Stamp Duty on Insurance Policies in India: Who Pays, How Much, and Why State Lines Matter

Every insurance policy issued in India is a chargeable instrument, and every one of them carries stamp duty. It is a small line, often a rupee-scale figure absorbed into the premium computation and never separately noticed by the buyer, which is exactly why it is worth a careful look. Small taxes that nobody reads are the ones that surface at the worst moment, and for stamp duty the worst moment is a contested claim in a courtroom.

The governing law is the Indian Stamp Act, 1899. A policy of insurance is an instrument specified in Article 47 of Schedule I to that Act, which makes it dutiable. The duty is not a tax on the risk or a levy like GST that scales with the premium in any large way; it is a stamp on the instrument, historically nominal for most classes and structured differently for marine cover.

The reason this deserves an explainer, when the amounts involved are trivial, is that almost everything a buyer or broker believes about insurance stamp duty is slightly wrong. They assume it varies wildly by state (mostly it does not, for the policy itself). They assume the insurer simply bears it (the statutory default is more nuanced). They assume an under-stamped policy is a housekeeping issue (in litigation it can be an admissibility problem). And they never think about the endorsement paperwork, which raises its own instrument questions.

This post walks the duty end to end: who bears it, how it differs by class, where state lines genuinely matter, why it appears in premium and endorsement documents, and the one real risk, which is what an unstamped instrument does to a claim in court.

The most useful correction first, because it reverses a widely held assumption. For most instruments, stamp duty is the classic example of state variation: the rate on an agreement, a bond, a lease, or a conveyance is fixed by each state, and the same document costs different amounts in Maharashtra, Karnataka, and Tamil Nadu.

A policy of insurance is not in that category. Under the constitutional division of powers, the rate of stamp duty on policies of insurance is a Union matter. Entry 91 of the Union List reserves to Parliament the power to fix stamp duty rates on a specified set of commercial instruments, including bills of exchange, cheques, promissory notes, bills of lading, letters of credit, transfer of shares, and policies of insurance. Because the rate sits with the Centre, it is prescribed uniformly in the central Indian Stamp Act, 1899 and its Schedule I, rather than left to each state to set.

The practical consequence is that the duty on the policy instrument itself does not swing from state to state the way an agreement's duty does. A fire policy is stamped on the same rate basis whether it is issued in Gujarat or West Bengal, because that rate is a national rate.

This matters for a buyer with operations across states or a broker placing a national programme: the instinct to worry about multi-state stamp-duty arbitrage on the policy is misplaced, because the arbitrage that exists in other instruments is largely absent here. Where state lines do matter, covered in a later section, is the machinery of stamping and the ancillary documents around the transaction, not the headline rate. Get the constitutional point straight first: the policy rate is set by the Centre, so it is one of the few duties that does not depend on which state you are in.

The common shorthand is that the insurer bears the stamp duty, and in practice that is what a buyer sees: the insurer issues the policy already stamped and folds the cost into its pricing and administration. But the statutory default is more precise than the shorthand, and the nuance occasionally matters.

Section 29 of the Indian Stamp Act, 1899 allocates, in the absence of an agreement to the contrary, who bears the expense of providing the proper stamp for various instruments. For insurance it draws a distinction that surprises people:

• For a policy of fire insurance, the expense falls on the person issuing the policy, that is, the insurer.
• For a policy of insurance other than fire, the default falls on the person effecting the insurance, that is, the insured.

So the tidy belief that the insurer always bears the duty is not quite the statutory position. For non-fire classes, the Act's default puts the expense on the insured, subject always to any agreement to the contrary.

Why does this rarely cause a problem in practice? Because the insurer is the party that physically issues the instrument and denotes the duty on it, so the insurer stamps the policy as a matter of course and prices for it. The allocation in Section 29 is about who bears the expense as between the parties, not about who does the mechanical stamping, and the market convention of the insurer stamping-and-pricing has smoothed the distinction into invisibility. The point to retain is narrow but real: if a dispute ever turns on who ought to have borne the duty, the answer is not automatically the insurer, and Section 29's fire-versus-non-fire split is where that question is decided.

Stamp duty on insurance is not a single figure; Schedule I structures it by class of cover, and the structure is worth understanding even though the amounts are small.

Marine insurance has historically been the class with the most elaborate treatment. Sea-insurance policies were dutied on a basis tied to the nature of the cover, distinguishing voyage policies from time policies and scaling with the sum insured or the voyage in the older schedule. Marine cover is the one class where the duty was designed to be more than a token, reflecting the centrality of marine insurance to commerce when the Act was framed.

Fire and other general insurance classes carry duty that is nominal in character, a small fixed stamp on the instrument rather than a meaningful proportion of the premium. For a corporate fire or property programme running to a large premium, the stamp duty is an immaterial fraction of the total, which is precisely why it disappears from attention.

Life insurance instruments have their own line in the schedule, again modest in the commercial-lines context this publication addresses, and mostly relevant to individual policies rather than a corporate risk programme.

Two cautions about the numbers. First, the rupee amounts in Schedule I have been amended over time and should be checked against the current schedule rather than quoted from memory. Second, the smallness of the duty is no reason to be careless, because the risk attached to an insurance stamp, as the litigation section explains, is out of all proportion to the amount. The buyer's takeaway is structural: marine is the class where duty was built to be substantive, the general classes carry a nominal stamp, and the exact figure for any class is a lookup against the live schedule.

Having said the policy rate is a national matter, it would be misleading to stop there, because state lines do matter to an insurance transaction, just not where buyers expect. Three places in particular.

