Two numbers sit at the centre of this, and they look like they cannot both be true.
Roughly 70% of angel investments return less than the money put into them. And diversified angel portfolios in the largest studies have returned 22 to 27% IRR.
Most people who look at angel investing latch onto one number and ignore the other. The ones who only see the 70% conclude it is gambling. The ones who only see the IRR conclude it is easy money. Neither is right, and the gap between the two is the entire subject.
Do angel investors actually make money?
On average, and across a large enough portfolio, yes.
The largest dataset on the question comes from Robert Wiltbank and Warren Boeker, who tracked 3,097 investments by 538 angels, including 1,137 exits and closures. The average return was 2.6 times the money invested over 3.5 years, an IRR of about 27%.
A later update by the Angel Resource Institute added 250 more investments and revised this to 2.5x over 4.5 years, with an IRR of around 22%.
Individual angel groups reporting their own numbers land in the same range. Tech Coast Angels reported 25% IRR across 247 outcomes between 1997 and 2022. Central Texas Angel Network reported 31% across 115 outcomes from 2006.
Those are good returns by most standards. They are also averages across whole portfolios, which is a different thing from what happens to any individual investment.
Why do most angel investments lose money?
Because the returns are concentrated in very few companies.
In the Wiltbank data, 52% of exits returned less than the money invested, while 7% returned more than 10x and generated 75% of all the dollars returned. The 2016 update put the share returning less than 1x at 70%.
The Angel Capital Association's April 2026 white paper reconfirms the shape: about 70% of investments return less than invested capital, and a small number of outliers drive 70 to 85% of total portfolio gains.
The distribution most practitioners quote looks roughly like this.
Outcome | Share of portfolio |
Goes to zero | ~40% |
Returns 1x to 3x | ~50% |
Returns 10x to 50x | ~8% |
Generational outlier | ~2% |
The instinct is to assume better judgment solves this. It does not. Professional venture funds, with full diligence teams and decades of pattern recognition, see failure rates in the same range. You will not out-pick your way to a portfolio without losses.
What you can control is whether your portfolio is wide enough for the 8% to appear in it at all
How many startups do you need in your portfolio?
More than most first-time angels build.
The recommended minimum varies depending on who you ask. Some put it at fifteen companies, others at thirty to fifty, and some experienced angels argue for fifty spread across three years.
They agree on the logic even where the number differs. If 8% of investments produce the returns, a portfolio of five has a meaningful chance of containing none of them. At twenty, the odds improve substantially. At fifty, you are playing the distribution rather than betting on individual picks.
Data compiled in 2026 points to portfolios of 15 to 25 startups performing best among individual angels, which sits at the lower end of that range and is achievable for most people investing seriously.
The practical implication is uncomfortable. If you can only afford three or four investments, you are not building an angel portfolio. You are making three or four bets, and the base rates say most of them lose.
How much money do you need to start angel investing?
More than the cheque size suggests.
The answer experienced angels give is uncomfortable. To do it properly, you want enough capital to make thirty to fifty investments and still be able to lose all of it, because if losing the money would frighten you, you will not take the right risks and you will not invest in enough companies.
The arithmetic works differently in India, where cheque sizes are smaller. Angel investments here commonly run from a few lakh to a crore, and syndicates let investors participate with less. A portfolio of 15 companies at ₹2 to 5 lakh each is a genuinely different commitment from the same portfolio at $25,000 a deal.
What does not change is the ratio. Whatever your cheque size, you need enough of them, and you need to be able to lose all of it without it affecting how you make decisions.
Our guide on investing in startups in India covers the entry points at smaller amounts.
How much of your net worth should go into startups?
The commonly recommended ceiling is 10% of investable assets.
The reasoning is the same as above. Angel investing is illiquid, high-variance, and slow. An allocation large enough to hurt when it goes badly distorts every decision that follows, including the decision to keep investing at all, which is precisely what you need to do to reach a portfolio size that works.
For first-time investors, the more conservative figure is 2 to 5%, rising once you have seen a few investments through a full cycle.
How long do returns take in India?
Five to eight years to a liquidity event, and often longer.
That is meaningfully slower than the US data, where average holding periods in the Wiltbank set were 3.5 to 4.5 years. Indian exits typically come through acquisition or secondary sale rather than IPO, and the median clusters around years six to nine.
This matters for how you read your own portfolio. In the first three years, you will see the failures, because companies fail quickly and succeed slowly. The paper value of your portfolio will look poor at exactly the point when nothing has gone wrong.
