Portfolio Construction: The Logic of Allocating Startup Investments

1. The rule that governs everything: returns are asymmetric

In early-stage investing, outcomes don't cluster around a mean. The recurring pattern is that a large share of investments fail or merely return capital, while a small minority produce most of the return.

Three practical consequences follow:

• Number isn't a luxury. A portfolio of three companies is a bet that one of them lands in the rare minority — a low-probability proposition.
• Partial loss isn't a flaw in the method. It's the structure of the asset class, not a deviation from it.
• Your stake in the winner matters more than the count of losers. Which makes follow-on reserves a decisive element.

2. Number of investments

There's no magic figure, but the logic is clear: the number that gives you a reasonable probability of containing at least one exceptional outcome. For an individual investor, that usually means a portfolio of dozens of small positions built gradually over years, not deployed at once.

Crowdfunding platforms and funds provide access to that count with smaller amounts per deal, which is among their strongest justifications for an individual investor.

3. Cheque size and follow-on reserves

Divide your allocation to this class into two parts:

• Initial capital: deployed into new investments.
• Follow-on reserve: held back to participate in later rounds of companies showing performance.

An investor who spends the entire allocation on initial positions finds themselves unable to increase their stake in the one company that worked — and their ownership dilutes round after round. The conservative rule is to reserve a meaningful portion of the total allocation for follow-on rather than deploying all of it on entry.

4. Diversification across three axes

Sector. Concentration in one sector means a single regulatory or technological shock hits the whole portfolio.

Stage. Very early stage carries higher risk and a longer horizon; later stage carries lower risk and lower multiples. Mixing the two balances the profile.

Vintage (year of investment). This axis is the most neglected and the most important. Spreading investments across several years protects you from concentrating the entire portfolio in a year when valuations happened to be high. Do not deploy a full portfolio within twelve months.

5. Horizon and liquidity

Startup investment is illiquid. There is no organised market for selling your stake at will; liquidity typically arrives through an exit event or a secondary transaction, and that can take many years.

Two practical consequences: don't invest money in this class that you may need within a few years, and don't value the portfolio on round valuations alone — a paper mark is not cash.

6. How much of your wealth?

The general principle is that this class receives a limited share of total assets, set according to the investor's capacity to absorb the complete loss of that portion without affecting their obligations. More important than the percentage itself is that it be fixed in writing beforehand — before an attractive opportunity appears and tempts you past it.

7. Tracking and documentation

For each investment, record: the entry thesis (why you invested), what must happen and by when, and the core terms. Review the log annually. This habit is what turns investing from a hobby into a method, and it's also what reveals your recurring errors.

Common mistakes

• A portfolio too small in number to withstand the return distribution.
• Spending the full allocation on entry with no follow-on reserve.
• Concentrating all investments in a single year.
• Investing money you may need soon in an illiquid asset.
• Investing without a written thesis, making it hard later to distinguish a bad decision from bad execution.
• Doubling down on the loser to justify the first decision.

Checklist

• Allocation as a share of total assets fixed in writing
• Target number of investments consistent with the return distribution
• Follow-on reserve defined and held back
• A multi-year deployment plan
• Sector and stage diversification
• Clear acceptance of illiquidity and a long horizon
• A thesis log reviewed annually

FAQ

Should I invest directly or through a fund?
Direct gives control and demands time and network; a fund gives diversification and expertise in exchange for fees and less control. Combining both is common.

When should I increase my position in an existing company?
When what your original thesis specified has come true — not when the company needs money.

How do I assess portfolio performance early?
By the operational progress of the companies rather than paper valuations. Value is realised only at exit.