Angel Investing for Founders

When there is a conversation about angel investing, the Revolut story always comes up.

In 2015, Balderton Capital led Revolut's seed round. They put in most of a $2.3m round, valuing the company at about £6.7m. At that point, Revolut was barely more than an idea. Ten years later, Balderton had taken roughly $2bn off the table and still held a stake worth billions more. Revolut is worth $75bn today, and those first cheques came back about 1,400 times over.

When founders talk about angel investing, this is the deal they have in mind as the most successful in recent history.

It's also the rare exception that many treat as the rule. In a 2007 study of 3,097 angel investments, 52% returned less than the money invested. Only 7% returned more than ten times the cheque. Revolut isn't what normally happens. It's the one we all remember, because nobody writes a story about the half that lost money.

Hardly anyone looks at those odds before they start. They've heard the success stories, they assume they'll be one of them, and the numbers say otherwise.

I learned what concentration costs the hard way. Between 2011 and 2014, I helped put together a $1bn infrastructure deal in Crimea (Ukraine): airport, ports, property, and farmland. It looked spread out and safe on paper. Then, in February 2014, Russia annexed Crimea and I lost everything overnight. What I think about now is how concentrated the whole thing really was under all that apparent spread. It was all in one country, resting on a single bet that everything else depended on. A lot of founder angel portfolios look exactly like that. They just haven't had their "Crimea moment" yet.

So this is how I think about angel investing: when it's worth doing, when it isn't, and what the numbers say once you move beyond the Revolut story.

Why Revolut is the wrong number to start from

When Balderton invested in Revolut, it was barely an idea. As seed deals go, it was nothing special. Most other 2015 seeds with the same belief behind them, the same cheque size, the same kind of founder, either went to zero or sold for one or two times the money in a quiet trade sale. Revolut didn't win because anyone picked it better than the rest. It won because of where it ended up: one of Europe's most valuable private companies. Almost everything alongside it went nowhere.

And we only ever hear about the winners. Every angel conversation circles back to Revolut, Uber, Stripe. The companies that quietly died never come up, and the quiet deaths are where most of the money went.

The research is clear about it. Wiltbank's 2007 study with the Kauffman Foundation examined 3,097 angel investments, and half of them returned less than the money invested. A UK update in 2009 found that 56% were underwater. The 2016 follow-up added another 250 deals and found nearly 70% losing money.

The headline return still looks fine: around 22% to 27% a year across the whole portfolio. That's the figure you see on the AngelList marketing pages. What they don't show you is how it's made. You lose money on most of the deals, and a tiny handful of huge winners drag the average back up. Those 7% to 9% of deals do all the work. Take them out, and the whole thing falls apart.

A few winners carrying everything is the whole game here, and that leaves you with two hard rules most people writing cheques never think through.

The first is about numbers. You need at least 20 to 30 bets to have a fair shot at landing one of those rare big winners. Five isn't a portfolio. Ten is barely one. If you can't see yourself making 20 to 30 of these over 3 to 5 years, you're gambling with extra steps, and you'd be better off not starting.

The second is about size. Most of these bets will go to zero, and you have to be able to lose the lot without it changing how you live. The usual rule of thumb is 5% to 10% of the liquid net worth can be allocated to alternatives (core-satellite approach), and angel investments are just part of that. Treat the 10% as a hard ceiling, not a goal. Most founders coming out of an exit belong somewhere between 0 and 5%. That top 10% slot is for the small group who know one field cold, can keep writing cheques steadily, and have the time to back 20 to 30 companies over 5 years without it taking over their life.

Most founders have neither the edge nor the time for that. And they're not putting in enough money to make all the work worth it anyway.

Why "founder angels" concentrate when they think they're diversifying

Picture a founder who sells their company for $30m. After tax, fees, and the earn-out, about $20m lands in their personal account. For the first time in their life, they're sitting on real cash. And within weeks, other founders start pitching them.

The story they tell themselves is that they're spreading their risk. I'm investing in 10 startups across 10 sectors. But look at the whole picture:

• $20m in cash from the sale, with $2m (10%) going into angel deals
• usually another $5m to $10m of shares in the company that bought them, still vesting or locked up
• often a new business they've started, or a full-time job they've taken on
• and now 8 to 10 angel cheques on top of all that

Take the cash out, and nearly everything left is the same kind of thing: young, private companies you can't sell quickly, where you either lose the lot or wait years for a payout. The angel cheques aren't spreading the risk. They're piling more of the same risk on top. It's diversification in name only.

