Running the Captive Tax Math: GIFT City's IFSC Tax Holiday and Zero-Rated GST Versus Onshore Section 37 for Indian Corporates in 2026

For most of the last decade, the Indian captive conversation stalled on a single wall: India still has no dedicated onshore captive licensing regime, and IRDAI's general-insurance entry capital sits at Rs 100 crore, which no corporate is going to lock up to insure only its own plants. So the practical captive has always been offshore, and the live offshore option for an Indian group is now GIFT City's International Financial Services Centre (IFSC).

What has changed the maths is the direction of recent Budget policy and the IFSC tax framework that now sits under it. The IFSC unit tax holiday gives a qualifying unit a 100 percent income-tax deduction on its income for a long run of consecutive years under Section 80LA, and the policy push around GIFT City has been to widen and lengthen that window to make the centre globally competitive. After the holiday ends, profits in the IFSC are taxed at a concessional flat rate well below the roughly 25 to 35 percent effective burden a normal Indian company faces.

That is a structural feature, not a marketing line. A captive is a long-life vehicle; a board approving it in 2026 is underwriting a multi-decade cash-flow story, and the after-tax cost of retained risk is what decides the structure. The broker who walks into that room with a premium quote and a slide on "control" loses. The broker who walks in with a long-horizon after-tax total-cost model wins the mandate.

This post is that model. We compare the two routes a CFO will actually weigh: keeping the risk onshore and deducting the premium (or self-insured cost) under Section 37, versus moving it into a GIFT City IFSC insurance vehicle and running it through the holiday and GST regime. The aim is a number, not a brochure.

Start with the route most Indian corporates are already on, often without calling it a strategy. They buy a conventional onshore policy from a GIC-Re-backed insurer, pay the premium, and deduct it as a business expense under Section 37(1) of the Income-tax Act. That deduction is real and immediate: a genuine, arm's-length insurance premium incurred wholly for the business is allowable, reducing taxable income at the company's marginal rate.

The limits matter more than the deduction. Three things quietly erode the onshore economics.

First, GST on premium. Commercial general-insurance premium attracts 18 percent GST. For a manufacturer or a hospital chain that cannot fully set off that input (large parts of insurance input credit are restricted or sit against exempt turnover), a meaningful slice of the GST is a dead cost layered on top of the premium.

Second, no retention of underwriting profit. When a corporate's loss ratio runs well below the priced rate, that surplus is the insurer's, not the buyer's. A good-risk corporate is effectively subsidising the pool. Over a soft phase of the cycle that is tolerable; through a hard market with rising fire and STFI rates, it is the single biggest leakage.

Third, the deduction is only as good as the premium is reasonable. If a group tries to inflate premium to a related insurer to shift profit, Section 40A(2) and transfer-pricing scrutiny bite. So Section 37 caps the benefit at the honest market rate.

The onshore baseline, then, is: immediate deduction at marginal rate, minus an 18 percent GST drag, minus all surrendered underwriting profit, minus zero ability to deduct retained risk you do not formally insure.

Now run the same rupee through a GIFT City IFSC insurance vehicle. The Indian operating company still pays a premium; the difference is where that premium lands and how it is taxed at each leg.

The operating company's premium remains an ordinary business expense, deductible under Section 37, exactly as before. So the parent leg is tax-neutral against the onshore baseline. The economics turn on the second leg, the captive's own profit.

A captive that prices its risk correctly and runs a sub-100 percent loss ratio earns an underwriting profit plus investment income on reserves. Onshore, that profit would be taxed at the insurer's rate and never returned to the group. Inside the IFSC, that profit sits in a unit eligible for the Section 80LA tax holiday: a 100 percent income-tax deduction on the unit's income across a long run of consecutive years, with a concessional flat rate thereafter. The retained underwriting margin that an onshore arrangement gives away is now captured and, for the length of the holiday, sheltered.

Then GST. Insurance and reinsurance services rendered by an IFSC unit to clients outside India, and many intra-IFSC services, are treated as export of services and zero-rated under GST, and services received by IFSC units enjoy broad GST exemption. The 18 percent drag that sits on a domestic premium does not attach in the same way to IFSC-routed reinsurance flows. For a large programme this GST differential alone can move the decision.

The vehicle itself matters. A corporate can either set up an IFSC Insurance Office (IIO) or, more commonly and far more cheaply, use a fronting insurer onshore that cedes the risk to a reinsurance vehicle in GIFT City. The IIO assigned-capital requirement is in the region of USD 1.5 million equivalent, an order of magnitude below the Rs 100 crore onshore wall, which is precisely why GIFT City, not Mumbai, is where the Indian captive actually lives.

Here is the comparison framed the way a board paper should frame it. Take a group with a Rs 40 crore annual property and liability programme and a normalised loss ratio of 55 percent. The figures below are illustrative ratios, not a quote; the structure is what matters.

Onshore route, per year:

• Premium paid and deducted under Section 37: Rs 40 crore, saving tax at roughly 25 percent, so a net cost of about Rs 30 crore before GST.
• GST at 18 percent, largely unrecoverable: roughly Rs 7.2 crore of additional sticky cost depending on input-credit position.
• Underwriting surplus surrendered (45 percent of premium less the insurer's costs and the cycle loading): retained by the insurer, returned to the group as zero.

GIFT City route, per year:

• Operating-company premium to the fronting insurer, deducted under Section 37: same Rs 40 crore deduction.
• Of the premium, the slice ceded to the GIFT City vehicle carries its underwriting and investment surplus into the IFSC income-tax holiday, so that surplus is retained by the group rather than lost.
• The reinsurance leg through the IFSC is zero-rated for GST rather than carrying the sticky 18 percent.

