Ask what angels actually earn and you get a number that sounds almost respectable next to public equities: around 2.5 times the money, over roughly five to seven years, which the studies tend to express as an internal rate of return - the annualised growth rate, smoothing out the long wait for an exit - somewhere in the mid-twenties per cent. It's the figure that launched a thousand syndicates. It is also, taken at face value, one of the more misleading statistics in private markets.
Not because it's wrong. Because of what an average does to a set of returns shaped like this one.
What's the headline return angels have earned?
The data we have comes from a handful of large studies that tracked diversified angel portfolios from cheque to exit - notably work by Robert Wiltbank in the United States and a parallel study he ran with Nesta on UK angels. The UK research, looking at British angels' realised exits, landed on a gross multiple of about 2.5x over an average holding period a little under four years for the deals that exited, with the bulk of the money coming back well beyond that. The American numbers sit in a similar band. Across both, the implied internal rate of return clusters in the mid-twenties per cent, before any tax relief and before fees.
Sit with the word gross. These are returns on capital that was actually deployed into deals that eventually resolved, measured before the costs of running a portfolio and before the tax reliefs that, in the UK, materially change the picture. They aren't net-of-everything numbers you could quietly assume for your own book.
The average angel return is real. The average angel does not earn it.
Why does the average mislead so badly?
Because the returns aren't spread evenly around it. Early-stage outcomes follow a power law: the great majority of individual investments return less than was put in, a chunk return nothing at all, and a small handful return many times the cheque and drag the whole average upward. The same studies that produce the 2.5x headline also report that something like half or more of individual exits came back below cost. The mean is buoyed by a thin tail of outliers - the one deal in twenty or thirty that returns ten, twenty, fifty times.
So the average describes a basket, not a participant. An angel who made a wide spread of investments and held to exit might have approached that distribution. An angel who made four bets and stopped almost certainly didn't - the most probable outcome of a tiny portfolio, given those base rates, is a loss, because the rare winner that carries everything never showed up. The headline figure is the reward for diversification and persistence, not a number that attaches to any single deal.
What do the return figures leave out?
Three things, mostly, and each one pulls in a different direction.
Survivorship and the living dead. Exit studies can only measure deals that exited. The companies that neither soared nor died - the ones that limped along for a decade, never sold, never wound up, quietly worth a fraction of what went in - are hard to capture, and tie up capital without ever showing as a loss. Self-reported data has the reverse problem: angels who did well are likelier to fill in the survey. Treat any clean headline as flattered.
Fees and the cost of access. Investing through a syndicate, platform or fund usually means management charges and a share of the upside - carry - paid to whoever runs it. Those come off the gross number. The 2.5x is what the deals produced, not what necessarily reached the investor's pocket after the people organising the round took their cut.
Tax relief, in the other direction. For UK investors who qualify, the SEIS and EIS reliefs work the gross figure the other way - returning part of every cheque up front, exempting qualifying gains, and cushioning the losses. They sit outside the studies' headline multiples, and for a UK taxpayer they're not a footnote. More on that below.
How long is the money tied up?
Longer than the multiple alone suggests, and that's the part the IRR is doing quiet work to disguise. A 2.5x return is a very different thing over four years than over ten, and real angel portfolios skew towards the latter. The winners that carry a book typically take somewhere between five and ten years to reach a sale, and a good many never get there. Until then the shares are illiquid - no daily price, no ready buyer, no way to mark your position to anything but hope.
The three-year minimum holding periods on SEIS and EIS are tax conditions, not return timelines. Holding for three years protects the relief; it says nothing about when - or whether - a company exits. Reading "three years" as a horizon for getting money back misreads the instrument.
Where do SEIS and EIS fit?
This is the lever that makes UK angel returns genuinely distinct from the raw American figures, and it operates entirely outside the gross multiples the studies quote. Two HMRC schemes - the Seed Enterprise Investment Scheme (SEIS) for the very earliest companies, and the Enterprise Investment Scheme (EIS) for the rung above - reward backing young, qualifying businesses, and they act on both tails of the distribution.
