Why you can’t just own your Delaware C-Corp with an Indian subsidiary (and the LLP route that fixes it)

Here’s a strange fact about Indian startups. The most common structure — a Delaware parent with an Indian subsidiary underneath it — is one that you, as an Indian resident founder, are not allowed to own directly. Not “shouldn’t.” Not “it’s complicated.” Not permitted — and unlike much of FEMA, there’s no approval window an individual can apply through to reopen it.

I’ve walked well over a hundred founders through this, and the reaction is always the same: wait, that can’t be right. It is right. And once you see why, the strange workaround almost every founder in this position uses — a small Indian partnership sitting between you and your own company — stops looking strange and starts looking inevitable.

Let’s build it up from the ground.

Who counts as a “FEMA resident” anyway?

Everything in cross-border structuring starts with one question: are you a person resident in India under FEMA? Not under income-tax law — FEMA has its own definition, and the two don’t always agree.

The short version: if you’ve been living in India, you’re a FEMA resident. The slightly longer version: you’re resident if you lived in India for more than 182 days in the preceding financial year — unless you’ve left India for a job, a business, or with the intention of staying out indefinitely, in which case you flip to non-resident the day you leave. It works the other way too: come back to India to live or work, and you’re resident, more or less immediately.

If you’re reading this from Bangalore or Gurgaon and you haven’t recently lived abroad, you’re a FEMA resident. The rest of this article is about what that means for owning a US company.

What does FEMA actually say about a resident owning a foreign company?

Less than you’d fear, and more than you’d like.

A resident individual can invest in foreign companies. The route is the Liberalised Remittance Scheme — currently USD 250,000 per person per financial year — and the rules for it live in Schedule III of the Overseas Investment Rules, 2022. Buying shares of a US startup, even founding one, is permitted.

But there’s one sentence in Schedule III that changes everything for founders. A resident individual may make ODI in an operating foreign entity only if two things are true: the entity is not in financial services, and it does not have a subsidiary or step-down subsidiary — where the resident individual has control in the foreign entity.

Read that again, because the qualifier is doing the work. It’s not “no subsidiaries ever.” It’s: if you control the foreign company, it can’t have a subsidiary under it. And “control” under FEMA is a low bar — 10% or more of voting rights clears it, as does the right to steer management or appoint the majority of the board. A founder with a founder-sized stake has control — that’s rather the point of being a founder. (Flip side: a genuinely small, non-controlling stake in someone else’s US startup is generally fine — though even a small unlisted holding is still ODI, with its own paperwork. More on that machinery below.)

And note the first condition too, because it catches a specific set of readers: if the company you’re building is itself in financial services — lending, payments, broking — the direct individual route is blocked even if there’s no subsidiary at all. Fintech founders hit a second, independent wall.

So the question becomes: does your Delaware company actually need a subsidiary?

Why does every Indian startup end up needing an Indian subsidiary?

Because your team is in India. That’s the honest one-line answer.

The standard Indian-founder playbook is: incorporate in Delaware because that’s where the investors and the market are, and build in India because that’s where the engineers are — usually at a fraction of the US cost. Almost every India-founded, US-incorporated startup I’ve worked with runs its dev and ops out of India. But a Delaware C-Corp cannot simply put Indian employees on its payroll. It has no Indian presence, no PF/ESI registrations, no ability to run compliant Indian payroll.

Founders try to shortcut this with contractors, and for the first few months that mostly works. Then three problems show up, roughly in this order:

First, the practical one: you can’t offer real employment. No offer letters from a proper Indian employer, no PF, no gratuity, no clean ESOP administration for the India team. Good people eventually ask for these things.

