ETMarkets NRI Talk| Want to invest in US IPOs? Ankita Pathak on what Indian investors need to know

US IPOs are increasingly attracting global investor attention, particularly when the companies involved are high-profile names in AI and technology.

But for Indian investors, the first question is often a practical one: can you actually invest in a US IPO such as Anthropic, and what does the process involve?

According to Ankita Pathak, Head – Global Investments at Ionic Asset, Indian residents are not prohibited from investing in overseas securities. Under FEMA and the RBI’s Liberalised Remittance Scheme, resident individuals can remit up to US$250,000 per financial year for permitted purposes, including buying foreign securities.


However, access to the IPO itself can be more complicated, as allocations tend to skew towards institutional investors and retail participation may be restricted based on the investor’s country of residence.

In this segment of ETMarkets NRI Talk, we speak with Ankita Pathak about what Indian investors need to know before looking at US IPOs, why retail allocations can be difficult to access, and how the dynamics of a high-profile listing can differ from simply buying a stock after it starts trading. Edited Excerpts -

Q) Anthropic has reportedly pushed back its IPO plans. Should investors see this as a red flag, or is postponing an IPO simply part of the normal listing process?

A) Postponing an IPO is far more common than the headlines suggest, and on its own it tells you very little. Companies delay listings for a handful of recurring reasons: the valuation on offer falls short of what the board and existing investors expect; the equity story has not yet convinced the institutions that would anchor the book; the original rationale for listing, whether that was raising capital, giving early investors liquidity or building a currency for acquisitions, has changed; or market conditions have shifted since the process began. A listing window is a moving target, and a company that can afford to wait for a better one usually will.

The useful distinction for investors is whether a delay is company driven or demand driven. If a business can keep raising private capital at valuations comparable to or better than what the public market would pay, deferring a listing is a choice, not a distress signal.

The private market for leading AI companies has been deep enough that the urgency to list is lower than it was for previous generations of technology companies. A delay becomes a red flag only when it coincides with slowing growth, a funding squeeze or a loss of strategic momentum, and none of those has been reported here.

In Anthropic's case, the broader backdrop also matters. The regulatory and public debate around AI safety and compliance is live, and a company going public inherits quarterly disclosure and scrutiny at exactly the moment that debate is unresolved.

It is notable that Sam Altman said in a recent interview that OpenAI has also deferred its listing plans, describing it as an ill advised moment to go public given concerns around AI safety. When the two most prominent names in the category independently reach the same conclusion, it reads less like a company specific problem and more like a sector wide judgement about timing.

Q) Are investors currently valuing AI companies on today's earnings, or on what they believe these companies could become five or ten years from now?

A) Nobody can say with precision what is priced in, and views differ widely. The common thread, in my view, is that current valuations are not discounting just the next one or two years of growth.

There is a medium to long term growth trajectory embedded in these multiples, which is why they are so sensitive to any change in the perceived length or durability of that runway. A small change in the assumed terminal growth rate moves the fair value of these businesses far more than a quarter's earnings beat or miss.

It also helps to remember that "AI companies" is not one bucket. Semiconductor and hardware names sit at the capital expenditure end of the build out and are being valued on visible order books and near term earnings.

Cloud and infrastructure players distribute that compute and are valued on utilisation and the durability of demand. Software companies consume it, and their valuations rest on whether AI expands their addressable market or compresses their pricing. Each layer is being underwritten on a different horizon.

Where investors disagree is on how long the growth runway lasts and how much of it any single layer actually captures. That dispersion of opinion, not the headline multiple, is what seemingly drives the volatility you see in these names.

Q) We have seen investors aggressively chase some of the biggest AI names. How do you distinguish genuine long term AI winners from companies benefiting mainly from the AI narrative?

A) Our starting question is who keeps the economics once the build out stops being purely a growth story. Right now a great deal of revenue is flowing to companies simply because capital expenditure is being spent, and that is very different from having a durable claim on the profit pool.

On the hardware side, plenty of companies are growing today because the infrastructure is being built. The long term winners, in my view, are those that end this phase with an installed base locked in, earning from servicing, replacement cycles, software attach and upgrades long after the initial build is complete.

The test is whether revenue persists when the capex curve flattens, not how fast it is growing while the curve is steep.

On the software side, the question is what AI integration actually does to the underlying business. Does it leave the company as a commodity that any competitor can replicate with the same models and the same APIs, or does adoption compound the franchise through a better product, better unit economics and stickier customers?

We look for evidence in gross margin trajectory, pricing power, net retention and whether proprietary data or workflow ownership creates a moat that the model layer alone cannot erode.

In general, I believe the genuine long term winners will be businesses that retain or even widen their competitive moat through this cycle. Companies that are merely adjacent to the narrative tend to show up in the numbers as revenue growth without margin expansion, or as customer wins without retention.

Q) We have seen the listing of SpaceX. What does the sharp move after listing, and the subsequent correction, tell us about buying into highly anticipated IPOs?

A) A blockbuster IPO is often as much a bet on the founder as on the financials. Investors buy into the vision and the execution track record, and that belief is what earns the company a premium valuation at listing. The premium holds only if delivery keeps pace with the promise.

If execution slips or guidance disappoints, the premium unwinds, and the correction can be sharp because there is little earnings support underneath to cushion it. In a mature business the multiple has a floor set by cash flows; in a story driven listing the floor is set by conviction, which is a far less stable foundation.

The mechanics of these listings amplify both legs of the move. They are typically heavily oversubscribed, so most investors receive a fraction of what they wanted and come into the aftermarket as forced buyers.

Free floats are thin, so modest demand produces outsized price moves. Lockups are staggered, which seemingly means supply arrives in waves over the following months rather than all at once, and each expiry is a potential overhang.

Sell side coverage only begins after the quiet period ends, so the early trading happens with very little independent analysis in the market. All of this contributes to the premium these companies command at listing, and all of it can reverse very quickly on disappointment.

The key consideration for investors is not to avoid anticipated IPOs, but to be honest about what you are buying in the first weeks. You are paying for scarcity and sentiment as much as for the business.

If you believe in the long term story, there is rarely a penalty for waiting until the float widens, the lockups pass and the first few quarters of public reporting give you something firmer to underwrite.

Q) Let us start from the basics: can an Indian resident actually invest in a US IPO such as Anthropic?

A) Nothing in Indian law prohibits it. Under FEMA, read with the RBI's Liberalised Remittance Scheme, a resident individual can remit up to USD 250,000 per financial year for permitted purposes, and that includes buying foreign securities. IPO allocations skew heavily institutional, and the retail tranches that do exist are usually restricted by country of residence.

(Disclaimer: Recommendations, suggestions, views, and opinions given by experts are their own. These do not represent the views of the Economic Times.)