Most angel investors would have done better in a tracker fund.
The research has said so for the better part of two decades, and most angels have smiled politely and carried on.
That sentence should annoy you. Good. It’s meant to. Because the interesting thing about angel investing isn’t that it’s risky, which everyone knows, but that the biggest driver of whether you make money is almost embarrassingly dull. It isn’t a nose for a winner. It isn’t your network. It isn’t the founder’s charisma across a coffee table. It’s hours. The number of hours you spend looking before you leap.
Robert Wiltbank and Warren Boeker worked this out in 2007, in a study funded by the Kauffman Foundation that remains the largest of its kind ever conducted: 539 angels, 3,097 investments, 1,137 exits. The headline is genuinely encouraging. As a group, these angels returned 2.6 times their money in about three and a half years, roughly a 27% annual return. That’s competitive with formal venture capital, and it embarrasses the old cliché that angels are rich amateurs setting fire to their money for fun.
But hidden inside that cheerful average is a fact that should change how you behave. Fifty-two per cent of individual investments returned less than the money put in. On any single deal, you’re more likely to lose than win. The 2.6x comes from a handful of investments doing spectacularly well and dragging everything else up behind them. This is the power law, and it’s the weather system every angel lives inside whether they admit it or not.
So the average is good and the median is a loss. What separates the angels who caught the average from the ones who got the median?
Diligence hours. And the gap is not subtle.
Wiltbank split the exits at the median of 20 hours of due diligence per deal. The angels who spent fewer than 20 hours averaged a 1.1x return over 3.4 years. That’s about 2.2% a year, which is to say they’d have been better off leaving the money in a building society. The angels who spent more than 20 hours averaged 5.9x over 4.1 years, an internal rate of return north of 43%. Push past 40 hours and the multiple rose again to 7.1x. The diligent angels weren’t dabbling either: a later survey by Forrester and colleagues of 277 angels put the mean diligence time at 60.6 hours per deal, around 34 hours once you strip out the outliers.
Read those two numbers next to each other again. 1.1x and 5.9x. Same asset class, same power law, same rough vintage. The only thing separating them is how hard the investor looked before writing the cheque.
The pattern isn’t a quirk of one American dataset. Wiltbank ran the study again in the UK for NESTA and the British Business Angels Association in 2009, under the title Siding with the Angels: 158 angels, 1,080 investments, 406 exits. British angels returned 2.2 times their money in just under four years, a 22% IRR, and the diligence effect showed up exactly as before. When UK angels spent under 20 hours on a deal, 60% of their investments returned less than 1x. When they spent more than 20 hours, that failure rate fell to 40%. A tracking update in 2016 added 250 more exits and the correlation held, if anything more strongly.
Twenty hours. That’s the line. And here’s the quietly damning part.
In 2025, Keith Arundale and Colin Mason published a study of Henley Business Angels, one of the UK’s established networks, in the International Entrepreneurship and Management Journal. Their verdict was blunt: “there is a lack of rigour in the due diligence that angels undertake.” The typical deal got five to ten hours of work. No external advisers. Much of the founder’s information taken on trust and never independently checked. And, in their words, “69% of angels were also influenced in their decision to invest by other investors in the company.” The gap Wiltbank named in 2007 has had eighteen years to close, and it hasn’t. Most angels are still investing below the line where the returns start to climb.
Which brings us to the two kinds of angel.
There is the old-school angel, and you know him, because there’s one in every syndicate. He’s been doing this for twenty years. He backs founders he “gets a good feeling about.” He invests in sectors he no longer works in but once did. He puts more money into the struggling companies because he can’t quite admit the first cheque was a mistake, and the sunk cost hums away underneath every decision. He does his diligence in the room, by reading the temperature of the other angels, most of whom are doing exactly the same thing. He’s not stupid. He’s often charming and occasionally right. But he has never once been benchmarked against another angel, so he has no idea which quartile he’s in, and the failures are quieter than the wins, so he remembers himself as a better investor than the numbers would allow.
