Late-stage companies are different

At seed stage, diversification raises both your expected return and reduces downside — a powerful dual effect that Koh & Othman (2020) confirm empirically across 10,665 real LP portfolios on AngelList. At Series C/D, the distribution has normalised (α > 3). The mean return is flat with pool size — adding more companies does not raise your expected return.

What you get instead is something equally valuable: dramatic variance reduction. At 50 names, your probability of losing money is under 0.1% in a normal market. Your 5th-percentile floor rises from 0× (single name) to ~1.34×. The range of outcomes tightens. Your outcome becomes more predictable. Watch the flat mean line in the pool-size chart to see this in action.

What this tool does

This dashboard simulates 10,000 possible ~4-year futures for a fund holding the number of late-stage companies you choose, drawing each company's outcome from a power-law distribution calibrated to late-stage venture data (Othman, 2019; PitchBook Q3 2024; Koh & Othman, 2020).

Adjust the sliders and hit Run simulation. The partial-loss slider is the most interesting: set it to 30% and watch the loss-probability curve shift — that is the post-2022 stressed environment, where 70% of Series D+ exits returned less than 1×. The pool structure cuts through it.

50
More holdings spread risk across more companies. ECLF targets 50–75.
13%
Share of holdings expected to return 0×. 13% is Series D+ per PitchBook Q3 2024.
3.0
Late-stage calibration: α = 3.0. Move toward 2.5 for seed-like tails; toward 5.0 for near-normal distributions.
0%
Default off (normal market). Set to ~30% for the post-2022 stressed environment.
Simulation Results
If I held my late-stage stock alone, I would have a chance of getting back less than my paper value. By placing it in this pool, my expected outcome lands between and , and my probability of losing money drops to .
Mean Gross MOIC
Average ending value across 10,000 simulations.
Median Gross MOIC
Middle outcome — half of paths did better, half worse.
Outcome Range (P5 → P95)
Near-floor to near-ceiling across all 10,000 simulated paths.
P(loss), MOIC < 1×
Probability the pool finishes below the original investment.

How returns change as the pool grows — Mean (flat), Median, 5th & 95th pct MOIC by pool size

Each point is the result of 500 simulated outcomes at that pool size. Watch the flat mean line — at α > 3, diversification compresses variance without raising expected return. The 5th-percentile floor rises; the 95th percentile compresses toward the mean. This is the late-stage power law at work.

Mean is flat with pool size at α > 3 — diversification reduces variance, not return.

Probability of losing money — by pool size

At any given failure rate and α, how often does a fund of n late-stage companies return less than 1× invested capital? Even a modest number of names dramatically cuts the chance of a loss. Toggle the partial-loss slider to see the post-2022 stressed regime.

Portfolio value distribution — 10,000 simulated paths

Each simulation traces one possible ~4-year fund outcome. The dark teal band shows where the middle 50% of outcomes land; the lighter band shows the middle 90%. The faint lines are 30 individual sample paths. Anything below the dashed line is a losing fund.

Company contribution breakdown — median-outcome path

For a single typical simulation, this shows where the fund's value comes from over time. At the late stage you will see far fewer breakout (≥10×) outcomes and far more moderate (1–3×) contributors — this is the late-stage power law behaving predictably, not disappointingly.

Full statistical breakdown

Complete distribution of the 10,000-path simulation. Percentiles describe where outcomes fall in the distribution; probabilities describe how often the fund clears specific return thresholds.

MetricValue
How this simulator works
For every simulation, each company in the fund draws an outcome at random. With probability equal to the failure rate slider, the company returns 0×. With probability equal to the partial-loss slider, it returns a uniform draw from 0.3×–0.7× (the post-2022 stressed-exit bucket). Otherwise it returns a draw from a power-law distribution with shape parameter α (capped at 30×). We repeat this 10,000 times and aggregate.

Default calibration
Failure rate 13% (PitchBook Q3 2024, Series D+) · α = 3.0 (Othman 2019, extended to late stage) · Partial-loss bucket off by default — turn on for post-2022 stress test · Holding period 4 years (Carta median Series C/D-to-exit)

Sources
Othman, AngelList (2019) · Koh & Othman, AngelList (2020) — empirical regime contrast from 10,665 LP portfolios · PitchBook Q3 2024, "VC Returns by Series, Part IV" · Treble Peak, Oct 2024 · Cambridge Associates US Venture Capital benchmarks, H1 2025

Important disclosures

Hypothetical and illustrative. The figures shown are simulated, not actual returns of any account, fund, or investor. No representation is made that any investor has achieved or will achieve results similar to those shown. Hypothetical performance has inherent limitations: it is prepared with the benefit of hindsight, does not involve real capital at risk, and may not reflect the impact of material economic and market factors.

Past performance is not indicative of future results. Model parameters are calibrated to PitchBook Q3 2024 Series D+ data and Othman (2019). The historical return distribution of the late-stage venture asset class is not a guarantee that future returns will follow the same distribution.

Model limitations. The simulation assumes equal-weighted holdings, independent outcomes across companies, no correlation between exits, no fees, no carried interest, no taxes, and no transaction costs. Net returns to an investor would be lower — potentially materially — after fees, expenses, carry, and taxes.

Risk of loss; illiquidity. Investments in private companies involve a high degree of risk, including the risk of total loss of capital. Private securities are illiquid and may not be readily transferable.

New product; no operating history. The Exceptional Company Liquidity Fund is a new product with no operating history. Structure and terms remain subject to change, and certain features (collateralized borrowing, secondary purchases, and §351 conversions) depend on regulatory, tax, and counterparty arrangements still being finalised.

Not an offer; not advice. This dashboard is for informational and illustrative purposes only. It is not an offer to sell, or a solicitation of an offer to buy, any security or interest in any fund, and is not investment, legal, accounting, or tax advice.

NAV milestones are targets, not guarantees, and are subject to market conditions and fund performance. The three additional liquidity paths (institutional debt access, secondary sale, and ETF conversion) may not be available to all pools or all investors, and are contingent on the fund reaching the stated NAV thresholds. Pricing for each liquidity path will be finalised at the time of the relevant transaction and is market-dependent. Nothing on this page constitutes a commitment, offer, or guarantee of liquidity.