DECODING THE TREND | Vol. 15 The End of Strata Escrows Programmable Rupee in Sub-Registrar Settlements

DECODING THE TREND | Vol. 15

The End of Strata Escrows Programmable Rupee in Sub-Registrar Settlements:

How Conditional e₹ Could Rebuild the Property Closing

**By Arindam Bose| BeEstates Intelligence | Finance & Funding | Vol. 15 | AUGUST 2026**

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

The ₹2 Crore Held Hostage

The buyer has signed. The seller is ready. The bank sanctioned the loan three weeks ago, and the sanction letter has already begun to smell faintly of expiry. Somewhere between this moment and the moment the property actually changes hands sits a gap that no brochure ever mentions: the money has to go somewhere it can be trusted, and in an Indian property transaction, trust has historically meant a third party.

That money is either sitting in a bank escrow account, moving through an RTGS instruction timed to registration, or riding on a demand draft physically carried into the sub-registrar's office. Whichever form it takes, the actual transaction risk lives in the same narrow window: the hours between funds leaving one account and the registry confirming a valid transfer. A document mismatch. A beneficiary account altered at the last minute. A registration server that crashes after biometrics are captured but before the index number generates. A seller who signs and then discovers, days later, that the buyer's cheque never cleared. None of this happens because Indian buyers and sellers are careless. It happens because the payment rail and the legal rail have never been the same rail — money settles on banking time, title settles on registry time, and nearly every closing-day failure of the last two decades has lived in the seam between the two.

In the next phase of India's property-finance digitisation, the most consequential change may not be how buyers borrow. It may be how they settle.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

PART ONE: THE ESCROW PROBLEM NOBODY PRICES

Escrow, in the textbook sense, is a neutral holding mechanism: a bank, trustee or lawyer holds funds until agreed conditions are satisfied, protecting the seller against a buyer who fails to fund and protecting the buyer against payment released before conditions are met. India has elements of this, but not a universal retail escrow system. Institutional deals, commercial transactions and RERA-mandated project collection accounts use formal escrow structures regularly. The ordinary resale apartment purchase — the transaction that accounts for the overwhelming majority of India's property volume — usually does not. It runs on something more improvised: a demand draft, a bank representative sent to physically witness execution, an RTGS instruction triggered around the moment of signing.

Whichever mechanism is used, it introduces the same three frictions. Operational dependency on an intermediary — the escrow bank, the lawyer, the developer's collection account — who becomes a single point of failure and a single point of delay. Manual reconciliation — someone has to confirm KYC, verify the beneficiary, check that funds cleared, and cross-reference that against a document execution that is happening in a different building. And release-timing risk — the interval during which money has moved but the legal event it was meant to fund has not yet been confirmed, or vice versa.

CURRENT SETTLEMENT MECHANISMS — WHAT EACH ACTUALLY SOLVES Mechanism

→ Main weakness Bank escrow

→ Conditionality is documented, but release is administered manually RTGS/NEFT at registration

→ Fast payment, but not inherently conditional on registry confirmation Cheque/demand draft

→ Physical proof, but slow clearing and real operational risk Developer collection account

→ Efficient for collections, not an independent buyer-protection structure Programmable e₹ (proposed)

→ Could lock and release value based on an authenticated registry event, not a manual instruction

None of this is uniform. Escrow use varies by state, by lender, by project type, and by whether a bank loan is involved at all. What is uniform is the underlying weakness: every one of today's mechanisms asks a human being, at some point, to manually confirm that a legal condition has been met before releasing money that has already left someone's account.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

PART TWO: MONEY THAT CAN WAIT

A normal bank transfer answers one question: has the payment been sent? A programmable digital rupee, by contrast, could be designed to answer a harder question: has the payment been sent only after conditions A, B and C actually occurred?

In simplified form, a conditional e₹ settlement instruction might read like this:

THE SETTLEMENT LOGIC IF buyer and seller KYC are valid AND lender disbursement is approved and locked AND stamp duty is confirmed paid AND the sub-registrar system generates a verified registration event AND applicable tax withholding is reserved THEN release the sale consideration to the seller's wallet ELSE keep the funds locked in the conditional settlement wallet

This is worth being precise about, because it is the single most misreported part of this story: the token is not the sale deed, and code is not the law. Sections 4 and 5 of the Information Technology Act, 2000 extend legal recognition to electronic records and digital signatures generally — but Section 1(4), read with the First Schedule of that same Act, explicitly excludes contracts for the sale, conveyance or transfer of interest in immovable property from that regime. A sale deed cannot be autonomously executed by a casual electronic signature or a smart contract; it still requires the wet-ink or state-managed biometric execution process mandated under the Registration Act, 1908 — a framework the central government has moved to replace with the Registration Bill, 2025, aimed at building formal legal foundations for online submission, e-certificates and paperless archiving, but which has not altered this core execution requirement.