Then there’s the new one. Younger, usually, though it’s a habit of mind more than an age. This angel has read the research and taken it personally. She knows the median angel underperforms the market, so she refuses to be the median angel. She invests where she has genuine expertise, because Wiltbank found that doubles your multiple, and yet half of all angel deals get made outside the investor’s area of knowledge, which is angels cheerfully throwing away their one structural edge. She’d rather hold fifteen or twenty companies than three, because the power law punishes small portfolios. AngelList’s analysis of nearly ten thousand investor portfolios found that a portfolio of 100 investments performed about 7% better each year than a portfolio of one, and that the returns follow a brutal power law in which roughly 1% of the winning bets return 22x or more. Miss those, and you’ve missed the point. And she treats her own time as the scarce input it is, because the difference between a 1.1x habit and a 5.9x habit is the whole game.
The gap between these two people, expressed in pounds, is where this gets uncomfortable.
The £340,000 habit
Take two angels. Each has £100,000 to deploy across a portfolio. One does it the ad-hoc way, under 20 hours a deal, on gut and warm introductions. The other does the research-informed thing, 20-plus hours a deal, inside her expertise, engaged with her companies. We’re not inventing their returns. We’re using Wiltbank’s own bands.
The ad-hoc angel earns the sub-20-hours multiple: £100,000 times 1.1 equals £110,000. A gain of ten grand across roughly three and a half years. Underwhelming, and, once you remember she could have earned around 7% a year in a cheap global tracker with none of the risk or the illiquidity, actually a loss in real terms. This is the opportunity cost the old-school angel never books, because he only counts the deals that worked.
The research-informed angel earns the over-20-hours multiple: £100,000 times 5.9 equals £590,000. A gain of £490,000 across about four years.
The raw distance between them, in terminal money, is £480,000. That’s the honest, undiscounted number, and it’s the one most people would stop at.
But we should be harder on ourselves than that, because the diligent angel’s money is tied up slightly longer, 4.1 years against 3.4, and a pound in four years isn’t a pound today. So discount both outcomes back to present value at the roughly 8% a pound could earn elsewhere. The diligent £590,000 becomes about £430,000 in today’s money. The ad-hoc £110,000 becomes about £85,000. The gap, adjusted for opportunity cost, is about £345,000.
Call it roughly £340,000. On a single £100,000 portfolio. Bought with nothing more exotic than the decision to look properly before investing.
Two things need saying plainly, because we’d rather be honest than impressive. First, that £340,000 is ours, a worked illustration extrapolated from Wiltbank’s quartile bands. The study does not state it. Wiltbank measured the multiples; the arithmetic to a portfolio figure, and the opportunity-cost discount, are our deduction, and we’ve shown every step so you can check it or disagree with it. Change the discount rate and the number moves: at 7% it’s nearer £360,000, undiscounted it’s £480,000. Second, 5.9x is itself an average sitting on top of a power law, so no individual diligent angel is promised it. The claim is narrower and sturdier than a promise. Across enough deals, the habit of thorough diligence has, in the largest datasets we have, correlated with returns several times larger than its absence. The gap is real even if any single outcome isn’t.
Now the obvious objection. Twenty to sixty hours per deal, done properly, across fifteen to twenty companies, is a second job. Most people who could be good angels don’t have that time, which is precisely why so many settle for the five-to-ten-hours, take-it-on-trust approach that Arundale and Mason found at Henley. The research tells you what works. It doesn’t hand you the hours to do it.
This is the problem The Startup Race was built to solve.
Where the diligence comes from
The usual sequence in angel investing runs backwards, and once you see it you can’t unsee it. Think about your highest-conviction follow-on, the one you wrote without hesitating. By the time you signed it you’d watched the company for a year or more: revenue you saw grow, a founder whose bad weeks you’d studied. The pitch was ancient history. You were backing behaviour you’d already seen. Then think about your first cheque into that same company, decided on a twelve-minute deck shown to nobody who’d actually bought the product. The deep evidence decides the second, smaller decision. The thin evidence decides the first, riskier one. There’s no good reason it should be that way round.
