Part IV — Digital Money

Chapter 24 · India's e-Rupee in Practice

Chapter 23 ended with a question worth a chapter of its own: a currency being adopted, or a pilot being administered? Four years of pilot evidence now exist — circulation, users, transactions, feature rollouts — and this chapter reads all of it, including the number nobody expected: circulation fell 24 per cent in FY26. The verdict, honestly reached, is more interesting than either the hype or the obituary.

📖 Chapter 24 of 35 ⏱ ~60 min read 🎓 Session 30 · CILO 2 Core 70% Stable 15% Volatile 15% 📝 Quiz · 30 MCQs

🎯 Learning Objectives

1. What the Pilot Is

One paragraph of recap, then the data. The wholesale pilot (e₹-W) began 1 November 2022; the retail pilot (e₹-R) on 1 December 2022. The design is Chapter 23's two-tier token model: the RBI issues, nineteen banks and a set of non-bank providers (CRED, MobiKwik among them) distribute wallets; wallets link to a savings account for KYC-free onboarding; a DBT beneficiary can onboard with just a mobile or Aadhaar number. The e-rupee pays no interest, charges no fees, comes in note denominations, scans both CBDC and UPI QR codes — at a UPI QR it settles on UPI timelines — and is legal tender at par. Everything else in this chapter is what happened when that design met the public.

One analytic tool before the numbers, because everything depends on it: stock versus flow. Circulation is a stock — how many e-rupees sit in wallets on one day. Transactions are a flow — how often e-rupees move in a period. A currency can have falling stock and healthy flow (money arrives, is spent instantly, leaves nothing in the wallet — exactly what programmable welfare money does), or rising stock and dead flow (incentives loaded wallets, nobody spends). Confusing the two is how both hype and obituaries get written. Hold the distinction through Section 2.

⚠ Volatile content — Reviewed: November 2026 · Next review: February 2027

2. The Numbers

The evidence, as of the RBI's Annual Report 2025-26 (May 2026), Deputy Governor statements through April 2026, and the FAQ of October 2026.

MetricValueAs of / period
Retail e₹ in circulation₹5.70 crore31 March 2023
Retail e₹ in circulation₹234.04 crore31 March 2024
Retail e₹ in circulation₹1,016.46 crore31 March 2025
Retail e₹ in circulation₹771.66 crore (−24%)31 March 2026
Retail users~1 crore (17→19 banks; non-banks added)April 2026
Cumulative retail transactions~15 crore, worth ~₹34,000 croreSince launch, to April 2026
Wholesale e₹ end-of-day balanceNil — auto-redemption intradayMarch 2026
UPI (comparator)~49 crore users; ~2,400 crore transactions in FY26Oct 2025 / FY 2025-26

Three readings before anyone spins them. First, the growth curve was real: 41× in FY24, 4.3× in FY25 — then the first annual fall, −24 per cent, in FY26. Second, the wholesale nil balance is not failure but design: institutions hold e₹-W only for the life of a trade and auto-redeem at day's end, so the stock is zero by construction while the settlement flow is invisible in this table. Third, the comparator line dwarfs everything above it: UPI moved roughly 2,400 crore transactions in FY26 alone; the e-rupee managed 15 crore in three years and four months. Both numbers are honest. Both need the next section.

📝
Note on sources: circulation figures are from RBI Annual Reports 2022-23 through 2025-26; the user and cumulative-transaction counts are from Deputy Governor T. Rabi Sankar's public statements (December 2025, April 2026). Pilot metrics are inherently provisional — treat every number here as a snapshot under review.

3. The Adoption Problem

Put yourself in the wallet of an ordinary user. UPI is already installed, already linked to the account your salary lands in, free, instant, accepted at every QR in the country. The e-rupee app asks you to do what? Load a separate wallet from your bank account — the money leaves your deposit (and the deposit's interest) to sit as tokens paying nothing — so that you can then pay at the same QRs you already pay at, with settlement that only differs (instant, wallet-to-wallet) at CBDC QRs, which almost no merchant displays. For that user, the e-rupee offers no advantage she can feel, one visible cost (lost interest, extra app), and a privacy promise (cash-like anonymity) that managed-anonymity has not yet delivered in any published threshold. Chapter 23 predicted the disuse failure mode; this is it in the wild.

The user counts confirm the diagnosis. One crore e-rupee users after three-plus years against roughly forty-nine crore UPI users. And the transaction counts carry a sharper stat the hands-on will make you compute: fifteen crore cumulative transactions across one crore users over roughly forty months is under one transaction per user per month — UPI's habitual users average more than that per day. Add the pilot's own history: the December 2023 push to one million transactions a day was an administered target, met through bank drives and incentives, not a demand curve. Usage built to hit a target is not usage that stays when the target leaves — and the distributor conflict Chapter 23 quoted (banks restraining a product that drains their deposits) never went away. None of this makes the pilot a failure. It makes the retail e-rupee, as a mass payment product, un-adopted — which forced the question the RBI then answered.

🌎
Real World: The RBI has itself flagged the go-slow: officials describe tracking global CBDC launches before any full rollout, and the August 2026 policy conference answer to "when does the pilot end?" was, in substance, that a pilot is exactly what it still is — technology, features and adoption all under test.

