Venture capital: funding rounds, valuation, the power law and what research knows about returns

In depth12 min read · 16 sources

Venture capital is equity for young, fast-growing companies that are not yet profitable and often have hardly any revenue. The investors are usually funds that buy minority stakes in several funding rounds and only get their money back through a sale or an IPO. This article explains rounds and valuation, works through a cap table over three rounds including the exit proceeds, shows why a few hits carry an entire fund, and summarizes the research on return and risk.

What venture capital is

A venture capital fund collects capital commitments from investors such as pension funds, foundations, insurers and funds of funds and invests them in start-ups over a few years. Unlike a buyout fund, it finances growth rather than takeovers and expects many of its holdings to fail. Money flows back almost exclusively through an exit: a sale to a buyer, an IPO or the sale of the shares to other investors. How capital calls, the J-curve and metrics such as TVPI and DPI work is explained in the article measuring private markets.

Gompers and Lerner (2001) describe venture capital as an intermediary that provides capital to young technology companies that would often not be financed at all otherwise, and that has developed its own mechanisms against information and incentive problems for each stage. Typical tools are staged financing, board seats and preferred shares with special rights. Kerr et al. (2014) see entrepreneurship as experimentation: the probabilities of success are low, extremely skewed and unknown before the investment.

Funding stages: from pre-seed to growth

Each round is meant to carry the company to the next milestone. If that succeeds, the price per share rises (up round); if not, a round at a lower price follows (down round), or the company folds.

Valuation: pre-money, post-money and dilution

In each round, founders and investors negotiate a pre-money valuation, the value before the new money. This and the number of shares before the round (calculated on a fully diluted basis, i.e. including all options) give the price per share. New shares reduce the percentage held by existing shareholders: this is dilution. It is not a loss as long as the value per share rises: the slice gets smaller, the cake bigger.

One detail with a big impact is the option pool, the shares reserved for employee equity. If the investors require the pool to be created before the round, it sits in the pre-money valuation and dilutes only the existing shareholders. The effective valuation of the founders' shares is then below the negotiated figure.

Contract terms: liquidation preference, anti-dilution, pro rata

Investors usually receive not common shares but preferred shares with special rights. Using real contracts, Kaplan and Strömberg (2003) showed that venture financings allocate cash flow rights, voting rights, board seats, liquidation rights and other control rights separately, and that this allocation changes from round to round. The most important terms:

• Liquidation preference 1× non-participating: in a sale or liquidation, the investor first gets back the capital invested, or converts into common shares and takes its share of the proceeds, whichever yields more.
• Participating preferred: the investor gets its capital back and then also shares in the remainder (“double dip”), sometimes up to a cap. Multiple preferences (2× or 3×) shift even more of the proceeds to the investors.
• Anti-dilution protection: after a down round, the conversion price of earlier investors is reduced, either all the way to the new price (full ratchet) or according to a weighted average.
• Pro rata right: the right to take part in later rounds to the extent that one's own stake stays the same. This allows funds to invest further in their winners.

Gornall and Strebulaev (2020) revalued 135 US unicorns (private companies with a reported valuation above 1 billion US dollars) taking their contract terms into account. The reported post-money valuations were on average 48% above fair value, common shares were overvalued by 56%, and 65 of the 135 companies lost their unicorn status after the correction.

Fund economics: why a few hits carry the fund

Because most start-ups fail and a few become very large, the proceeds of a venture fund are extremely skewed. This is known as the power law. Hall and Woodward (2010) show this for founders: almost three quarters of the founders of venture-backed companies received nothing at exit, a very few more than a billion US dollars, and the average was 5.8 million US dollars. The stock market shows the same pattern in milder form: according to Bessembinder (2018), the best 4% of listed US companies account for the entire net gain of the US stock market since 1926 (see few big winners). An example with made-up, round numbers:

Mathematically, it therefore pays to look for companies that can return the whole fund on their own, and to invest further in winners via pro rata rights. The fees, by contrast, flow with certainty: according to Metrick and Yasuda (2010), about two thirds of the expected revenue of private equity fund managers comes from fixed components unrelated to performance. Buyout managers grow their funds faster after successes than venture managers do: the venture business is harder to scale.