On the way in, SEIS gives 50% income tax relief on up to £200,000 invested per tax year, and EIS gives 30% on up to £1,000,000 per tax year - or £2,000,000 where at least £1,000,000 goes into knowledge-intensive companies (broadly, those doing heavy research and development). Each carries a three-year minimum hold, and relief on either can be carried back to the previous tax year. So before any company has succeeded or failed, a meaningful slice of every qualifying cheque is already back.
On the way out, the asymmetry continues. Gains on SEIS and EIS shares are exempt from capital gains tax if the shares were held for three years or more and income tax relief was received - so the rare outliers that carry a portfolio aren't taxed on exit. SEIS adds a capital gains reinvestment relief worth 50% of a gain reinvested into SEIS shares, on up to £100,000 of investment a year; EIS offers capital gains deferral relief. And when a company fails, loss relief lets you set the capital lost, after deducting the income tax relief already given, against income or capital gains. The true cost of a wipe-out is well below the headline cheque.
It's not unconditional. The schemes need advance assurance from HMRC before the investment, and the company issues SEIS3 or EIS3 certificates afterwards for the investor to claim; you generally have to be a UK taxpayer for any of it to be worth anything. The companies have to qualify, too - SEIS is for firms trading under three years with fewer than 25 full-time-equivalent staff, and EIS sits within larger ceilings (gross assets up to £30 million before the share issue, an annual raise cap of £10 million, fewer than 250 employees, shares issued within seven years of first commercial sale), with higher limits for knowledge-intensive companies. Some of those figures shifted from 6 April 2026, so check the current numbers in HMRC's guidance for the Enterprise Investment Scheme and the tax relief available to investors.
For a higher-rate UK taxpayer, the reliefs don't tweak the after-tax return so much as reshape it - taking part of the cheque off the table at the outset, leaving the winners untaxed, and sharing the cost of the failures with the Treasury. That's a very different game from the gross 2.5x the studies measure. It doesn't change the distribution underneath, though. Most companies still struggle; the relief makes a loss cheaper, not a bad business good.
None of which tells you what your own portfolio would do, and we wouldn't pretend otherwise. The historical numbers describe what large, diversified, persistent baskets of angel money have done over long horizons - not what the next cheque, or the next decade, holds. This is general information, not financial advice; consider taking FCA-regulated advice before making any investment.
Frequently asked questions
What return have angel investors historically earned?
The most-cited angel return studies, run over diversified portfolios held to exit, have reported gross returns in the region of roughly 2.5 times the money invested over an average holding period of around five to seven years, which the studies translate into an internal rate of return broadly in the mid-twenties per cent before tax and fees. Those are averages across large baskets of investors and deals, not a return any single angel can expect. The spread underneath the average is enormous, most individual investments return less than was put in, and the headline figures exclude tax reliefs, fees and the cost of failed companies that never formally close.
Do most individual angel investments make money?
No. Across the major studies, the majority of individual angel investments return less than the amount invested, and a large share return nothing at all. The positive headline portfolio numbers come almost entirely from a small minority of investments that return many times the original cheque. This is the normal shape of early-stage returns rather than a sign that something has gone wrong, and it is why the studies measure diversified portfolios rather than single deals.
How long is angel investment capital typically tied up?
Several years, and often longer than people expect. The return studies report average holding periods to exit of roughly five to seven years, and many successful companies take a decade or more. Angel shares are illiquid, with no daily price and usually no ready buyer, so capital is generally committed until a company is sold or, far more rarely, floats. The three-year SEIS and EIS minimum holding periods are tax conditions, not a guide to when money actually comes back.
How do SEIS and EIS affect angel returns?
The UK tax reliefs change the after-tax picture for investors who qualify, though they sit outside the gross return figures the studies usually quote. SEIS offers 50% income tax relief and EIS 30%, each on qualifying investments held for at least three years, so part of every cheque is returned up front. Gains on shares held three or more years with relief can be exempt from capital gains tax, and loss relief reduces the cost when a company fails. These reliefs lower the cost of being wrong and leave qualifying gains untaxed, but they do not change the underlying distribution of outcomes. This is general information, not advice; the relief depends on your circumstances.