Site icon HexGn

GIFT City, SEZ or STPI: India’s Location Incentives, Explained Simply

GCC & INDIA India’s incentives, explained HEXGN INSIGHTS · 09

Every India-entry deck reaches the slide with three acronyms — SEZ, STPI, GIFT — and most rooms respond the same way: someone remembers a tax holiday from 2009, someone else has read a GIFT City headline, and the decision drifts toward whichever regime sounds most generous. The regimes are real and worth money; they are also widely misunderstood, because each was built for a different era and a different problem. This guide explains what each actually offers a new entrant today, charts the honest comparison, and states the rule that keeps incentive decisions from corrupting location decisions.

The idea in brief. STPI is a lightweight registration with modest benefits and full location freedom; SEZs offer campus-grade infrastructure and remaining indirect-tax advantages, bought with zone-boundedness and compliance weight; GIFT City’s IFSC offers India’s strongest fiscal package — including a ten-year tax holiday — but only for work that genuinely fits its financial-services framework, in a talent market still maturing. None of the three should ever pick your city: incentives are tiebreakers between talent-equivalent options, and the regimes evolve — verify current rules at the primary sources before committing.

Why three regimes exist at all

Each regime answers a different decade’s policy question. STPI (1991) asked: how do we let software exporters operate anywhere and still get duty and process benefits? SEZs (2005 Act) asked: how do we build export enclaves with world-class infrastructure and fiscal privileges? GIFT City’s IFSC (operationalised through the 2010s, unified under the IFSCA in 2020) asks the most ambitious question: can India host the offshore financial centre that Indian finance itself uses — competing with Singapore and Dubai — with a regulatory island and incentives to match? Read as answers to different questions, the three stop being rivals and become tools.

STPI: the veteran, correctly sized

The Software Technology Parks of India scheme powered the original export boom, and its famous income-tax holiday shaped a generation of assumptions — but that holiday sunset years ago. What remains is genuinely useful and genuinely modest: duty benefits on imported equipment, a simple export-reporting track, and — the underrated feature — complete location freedom: an STPI unit operates from ordinary offices in any city, including tier-2 locations no SEZ serves.

Right-sized for: entrants who value flexibility over fiscal engineering; centres headed for tier-2 cities; anyone whose business case works without incentives — which, per the rule below, should be everyone.

SEZ: real benefits, real constraints, evolving rules

The Special Economic Zone framework built the campuses that define India’s GCC skyline — and for early entrants, delivered substantial direct-tax holidays. For a new entrant the picture is narrower: the direct-tax sunset has passed, while indirect advantages persist (duty-free capital goods, GST treatment on qualifying supplies) alongside the practical draw — plug-and-play, campus-grade infrastructure engineered for exactly this industry.

The constraints are equally concrete: you locate inside the zone; compliance is heavier (customs-bonded logic, movement rules); and hybrid-work rules for SEZ units — a live policy issue since the pandemic — have required successive relaxations. Policy is genuinely in motion: reform proposals, including partial de-notification flexibility, surface regularly. The planning posture: treat SEZ benefits as real but dynamic, and price the compliance overhead honestly against your operating model — particularly if flexible work is central to your talent proposition (and in India’s market, it is; see our attrition analysis).

GIFT City: the strongest package, for the right mandate

Gujarat International Finance Tec-City is a different kind of object: a purpose-built international financial services centre with its own unified regulator (IFSCA), treated for financial regulation as offshore. For qualifying IFSC units the fiscal package is the strongest in India — headline: a 100% income-tax holiday for any ten consecutive years of the first fifteen, plus GST and duty advantages and a single-window regulator that behaves like the pitch deck says.

The honest boundaries: the package attaches to qualifying financial-services activity — banking, insurance, asset management, fintech, and notably the technology and global in-house centres serving them — not to generic software export. And GIFT’s local talent pool, while growing quickly with marquee tenants, is a fraction of a metro hub’s; most serious operators run a hub-and-spoke: the regulated, qualifying work in GIFT capturing the incentive, the deep engineering bench in Bengaluru or Hyderabad. For a financial-services GCC, the arithmetic deserves genuine attention; for anyone else, GIFT is usually the wrong question.