Underneath, three things are usually going on at once, and only one of them is a good reason.

The good reason is that they know the field and want to put that to work. They spent 10 years building in fintech. They can read these deals; the good ones come to them, and the founders they back are better off for having them involved. If that's what's driving it, angel investing can earn its place.

The second reason is social. Friends are raising money. Other founders are raising money. Saying no feels like letting people down; saying yes feels like being a good member of the club. Half of that cheque is about the relationship, not the investment.

The third one sits underneath the other two: they can't sit still with cash. $20m parked in safe government bonds feels like a waste. Every cautious, sensible choice feels like admitting the game is over. Writing cheques scratches that itch. It feels like doing something. And doing something for its own sake is what wrecks most founders' wealth in the first year or two after the money lands, which is the whole point of Why Smart Founders Make Terrible Investors.

Two of those three are bad reasons. The honest thing is to work out which one is driving you. If it's mostly that you know the field and you can do this properly and steadily, fine. If it's mostly the friends or the restlessness, the answer is no. And saying no early saves you years and a lot of money.

What separates the best angels from everyone else

The angels who consistently make top returns all share one thing, and it isn't luck, getting first look at the best deals, or making more bets than everyone else.

Its depth and access to the right deals. They go deep in the handful of areas they know well from working in them, and they mostly leave everything else alone.

Naval Ravikant didn't write cheques into everything that crossed his desk. His strong patch was consumer mobile and social apps, roughly 2010 to 2014, and it came from years of building Epinions and AngelList, which gave him a read almost nobody else had. Twitter, Uber, Stack Overflow, Postmates. That's betting heavily in one area he knew, not spraying money around.

Jason Calacanis ran internet-media businesses before he ever invested. Building Weblogs and ThisWeekIn taught him how marketplaces work, which got him into Uber early.

Ron Conway, through SV Angel, sat right at the centre of the Bay Area. He knew everyone, so he could see which deals were filling up and which weren't as they were happening.

Charlie Songhurst has backed close to 500 companies, and he treats it less like gambling and more like mapping out the world piece by piece, focusing on new areas where even the experts don't have the answers yet, and there aren't many other bidders.

Four different styles, but the same thing underneath every one of them. Each of them goes deep into an area they've worked in, backs many companies, and helps founders in ways that come straight from their own experience.

The typical angel does the opposite. They write $25k cheques into consumer apps, then deeptech, then fintech, then biotech, mostly because those deals happened to come through their network. There's no depth because they've never spent real time in any of those markets. The Wiltbank data puts these investors near the bottom of the pile, closer to losing money than to doing well.

For a founder fresh out of an exit, the lesson is narrow and a bit uncomfortable. Your experience only helps you when the deal in front of you looks like the business you ran. A SaaS founder being shown another SaaS company has a real edge. That same founder, being shown a battery startup, has none, and probably less than none. A founder's confidence in a market you don't understand can lead to wrong answers that sound convincing. So the actual size of your edge is much smaller than you'd like to think.

Where the deals come from, and why most founders should use platforms

There are basically five ways a founder finds angel deals:

• Your own network. The best deals if your network is strong, but slow, and limited to the kinds of people you already know.
• Angel groups like the UK Angel Investment Network or Cambridge Angels. You get organised deal flow and other people helping with the homework, but the quality is mixed, and the money moves slowly.
• Syndicates on AngelList, Seedrs, or Republic. One experienced angel leads the deal and does the work; everyone else puts in smaller amounts alongside them.
• Accelerator demo days (Y Combinator, Techstars, Entrepreneur First). A polished, fast-moving lineup, with everyone fighting for a slice of the good ones.
• Inbound, you build up over the years. You write in public, get known for something, and founders start coming to you.

For most founders putting $1m to $2m to work over 4 to 5 years, the sensible answer is mostly syndicates, with a few of your own deals on the side when something fits you well.

