Build a venture fund return model in your browser. Set your fund size, check size, ownership, reserves, fees, carry, and a power-law outcome distribution β and watch Gross/Net TVPI, DPI, and Net IRR update live.
40% reserved for follow-ons
$158M total value
after $10M fees + $22M carry
over a 10-year fund life
Drag the sliders or type a value. Everything recomputes instantly.
Total committed capital.
First check into a new company.
Ownership you aim for at entry.
Share of the fund kept for follow-ons.
Charged each year over the fund life.
GP share of profits above paid-in.
Years from close to wind-down.
The power law: what % of portfolio companies return each multiple. Sum: 100%
Each bucket a different share of the portfolio.
The 10x+ winners drive 79% of total value.
Value dips early from fees, then distributions (DPI) ramp from year ~4 to catch up to net TVPI.
Simplified model for illustration. Gross multiple = Ξ£(% Γ multiple). Fees = fee% Γ years Γ fund size. Carry = carry% Γ profit above paid-in. Net IRR = (netΒ TVPI)^(1/years) β 1. The J-curve is illustrative.
A venture fund model answers one question: given a strategy and a realistic spread of outcomes, what will this fund return to its LPs? It chains together three things β portfolio construction (how many companies, at what check and ownership), the outcome distribution (the power law of wins and losses), and fund economics (management fees and carried interest).
Step 1 β Portfolio construction. Carve out a follow-on reserve (commonly 40β50% of the fund), then divide what's left by your average initial check to get the number of new companies you can back. Fewer, bigger checks mean higher ownership but fewer shots on goal.
Step 2 β Outcome distribution. Venture returns are not normally distributed β they follow a power law. A realistic portfolio writes off ~half its companies, sees a band return 1β3x, and relies on one or two outliers returning 10xβ50x to carry the whole fund. The gross multiple is the weighted average of those outcomes.
Step 3 β Fund economics. Management fees (typically 2%/year over the fund life) and carried interest (typically 20% of profits) sit between the gross result and what LPs actually keep. That gap is why net TVPI is always lower than gross β and why the outcome distribution has to be strong enough to clear the fees before any carry is earned.
Fund modeling (or fund construction modeling) is the process of projecting how a venture fund will perform before β or while β it is deployed. You start with the fund size and a strategy (average check, target ownership, reserve ratio) to estimate how many companies you'll back, then layer an outcome distribution (how many will fail, return capital, or be big winners) and the fund's economics (management fees and carried interest) to estimate returns to LPs as TVPI, DPI, and IRR.
TVPI (Total Value to Paid-In) measures total value created per dollar invested; a top-quartile early-stage fund typically targets 3x+ net TVPI, with 2x considered solid and below ~1x a loss. DPI (Distributions to Paid-In) is cash actually returned β strong funds reach 1x DPI before re-investing further and aim for 2β3x by wind-down. Net IRR for a strong early-stage fund is roughly 20β30%+; anything sustainably above 25% net is excellent.
Because venture returns follow a power law, most funds need enough 'shots on goal' to catch an outlier. Seed funds commonly make 25β40+ initial investments, while concentrated Series A funds may make 15β25 with larger checks. The math is simple: initial investments β fund size Γ (1 β reserve %) Γ· average initial check. Reserving 40β50% for follow-ons reduces the number of new names but lets you double down on winners.
The power law describes how a small number of investments generate the vast majority of a fund's returns. In a typical portfolio, ~50% of companies return little or nothing, a middle band returns 1β3x, and one or two outliers returning 10xβ50x+ drive the entire fund. This is why fund modeling weights the outcome distribution so heavily β your fund's result is dominated by the tail, not the average company.
Built and maintained by Trace Cohen Β· @Trace_Cohen Β· t@nyvp.com. This tool is for illustration and education β not investment advice.