Published data says both matter, on different timescales. In the best-known study of angel returns (Wiltbank and Boeker, 2007), the top 10% of exits produced 75% of the money returned, so any single outcome is mostly luck. Across a portfolio, measurable habits such as deeper due diligence, sector expertise and staying involved line up with much better multiples.
The luck-versus-skill question matters because it changes how you should behave. If angel returns were pure luck, the right strategy would be to write as many small checks as possible and spend no time on anything else. If they were mostly skill, a careful investor could hold a concentrated portfolio and expect to beat the average. The research points somewhere in between, and the details are useful.
What does the best-known angel returns study show?
The best-known dataset is "Returns to Angel Investors in Groups" by Robert Wiltbank and Warren Boeker, published in 2007 by the Kauffman Foundation and the Angel Capital Education Foundation. It covered 539 angels in 86 angel groups and 1,137 exits, most of them after 2004. An exit here means an acquisition or IPO, or a closure that returned less than the money invested.
The headline numbers:
The average return was 2.6 times the money invested over an average hold of 3.5 years, roughly a 27% internal rate of return.
52% of exits returned less than the capital invested.
7% of exits returned more than ten times the money, and the top 10% of exits accounted for 75% of all the dollars returned.
Looking at investors rather than deals, 39% of the angels had an overall multiple below 1x, while 61% made money across their exits.
That last point is the first clue about luck. More than half of individual deals lost money, yet most investors came out ahead, because a few outliers carried each portfolio. A single deal is a lottery ticket with a skewed payout. A portfolio is something you can actually reason about. For a fuller look at what those averages mean for an individual angel, see what returns you can expect from angel investing.
Where does skill show up in the data?
The same study asked investors how they worked, and three habits stood out:
Due diligence time. The median angel spent 20 hours on diligence before investing. Exits where investors spent more than the median returned 5.9x overall, against 1.1x for those below it. The top quartile, above 40 hours, saw 7.1x.
Industry expertise. About half of the investments had nothing to do with the investor's own industry experience. Multiples were twice as high for the investments that did.
Participation after the check. Angels who interacted with a company a couple of times a month, through mentoring, introductions or monitoring, saw 3.7x over four years. Those who checked in a couple of times a year saw 1.3x.
One habit pointed the other way: follow-on investments in existing portfolio companies were associated with lower performance.
These are correlations from self-reported data, and the authors are careful not to claim more. Investors may have spent longer on deals they already liked, or stayed closer to companies that were going well. Still, the pattern is consistent: things an investor controls line up with materially different results. That is what skill looks like in a noisy field.
Does past success predict future success?
If skill were a large part of outcomes, winners should keep winning. Two well-known papers test this, one on founders and one on investors.
Paul Gompers, Anna Kovner, Josh Lerner and David Scharfstein studied venture-backed entrepreneurs in a paper first circulated as "Skill vs. Luck in Entrepreneurship and Venture Capital" and published in the Journal of Financial Economics in 2010 as "Performance Persistence in Entrepreneurship". Defining success as going public, they found a founder whose previous venture-backed company succeeded had about a 30% chance of success in the next one, against 18% for first-time founders and 20% for founders whose previous company failed. Prior success carried forward; prior failure barely hurt.
Ramana Nanda, Sampsa Samila and Olav Sorenson looked at venture firms in "The Persistent Effect of Initial Success" (Journal of Financial Economics, 2020). Each extra IPO among a firm's first ten investments predicted a higher IPO rate on its later deals, by as much as 8%. Their explanation is not a sixth sense for picking. Early winners mostly invested in the right places at the right time, and the reputation that came from those wins gave them access to better later-stage deals.
That is the most practical finding in this literature. Early luck can turn into lasting advantage through access, which looks like skill from the outside. It also means an angel's reputation and deal flow are part of the skill, not separate from it. The lessons on creating your own luck are about exactly that loop.
How should an angel act on this?
Michael Mauboussin's book "The Success Equation" describes activities on a continuum from pure luck to pure skill, and notes that the more luck matters, the larger the sample you need before results tell you anything. Early-stage investing sits well toward the luck end for any one decision. A few rules follow from that:
Build a real portfolio. With most of the money coming from a small share of deals, a handful of investments is not enough to judge yourself or to give the outliers a chance to appear.
Judge decisions, not outcomes. A company that failed may have been a good bet; a breakout may have been a bad one that got lucky. Write down why you invested at the time and compare later.
Spend effort where the data says it pays. Diligence hours, sectors you know and real involvement after the check are the habits tied to better multiples.
Expect your track record to take years to mean anything. Losers show up early and winners late, so a young portfolio almost always looks worse than it is.
My own portfolio has 234 angel investments, including OpenAI, Anthropic and Hugging Face, and Lenny's Newsletter ranked me the most active angel investor of 2024. Numbers like that describe volume and access, not proof of picking skill, and I think every angel should read their own record the same way. The reflections on luck in startup success and the lessons from investments that returned 100x go further into how outliers happen. More on portfolio strategy and diligence is in the angel investing section.
Frequently Asked Questions
Q:Is angel investing mostly luck?
Any single investment is mostly luck, because a small share of exits produces most of the returns. Across a portfolio, the evidence shows habits that line up with better results, such as more due diligence, investing in sectors you know and staying involved. Skill shows up in aggregate, not deal by deal.
Q:How many angel investments do you need before skill shows?
There is no published threshold, but the skew in returns means a handful of deals tells you very little. Plan to build a portfolio over several years before judging yourself on it. Losses also arrive faster than wins, so early records understate long-term results.
Q:Do more experienced angels earn better returns?
Experience in the industry of the company matters in the Wiltbank and Boeker data: multiples were about twice as high for investments related to the investor's expertise. General investing experience is harder to isolate. The venture research also suggests that experience brings better access to deals, which itself improves outcomes.
Q:Can following famous investors replace skill?
Co-investing with strong leads can improve access, but it does not remove luck, and the best-known names do not get every allocation. The Gompers study found top-tier venture backing added little for founders who had already succeeded. Use a lead's involvement as one signal, not as the decision.
Q:Why do failures show up before successes?
In the Wiltbank and Boeker data, exits that lost money came after about three years on average, while exits above 30x took about six. Failing companies run out of money quickly, while winners keep raising and stay private longer. A portfolio's early numbers are therefore biased toward its losers.