Most writing about angel investing treats syndicates as the soft option. Real angels do their own homework and write their own cheques. That only makes sense if your time is free, and after an exit, it's the opposite of free. At this point, your attention is harder to come by than your money. Running your own set of 20 to 30 deals over 5 years is a real job: finding the companies, checking them out, sorting the paperwork and the lawyers on every deal, keeping track of it all, and deciding on follow-on rounds. Most founders badly underestimate how much work that is, and then quietly let the whole thing fall apart.

A syndicate puts most of that weight on the lead instead. They do the digging, sort the terms and the legal side, and handle the follow-on rounds. You read their write-up and say yes or no. They take a cut of the profits, usually 15% to 20%. Compared to a situation where you'd have done all that work yourself in no time, that's pricey. Compared to real life, where you'd never have done it at all, it's cheap.

It's the same idea behind the lean family office. You can now get the kind of quality that used to be reserved for big institutions, using software and a handful of specialists, for a tiny fraction of the cost. Angel investing is one of the clearest examples. The platforms hand you a spread of deals, leads who've already done the homework, and a way to skip most of the grunt work that ten years ago would have needed a $100m+ family office. Founder's Guide to Building a Private Investment Office goes deeper on this.

What's worth checking, and what to ignore

At the earliest stages, most of the checking people do is for show. The team is half-hired, the market is still a guess, and the product is half-finished. Spending three weeks calling references for a company that has eight customers tells you next to nothing.

Four things are worth real attention:

• The founder and the team. Does this person learn 3 to 5 times faster than most people? The fast learners get through every problem they hit; everyone else stalls. You can hear it in the way they talk about a past failure, describe their team, and think about risk.
• Why now, and not five years ago. Is there a real shift opening this up: a rule change, a new technology, a change in how people behave? A clear, honest answer usually means they've thought hard about how the whole thing works.
• The shape of the early growth, not the headline number. At this stage, $20k a month in revenue growing 30% month on month, with 90% of customers sticking around, beats $200k a month that's flat and leaking customers. The direction matters more than the size.
• How sensibly they're raising. How much money are they asking for, and what are they trying to prove with it? A huge round at a rich price is usually a warning sign, either nerves about running out of cash or pressure that's already priced in.

And three things founders worry about too much:

• The exact entry price, as long as it's in a normal range. Getting in at $8m versus $12m matters far less than people think. The gap between the company that makes it and the one that dies is so huge that paying a bit more on the way in barely moves your result. You lose your money on the failures, and you make it on the winners.
• Control of the board at this cheque size. Under $250k, you have no real say on the board anyway, and pushing for a board seat or veto rights mostly annoys the founders and buys you no real protection.
• Fancy legal side deals. At this size, all the extra complexity does is run up legal bills without protecting you. A standard SAFE with the right to follow on and the right to information covers most of what you need.

There are only three terms worth pushing for. The right to put more money into your winners at the next round. The right to see the company's numbers, so you spot trouble before the next raise rather than after. And a clean payout order if the company is sold, so a later, bigger investor can't quietly push you to the back of the queue.

What the tax breaks do, and what they don't

A few countries give you tax breaks for backing early-stage companies. They change the maths enough that they're worth understanding in general, not as a reason to put a specific amount into a specific deal.

In the UK, the EIS and SEIS schemes roughly work like this. You get income-tax relief on what you invest in companies that qualify. If one of those investments does well, you pay no capital-gains tax on the profit. And if one fails, you can claim a good chunk of your money back against your tax bill. So your winners grow tax-free, and your losers cost you a lot less than they otherwise would.

In the US, the QSBS rules (Section 1202) can let you pay no capital-gains tax at all on shares in small companies that qualify, as long as you hold them long enough and the company and the size of the gain tick the right boxes.

The difference is substantial. For a UK angel using EIS, or a US angel using QSBS, the same win can be worth 30% to 50% more after tax than it would be without the scheme. What none of this does is change the basic game underneath. You still need 20 to 30 bets. A few winners still carry the whole thing. Most deals still fail. The schemes make your winners worth more, and your losers hurt less, but they don't change the odds of any one deal working.

The details keep changing. The UK has reworked SEIS and EIS several times over the past decade, and QSBS regularly comes up for review in the US. Check the current rules with a tax adviser before you put money in, and make sure you're investing in a way that qualifies. Any article quoting exact relief percentages without a date on it is probably a year or more out of date.

