Venture Capital Fund Portfolio Construction

Key Takeaways

Fund size sets the constraints: The amount of capital raised influences check sizes, company count, ownership targets, entry stage, and reserve capacity.
Diversification has a limit: More companies reduce single-company exposure, but positions that are too small may prevent successful investments from meaningfully affecting fund returns.
Ownership changes over time: Dilution, option pools, follow-on financings, and down rounds can make exit ownership materially different from the initial stake.
Construction should connect to LP outcomes: The model should translate into plausible net MOIC, net IRR, capital calls, and distribution timing.

A Public Example

Cambridge Associates' private-investment benchmark work organizes performance by asset class, vintage year, sector, and geography.

Vintage year matters: Venture outcomes are shaped by the market environment in which capital is deployed and exited.
The practical lesson: Portfolio construction should avoid overloading one vintage, one strategy, or one market cycle just because a manager is available today.
Useful number: Cambridge Associates cautions that private-fund performance may need about five to six years before relative rankings become meaningful.

Start With the Fund Size

Deployment example: A $50 million fund targeting 25 companies and reserving 50% for follow-ons has $25 million for initial checks, or an average of $1 million per company.

Fund size shapes almost every construction decision. As a fund grows, the manager may need to change how it invests.

Larger checks: More capital may be deployed into each company.
Later entry: The manager may move toward larger, more mature financing rounds.
Lower ownership efficiency: Competitive or later-stage rounds may provide less ownership for each dollar invested.
More companies: The portfolio may expand to absorb the additional capital.

LPs should ask whether the stated strategy still works at the proposed fund size.

Company Count and Ownership

Ownership decay: A $2 million investment in a $10 million round initially represents 20% of that round. Two later financings that each dilute holders by 20% reduce a 20% company stake to 12.8% before any pro rata investment.

There is no universally correct company count. The trade-off depends on position size, expected losses, and the number of investments that must drive returns.

More companies: Can reduce company-specific risk, but a large number of small positions may make it difficult for winners to move the fund.
Fewer companies: Can increase the impact of strong selection, but one or two losses may materially impair returns.

Visual analysis

Portfolio Construction Levers

Initial checks, follow-on reserves, fees, and liquidity buffers compete for the same fund capital.

Priority map Illustrative example

Initial investments57%

Follow-on reserves28%

Fees and expenses10%

Liquidity buffer5%

View chart data and assumptions

Data and assumptions for Portfolio Construction Levers
Area	Treatment
Initial investments	57%
Follow-on reserves	28%
Fees and expenses	10%
Liquidity buffer	5%

Illustrative allocation of committed capital. Actual construction depends on fund documents, strategy, and reserve policy.

Reserve Policy

Reserve capacity: If that same $50 million fund holds $25 million in reserves across 25 companies, it has an average of $1 million per company for follow-ons. Concentrating reserves in 10 winners would raise the average available to $2.5 million.

Reserves should reflect the stage and likely follow-on needs of the strategy.

Seed funds: May need disciplined reserves to support companies that begin to break out.
Later-stage or secondary strategies: May reserve less when positions are closer to possible liquidity.
Decision rules: The manager should define how reserves are allocated before the portfolio comes under stress.

Dilution and Ownership Decay

Initial ownership is not final ownership. Model the stake expected at exit rather than relying only on the entry percentage.

Option-pool expansion: New employee equity can dilute existing shareholders.
Later financings: Each round may reduce ownership unless the fund invests additional capital.
Pay-to-play provisions: Investors that do not participate may lose rights or economic position.
Insider and down rounds: New capital under difficult conditions can materially change the ownership structure.

Loss Ratio and Outlier Dependency

Return asymmetry: Carta notes that venture-company outcomes can reach 100x or even 1,000x, while many investments return zero. A construction model should show exactly how many outliers it requires.

Venture portfolios expect losses, but the return case should not depend on an implausible outcome.

A simple sensitivity analysis can ask:

Loss capacity: How many investments can fail before the fund misses its target?
Top-company delay: What happens if the largest position takes longer to exit?
Valuation reduction: How does a material markdown change expected fund returns?
Lower exit: What happens if the leading company exits at half the base-case value?

Concentration Limits

Company count alone does not show whether a portfolio is genuinely diversified. Review concentration from several angles.

Cost and fair value: A position may become much larger as its valuation rises.
Sector and theme: Different companies may still rely on the same demand drivers.
Geography: Regulatory, currency, and market conditions can create shared exposure.
Sponsor and financing cycle: Several investments may depend on the same investor network or capital environment.
Exit market: A portfolio can hold many companies while relying on a narrow route to liquidity.

Connect Construction to LP Outcomes

The portfolio model should translate into results that matter to an LP:

Net MOIC: The total value expected after fees, expenses, and carried interest.
Net IRR: The annualized return implied by the timing of calls and distributions.
Capital-call pacing: When and how quickly the LP may need to fund its commitment.
Distribution timing: When realized proceeds may reasonably return to investors.

See MOIC vs IRR and return sensitivity for related frameworks.

Public deal case study

CalPERS: construction rules before individual selections

CalPERS' June 2024 policy set private-equity strategy ranges and delegated transaction limits. It is a useful public example of portfolio construction operating as a rule set rather than a list of preferred managers.

6% Venture target

The permitted venture-capital range was 0% to 12% of private-equity NAV.

25% Growth target

Growth and expansion carried a 5% to 30% range.

$1.5B CIO co-investment authority

Commitments above delegated limits required committee approval.

What it shows: The specific limits are CalPERS' and are not a model allocation for other LPs. The transferable point is that strategy weights, ranges, delegation, and exception handling should be defined before attractive transactions arrive.

Primary sources: CalPERS, June 2024 investment policy. Public transaction evidence only; this is not represented as a Frontierspace investment or result.

Frequently Asked Questions

How many companies should a venture fund hold?

Short answer: There is no single correct number. Company count should be assessed alongside ownership targets, reserve policy, stage, fund size, sector exposure, and the return contribution required from outliers.

Why do venture funds maintain reserves?

Short answer: Reserves allow follow-on investment in selected companies and can protect ownership through later rounds. They can also compound mistakes if follow-on decisions lack discipline.