70% partial-loss-to-zero (the company fails or returns less than invested), 20% modest returns (1–3x exit), and 10% home runs (10x+ exit). The home-run exits drive the entire fund's return — without them, the fund delivers below-public-equity. This 'power-law' pattern is structural; VC GPs explicitly target portfolios where one or two outsized exits compensate for many failures.
Fund-selection skill matters more in VC than any other alternative asset class. Top-decile VC funds (typically the established Sand Hill firms — Sequoia, Andreessen Horowitz, Benchmark, Greylock, Accel) routinely deliver 25%+ net IRRs over 20-year track records. Bottom-quartile VC funds underperform public equity by 5%+ per year. The dispersion is far wider than PE, where top vs. bottom quartiles differ by 5–8% rather than VC's 15%+. Access to top-decile VC is gated; minimum commitments at the top firms run $5M+ and are restricted to long-term institutional LPs.
QSBS (§1202) intersects materially with VC. Qualifying VC investments are original-issue C-corp stock from companies meeting the §1202 gross-assets and active-trade requirements, and two regimes now coexist based on acquisition date. Stock acquired on or before July 4, 2025 keeps the original regime: a $10M-or-10×-basis exclusion cap, a 5-year cliff holding period (all-or-nothing at 5 years), and a $50M gross-assets test. Stock acquired after July 4, 2025 (under OBBBA, P.L. 119-21) uses a larger, tiered regime: a $15M-or-10×-basis cap, a tiered holding-period exclusion (50% excluded at a 3-year hold, 75% at 4 years, 100% at 5 years), and a $75M gross-assets test. Top-decile VC GPs structure their portfolio companies to preserve QSBS eligibility. LPs in QSBS-friendly funds can stack multiple non-grantor trusts to multiply the exclusion (each trust gets its own cap). For UHNW households, VC + QSBS + non-grantor-trust stacking is one of the most powerful tax-arbitrage structures available.
Retail-accessible VC is limited but growing. AngelList syndicates allow accredited investors to invest in single-deal VC opportunities. Listed venture-capital firms (BDCs, listed-VC interval funds) provide market-priced retail exposure but with the structural drag of public-vehicle costs. True direct-VC access remains gated to institutional-and-UHNW LPs.
Synthetic UHNW households should include VC fund positions at realistic frequency (~30% of $25M+ households). Each VC position tracks: commitment, called-to-date, distribution history, fund stage focus (early/late), QSBS eligibility status. The skewed return distribution means most VC positions show losses or modest gains; a small fraction show extraordinary returns. Test scenarios should include the home-run exit case (a single portfolio company driving the fund's entire return).
Common pitfalls
- Modeling VC returns as a normal distribution — they're power-law; mean-variance models produce wrong outputs.
- Aggregating VC and PE returns into a single 'private equity' bucket — they have very different risk and return profiles.
- Forgetting QSBS qualification on VC investments — many VC fund-of-fund LPs miss the §1202 stacking opportunity.
- Treating fund-vintage as the only diversification dimension — stage-diversification (Seed + Series A + Series C) and sector-diversification matter equally.
Examples
VC fund makes 30 investments at $5M each ($150M total deployed). Outcomes: 12 companies fail (write-off): −$60M. 10 companies return capital or modest profit (avg 2x): +$50M. 6 companies modest success (3–5x): +$120M. 2 companies large success (10x): +$80M. Net gain across the book: −$60M + $50M + $120M + $80M = +$190M. Total value returned = $150M deployed + $190M gain = $340M, a 2.27x gross multiple ($340M / $150M). Net of fees and carry (2-and-20 over fund life): roughly 1.8x. The 8 top-tier deals (the 3–5x and 10x buckets) generate essentially the entire gain; the 12 write-offs and 10 modest-return deals together roughly return capital — the concentration that defines the power-law.