The Startup Race turns it the right way round. Fifty founders, each already at £5,000 or more in annual recurring revenue, which is about £96 a week. The qualifying test is deliberately unglamorous: if you can’t get strangers to pay you £96 a week, you’re not ready to race. They then compete for fifty-two weeks on a public league table. Every figure is verified weekly, pulled straight from Xero or QuickBooks, so nothing is self-reported and a coach can’t fake it. What you’re watching is whether a founder can sell to customers, week after week, for a year. Performing across a boardroom table is a different skill, and the two barely move together.
By the time the race ends, an investor has watched twenty-plus hours’ worth of genuine evidence accumulate, gathered passively, across fifty founders at once, before committing a pound. That’s the point. The Race takes the thing Wiltbank found high-performing angels do by grinding through diligence deal by deal, and produces it at a scale no individual could reach. It institutionalises the 20-hour line.
There’s a second engine running alongside the revenue table. Thirty years of UK angel research, much of it by Scotland’s Colin Mason, shows that angels don’t score a company and average its strengths against its weaknesses. They eliminate. One fatal flaw in a single dimension, unassigned IP, a messy cap table, no SEIS assurance, and the company is rejected regardless of how good the revenue looks. So every racer works through an Investment Readiness Constellation of eight binary milestones, each clearing one of the deal-killers that sinks an otherwise strong founder: a clean cap table, IP assigned to the company, a founders’ agreement with vesting, SEIS Advance Assurance, documented customer evidence, a defensible valuation, an honest weakness-disclosure, and a recorded Q&A stress test with real angels. Revenue rank tells you who’s executing. The Constellation clears the things that would end the conversation whatever the traction.
The whole progression sits on a revenue-first methodology, the Four Sequential Problems, which insists you can’t skip levels: awareness, then founder/business-model fit, then systematic revenue, then scaling under competitive pressure. Solve problem three before problem one and you’ll execute brilliantly on the wrong thing. It’s the same methodology that produced winners growing revenue 12x, 17x and 44x during earlier races at £1,000 and £10,000 stakes.
For the angel, this changes the nature of the decision. You commit to the founding syndicate before the race starts, then watch fifty-two weeks of verified performance, and the capital deploys into the proven winner only once the winner is known. It’s closer to a follow-on than a first cheque, which is to say it’s the good kind of decision, the one you’d normally only get to make on your best-known company. Contributions start at £5,000, the fee model is membership and admin only with no carry or commission, the same model Archangels and Cambridge Angels use, and The Startup Race Ltd puts its own £10,000 into the winner on identical terms. And because SEIS caps the downside hard, the maths that frightens sensible people gets a lot less frightening: 50% income tax relief on the way in, capital gains free on the way out if it works, and loss relief if it doesn’t, so an additional-rate taxpayer’s worst case on a £10,000 investment is a loss of about £2,750 rather than the full ten grand.
Getting through the gate: how to qualify as an investor
Here’s the catch that stops a lot of good would-be angels before they start. To receive angel investment promotions in the UK, you have to be certified as a high net worth individual or a self-certified sophisticated investor. But one traditional route to becoming sophisticated is to have already been investing, which is a neat little trap: you can’t get the promotions until you’re sophisticated, and the old path to sophistication assumed you were already in the game.
You’re not as stuck as that sounds. Under Article 50A of the Financial Promotion Order, as it stands after the rules were reinstated on 27 March 2024, you can self-certify as a sophisticated investor if any one of the following is true, and there are two routes here that a motivated newcomer can reach deliberately.
- You’ve been a member of a business angel network or syndicate for at least the last six months. Join one, give it six months, and this route opens on its own.
- You’ve made two or more investments in an unlisted company in the last two years. This is the door a platform like Seedrs (now trading as Republic Europe) opens. Two genuine, considered equity investments in early-stage companies, made through a regulated crowdfunding platform, put you legitimately through the Article 50A gate. The word that matters is considered. This means real, thought-through investments in companies whose product you understand and whose numbers you’ve read, in the spirit of the 433 early Revolut backers who put in an average of about £2,152 each on Crowdcube in 2016 and, at Revolut’s later valuations, are sitting on returns of many thousands of per cent. It does not mean two token £10 punts made to tick a box. The certification has to be signed honestly, and the promoter has to have reasonable grounds to rely on it, so the substance needs to match the statement.
