Portfolio Construction for VC Funds: How Many Bets and How Much Per Bet

The most consequential decision a fund manager makes isn't which company to invest in. It's how to construct the portfolio. How many companies. How much per company. How much to reserve for follow-ons. These decisions, made before a single deal is done, determine the fund's return profile more than any individual investment.

Most emerging managers get this wrong. They either spray too thin (too many small bets), concentrate too heavily (not enough bets for the power law), or botch their reserves (no dry powder when winners need follow-on). Let's build a portfolio model from scratch.

The Power Law: Why Portfolio Construction Matters

Venture capital returns follow a power law distribution. In a typical fund, 1-2 investments generate the majority of total returns. The rest are a mix of modest winners, breakeven outcomes, and total losses. Data from Cambridge Associates shows that in top-quartile funds, the single best investment often returns more than all other investments combined.

This has a profound implication for portfolio construction: you need enough bets to give yourself a reasonable chance of finding that one outlier. But you also need enough concentration that the outlier meaningfully moves the fund's return. Invest in 100 companies with tiny checks, and even a 1000x return on one investment barely matters. Invest in 5 companies, and you might miss the outlier entirely.

How Many Investments by Fund Stage

Pre-seed/Seed funds: 25-40 companies. At the earliest stages, failure rates are highest and outcomes are most unpredictable. You need a wider portfolio to increase your chances of catching a breakout. Seed funds accept that most investments will fail in exchange for the possibility that 2-3 return the entire fund.

Series A funds: 15-25 companies. At Series A, you have more data — product-market fit signals, early revenue, team cohesion. This allows for slightly more concentrated bets. Larger check sizes require larger ownership targets, which means fewer but more conviction-driven investments.

Growth funds: 10-15 companies. At growth stage, the risk profile is different. Companies have proven business models and significant revenue. Failure rates are lower, but so are the return multiples. Growth funds make fewer, larger bets and target 3-5x returns rather than the 10-100x that early-stage funds chase.

Check Size Math: Building the Model

Let's model a $50M seed fund. Start with the total fund size and subtract management fees. Over 10 years at 2%, that's roughly $8M in fees, leaving $42M for investments. Now decide your allocation: how much for initial checks versus follow-on reserves.

A common split is 40-50% for initial investments and 50-60% for reserves. Let's use 50/50: $21M for initial checks, $21M for follow-ons. At $1.5M average initial check size, that's 14 companies. Too few for seed. Reduce check size to $800K, and you get 26 companies with $21M in reserves. Now you have enough bets.

A larger $100M seed fund might write $1.5M initial checks into 25 companies ($37.5M) and keep $44.5M in reserves. Or it might write $2M checks into 20 companies ($40M) and reserve $42M. The math must work backward from both the number of bets you need and the ownership you want.

Reserve Strategy: When to Double Down

Reserves are where portfolio construction gets interesting. The typical reserve ratio is 50-60% of total investable capital reserved for follow-on investments. This means for every dollar you invest initially, you're holding back $1-1.5 to invest in the winners later.

The decision of when to deploy reserves is one of the hardest in venture. The instinct is to support all your portfolio companies. The math says you should concentrate follow-on in your best performers. If 5 of your 25 companies are clearly breaking out, put most of your reserves there. The ones that are struggling need a different kind of help — not more capital from you.

Under-reserving is a common mistake for first-time managers. They deploy too much in initial checks, then can't follow on when their best companies raise follow-on rounds. They get diluted down in the winners and over-concentrated in the losers. The opposite mistake — over-reserving — means you're sitting on too much dry powder and not making enough initial bets.

Ownership Targets by Stage

Ownership targets determine your return math. If you own 10% of a company that exits at $1B, your fund gets $100M. If you own 1%, you get $10M. On a $50M fund, $10M is 0.2x — barely noticeable. $100M is 2x — fund-returning.

Typical ownership targets: seed investors aim for 10-15% initially. Series A investors target 15-25%. Growth investors target 5-15%. These targets account for dilution — if you buy 15% at seed, you'll be diluted to 8-10% by Series B unless you actively follow on to maintain ownership.

Concentration vs Diversification

There are two valid but very different philosophies. The diversified approach (Tribe Capital, SV Angel) spreads bets across 30-40+ companies, accepting that most will fail but maximizing the chance of catching an outlier. The concentrated approach (Benchmark, Founders Fund) makes 10-15 high-conviction bets, accepting higher variance but generating larger multiples when they hit.

Both can work. The diversified approach produces more consistent returns with less variance. The concentrated approach produces more volatile returns but higher potential upside. Your choice should depend on your fund size, your LP base (endowments tolerate variance better than family offices), and your own conviction in your ability to pick winners.

Recycling: Stretching the Fund

Recycling means reinvesting early returns back into new investments instead of distributing them to LPs. If a portfolio company gets acquired in year 2 and returns $3M, you can reinvest that $3M into new deals. Most LPAs allow recycling of 10-20% of committed capital during the investment period.

Recycling is a double-edged sword. It increases your deployed capital — a $50M fund might deploy $55-60M total. But it delays distributions to LPs, extending the time before they see cash. Use recycling strategically: reinvest returns from quick exits and small wins, and let the big exits flow through to LPs.

Common Portfolio Construction Mistakes

Over-reserving: Holding 70% in reserves means you're only making initial investments with 30% of your fund. That's not enough for adequate diversification unless your fund is very large. You'll end up with too few initial bets.

Under-reserving: Deploying 80% in initial checks feels great in year 2 when you're doing lots of deals. It feels terrible in year 4 when your best company raises a Series A and you can't participate. You watch your ownership get diluted from 12% to 6% because you had no dry powder.

Chasing ownership: Insisting on 20% ownership at seed when the round is oversubscribed and the founder is only offering 7-8%. You pass on what becomes a $5B company because you couldn't get your target ownership. The return on that 8% would have been fund-defining.

Portfolio drift: You told LPs you'd do fintech seed deals. By year 3, half your portfolio is climate tech Series A. This isn't evolution — it's drift. Stick to your thesis or formally communicate the pivot to LPs.

Building Your Portfolio Model

Every fund should have a written portfolio construction plan before making their first investment. It should specify: target number of initial investments, average initial check size, reserve ratio, follow-on criteria, ownership targets, and sector allocation. Revisit this plan quarterly and measure actual deployment against the model.

Model different scenarios with the Fund Economics Simulator at /tools — plug in your fund size, check sizes, and reserve ratio to see how the math plays out. The Portfolio Construction module at /academy/portfolio-construction goes deeper into the theory and practice. For a comprehensive guide to building your first fund, start with the Emerging GP learning track at /learn/emerging-gp.