The captive does not beat onshore on day-one cash; it beats it on retained surplus compounded over the holiday. In a group running consistently good loss experience, retaining even half the surrendered underwriting margin, sheltered across the holiday years and earning investment income inside the IFSC, is the difference that pays back the setup and running cost (actuarial, audit, IIO or fronting fees, IFSCA compliance) several times over.

The honest caveat, which a credible broker states first: this only works for a group with good, stable loss experience and enough premium volume to justify fixed running costs, typically a programme north of Rs 20 to 25 crore. A volatile-loss or sub-scale group keeps more value buying conventional cover and deducting it. The captive is a tax-and-retention play for the well-run risk, not a universal answer.

A clean model can still die on regulatory detail, so build these into the recommendation rather than discovering them after the board says yes.

Fronting is not frictionless. Indian general business written for an Indian risk usually has to be placed through an admitted onshore insurer first; that insurer fronts and then cedes to the GIFT City vehicle. IFSCA's framework leans on maximising retention within India and applies cession and retrocession limits, so a group cannot simply pass 100 percent of every risk offshore at will. The fronting insurer charges a fee (commonly a few percent of premium) for lending its paper and its solvency. That fee is a real line in the model and must be netted against the holiday benefit.

Order-of-preference and obligatory cessions still apply to the Indian leg. GIC Re's obligatory cession and the domestic order of preference shape how much risk genuinely reaches the captive on Indian business. The captive's clean run is easiest on the group's international exposures and on layers above the domestic retention.

Substance and pricing have to be real. The premium ceded must be arm's-length and actuarially supported; IFSCA expects genuine insurance operations, and Indian transfer-pricing rules apply to the related-party cession. A captive priced to dump profit offshore invites both a transfer-pricing adjustment and a challenge to the very deduction the structure relies on. The discipline that makes the tax work is the discipline of pricing the risk honestly.

GAAR and economic substance sit in the background. The General Anti-Avoidance Rules can be invoked where an arrangement's main purpose is a tax benefit without commercial substance. A captive with real risk transfer, real claims, real reserves and a real board defends itself; a paper vehicle does not. Brokers should frame the captive as a risk-management decision that happens to be tax-efficient, never the reverse, because that framing is also the legal defence.

A specialist earns trust by saying when their product is wrong for the client. The GIFT City captive loses to plain onshore cover in several common situations, and brokers should screen for them early.

Sub-scale premium. Below roughly Rs 20 crore of annual premium, the fixed running cost of a captive (actuary, auditor, fronting fee, IFSCA filings, board and management time) eats the retained-surplus benefit. A mid-market manufacturer is almost always better off optimising its self-insured retention on a conventional policy than building a vehicle.

Volatile loss experience. The entire captive thesis is retaining underwriting surplus. A group whose loss ratio swings from 30 percent to 130 percent across years is not retaining surplus; it is retaining tail risk it must capitalise. For that profile, transferring volatility to a commercial insurer and deducting the premium is cheaper and calmer.

Heavy domestic-only exposure. Because fronting, obligatory cession and the order of preference constrain how much Indian risk reaches the captive, a group with no international footprint sees a thinner net benefit than a multinational that can route global property and liability through the IFSC vehicle.

Short investment horizon. The holiday rewards a long hold. A group that may be acquired, restructured or wound down inside five to seven years will not compound enough sheltered surplus to clear the setup cost.

The reading exercise is simple. Pull three to five years of loss ratios, the total insurable premium spend, the domestic-versus-international split, and the group's expected life. If loss experience is good and stable, premium clears the scale threshold, a real share of risk is international, and the horizon is long, the GIFT City captive is likely the cheaper after-tax home for the risk. Miss any of those and the honest recommendation is to stay onshore and sharpen the conventional programme.

Translate all of this into a deliverable, because a CFO does not buy analysis, they buy a decision-ready document. A captive feasibility note that wins approval has five parts, and a broker can build it without waiting for a consultant.

1. A five-year loss-and-premium history for the group's whole programme, normalised for one-off catastrophe years, so the board sees the true retained-surplus opportunity rather than a single lucky or unlucky year.
2. A side-by-side after-tax total-cost table (described in words, since boards read the bottom line): onshore Section 37 cost net of the GST drag and surrendered surplus, versus the GIFT City route net of fronting fee, running cost and holiday-sheltered retained surplus, projected across the full holiday window the unit can claim.
3. The structure choice, IIO versus fronted reinsurance vehicle, with the assigned-capital and fronting-fee implications spelled out, plus a clear statement of which risks can realistically be ceded given cession and retention rules.
4. The compliance and substance plan: arm's-length pricing supported by an actuary, transfer-pricing documentation, IFSCA filings, and a governance structure that defends the arrangement against GAAR and against any challenge to the underlying deduction.
5. The kill criteria: the explicit thresholds (premium scale, loss volatility, horizon) at which the group should not proceed, which paradoxically is what makes the board trust the recommendation to proceed.

Deliver that and the broker has moved from quoting premiums to advising on the group's risk-financing architecture. That is a different relationship, a stickier mandate, and the one place where the GIFT City framework genuinely rewrites the broker's role. The long-dated IFSC holiday does not just change a captive's economics; it gives the broker a reason to be in the boardroom with a multi-decade model rather than in procurement with a renewal quote.