What Japanese Companies Actually Search For Before Entering India: Notes From the Advisory Desk

Most articles on "how to start a company in India" are written for a generic foreign investor — American, European, doesn't matter. Japanese companies rarely search for those. Over the years, the questions that land on my desk from Osaka, Nagoya, and Tokyo HQs are narrower, more specific, and often shaped by something a JETRO seminar or a sogo shosha's in-house legal team has already half-answered. This piece goes topic by topic through the things Japanese promoters and their Indian country managers actually type into a search bar — not the textbook FDI checklist.

I'll flag upfront: some of these areas have genuine grey zones. Where that's true, I've said so rather than papering over it.

1. Liaison Office or Wholly Owned Subsidiary — which one first?

This is, without exaggeration, the single most common opening question from a Japanese HQ. The instinct — shaped by a very Japanese approach to risk — is to enter cautiously: set up a Liaison Office (LO), watch the market for two to three years, then convert to a Wholly Owned Subsidiary (WOS) once the board in Tokyo is comfortable.

The practical issue: an LO cannot invoice, cannot undertake any commercial or industrial activity, and exists only to represent the parent, gather market information, and act as a communication channel. RBI renews LO approval typically for three years, and by the second renewal cycle, I've seen liaison offices get flagged in AD bank compliance reviews for functioning closer to a liaison-plus-quiet-sales-support role than what the original approval described. That mismatch, if caught, is a headache during renewal.

My honest advice, given more often than not: if the Japanese parent already knows it wants manufacturing or a joint venture with an Indian distributor within 18–24 months, skip the LO stage and go straight to WOS via the automatic route. The LO route makes sense mainly when the parent genuinely doesn't know yet whether India will have offtake for its product — pure market reconnaissance, nothing more.

A client from a Nagoya-based auto components manufacturer once asked me whether their LO could "help identify potential customers and share pricing information informally" while they decided on WOS timing. That crosses the line into liaison overreach — RBI's own FAQ is explicit that LOs cannot engage in any activity of a trading, commercial, or industrial nature, and pricing discussion edges toward commercial negotiation. I told them plainly: keep it to market research and parent-company representation, or convert to WOS now.

2. Permanent Establishment risk from seconded Japanese employees

This is the one I'd put at the top of the list for underappreciated risk. Japanese companies second expatriate staff to India far more heavily, and for far longer tenures, than most other foreign investors — often five, seven, even ten years for a plant head or technical adviser. That secondment pattern is exactly what creates India Permanent Establishment (PE) exposure for the Japanese parent under Article 5 of the India-Japan DTAA, particularly the "service PE" and "dependent agent PE" tests.

If the seconded employee continues to be on the Japan parent's payroll, receives instructions from Japan on substantive matters, or has authority to conclude contracts on the parent's behalf while sitting in the Indian office, tax authorities have a reasonable basis to argue the Japanese parent has a taxable presence in India — separate from the Indian subsidiary's own tax position. This has actually been litigated; cases involving Japanese trading houses and India PE exposure aren't hypothetical.

The fix isn't complicated in principle — transfer the secondee onto the Indian entity's payroll, route their reporting line functionally through the Indian entity, and keep the secondment agreement narrow — but it requires the Japan-side HR and the India-side company secretary to actually talk to each other, which in my experience is where it breaks down. Japanese HQs tend to keep expat payroll centralized in Japan for pension and benefits continuity reasons, and that's precisely the structure that creates PE risk.

One manufacturing client kept their plant GM on Japan payroll for four years "for pension continuity," while he signed local vendor contracts and represented the company at Indian regulatory hearings. That's a textbook PE fact pattern. We restructured it — dual employment agreement, India-sourced salary component, narrower authority — but it should have been addressed on day one, not year four.

3. The India-Japan Social Security Agreement — and what it does NOT cover

Japanese companies search for this because seconded employee cost is a genuine budget line, and the India-Japan Social Security Agreement (SSA), in force since October 1, 2016, is one of the more favourable ones India has signed. Under the detachment provision, a Japanese employee seconded to India can remain solely covered under Japan's pension system — with a Certificate of Coverage (CoC) obtained before arrival — for up to five years, extendable further in genuinely unforeseen cases. That avoids double social security contributions on both sides.

