India Tokenised ₹1,025 Crore of a ₹53 Lakh Crore Bond Market. The First Issue Had 18 Buyers.

On 10 September, at the Global Fintech Fest in Mumbai, SEBI and the Reserve Bank of India launched Demat 2.0 — corporate bonds issued as native digital tokens on permissioned distributed-ledger infrastructure, with the cash leg settled in the RBI's wholesale digital rupee. CoinDesk's headline said India had begun tokenising a $620 billion corporate bond market.

Three issuers have used it so far, for a combined ₹1,025 crore, reported at about $107.2 million. REC Ltd, the state-owned non-bank lender, went first on 7 September with ₹500 crore sold to 18 investors. Larsen & Toubro followed on 9 September with ₹500 crore to four investors. IIFL issued ₹25 crore the same day, to one.

Eighteen, four, one. That is the entire market being described.

The Ratio, Done Carefully

This next calculation is DrafterDaily's arithmetic, and it is worth showing rather than asserting, because the conversion matters more than it should.

₹1,025 crore against India's outstanding corporate bond stock of roughly ₹53.64 lakh crore is about 0.019%. On the reported dollar figures — $107.2 million against $620 billion — it comes to about 0.017%.

Those two ratios do not agree, and the reason is instructive. The reported conversions imply different exchange rates: ₹1,025 crore at $107.2 million implies roughly ₹96 to the dollar, while ₹53.64 lakh crore at $620–627 billion implies roughly ₹85–86. Doing the division in rupees sidesteps the problem entirely, which is why we have. Either way the answer lands in the same place: less than one fiftieth of one percent of the market has touched this system.

The market-size framing is doing the opposite of useful work. It invites the reader to believe something large has occurred. What occurred is small, deliberate, and — this is the part the coverage skips — mechanically important for a reason that has nothing to do with size.

Why the Central Bank Had to Be in the Room

Start with what settlement risk is, because it is the thing being removed.

In a conventional Indian corporate bond trade, delivery of the security and payment of the cash are separate events, handled by separate systems, on a T+1 cycle. Between the moment a trade is agreed and the moment both legs complete, each side is exposed to the other: the buyer may have paid and not yet received, or the seller may have delivered and not yet been paid. Clearing corporations exist to stand between the two parties and absorb that exposure, which is why they collect margin, maintain default funds and impose membership requirements. The risk is not eliminated by any of that. It is intermediated, collateralised and priced.

What atomic delivery-versus-payment actually changes

Delivery-versus-payment is atomic when both legs are conditioned on each other inside a single transaction: either the bond token moves and the payment moves, or neither does. There is no window in which one side has performed and the other has not, because that intermediate state is not representable in the system. Settlement risk is not reduced or collateralised here. It stops existing as a category.

That only works if both legs sit on infrastructure capable of enforcing the condition jointly. The securities leg is the easy half — a token on a ledger. The cash leg is the hard half, and it is the reason this required a central bank rather than an exchange.

Central bank money, not a stablecoin, not a tokenised deposit

The distinction is not pedantic; it is the crux of the design. A payment in commercial bank money is a claim on a commercial bank. Settling with it means the buyer's bank must be good for the money at the moment of settlement, so a residual credit exposure to that bank persists inside a system advertising that it has removed credit exposure. A stablecoin has the same structural problem with a less regulated issuer: you have swapped settlement risk for issuer risk, which is not obviously an improvement.

Central bank money carries no issuer credit risk within its own domestic system, because the issuer is the entity that defines the currency. Settling the cash leg in wholesale CBDC is what lets the atomicity guarantee hold all the way down, rather than resting on an assumption about some bank's solvency. That is why the RBI's participation is the substantive fact about Demat 2.0, and the $620 billion is not.

The practical effect reported so far is prosaic and real: funds that took two to three days to reach investors under the conventional cycle arrive immediately, which means capital can be redeployed rather than sitting in transit.

What a Pilot With No Secondary Market Cannot Tell You

Now the limits, which are substantial and are mostly acknowledged in the coverage without being weighted properly.

• There is no secondary trading. Bonds have been issued and settled. They have not been traded between holders on this infrastructure. Secondary trading is a planned later phase.
• There is no retail access. The pilot is institution-only and runs inside India's existing regulatory perimeter. Retail participation is also a later phase.
• Corporate actions are not automated. Smart-contract handling of coupons and redemptions is described as a future capability, not a shipped one.
• The infrastructure is permissioned. The ledger is private, built and operated by regulated market-infrastructure institutions with implementation support from NPCI. It is not a public blockchain and inherits none of the properties of one.

The most common claim made about tokenised bonds is that they improve liquidity. Nothing in this pilot tests that. Liquidity is a secondary-market property, and the secondary market does not exist here yet. Issuance working is evidence that issuance works. It is not evidence that a market works, and the two are routinely conflated.

One further detail, minor but worth getting right: REC's issue settled on 7 September, three days before the 10 September launch event. That sequence is ordinary for a pilot — you demonstrate something that has already run — but it means the launch date and the first-issuance date are two different facts, and coverage that merges them creates a false chronology.

What to Watch

• The investor count. Eighteen, then four, then one is not a trend line; it is three data points from a controlled pilot. If a later issue clears fifty or a hundred allocations, the buyer base is genuinely widening. If it stays in the low double digits, this remains a demonstration.
• The first secondary trade. It is the only thing that will test the liquidity claim. Until it happens, the claim is untested rather than disputed.
• Automated corporate actions. A coupon paid by smart contract without manual reconciliation would be the first evidence that tokenisation reduces operational cost, not just settlement risk.
• Whether other central banks copy the cash leg. The portable idea here is wholesale CBDC as the settlement asset for tokenised securities. That travels in a way India's specific bond-market structure does not.

Demat 2.0 is a small thing engineered correctly, described by a number that makes it sound like a large thing that is not yet true. The $620 billion is the size of the room. About ₹1,025 crore has walked into it.