What a programmable token can legally do is automate the payment consequence of a legally completed event. It cannot legally execute the event itself. The deed remains the legal instrument. The sub-registrar remains the statutory authority. The digital rupee becomes, at most, the settlement layer sitting underneath those institutions — not a replacement for them.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

PART THREE: THE SUB-REGISTRAR BECOMES AN ORACLE

The hardest problem in a programmable property settlement is not the token. It is the trigger.

In smart-contract language, an "oracle" is a trusted source that tells a code-based system whether a real-world event has actually happened. For a retail payment, that source can be a merchant's point-of-sale confirmation. For a property closing, it cannot be a broker's WhatsApp message or a lawyer's email — it has to be an authenticated, tamper-evident signal from the one institution with statutory authority over the event: the sub-registrar.

That single requirement opens a chain of questions India's registry infrastructure has not yet had to answer. What digital event, precisely, constitutes "registration complete" — the biometric capture, the document upload, or the final index-number generation, and what happens if the server fails between those steps? Which institution is authorised to feed that event, cryptographically signed, into a settlement system outside its own walls? If a registry accepts a document that later turns out to be defective — a forged power of attorney, a document accepted in error — who bears liability for a payment already released against it? And critically: can a release be paused if a court injunction or a competing claim surfaces in the window between execution and settlement, or does the architecture treat the registry event as final the instant it fires?

None of these are cryptography problems. They are governance problems, and Karnataka's own registration architecture is a useful illustration of exactly how far apart the pieces currently sit. The state's Kaveri Online Services 2.0 platform has digitised pre-registration data entry, appointment scheduling, e-stamping and access to digitally signed encumbrance certificates, cutting in-office turnaround to roughly fifty minutes and drawing in real time on the Bhoomi rural land-record base and the e-Swathu/e-Aasthi urban property-card system. It is, by any regional comparison, one of India's more advanced registry environments. But by the state's own description, the system remains "loosely coupled" — Kaveri verifies property particulars, but it does not currently broadcast an outward financial-execution signal that any external banking rail can act on. Registration is digital. Settlement is still a separate, privately conducted banking event that the registry has no visibility into.

The difficult part is not programming the rupee. The difficult part is deciding which institution has the authority to tell the rupee that truth has occurred — and building the legal and technical scaffolding to make that authority both auditable and reversible when it gets something wrong.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

PART FOUR: THE ₹1 CRORE CLOSING — A FORENSIC WALK-THROUGH

Consider a resale apartment in Bengaluru, priced at ₹1 crore. The buyer contributes ₹20 lakh in personal funds. A bank finances the remaining ₹80 lakh through a home loan.

Stage One — Buyer contribution. The buyer moves ₹20 lakh into a designated, purpose-bound settlement wallet. The funds are tagged to a specific property, seller and registration reference and cannot be freely withdrawn or redirected.

Stage Two — Lender contribution. The bank credits ₹80 lakh into the same controlled structure. This is where the architecture solves a real, currently expensive regulatory problem. RBI Master Directions require banks and housing finance companies to perform strict end-use monitoring of home-loan disbursals specifically to prevent fund diversion — loan tranches cannot simply land in a buyer's personal account, which is why banks today issue physical demand drafts or send a legal-panel representative to sit at the sub-registrar's office until a deed is signed, watching for the exact moment execution completes before authorising release. A purpose-bound e₹ tranche, cryptographically earmarked and mathematically incapable of reaching anywhere except the authenticated seller's wallet, replaces that manual physical vigilance with an automated compliance trail — a meaningfully cleaner control for a bank's risk desk than a representative standing in a crowded registration hall.

Stage Three — Conditions attach. The settlement rules encode buyer and seller PAN/KYC status, the property reference, stamp-duty payment confirmation, the lender's lien notation, and the applicable tax-withholding treatment. Since April 1, 2026, that withholding obligation runs under Section 393(1) of the Income Tax Act, 2025 — the restructured successor to the old Section 194-IA one-percent TDS rule — with the buyer remaining the responsible "transferee" and filing shifting to the new common Form No. 141 in place of the legacy Form 26QB. For this transaction, the settlement logic would need to reserve roughly ₹1 lakh for that withholding rather than releasing the full ₹99 lakh net figure to the seller in one undifferentiated payout.