Second, the tax-shaped one — and here the detail matters. A team in India working for a US company with nothing in between can constitute a permanent establishment (PE) of the US company in India, which means Indian tax on the profits attributable to that operation, plus the joy of arguing about how much that is. This is the live risk for an early-stage company. The fix is not to hide the team; the fix is the subsidiary: a properly set-up Indian entity that employs the team and bills the parent for its services at arm’s length is the standard, accepted way to contain PE exposure — the structure is the mitigant, not the threat. There’s a bigger cousin of this risk called POEM — if a US company’s effective management genuinely sits in India, the whole company can be treated as an Indian tax resident — but current guidance carves out companies below ₹50 crore turnover, so for most readers of this article it’s a “when you’re at real scale” problem, not a day-one one. (Thresholds and treaty analysis are CA/tax-lawyer territory — treat this as the shape of the risk.)

Third, the housekeeping one: IP created by loosely-papered Indian contractors is a diligence mess waiting for your Series A.

The clean fix for all three is the same: the Delaware parent forms a wholly-owned subsidiary in India — an Indian Pvt. Ltd. — which employs the team, runs payroll, and invoices the parent at arm’s length. Which is exactly how you arrive at the classic structure: US parent → India subsidiary.

So what’s the catch with the US-parent-India-subsidiary structure?

You already know the catch. Walk the loop:

You’re a FEMA resident. You control your Delaware C-Corp. The C-Corp needs an Indian subsidiary to hire your team. And Schedule III says a resident individual can’t hold ODI in a foreign entity that has a subsidiary where that individual has control.

The structure every Indian startup needs is the structure an Indian founder is barred from holding directly. That’s the trap, stated plainly. It isn’t a loophole someone forgot to close — the OI Rules also police the round-trip (foreign money coming back into India) with a separate two-layer rule. The system is built to see, and constrain, exactly this shape.

How does the LLP route fix it?

By changing who the investor is.

The Schedule III restriction applies to a resident individual. But FEMA’s OI framework treats an Indian entity — a company, or a Limited Liability Partnership under the LLP Act, 2008 — as a different kind of investor with a different rulebook (Schedule I). An Indian entity making ODI into a foreign company that has step-down subsidiaries is permitted under the automatic route, subject to its own conditions — a financial-commitment ceiling of 400% of the entity’s net worth, bona fide business activity, and the standard reporting. (One condition carries over unchanged: the two-layer round-trip rule below binds the LLP’s investment exactly as it would bind yours. The LLP removes the individual bar; it doesn’t switch off the layering limit.)

So the workaround: instead of you holding the Delaware shares, your LLP holds them. You form an LLP, the LLP subscribes to the C-Corp’s founder stock, and the C-Corp is then free to set up its Indian subsidiary. The individual-level restriction doesn’t trigger, because the individual is no longer the direct investor.

In practice it looks like this. Say the founders are Mayank and Sumit. Each forms his own LLP — Mayank LLP and Sumit LLP. An LLP needs a minimum of two partners, so each founder holds 99.99% of his LLP and parks the remaining sliver with someone he deeply trusts — typically a parent or spouse. One detail that matters more than it looks: that sliver partner must be an Indian resident. Put a non-resident sibling in that seat and the LLP itself acquires a foreign-investment character, which quietly undermines the whole premise of using a clean Indian entity as the investor. Then: Mayank LLP buys Mayank’s shares in the Delaware company; Sumit LLP buys Sumit’s.

Why one LLP each, instead of one shared LLP? Because a shared LLP welds your shareholdings together. If the co-founders fall out — and pretend all you like, it happens — untangling who owns what inside a single LLP is a second dispute layered on top of the first. Separate LLPs mean each founder’s stake, exit proceeds, tax position, and filings stay cleanly his own. It also keeps future dilution, secondaries, and estate questions per-founder instead of per-committee.

How does the money actually move from the LLP to the Delaware company?

This is where founders meet three new acronyms — ODI, AD bank, UIN — and where the process becomes less about law and more about patiently feeding a bank exactly what it asks for.

When your LLP subscribes to shares of a foreign company, that’s an Overseas Direct Investment. ODI transactions don’t go through any random branch: they’re routed through an AD Category-I bank — an RBI-authorised dealer — that you designate. The AD bank is the gatekeeper: it checks the paperwork, files the reports, and is itself on the hook with RBI if it remits money without them.