One thing to keep in mind, though: tax considerations tilt the maths in your favour, but they can't save a bad portfolio. A 50% better after-tax return on five deals that all go to zero is still a total loss. The schemes make a good set of bets better. They can't turn a bad set of bets into a good one.

Should this be in your portfolio?

A short test, before any money goes near a startup. Three questions, and you have to answer them the way things are, not the way you'd like them to be.

First, the edge question. Do you know the kind of business you'd be backing? Saying I've been a founder, so I understand that startups don't count. The real test is whether you've spent years working in the exact area these deals come from. Built a consumer mobile app and are showing a consumer mobile app? Yes. Built a consumer mobile app and are showing a fancy AI venture? No. Answered honestly, this one question rules out around two-thirds of the people who think they should be doing this.

Second, the time question. Can you give this 4 to 6 years of steady, patient work without it taking over your life? Backing 20 to 30 companies means roughly five years of finding deals, checking them, making decisions, and handling follow-on rounds. If your life after the exit already includes a new job, young kids, or a project you care about, that second job is the one that loses. It quietly shrinks down to a few cheques you barely keep an eye on.

Last, the size question. Is all of this inside a slice of your liquid wealth, somewhere around 5% to 10%, that you could lose completely without it changing how you live? The number people throw around in the angel world is 5% to 10% at max. Treat the 10% as a hard ceiling, not something to aim for (and remember, this is NOT INVESTMENT ADVICE). If losing the whole lot would change anything that matters in your life, you've put in too much. Pushing all the way to 10% because someone called it the norm is how people end up too exposed.

Three honest yeses, and angel investing might be worth doing. Anything less, and the smart move is to put the money to work somewhere else.

A worked example at $10m liquid

Say you get a yes on all three. You've got $10m in liquid assets, and you've set aside $1m for this, right at the 10% ceiling. There are three ways to put it to work, and they're worth comparing.

Path A: do it all yourself. Write 25 to 30 cheques of $30k to $40k each, finding and checking every deal personally over 4 to 6 years. This makes the most of your edge, and it's also by far the most work. It only makes sense if your edge is genuine, the good deals keep coming, and you have the time.

Path B: go through syndicates. Same $1m, same 25 to 30 companies, but you put your money behind one or two leads you trust on AngelList, Seedrs, or similar, and pay them 15% to 20% of the profits. You do maybe 70% to 80% less of the work than in Path A, and you get roughly the same spread. For most founders who want to be in this without running their own little fund, this is the one.

Path C: put the money into a small venture fund. Hand your $1m to one or two seed-stage funds and let them invest it. Professionals make all the calls; you do nothing day to day, and your own experience never comes into it. You pay a yearly fee (usually about 2%) and a share of the profits (usually 20%). This is the cleanest option for founders who want a piece of this without having the edge or the time to do it themselves. Also, given the additional fees and their impact on final returns, make sure you understand what you are getting into and how the math works.

The bare minimum that still counts as serious sits inside Path B: 10 to 15 cheques through a single syndicate over 2 years, at smaller sizes of $30k to $50k, plus the odd deal of your own when something clearly lands in your wheelhouse. Anything less than that is a hobby, not a portfolio.

One last thing, because this is where founders come unstuck. Saying no to another founder who's pitching you is awkward. Most people say yes too often, and then handle the next round badly when it comes around. The cleanest no is about your own limits, not their company: I'm only backing companies in one specific area right now, and yours is outside it. Happy to introduce you to a few angels who do invest in your space. That keeps both the friendship and your discipline intact. Decision Architecture: Building Your Personal Investment Committee goes into the wider picture.

Most founders who start angel investing after an exit do it for the wrong reasons, at the wrong size, piling more risk onto a pile that's already too big instead of spreading it out. The few who do it well treat it as a real investment decision, not a favour to their friends. They stick to an area they know, keep at it steadily, and let the platforms do the heavy lifting. For everyone else, a small venture fund, buying into existing positions, or simply holding cash would do more good.

Holding cash on purpose is its own kind of discipline. Not every dollar needs a job to do. And for most founders, the best angel cheque they'll ever write is the one they talked themselves out of.