Skip to content

Investor Lens

Angel investor, the unicorn math needs every winner to grow 400x from your entry valuation

Petri Lehmuskoski ·

Also available as a podcast episode → Listen on Spotify · Podcast

We passed on a €15M SAFE because no trade sale could pay it back

We were offered a SAFE (a contract, not a deal) at a €15M valuation. The cap was €17M and the floor €13.5M, so nobody could tell us what ownership the money would buy. The company had a few paying customers and a "strong" pipeline. It was not for us. The exit mathematics did not work for a trade sale exit.

Many angels would have signed it. In my experience most angels do not make that math, or do not understand it, and SAFE notes are pushing entry prices even higher. So here is the math, for an angel who backs a company on the unicorn path and holds the shares until the exit.

Ten rounds of dilution leave the first angel 13% of the stake

A company on the unicorn path raises money often. In my experience ten rounds, bridges included, is typical for these companies, and the number of rounds is often driven by the VCs' need to invest more.

Make the angel's round the first of the ten, and let each of the nine later rounds sell 20% of the company. Each round leaves every existing shareholder with 80% of what they held before. After nine of them the angel holds 0.8 to the ninth power of the original stake: 13.4%. Most angels do not make follow-on investments, so that is the stake they take to the exit. Of every euro of value the company creates after the angel's entry, the angel keeps about thirteen cents.

Then add the holding period. Ownership in unicorn-path companies always runs 10 to 15 years. I use ten years in every number below. Real holds are usually longer, and a longer hold makes every number worse.

Preferences cost the angel nothing until the exit falls below the last round

Dilution shows on the cap table. The second cost sits in the shareholders' agreement: the liquidation preference, the right of an investor to take its money back before anyone else shares the rest.

Take the common structure. Every round carries a 1x preference, the latest round is paid first, and nobody participates on top of the preference. At an exit at or above the last round's price, every investor does better converting to ordinary shares, and the preferences cost the angel nothing. Below that price, the investors take their money back in reverse order, and the angel stands at the bottom of the stack. Together the preferences equal everything the company ever raised. In my model that is 40 to 60% of the last round's valuation, depending on how fast the price rose between rounds. An exit at that level leaves the angel close to nothing, however much the company grew from the angel's entry.

Seniority is not rare. In Fenwick & West's survey of Silicon Valley financings during the 2008 and 2009 downturn, 40 to 49% of rounds made the new investors senior to the earlier ones. Preference multiples were rarer: 11 to 28% of those senior preferences were above 1x, most of them between 1x and 2x.

The multiples still matter, because they move the line. In a stress case with a 3x preference on the last two rounds, an exit at exactly the last round's price pays the two preference layers 90% of the price. The angel keeps 16% of a plain pro-rata share.

Late-round valuations are inflated by the terms behind them

A late investor protected by preferences can pay a high headline price, because the money comes back first.

Will Gornall and Ilya Strebulaev measured this in 135 US unicorns (Journal of Financial Economics). Reported post-money valuations averaged 48% above fair value once the investors' protective terms were priced in. Ordinary shares were overvalued by 56%, and 65 of the 135 companies lost unicorn status after the correction.

For the angel this is the dangerous case. An exit at fair value lands at about two-thirds of the last round's price, below the line where preferences start to work. With 1x preferences the angel then keeps 70 to 87% of a plain pro-rata share in my model. In the 3x stress case it keeps nothing. The company can do well and the angel still get nothing, because its last round was priced on the terms.

A 20% portfolio needs every winner to grow 400x from your entry valuation

One deal is the wrong unit. Angels need to invest and calculate at portfolio level, because a handful of exits carries the whole return. In Robert Wiltbank and Warren Boeker's 2007 survey of 539 US angels for the Kauffman Foundation, 7% of exits returned more than ten times the money and produced 75% of all the cash.

So take a portfolio of ten companies, built on that angel data. Six lose everything. Three return the money around year five. One winner carries the portfolio at year ten. An angel should aim at 20 to 30% a year at portfolio level, because results always come in lower than the target.

At a 20% target the winner has to return 54 times the money. At 30%, 127 times. The angel holds 13.4% of the original stake at the exit, so the company has to grow about 400 times from the angel's entry valuation for 20%, and about 950 times for 30%. That is dilution alone. If the exit lands at the last round's price in the 3x stress case, the 20% target needs the company to grow about 2,600 times.

A unicorn-chasing pre-seed fund puts its own version of this bluntly: one check has to return the whole fund. Most angels run no version of it at all.

At a low enough entry price, the multiples become possible

The entry valuation turns the multiple into an exit price. Required exit equals entry valuation times 400. At a €2M entry that is a €0.8B exit. At the €15M SAFE it is €6B, before any preference has taken its share.

I have followed this math from entry to cash on both paths: unicorn-path companies where dilution and the terms took the return, and trade-sale companies where a low entry price made a high multiple possible. On both paths, what we paid at entry decided which multiples were possible. I have watched that over 18+ years as a startup investor.

This is not an argument against unicorn bets. Some companies need venture-scale capital to be built at all, and for an angel who enters them early and cheap, these multiples can happen. The price is where the math breaks. When an angel pays €15M for a company with a few paying customers, the required exit lands in the billions.

The rounds, the terms and the years in this math are set by other people after the angel has signed. The entry valuation is the one number the angel sets, and it is set on the day the money goes in.

The model is mine: ten rounds of 20% dilution, a ten-year hold, no follow-on investment, 1x stacked preferences with a 3x stress case, and a ten-company portfolio shaped on the Kauffman angel data. The figures are illustrations of the arithmetic, not observed returns. The company offering the SAFE is not named.

More scars than trophies.