That shape has a name, the J-curve, and understanding it is what stops investors from concluding too early that the asset class does not work.
What do dilution and tax take out?
Two things that almost no returns article accounts for.
Dilution. Your ownership falls with every round after yours. Two subsequent rounds at 28% dilution each leave you holding roughly 52% of your original stake. A headline exit multiple calculated on your entry percentage overstates what you actually receive, sometimes by half.
Tax. Gains on unlisted equity held for more than 24 months are taxed in India as long-term capital gains at 12.5% without indexation. Held for less, they are short-term and taxed at your slab rate, which for most angel investors is considerably higher.
So a 5x on paper, after two rounds of dilution and LTCG, is not a 5x in your account. Modelling both before you invest gives you a far more honest picture than the multiple alone.
How reliable is the returns data?
Less than the confidence with which it gets quoted.
The Wiltbank dataset, which most angel return figures ultimately trace back to, carries two acknowledged limitations. The researchers collected data from angels still participating in angel groups, which means angels who lost money and left are largely missing. And there is self-selection, because investors with good outcomes are more willing to share their numbers than investors without.
The researchers themselves flagged this and tested for it, but the effect cannot be removed entirely. The honest reading is that the 22 to 27% IRR range is probably an optimistic estimate of what a typical angel experiences, and that the failure rates are, if anything, understated.
That does not make angel investing a bad asset class. It makes the case for diversification stronger rather than weaker.
What actually improves your odds?
Three things, all within your control.
Portfolio width. The single largest determinant. Everything else is secondary to having enough investments for the outliers to appear.
Deal flow quality. You can only invest in what reaches you. Angels who see fifty good companies a year make different decisions from those who see five.
Investing alongside others. Data from 2026 suggests community-backed angels achieve around 2.3x higher returns and 40% faster exits than solo investors. Some of that is deal flow, some is shared diligence, and some is simply that a group makes it easier to keep investing through the years when nothing is returning.
There is a related argument for new angels specifically: do ten small syndicate investments before investing directly, so you learn the shape of the asset class with less money at risk.
That is broadly how PedalInvest works, and it is also why we say plainly that the first cheque should be small enough that being wrong teaches you something.
What this means for a new angel
The returns are real, and they are not evenly distributed.
If you invest in three companies, the base rates say you most likely lose money. If you invest in twenty, across several years, with a reasonable share of your portfolio and the patience to wait out the J-curve, the historical data says you have a reasonable chance at returns that beat most other asset classes.
The difference between those two outcomes is not stock-picking skill. It is portfolio construction, and it is decided before you look at a single deal.
Key takeaways
Around 70% of angel investments return less than the invested amount. In large studies, diversified portfolios have returned a 22% to 27% IRR.
About 7 to 8% of investments produce most of the returns, which is why portfolio width matters more than individual selection.
Recommended portfolio sizes range from 15 to 50 companies. Fewer than ten means you probably miss the outlier entirely.
Indian exits typically take five to eight years, longer than the 3.5 to 4.5 years in US datasets.
Dilution and a 12.5% LTCG tax both reduce the realized return below the headline multiple. Neither is shown in most returns data.
Frequently asked questions
Do angel investors make money?
On average and across a diversified portfolio, yes. The largest studies show 2.5 to 2.6x returns over 3.5 to 4.5 years, or 22 to 27% IRR. Individual investments lose money far more often than they win.
What percentage of angel investments fail?
Around 40% go to zero and roughly 70% return less than the capital invested. Approximately 7 to 8% return more than 10x and generate most of the portfolio's gains.
How many startups should an angel investor back?
Commonly cited minimums range from 15 to 50. Data from 2026 suggests portfolios of 15 to 25 companies perform best for individual angels. Fewer than ten makes missing the outlier likely.
How much money do you need to be an angel investor?
Enough to make many investments and lose all of it without changing your behaviour. In India, syndicates allow participation at a few lakh per deal, but the portfolio logic still requires fifteen or more investments.
How long does it take to see returns from angel investing?
Five to eight years in India, sometimes longer. Failures surface within the first two or three years while successes take considerably longer, which is why early portfolio performance looks worse than it is.
How is angel investing taxed in India?
Gains on unlisted equity held for more than 24 months are taxed as long-term capital gains at 12.5% without indexation. Shorter holdings are taxed as short-term gains at your applicable slab rate.
Is angel investing better than the stock market?
The historical IRR range compares well, but the risk profile is entirely different. Angel investing is illiquid, concentrated in a few outcomes, and takes years to resolve. It suits a small share of a portfolio rather than a replacement for public market exposure.