4. The Pivot to Purpose

Re-read the FY26 fall with Section 4's new eyes. In the same year circulation dropped 24 per cent, the RBI ran its most aggressive feature expansion: programmable direct-benefit transfers became the flagship — Odisha's Subhadra paying 88,000 women, PDS food subsidies in Gujarat, Puducherry and Chandigarh redeemable only at fair-price shops, tenant-farmer credit fenced to input stores. Programmable welfare money is designed to be spent — it arrives, does its job, exits. A falling stock with a working flow is not a bug in that design; it is the design. The RBI's revealed strategy, after three years of finding no organic retail demand, is use over holding: be the infrastructure for specific jobs rather than the wallet everybody keeps.

The rest of the pivot is a portfolio. Offline: NFC- and SMS-based solutions in test, the one capability no UPI-linked product offers, aimed precisely at the connectivity gaps UPI cannot serve. Wholesale tokenisation: the Unified Markets Interface went live, tokenised certificates of deposit settled in e₹-W, and corporate-bond tokenisation settlement followed in September 2026 — the quiet track where pilot success is measured in basis points. Cross-border: an MoU with the Monetary Authority of Singapore, talks with the UAE central bank, and seats in BIS Innovation Hub projects — the subject Chapter 25 takes up fully. Ecosystem: a retail sandbox since October 2025 so fintechs build the use cases the RBI will not. Judge the portfolio as a portfolio: the retail mass-adoption bet has not paid; the purpose-bound bets are paying in the only currency pilots trade in — working evidence.

5. Reading the Evidence Honestly

The session's method, and the standard this book applies to every pilot claim it will ever print. Four tests:

  1. Stock or flow? Name which one a number measures before interpreting it. (Circulation: stock. Transactions: flow. A fall in one is not a fall in the other.)
  2. Adoption or administration? Ask what would have happened without the push. Bank-drive transaction days, incentive-loaded users and DBT-credited beneficiaries are administrative demand. A user who loads her own wallet in month nine is adoption.
  3. Against what counterfactual? UPI is the counterfactual for retail payments; RTGS plus collateral arrangements for wholesale. The e-rupee's honest comparisons are the jobs those rails do badly — offline, purpose-bound, no-counterparty settlement — not the jobs they do brilliantly.
  4. What would change the verdict? State the evidence that would flip your reading, in advance, so the reading is falsifiable. (For adoption: sustained voluntary load-and-spend by non-DBT users; merchant CBDC-QR pull; offline usage growing where UPI cannot reach.)

Applied here: the steelman for success is real — the pilots have tested exactly what pilots exist to test (technology, architecture, accounting, use cases), the welfare applications work at state scale, and the wholesale track is quietly succeeding on its own metric. The steelman for failure is equally real — no organic retail demand after three and a half years, one transaction per user per month, and a distribution channel whose incentives point the other way. The honest verdict is not "success" or "failure" but a sentence this chapter has earned: the e-rupee has found users for what it can do that UPI cannot, and has not found users for anything else. Whether that sentence describes a niche or a foundation is what the FY27 pilots will decide.

📋 Stable content — Reviewed: November 2026 · Next review: May 2027

6. Where This Goes

The declared FY27 agenda, from the Annual Report 2025-26: widen the DBT pilots across states and schemes; open a CBDC and Asset Tokenisation sandbox; run bilateral and multilateral cross-border pilots; keep testing offline. Around it, the posture is explicit patience — track the global launches, watch what the digital euro's law-first path and other jurisdictions' rollouts teach, and keep the pilot a pilot. Two things would force the pace: a working privacy threshold that gives the retail token its cash-like edge, or a cross-border corridor that only CBDC settlement can serve cheaply. The first is an internal decision the RBI has not taken; the second is Chapter 25's subject.

🛠 Hands-On: Pilot Evidence Review

You are given the raw pilot evidence. Review it the way an analyst briefs a board: compute, classify, verdict.

  • Circulation (31 Mar): ₹5.70 cr (2023) · ₹234.04 cr (2024) · ₹1,016.46 cr (2025) · ₹771.66 cr (2026)
  • Users: ~1 crore (Apr 2026); UPI: ~49 crore users (Oct 2025)
  • Cumulative retail transactions: ~15 crore worth ~₹34,000 crore, Dec 2022–Apr 2026 (~40 months)
  • UPI FY26: ~2,400 crore transactions; habitual users average multiple transactions daily
  • FY26 events: DBT pilots expanded (Subhadra 88,000 women; PDS in 3 states/UTs); UMI live; tokenised CDs; circulation −24%

Steps

  1. Compute the year-on-year circulation growth rates (FY24, FY25, FY26), and the per-user per-month transaction frequency for the e-rupee against UPI's rough per-user monthly frequency (use FY26 volume ÷ 49 crore users).
  2. Give the stock-flow reading of the FY26 circulation fall: list every mechanism that shrinks the stock while the pilot still "works".
  3. Classify each evidence line above as an adoption signal or an administrative signal, with one sentence of justification each.
  4. Write the verdict paragraph (150 words): pilot succeeding, failing, or the third thing — and name, in advance, two pieces of evidence that would change your verdict.
View Solution / Walkthrough

