Startup investing

To invest in a private startup fund in India today, you need at least Rs 1 crore in a single cheque. To invest in an angel fund, since SEBI’s September 2025 rule change, you need to be an accredited investor: Rs 2 crore of annual income, or Rs 7.5 crore of net worth with half of it in financial assets. Those are not typos. They are the actual entry price of the asset class people ask me about most.

I work at Rainmatter, Zerodha’s fintech fund. We invest in early-stage companies, so I see startup investing from the inside: the ones that quietly die, the ones that slowly find their footing, the rare one that breaks through. The view from here looks nothing like the highlight reel on your feed, and the question “how do I invest in startups” usually comes from someone who has only seen the reel.

What you are actually signing up for

First, illiquidity measured in years. An early-stage fund runs 7 to 10 years, and often gets extended. Your money goes in during the first few years, the companies spend the middle years trying to grow, and exits come at the end, if they come. When we make an investment at Rainmatter, the mental horizon is 5 to 7 years at minimum. There is no sell button if you need the money in 18 months.

Second, failure rates that would horrify a stock investor. A good fund expects roughly 2 or 3 of 20 companies to return most of the money, a handful to return a little, and the rest to die. The math only works across a diversified portfolio with real due diligence. It does not work if you put your savings into two startups a friend told you about. Picking one startup and being right is closer to lottery than to investing, which is why the ticket sizes assume you can afford to lose the ticket.

Third, valuations you cannot check. A listed stock has a price that updates every second. A startup’s valuation is whatever someone last agreed to pay, possibly 18 months ago, in a different market, in a round with terms you cannot see. Liquidation preferences, ratchets, and stacked preference shares mean the headline valuation and what your share is actually worth can be very different numbers.

Why the gates are where they are

The accreditation thresholds look elitist until you run the failure math above. SEBI’s position, roughly stated: a person with Rs 7.5 crore can absorb a total loss on a Rs 25 lakh angel cheque without their financial life derailing. A person whose entire savings are Rs 25 lakh cannot. The gate is not there to keep the good returns for the rich. It is there because the product’s worst case is zero, and the regulator wants only people who can afford zero. You can argue about where the line is drawn, and many do. The direction of it is hard to argue with.

What a retail investor can actually do

If you are below the thresholds, the honest options are these. Listed proxies: holding companies and listed firms with meaningful startup portfolios give you indirect exposure with daily liquidity and a quarterly balance sheet. Startup IPOs: the point where a private company finally lets you in, with audited financials, a DRHP you can read, and a price discovered in public, though by then the steepest part of the growth curve is usually behind it. And your own career: for most people, the highest-return startup investment available is working at one, or building one. That is also the only one where your effort changes the outcome.

The unlisted-share brokers sitting outside this framework will happily sell you pre-IPO shares in small lots. No SEBI registration protects you there, the pricing is opaque, the seller knows more than you, and the shares often come with transfer restrictions. The spread between what those platforms charge and what the shares are worth is where the real fee hides.

Read: Weather derivatives

The democratisation of startup investing is a nice slogan, and every few years someone launches a platform promising it. Until the failure math changes, the gate stays sensible. The asset class is not hiding returns from you. It is hiding losses from people who cannot afford them.