Corporate VC, venture debt and secondaries

• Corporate venture capital (CVC): large companies invest in start-ups through their own units, often for strategic reasons. Chemmanur et al. (2014) found that companies backed by corporates are younger, riskier and less profitable, but produce more patents than companies with independent investors.
• Venture debt: loans to start-ups that already have venture capital, often with warrants on shares as additional compensation. According to Hochberg et al. (2018), this market depends on saleable patents as collateral and on the commitment of the equity investors: after the 2000 crash, lenders mainly continued to finance start-ups whose investors had recently provided money and could credibly put in more.
• Secondaries: investors sell fund interests before the end of the term, or founders and early investors sell shares directly. Prices can differ considerably from the reported value (more under private equity).

What the research shows

• Selection bias: start-up valuations are only observed at events such as new rounds, sales or IPOs, and these occur more often at successful companies. Cochrane (2005) corrected for this and lowered the mean log return from 108% to 15% a year and the intercept in the market model from 92% to −7%. Because of the enormous dispersion, arithmetic means remain high: the correction lowered the mean arithmetic return from 698% to 59% and the arithmetic alpha from 462% to 32%.
• Korteweg and Sorensen (2010) developed a more general model of this dynamic selection; with the correction, alphas come out considerably lower and risks higher than in earlier studies.
• Fund returns: according to Kaplan and Schoar (2005), private equity funds, including venture funds, earned on average roughly the return of the S&P 500 after fees, with wide dispersion. Using cash flow data from almost 1,400 US funds, Harris et al. (2014) show that venture funds outperformed the stock market in the 1990s and lagged it in the 2000s.
• Persistence: according to Harris et al. (2023), the performance of venture funds from the same manager remains remarkably persistent, though weaker after 2000; for buyout funds, persistence has declined markedly.
• Limits of the model: Lerner and Nanda (2020) criticize the fact that only a narrow band of technologies fits institutional venture funds and that a few investors steer a large share of the capital.

For advanced readers: selection, skewness and diversification

• How the bias arises: a valuation is only observed when an event occurs, and the probability of that event rises with the value. If you average observed returns, you estimate E[r | observed] instead of E[r]. Cochrane (2005) and Korteweg and Sorensen (2010) estimate the selection jointly with the returns.
• Arithmetic and geometric: the gap between 59% and 15% shows the enormous dispersion; for an investor with only a few holdings, the geometric return is what counts (see volatility drag).
• Diversification: if each start-up independently has a 5% chance of a 25× hit, a portfolio with n holdings contains at least one with probability 1 − 0.95^n: 40.13% with 10 holdings, 64.15% with 20 and 87.15% with 40. Our assessment: independence is optimistic, because valuations and exit opportunities depend on the shared market environment.
• Smoothing: fund reports value holdings at the last round or using a model; measured volatility, correlation and beta therefore come out too low. For unsmoothing, for example following Geltner (1993), see measuring private markets and private credit.

Access for private investors

Venture funds are mostly aimed at institutional investors. For private investors, the main routes are the following. Whether and how much of them suits you depends on your horizon, liquidity needs and capacity to bear losses (this is not investment advice):

• Listed private equity companies: tradable daily, but the share price can deviate considerably from the reported net asset value and fluctuates with the stock market.
• Business angels and equity crowdfunding: direct stakes in individual start-ups. Given the power law, every single stake carries a risk of total loss, and broad diversification is hardly achievable with small amounts.

Risks, limits and common mistakes

• Total loss and concentration: most holdings return little or nothing; without broad diversification, chance decides.
• Illiquidity: capital is tied up for many years and called as needed; selling before the end of the term is only possible on the secondary market (see illiquidity premium).
• Valuations are not prices: interim values are based on the last round or on models. As Gornall and Strebulaev (2020) show, post-money valuations overstate the value of common shares.
• Samples: databases mainly contain funds and companies that report (survivorship bias); results from older vintages can only be transferred to a limited extent.

How Portfolio Lab calculates it

Portfolio Lab has no time series for venture capital: illiquid segments such as direct private equity and venture capital are not available as a freely accessible monthly series. As a listed proxy, Asset classes offers Listed private equity, chained from two funds that invest in listed private equity companies. This series fluctuates like equities and does not reflect how venture funds are valued. In Look-through, the taxonomy assigns funds with “Venture” in their name to the category Private markets → Venture capital, with the risk driver Illiquidity premium (private equity & VC) and equity beta as a second driver; because the rule only looks at the word, venture debt funds end up there too. The app does not calculate cap tables, liquidation preferences or unsmoothing. More in the asset classes methodology and the look-through methodology.

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.