Stage Four — Registration. At the sub-registrar's office, both parties complete the legally required biometric verification and deed execution. The registry validates stamp duty and generates the final registration event — the moment the registry becomes the oracle described above.

Stage Five — Conditional release. Once that authenticated event is received, the seller's portion becomes available, the TDS amount is ring-fenced and routed toward compliance, the lender receives a settlement confirmation completing its end-use monitoring obligation, and every party retains a time-stamped settlement receipt.

WHAT THIS COULD PLAUSIBLY REDUCE — AND WHAT IT CANNOT TOUCH Could reduce

→ Cannot solve Manual escrow-release delays

→ Defective historical title Payment-before-registration risk

→ Forged or wrongly accepted registry documents Beneficiary-account diversion

→ Court disputes over ownership Reconciliation burden across banks and registries

→ Municipal mutation and record-update delays Diversion of loan funds before closing

→ Legal liability arising from a bad sale

The gain, if this were ever built, is not merely speed. It is atomicity — the legal completion event and the financial settlement event occurring together, rather than being coordinated manually across two systems that currently cannot see each other.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

PART FIVE: WHY REAL ESTATE IS THE HARD CASE

It is worth stating plainly why this is a materially harder problem than programmable welfare payments or purpose-bound corporate disbursements, where India's e₹ pilots have made their genuine, verifiable progress.

Land governance sits with the states; monetary rails sit with the RBI. That single jurisdictional split means every one of India's roughly two dozen major state registration systems — Karnataka's Kaveri, Maharashtra's IGR, and the rest — would need its own secure, authenticated integration with a central banking rail, and none of them currently maintain that kind of outward-facing financial API. Land registries, municipal mutation records, cadastral maps and the judicial system that resolves title disputes all run on separate, largely non-interoperable technology stacks built at different times for different purposes.

There is also a genuinely new risk this architecture would introduce rather than solve: money that is correctly locked but cannot be released because one government system — the registry server, the mutation database, the biometric verification service — goes offline or drops the authentication signal at the wrong moment. Because e₹ mimics physical cash, RBI's own framework treats settlement as achieving instantaneous finality upon transfer; unlike commercial banking rails, there is no native mechanism for a routine administrative reversal if a smart contract misfires or acts on an erroneous signal. A dispute at that point does not get resolved by a phone call to a bank branch. It gets pushed to civil court — which, given that land and property disputes already account for roughly two-thirds of pending civil cases in Indian district courts and take an average of seven to twenty years to resolve, is not a forum anyone building this system should want to rely on as a routine backstop.

Retail wallet architecture compounds the problem at a more basic level. Current e₹ retail wallets operate under conservative transaction and balance ceilings designed for everyday cash-equivalent use, nowhere close to the scale a ₹1 crore — let alone a multi-crore — property settlement would require. Before any registry-linked closing becomes viable, the RBI and participating banks would need institutional-grade wallet structures, higher conditional-settlement thresholds, and clearer consumer-protection and dispute-resolution rules for exactly the kind of interrupted-transaction scenario described above. None of that currently exists in the public pilot architecture.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

PART SIX: THE FRAUD IT SOLVES, AND THE FRAUD IT CANNOT TOUCH

This is the distinction that determines whether this idea is written about honestly or hyped irresponsibly, and it deserves to be stated as its own structural boundary rather than a caveat buried in a paragraph.

Title fraud is an upstream data problem: forged sale deeds, impersonation of absent or deceased owners, fabricated powers of attorney, hidden legal heirs, defective historical chains of ownership. If a sub-registrar's office processes a forged document and the registry produces a valid-looking digital confirmation, a programmable rupee will do exactly what it was built to do — release the money efficiently, to the wrong person, with perfect precision. The currency cannot independently verify historical lineage. This is the oracle trap: the system is only ever as trustworthy as the event that triggers it.

Settlement fraud is a downstream execution problem: a buyer's RTGS cancelled after a seller has already signed, a cybercriminal altering beneficiary account details in the window before a transfer clears, funds released before a document is actually registered, or simply a reconciliation failure between two disconnected systems. This is precisely what a purpose-bound, registry-triggered token is designed to close — by making it structurally impossible for locked funds to move anywhere except an authenticated destination, and only after the registry confirms the condition that was supposed to unlock them.