The part that trips people up: the SSA's totalization benefit applies to the EPS (Employees' Pension Scheme) component, not the EPF (Provident Fund) itself. An international worker under the agreement can withdraw their full EPF balance on ceasing employment regardless of the totalization rules — but EPS eligibility (which otherwise needs 10 years of contribution) is what gets "topped up" by counting Japanese coverage years. If you're budgeting secondment cost assuming blanket exemption from all Indian social security heads, you'll be off — EPF contribution obligations don't simply disappear without the CoC-backed detachment claim being filed correctly and in time.

A CFO in Yokohama asked me point-blank: "So our seconded staff pay nothing into Indian social security?" Not quite — if the CoC is filed properly before secondment begins, they're exempt from ongoing EPF/EPS contribution as a "detached worker," but this needs paperwork lodged with EPFO, not assumed by default. Miss the filing window and the exemption doesn't retroactively apply.

4. Repatiating profits — why the India-Japan DTAA rate matters more here than in most treaties

Dividend, interest, and royalty/technical service fee withholding under the India-Japan DTAA is capped at a flat 10% — under Articles 10, 11, and 12 respectively — regardless of shareholding percentage. That flat structure is actually more favourable than several of India's other major treaties (the India-US treaty, for instance, tiers between 15% and 25% depending on shareholding), which is why Japanese parent companies routing dividend repatriation directly, rather than through an intermediate holding jurisdiction, often don't need the layered structures that American or European groups sometimes use.

Where I still see Japanese clients overpay: not filing Form 10F and the Tax Residency Certificate correctly with the Indian AD bank before remittance, defaulting the withholding to India's domestic 20% (plus surcharge and cess) rate instead of the treaty's 10%. It is entirely avoidable paperwork friction, but it happens often enough that I now build a repatriation documentation checklist into the WOS engagement from month one, rather than waiting until the first dividend declaration.

A Tokyo-headquartered client's finance team once asked whether royalty payments for technology licensing to their India JV needed separate treaty relief from dividend relief. Yes — Article 12 (royalties/FTS) and Article 10 (dividends) are assessed independently, both capped at 10%, but each needs its own supporting documentation and, for royalties, arm's-length transfer pricing justification under Indian transfer pricing rules (see point 6 below) — treaty relief alone doesn't satisfy the TP requirement.

5. Japanese Industrial Townships (JITs) — and whether they're actually worth it

This is a very Japan-specific search term that almost no other nationality of investor uses. India has around a dozen JITs developed in coordination with METI and JETRO — Neemrana and Ghiloth in Rajasthan, Mandal in Gujarat, Sri City in Andhra Pradesh, Supa in Maharashtra, Tumkur in Karnataka, Pithampur in Madhya Pradesh, Greater Noida in Uttar Pradesh, MET City in Jhajjar (Haryana), and clusters around Chennai in Tamil Nadu, among others. Several sit along the Delhi-Mumbai Industrial Corridor, where JBIC and JICA have direct financing involvement — which is part of why the townships feel more "Japan-calibrated" than a generic Indian industrial estate.

Neemrana remains the flagship — upward of 48 Japanese companies including Daikin and Nissin Brake, with cumulative investment well into the billions of dollars, and a country-specific zone with dedicated infrastructure and, in practice, enough of a resident Japanese community that day-to-day life is easier for expat families than a first-time posting to, say, an unfamiliar Tier-2 Indian city.

The honest caveat: JITs are strongest for manufacturing operations that need ready-developed industrial land, plug-and-play utilities, and geographic clustering with other Japanese suppliers (a real advantage for auto-ancillary units that sell into a Japanese OEM's local supply chain). They are not automatically the right answer for a services, IT, or R&D-focused entity — for those, a JIT often adds real estate cost and location constraints without the ecosystem benefit. I've had clients default to "put it in Neemrana because that's what everyone does" when their actual business — a software delivery centre — would have been far better served by Bangalore or Pune.

6. Transfer pricing on royalty and technical fee payments to the Japan HQ

Japanese manufacturing subsidiaries in India routinely pay royalty for technology/know-how licensing and management/technical service fees to the parent — and this is one of the most heavily scrutinized categories in Indian transfer pricing assessments generally, not just for Japanese groups, but Japanese structures get extra attention because the royalty rates are often set centrally by HQ policy across all overseas subsidiaries without India-specific benchmarking.