The sequence, in practice:

One practical note from experience rather than from the rulebook: banks strongly prefer the sequencing above — LLP first, then the US entity, then ODI — and many insist the ODI remittance be the first meaningful money into the new US account. None of that is statute; all of it is how AD banks behave, and fighting your bank’s checklist is a losing game. Budget one to two months for the whole loop and be pleasantly surprised if it’s faster.

What does the US-side incorporation actually involve?

The Delaware side is the easy half, which surprises people. The standard stack:

Incorporation. File the certificate of incorporation in Delaware — directly, or via Stripe Atlas or Clerky, which exist to make this a form-filling exercise. You’ll authorise a standard pool of shares (10 million is the convention) with a tiny par value.

The post-incorporation bundle. Incorporation gives you a shell; the first board actions give it organs: bylaws, initial board consent appointing officers, founder stock purchase agreements (with vesting — typically four years, one-year cliff), and CIIA/PIIA agreements — the confidential-information and invention-assignment agreements under which everyone working on the company, founders first, assigns their IP to it. Investors will read these in diligence before they read your deck. Then the EIN from the IRS, and the US bank account.

Issuing founder shares — the cross-border twist. In the LLP route, the stock purchase agreements are with the LLPs, and the subscription money is the ODI remittance from the timeline above. Shares get issued once the money lands, so your cap table and your FEMA paper tell the same story.

The 83(b) question. US founders holding vesting stock directly routinely file an 83(b) election with the IRS within 30 days of the stock purchase, to be taxed at grant (when the shares are worth pennies) rather than at each vest. Whether and how any of that applies when an LLP is the stockholder is genuinely unsettled — the analysis changes when the founder isn’t the direct holder, and it may not map at all. Get a US tax attorney’s answer for your specific setup before the 30-day window closes, because it cannot be fixed afterwards.

The IP tail. Founders usually start building months before any entity exists. Who owns that code, and how it gets cleanly into the Delaware company across an international border, is a real question with real diligence consequences — and it deserves its own article rather than a paragraph. Parking it.

What else should be on your radar?

Four things that don’t fit neatly above but will find you eventually.

Pre-incorporation expenses. The AD-bank directions allow remittance of up to USD 100,000 per foreign entity for pre-incorporation expenses — useful for Delaware fees and setup costs before the ODI proper. For a resident individual this counts against the LRS limit.

The two-layer rule. Once your structure “round-trips” — the foreign entity invests back into India — the OI Rules cap it at two layers of subsidiaries, and as noted above, this binds the LLP’s investment too. Delaware parent → Indian subsidiary is fine. Start adding layers and you hit Rule 19(3). Design flat.

TCS on remittances. Outward remittances interact with Tax Collected at Source, and the thresholds and rates have been a moving target across recent budgets. Do not take a number from a blog post (including this one) — confirm the current rate for your remittance purpose with a CA at the time you remit.

The clock never stops. The APR is annual, forever, for as long as the LLP holds the shares. Founders treat the ODI as a one-time gate; RBI treats it as a subscription.

The short version

You can’t hold your own US-parent-India-subsidiary structure as a resident individual — Schedule III forbids it where you have control, and no approval route reopens it. Every real startup needs the Indian subsidiary: for hiring, and because an arm’s-length subsidiary is how you contain PE risk rather than create it. So the shareholding goes through one Indian LLP per founder — a market-practice structure you set up with FEMA-aware counsel, not off a blog post — properly capitalised, investing as ODI through an AD bank against a UIN, with Form FC now and an APR every December as the ongoing price of admission. The Delaware side is genuinely the easy part.

Which is exactly why the order matters: before you click “incorporate” on Stripe Atlas or Clerky, make sure your India-side ownership structure is ready. Setting it up in the right sequence is a few weeks of paperwork; untangling it after the shares are already in the wrong hands is a different kind of project.