The machinery of stamping. How duty is physically paid and denoted (physical stamps, franking, or electronic stamping through a state-operated system) is administered at the state level. States have moved to e-stamping through central record-keeping arrangements operated in each state, and the mechanics, the portals, the vendors, and the workflow differ from state to state even where the underlying rate does not. An insurer or a large buyer operating nationally deals with several stamping systems, one per state of operation, and that operational variation is real even though the policy rate is uniform.

Ancillary instruments. An insurance transaction rarely travels alone. Financing of an insured asset brings hypothecation deeds and agreements; a performance obligation brings a bond or guarantee; a settlement brings a release or indemnity. These ancillary instruments are not policies of insurance, so their stamp duty is state-rate, and they carry exactly the state-to-state variation the policy itself lacks. The agreed-bank-clause and financier-interest paperwork that accompanies financed-asset cover is where a buyer meets state stamp variation in practice, not on the policy.

Place of execution and use. Which state's stamp law applies turns on where an instrument is executed and used, and one executed in a low-duty state but relied on in a higher-duty state can attract a top-up on the difference. For the policy this is muted by the national rate; for the ancillary instruments above it is a live consideration.

The honest summary: the policy is the wrong place to look for state variation, and the documents around it are the right place. A buyer who worries about the former and ignores the latter has the risk exactly inverted.

Two documents make stamp duty visible to a buyer who is paying attention: the premium computation and the endorsement.

On the premium computation, stamp duty appears because the insurer, having stamped the instrument, accounts for the cost. On most commercial policies it is a negligible line beside the premium and the 18 percent GST, which is why it is easy to overlook. But it is genuinely part of the cost of issuing the instrument, and a finance team reconciling a policy schedule to a debit note will occasionally see it broken out. It should not be confused with GST: GST is a tax on the supply of insurance services and is potentially creditable; stamp duty is a duty on the instrument and is not a GST input. They are different taxes with different treatment.

On endorsements, the question is more interesting and more often mishandled. An endorsement alters the policy, and whether a particular endorsement is itself a separately chargeable instrument depends on its nature. A minor administrative correction is one thing; an endorsement that effects a fresh chargeable transaction can raise its own stamp question. Most routine endorsements do not, but the buyer who assumes no endorsement ever attracts duty is making an assumption rather than a determination. The safer posture is to treat significant endorsements, particularly those that change the sum insured materially or add a new insured interest, as documents whose stamp position is worth confirming rather than assuming.

The practical point is that stamp duty is not only an issuance-time event: it travels with the instrument, and a policy properly stamped at inception can still accumulate an under-stamped endorsement later if nobody is watching the endorsement stream.

Everything above is low-stakes until a claim is disputed and the policy goes to court. Then the smallest tax on the schedule becomes a potential procedural weapon, and this is the reason the duty is worth getting right.

The mechanism sits in Section 35 of the Indian Stamp Act, 1899: an instrument that is not duly stamped is, as a general rule, not admissible in evidence for any purpose, and cannot be acted upon or registered. Read that against a contested claim. The policy is the central document a claimant relies on to prove the contract of insurance. If that instrument is unstamped or insufficiently stamped, an opposing party has a route to challenge its admissibility, turning a coverage dispute into a preliminary fight about whether the policy can even be put before the court.

The Act does provide a cure. An insufficiently stamped instrument can generally be admitted on payment of the deficient duty together with a penalty, which can run to a substantial multiple of the deficient amount, after which the instrument is treated as duly stamped for the purpose. So an under-stamped policy is usually not fatal, but it is a delay, a cost, and an avoidable vulnerability injected into a claim at the moment the insured most needs the document to work cleanly.

The exposure is asymmetric. Correct stamping costs almost nothing; incorrect stamping also costs nothing, right up until the policy is contested, when it costs a procedural detour, a penalty, and an advantage handed to the other side. That asymmetry, tiny cost to get right against real cost to get wrong at the worst moment, is the entire practical case for caring about a tax most buyers never read.

The response to all of this is light-touch, because the duty is small and the insurer does most of the work. A short discipline for a corporate buyer or a broker:

1. Confirm the policy is stamped at issuance. The insurer issues the instrument stamped, but confirm the policy document as received bears proper stamping rather than assuming it, a two-minute check against a disproportionate downside.
2. Do not chase state arbitrage on the policy. The policy rate is national, so there is nothing to optimise across states on the policy itself. Spend that attention on the ancillary instruments instead.
3. Watch the ancillary documents. Hypothecation deeds, bonds, guarantees, agreements, and financier-interest endorsements are state-rate instruments and carry the variation the policy lacks. On a financed-asset programme, these are where stamp exposure actually lives.
4. Treat material endorsements as stampable until confirmed otherwise. An endorsement that materially changes the sum insured or adds an insured interest is worth a stamp check. Routine administrative endorsements generally are not, but the assumption should be verified for the significant ones.
5. Fix under-stamping before, not during, a claim. If a stamping deficiency is ever noticed, cure it through the impounding-and-payment route proactively rather than waiting for an opposing party to raise admissibility in litigation. The penalty is the same; the timing is entirely in the buyer's favour if done early.
6. Keep stamping evidence with the policy file. Retain proof of proper stamping alongside the policy, so that if admissibility is ever challenged, the answer is a document rather than a scramble.

Stamp duty on insurance is a small, mostly national, insurer-administered duty. It becomes a problem only through neglect, and only in litigation, the combination that makes a cheap compliance habit worth having.