Form 3CEB filing, a contemporaneous transfer pricing study benchmarking the royalty rate against comparable uncontrolled transactions, and clear documentation of what the royalty is actually paying for (brand, process know-how, technical support, or a bundle) are non-negotiable. I've seen assessment orders challenge royalty deductions specifically because the agreement described a bundled fee without disaggregating the components — making it hard to benchmark any single piece against comparables.

A client's Japan-side controller once told me their royalty rate was "fixed globally at 3% of net sales, same as our Vietnam and Indonesia subsidiaries." That's precisely the kind of one-size-fits-all approach an Indian TP officer will test — comparables in Vietnam or Indonesia don't establish an arm's length rate for the Indian market. We ended up commissioning an India-specific benchmarking study; it cost more upfront but closed off a real audit exposure.

7. Joint Venture vs. Wholly Owned Subsidiary — after the JV era

Older-generation Japanese investment in India — going back to the Maruti-Suzuki template — leaned heavily on JVs with Indian promoter families. That template has largely reversed since FDI policy opened up 100% automatic-route ownership in most manufacturing sectors; most new Japanese entries I advise on today choose WOS specifically to avoid the governance friction that defined a lot of 1990s and 2000s-era India-Japan JVs — disputes over board control, technology transfer pace, and profit reinvestment versus dividend preference have played out publicly enough (Maruti-Suzuki's own board disputes being the most visible example) that risk-averse Japanese boards now often prefer full control from day one, even at the cost of losing a local partner's market access and relationships.

Where JVs still make sense: sectors where FDI caps remain below 100% (a shrinking list, but still relevant in a few areas), or where the Indian partner brings genuinely difficult-to-replicate distribution or government relationships — defence-adjacent manufacturing being the clearest example.

8. GST and customs duty on CKD/SKD imports during the manufacturing ramp-up phase

Japanese manufacturers — automotive and electronics particularly — almost always begin Indian production with Completely Knocked Down (CKD) or Semi Knocked Down (SKD) kit imports before localizing components, and the customs duty differential between CKD, SKD, and fully localized production is steep by design (India's tariff structure is built to push localization). What gets searched here specifically is the interaction between the India-Japan Comprehensive Economic Partnership Agreement (CEPA) preferential tariff lines and the CKD/SKD duty structure — CEPA concessions apply to specific tariff headings, and they don't uniformly override the higher CKD/SKD duty slabs that apply irrespective of country of origin for certain automotive and electronics categories.

This is a genuine grey area in practice: the interplay between CEPA rules of origin certification and CKD-specific duty notifications needs a customs specialist reading the exact HS codes involved, not a general assumption that "CEPA gives us a lower rate." I flag this honestly to clients rather than promising a blanket saving — the actual number depends entirely on the specific tariff lines your kit components fall under.

9. Talent retention and the works-council culture gap

Not a legal or tax topic, but it's searched constantly, usually a year or two after the Indian entity is operational and the Japan HQ is puzzled by attrition numbers that would be considered a crisis back home. Indian white-collar attrition in manufacturing and engineering roles commonly runs materially higher than what Japanese HR planning models assume, and the consensus-driven, seniority-weighted decision culture that works domestically in Japan often frustrates mid-level Indian managers used to faster individual advancement.

I'm not a HR consultant and won't pretend to solve this in a compliance article, but from a pure governance-document standpoint: I do encourage clients to build compensation review cadence and promotion criteria into the Indian entity's HR policy documentation independently of the Japan parent's global HR framework, rather than importing it wholesale. It's a small thing that avoids a much bigger retention problem eighteen months in.

10. Protecting IP and trademarks before, not after, technology transfer

Last on this list but should really be first chronologically: Japanese companies are, in my experience, more relationship-trusting and less contractually defensive than Western counterparts when a JV or licensing arrangement is with a long-standing Indian business family — sometimes to their detriment. Trademark and patent registration in India should be completed before any technical know-how, drawings, or process documentation changes hands, not after the commercial relationship is finalized. India's "first to file" trademark regime means a domestic party can, in theory, register a confusingly similar mark before the Japanese entity does, and untangling that after the fact is slow and expensive.

A closing honesty note

Several of the areas above — PE risk from secondment, the CEPA/CKD duty interplay, and royalty rate benchmarking — don't have a single clean answer that applies to every company. They depend on the specific structure, the specific HS codes, and the specific royalty agreement language. If anything in this piece sounds too neatly resolved for your situation, it probably needs a closer, fact-specific look rather than a generic answer — including from us.

Learn More- https://accorppartners.com/services/incorporation/india-incorporation