The comparison, charted

Incentive strength is mandate-shaped Indicative fiscal-benefit strength for a NEW entrant, indexed (FS-qualifying for GIFT) GIFT City IFSC90SEZ (indirect)45STPI20 Illustrative model — HexGn analysis; parameters described in the text.

One caution on reading the chart: benefit strength is conditional — GIFT’s towering bar applies only to FS-qualifying mandates, SEZ’s middle bar assumes export-dominant revenue and campus tolerance, STPI’s modest bar applies almost unconditionally. A chart of your eligible benefits, for your mandate, is the one that matters — build it with advisors before any of these numbers move a decision.

Exhibit: the regimes side by side

STPI SEZ GIFT City (IFSC)
Era & design goal 1991 — export from anywhere 2005 — export enclaves 2015+ — offshore financial centre
Fiscal benefit today (new entrant) Modest (duty/process) Moderate, indirect Strongest — 10/15-yr tax holiday (FS-qualifying)
Location freedom Full — any city Zone-bound Zone-bound (GIFT)
Compliance weight Light Heavier, evolving Specialised, single-regulator
Talent-pool depth today Wherever you choose Major hubs Growing from small base
Best-fit entrant Flexibility-first builds Large export campuses FS institutions & their tech centres

The rule that keeps companies out of trouble

Now the discipline the whole analysis exists to serve — and the chart that expresses it:

What should carry the location decision HexGn recommended decision weights, % (normative, not survey data) Talent supply & retention60%Total cost25%Incentive regimes10%Other factors5% Illustrative model — HexGn analysis; parameters described in the text.

Incentives are a tiebreaker, never a driver. The reasoning is arithmetic, not aesthetic: a tax holiday operates on profit margins, while talent supply, retention and ramp operate on the entire cost and output base — the levers examined across this series (cost model, attrition, city choice) move sums that dwarf any regime’s benefit. A centre in the wrong city with the right incentive is a discounted mistake. The professional sequence: shortlist cities on role-level talent evidence; then let regimes optimise among talent-equivalent options; then verify current rules — all three regimes evolve, and this article, like every secondary source, ages from the day of writing.

What could go wrong

The tail-wags-dog build: mandate contorted to fit a regime’s qualifying rules — engineering teams squeezed into an FS wrapper, or growth plans bent around zone boundaries. If the fit requires contortion, the benefit is telling you it was not for you. Grandfather-clause confusion: benefits your advisors enjoyed as 2010 entrants are not your benefits; insist on current-law analysis. Policy drift: hybrid-work rules, de-notification frameworks and IFSC scope all move; assign someone to own regime-watch as a standing item, not a one-time memo.

A decision agenda

  1. Write the mandate first: what work, what revenue model, what growth shape. Eligibility flows from mandate, not preference.
  2. Shortlist cities on talent evidence alone (role-city mapping, per our location analysis).
  3. Within the shortlist, model eligible benefits per regime — current law, your numbers, advisor-verified.
  4. Price the constraints (zone-boundedness, compliance, hybrid rules) as costs in the same model.
  5. Decide, document the assumptions, and diarise an annual regime review.

The fourth layer: state incentives, read like term sheets

Beneath the three national regimes runs a fourth layer this guide would be incomplete without: state-level GCC and IT policies, which have quietly become the most negotiable money in the stack. Multiple states now publish dedicated policies offering capital subsidies, payroll and training reimbursements, stamp-duty waivers, rental support and single-window facilitation — with several explicitly targeting capability centres, and the emerging-city states bidding hardest (the tier-2 analysis shows why).

Reading a state policy like a term sheet, four clauses matter:

Practically: shortlist cities on talent first, then request the current policy packet from each state’s investment-promotion agency (Invest India brokers the introductions willingly — courting you is their mandate), and negotiate — the published policy is frequently a floor, not a ceiling, for projects with credible headcount plans. Then apply this article’s rule one final time: a state subsidy is the tiebreaker’s tiebreaker, never the reason your engineers live somewhere their community does not.

