Building an early-stage investment portfolio

Most people approach early-stage investing as a picking problem. They read about diligence, learn to interrogate a deck, and treat the quality of individual selection as the whole discipline. Selection matters. But in an asset class where a small number of positions account for most of the return, and where a substantial proportion return nothing at all, how you construct the portfolio does at least as much work as which companies you choose.

This article covers the construction decisions: how much to allocate, how many positions, at what size, over what period, and what to hold back. It assumes you are already comfortable assessing an individual company — if not, start with assessing risk in early-stage companies.

Start from dispersion

The defining characteristic of early-stage investing is not that the average outcome is poor. It is that the outcomes are wildly unevenly distributed. A portfolio's result is usually determined by a small number of positions, and there is no reliable way to identify in advance which ones they will be — if there were, everyone would hold only those.

Two consequences follow, and almost everything else in portfolio construction is downstream of them. First, you need enough positions for the distribution to have a chance to work in your favour. Second, no single position should be sized such that its failure changes your life, because a meaningful proportion of them will fail.

Deciding the allocation

Work top-down. Decide what proportion of your total investable assets belongs in unquoted early-stage companies as a whole — an allocation question your adviser is better placed to answer than any article — and treat that as the boundary. Then divide it into positions rather than deciding position by position and discovering your exposure afterwards.

Doing it in that order prevents the most common accident in this asset class: an investor who intended to make three investments makes eleven over four years, each individually reasonable, and ends up far more exposed than they would ever have chosen deliberately.

How many positions

Reserves are the decision most often skipped. It feels inefficient to hold cash while good opportunities pass, but a portfolio fully deployed at entry has no way to increase exposure to the companies that are working — which is the only information advantage a portfolio actually generates over time.

Diversify across what, exactly

Number of positions is only one dimension. Two portfolios of fifteen companies can carry very different risk if one is fifteen fintech seed rounds made in the same year.

• Sector: different demand drivers and different regulatory exposure.
• Stage: earlier positions offer more upside and more failure; later ones cost more per unit of ownership but carry more evidence.
• Vintage: spreading entry across years spreads exposure to valuation environments.
• Business model: recurring software, marketplace and transactional businesses fail for different reasons and at different points.
• Source: several routes into opportunities rather than a single network or channel.

Our own portfolio illustrates the sector and model point rather than the diversification advice: threat intelligence, attractions management, an SME marketplace, commercial finance and investment software are all different businesses with different failure modes. You can see them on the portfolio page.

Where SEIS and EIS fit

For UK investors, the tax schemes are a portfolio-construction tool as much as a tax one. Because SEIS and EIS reduce the effective cost of a position and provide loss relief on failures, an investor using them can often hold a wider spread for the same effective exposure. In an asset class defined by dispersion, being able to hold more positions is a structural advantage, not a marginal one.

There are construction implications specific to the schemes. Annual investor limits shape pacing. The three-year minimum holding period sets a floor on your time horizon. Carry-back to the previous tax year gives some flexibility on timing. And the reliefs are only useful to the extent you have a tax liability to set them against. The mechanics are covered in SEIS and EIS explained for investors.

One warning worth repeating: never let the tax treatment drive the selection. A relief on a company that should not have been funded is a smaller loss, not a gain.

The part after the investment

A portfolio is not a set of decisions made once. Over a holding period measured in years, the work that matters is unglamorous:

1. 01Read the updates. Companies that stop reporting are telling you something.
2. 02Track your actual exposure by sector, stage and vintage — not the exposure you intended.
3. 03Decide follow-on participation deliberately, against the evidence, not against your attachment to the company.
4. 04Keep your SEIS3 and EIS3 certificates and mind the three-year conditions.
5. 05Revisit, annually, whether your total early-stage allocation still reflects your circumstances.

Venture-built companies in a portfolio

Companies created through a structured build can be a useful component of an early-stage portfolio because the stage a company has reached is unusually legible — the seven-stage process defines what has been evidenced before each funding point, which helps with pacing and sizing decisions that are otherwise guesswork.

They are a component, not a portfolio. Ventures from a single builder share people, methods and judgement, so exposure to them should be treated as correlated even where the sectors differ. The investor overview sets out how private, institutional and corporate investors access these companies through GCV Invest, and how GCV Labs and GCV Invest work together explains the structure behind that.

The discipline, in the end, is simple to state and difficult to hold: decide your allocation before you fall in love with a company, spread it wider than instinct suggests, pace it over years, keep something back, and only invest what you can leave alone.