The scale of the upstream problem is why this distinction matters so much. Independent parliamentary and policy research places land and property disputes at roughly two-thirds of all pending civil litigation in Indian district courts, with an average dispute lifecycle stretching close to two decades from a local revenue office to final resolution at the Supreme Court — where property matters already account for roughly a quarter of the docket. Research cited in parliamentary discussion has put the number of people directly affected by land conflicts at over 7.7 million, with more than $200 billion in infrastructure and real-estate investment tied up or threatened across upwards of 2.5 million hectares, and individual litigants losing an estimated ₹50,000 crore a year in aggregate legal costs, travel and lost wages. None of that number moves because a payment rail got faster. A programmable rupee sits entirely on the right side of that ledger — it is a closing-day tool, not a title-reform tool — and conflating the two is the single most common and most damaging error in how this technology tends to get reported.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

PART SEVEN: WHO ACTUALLY WINS

Buyers gain reduced counterparty and release-timing risk — the fear of releasing money before a deed is genuinely registered, or of a lender's tranche landing somewhere it shouldn't, is precisely what a correctly sequenced token removes. Sellers gain faster certainty that consideration has actually cleared before they sign away title, rather than trusting a demand draft or a promised transfer. Lenders gain a materially tighter and cheaper end-use monitoring mechanism than the current practice of physically stationing legal representatives at registration offices. State governments gain a cleaner, more auditable settlement trail alongside their existing stamp-duty and registration-fee reconciliation — which, in a market like Karnataka's, where the state collected roughly ₹25,493 crore in registration revenue across 22.88 lakh documents in FY 2025–26, is not a trivial administrative benefit. Developers gain the ability to tie construction-linked collection-account withdrawals to certified progress milestones rather than manual bank release, a genuine RERA-adjacent use case that may mature faster than the resale-transaction case.

The party most exposed is the purely custodial layer — a bank's back-office reconciliation function, or a standalone escrow agent whose entire value proposition is holding money until told to release it. That is close to exactly the job a correctly sequenced programmable token is built to do more cheaply and with a shorter exposure window. What does not disappear is judgment work. A property lawyer's actual value was never custody of funds — it was reading a title report, spotting a defect, advising a client not to proceed. If anything, a settlement architecture that removes manual reconciliation friction raises the relative importance of getting legal diligence right before a token ever locks, because there is less of a natural pause left in the process during which a careful advisor might otherwise catch a problem.

⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡⬡

THE TOOLKIT: WHAT TO WATCH FOR BEFORE TREATING ANY OF THIS AS LIVE

Separate a registry-digitisation headline from a payment-integration headline. A blockchain or DLT announcement out of a state registration department is almost certainly about tamper-resistant title records, not about e₹ settlement — check which half of the architecture is actually being piloted before assuming the other half is too.

Watch the RBI's CBDC and Asset Tokenisation Sandbox specifically, not aggregator coverage. That framework is the stated channel through which a registry-linked use case would first surface officially, and sandbox admission is not the same as production deployment.

Check the sequencing before trusting any pilot description. A design only solves the escrow problem if the registry confirmation triggers the payment release — not the reverse. A pilot where payment settles ahead of or independent of registration has recreated escrow's central risk under a different name.

Track wallet and transaction ceilings, not headline adoption numbers. Retail e₹ wallet and per-transaction limits that sit far below a typical property transaction value are a hard, practical constraint on when this becomes usable for anything beyond token or booking amounts — regardless of how mature the underlying technology gets.

Hold the line between title fraud and settlement fraud in every claim you read. "Eliminates fraud" is not a claim current evidence supports. "Reduces closing-day payment diversion and reconciliation risk" is defensible, and it is where the real efficiency case actually lives.

There is no confirmed RBI or state-government programme currently operating a programmable e₹ property settlement inside a sub-registrar office. That is not a caveat to bury at the end of this piece — it is the honest state of the story. What exists instead are two halves of an architecture being built in separate rooms by separate arms of the state: a payment rail that is real, legal tender, and expanding — and a registry-digitisation effort that is real, genuinely advanced in states like Karnataka, and still structurally unable to talk to a bank. The interesting question is not whether these two halves are already joined. They are not. The interesting question is which state signs the paperwork to put them in the same room first — and whether it starts, sensibly, with the cleanest transaction it can find: a single buyer, a single seller, a bank-financed resale flat, standardised documentation, and no active litigation anywhere in the chain.