- You’ve worked, in the last two years, in a professional capacity in private equity or in the provision of finance to small and medium enterprises.
- You are, or have been in the last two years, a director of a company with an annual turnover of at least £1 million.
There’s also the high net worth route under Article 48, which doesn’t ask about experience at all, only means: an income of at least £100,000 in the last financial year, or net assets of at least £250,000. Net assets here exclude your main home (and any loan secured on it), your pension, and your rights under insurance contracts, so it’s £250,000 over and above the house and the pension.
A practical note on all of this. Since 31 January 2024 the actual statement you sign is a tick-box form that asks you to state how you meet the criterion you’re claiming, the number of unlisted investments, your income to the nearest £10,000, your net assets to the nearest £100,000, rather than the old signature at the foot of a paragraph. Plenty of firms still show the superseded wording on their websites, which is a small compliance trap in itself. The Startup Race operates within these exemptions as a private club of certified investors under Articles 48, 50A and 51, and once you’ve expressed interest you get the current self-certification form straight away.
For the new breed of angel, the two deliberate routes matter most. Six months in a syndicate, or two considered investments through Seedrs, and you’re through the gate, ready to do the thing the research actually rewards.
The uncomfortable summary
The old-school angel is trapped by three things: the sunk costs he won’t write off, the social group he decides with rather than against, and the returns he never benchmarks. He does five to ten hours of diligence, takes the rest on trust, and tells himself a flattering story about the results.
The new angel does the boring thing that works. She looks for twenty hours before she leaps, invests where she knows something, spreads her bets wide enough to catch the power law, and stays out of the follow-on trap that Wiltbank flagged in both his studies (1.4x for angels who followed on, 3.6x for those who didn’t). And where she can, she borrows a structure that does the looking for her, at a scale her own diary could never manage.
That’s the whole argument, and it’s roughly £340,000 long.
Appendix: the £340,000 working, shown in full
Inputs, all from Wiltbank & Boeker (2007): sub-20-hours diligence produced an overall multiple of 1.1x over an average hold of 3.4 years; over-20-hours diligence produced 5.9x over 4.1 years.
Step one, raw terminal values on a £100,000 portfolio. The ad-hoc angel: £100,000 × 1.1 = £110,000. The research-informed angel: £100,000 × 5.9 = £590,000. Raw undiscounted gap: £480,000.
Step two, the opportunity-cost adjustment. Because the two portfolios are held for different lengths of time, discount both back to present value at 8%, roughly the long-run return of a cheap global equity tracker. The diligent £590,000 becomes £590,000 ÷ 1.08^4.1 ≈ £430,000. The ad-hoc £110,000 becomes £110,000 ÷ 1.08^3.4 ≈ £85,000. Present-value gap ≈ £345,000, which we round to roughly £340,000.
Sensitivity: at a 7% discount rate the gap is about £360,000; undiscounted it is £480,000. So £340,000 is the conservative, opportunity-cost-adjusted statement of the gap rather than the largest defensible figure.
Three flags we’d rather you saw than missed. The £340,000 is our deduction, worked from Wiltbank’s bands; the study doesn’t state it. The 5.9x is a power-law average, promised to no individual. And the diligence-to-returns link is a correlation that replicates across the 2007 US, 2009 UK and 2016 datasets, though researchers have noted the hours may partly be a marker of more engaged investors rather than the sole cause.
This article is for information, not investment, tax or legal advice. Angel investing is a high-risk asset class. Your capital is at risk and you could lose all of it. Tax reliefs depend on individual circumstances and may change. The regulatory position described is current as at the rules reinstated with effect from 27 March 2024.
This article is general commentary on UK early-stage investment dynamics. It is not investment advice, a recommendation, or a financial promotion. Founders considering fundraising, and investors considering an angel commitment, should seek independent professional advice before acting on anything described here.