A worked comparison: a financial-services centre, two designs

The regimes become vivid in a concrete case — composite, with round numbers chosen for arithmetic clarity. A global asset manager plans a 300-person India centre: 120 roles in fund operations and risk (squarely FS-qualifying), 180 in platform engineering and data.

Design A — everything in Bengaluru, STPI registration: deepest talent for all 300 roles; modest incentives; the full metro cost-and-competition structure across the board. Simple, robust, incentive-light.

Design B — hub-and-spoke with GIFT: the 120 qualifying roles in a GIFT IFSC unit capturing the ten-year holiday on that unit’s profits; the 180 engineering roles in Bengaluru where their talent actually lives. The arithmetic that surprises boards: because the tax holiday operates on the qualifying unit’s profit margin while talent economics operate on the entire cost base, Design B’s fiscal gain is meaningful but second-order — typically low-single-digit percentages of total centre economics — while its operational cost (two sites, split culture, transfer-pricing discipline between the units) is real and permanent.

For this composite, Design B still wins — the qualifying block is large and genuinely FS-shaped, and GIFT’s regulator adds non-fiscal value for regulated activity — but it wins narrowly, and only because the talent map supported the split anyway. Shrink the qualifying block to 40 roles, or make the engineering talent the scarce constraint, and Design A wins outright despite surrendering the holiday. That sensitivity — the decision flipping on talent shape while the incentive stays constant — is the tiebreaker rule demonstrated with arithmetic rather than asserted.

Questions boards ask about the regimes

“Our advisors mention ‘grandfathered’ SEZ benefits — can we get those?” No. Grandfathering protects entrants who qualified before sunset dates; a new unit receives the current framework only. Any model in your deck showing the old direct-tax holiday for a new SEZ unit is describing someone else’s past, not your future.

“Is GIFT City only for banks?” No — the IFSC framework spans banking, insurance, funds, market infrastructure, fintech and, importantly for this series, technology and global in-house centres serving financial services. The qualifying question is the activity’s substance; IFSCA’s published frameworks are the primary source, and they are unusually readable as regulators go.

“Do these regimes constrain hybrid work?” Differently: STPI effectively not; SEZ units have needed evolving permissions for work-from-home arrangements (rules have relaxed repeatedly — verify current state); GIFT operates its own framework. If flexible work is central to your talent proposition — and our attrition research argues it is — this belongs in the constraint pricing, not the footnotes.

“Can we combine regimes?” Yes, and sophisticated footprints often do: a GIFT unit for qualifying FS work, an STPI-registered office in a tier-2 spoke, a SEZ campus for the export-heavy scale function. Each unit’s registration follows its own activity and location — which is precisely why the mandate-first sequence in the decision agenda matters.

“How stable are these frameworks politically?” Directionally stable, detail-dynamic. Export promotion and GIFT’s success are multi-government national priorities; the specifics (rates, sunset dates, WFH rules) move within that stability. Hence the standing regime-watch: the strategy survives rule changes; specific arithmetic may not.

How the regimes evolved — a compressed timeline

The regimes make more sense as history than as brochure, and the compressed timeline doubles as a caution about assuming stasis:

The pattern across three decades: regimes launch generous, mature narrower, and are periodically superseded by newer vehicles aimed at newer policy goals. A location strategy built on any single year’s arithmetic inherits that lifecycle — which is the timeline’s argument, one more time, for talent-first decisions with incentives as the annually-reviewed tiebreaker.

Methodology & data notes

Regime descriptions reflect the publicly documented frameworks at the linked primary sources; the benefit-strength chart is an indicative synthesis (indexed, mandate-conditional), not a tax computation. Nothing here is tax or legal advice — the regimes’ details change, and current-law verification with qualified advisors is part of the method, not a disclaimer.

References & further reading

HexGn’s location work runs exactly this sequence — talent evidence first, regime optimisation second — so incentives sharpen good decisions instead of financing bad ones.

Exit mobile version