1. Arithmetic. Circulation: FY24 = 234.04/5.70 ≈ 41×; FY25 = 1,016.46/234.04 ≈ +334% (4.3×); FY26 = 771.66/1,016.46 ≈ −24.1%. Frequency: e-rupee 15 crore transactions ÷ 1 crore users ÷ 40 months ≈ 0.375 transactions per user per month — roughly one transaction every two to three months. UPI: 2,400 crore ÷ 49 crore users ÷ 12 months ≈ 4 per user per month on the crude average — but the honest caveat is that averages hide concentration: a large majority of UPI users transact weekly or daily, so the gap between habitual usage is far wider than 10×; and the e-rupee count mixes cumulative volume with a point-in-time user base, so 0.375 is an upper-bound flavour, not a precision figure. State the caveat; it does not rescue the number.

2. Stock-shrink mechanisms. (a) Programmable DBT money spent fast and redeemed out — welfare flow leaves no stock. (b) Dec 2023-style target-driven loads unwinding — incentive balances redeemed once drives end. (c) No-interest design: any user who compares keeps balances in deposits, loading only to spend. (d) The base effect: FY25's ₹1,016 crore was itself partly administrative, so FY26 normalises it. All four shrink circulation without the settlement flow failing.

3. Classification. Circulation series: mixed signal — administrative growth, then normalisation (neither proves adoption). One crore users: administrative — bank-onboarded, DBT-credited, pilot-gated. 15 crore cumulative transactions: administrative-heavy — includes target drives and DBT flows. UPI comparator: the counterfactual, not a signal. FY26 DBT expansion: adoption signal for the purpose-bound design — states chose the instrument for a real problem. UMI/tokenised CDs: adoption signal on the wholesale track — institutions integrated voluntarily for settlement efficiency. Circulation fall: neither — it is a stock artefact of a flow-first strategy.

4. Model verdict. "The e-rupee pilot is succeeding at being a pilot and failing at being a habit. The technology works, the wholesale track adds settlement efficiency institutions chose, and purpose-bound welfare delivery now runs at state scale. But after 3.5 years there is no evidence of organic retail demand: one transaction per user per quarter, one crore onboarded against UPI's forty-nine crore, and a distributor chain whose incentives oppose the product. The strategy has honestly pivoted from 'digital cash for everyone' to 'infrastructure for jobs nothing else does'. The verdict flips toward success on two pieces of evidence: (1) sustained voluntary load-and-spend by non-DBT users over two consecutive quarters, and (2) published offline adoption data showing usage in connectivity-dark districts where UPI cannot reach. It flips toward failure if FY27 DBT pilots stall or the wholesale track reverts to legacy settlement."

Key Takeaways

1

Stock is not flow: circulation measures one day's wallet balances, transactions measure movement — the FY26 fall is a stock fact about a flow-first strategy.

2

The adoption problem is structural: a separate no-interest wallet offers nothing UPI users can feel, and the banks distributing it profit from its failure.

3

One transaction per user per month against UPI's daily habit is the number that separates administration from adoption.

4

The pivot is real: programmable DBT, offline, wholesale tokenisation and cross-border corridors — use over holding, infrastructure over wallet.

5

The honest verdict: the e-rupee has found users for what it can do that UPI cannot, and none for anything else. FY27 decides whether that is a niche or a foundation.

Test Your Understanding

Choose one best answer per question. The evidence table in Section 2 and the hands-on data are fair game for any item. Aim for 24/30 before moving to Chapter 25.

1. The e-rupee pilots began:

2. Circulation versus transactions, in stock-flow terms:

3. The retail circulation series, in order:

4. The number nobody expected, in FY26:

5. The wholesale e₹ end-of-day balance is nil because:

6. The user-count gap:

7. Fifteen crore cumulative transactions over ~40 months across ~1 crore users works out to:

8. For an ordinary UPI user, the e-rupee wallet's offer is:

9. Instant wallet-to-wallet settlement applies:

10. The December 2023 one-million-transactions-a-day push was:

11. The distribution-channel problem that persists:

12. The RBI's revealed strategy after finding no organic retail demand:

13. The stock-flow reading of the FY26 fall alongside the DBT expansion:

14. The flagship DBT pilots, correctly matched:

15. The one capability the pivot offers that no UPI-linked product can:

16. On the wholesale track, the Unified Markets Interface enabled:

17. The cross-border partnerships in the pilot portfolio include:

18. The retail sandbox, opened October 2025, exists so that:

19. The honest counterfactual discipline says the e-rupee should be compared with:

20. The adoption-versus-administration test asks:

21. Evidence that would flip the verdict toward adoption:

22. The chapter's one-sentence honest verdict:

23. The two things that would force the pace toward full launch:

24. The declared FY27 agenda includes all EXCEPT:

25. The RBI's declared posture on full rollout:

26. Onboarding under the pilot design:

27. Wallet providers as of late 2026:

28. September 2026's wholesale milestone:

29. Stating your flippable evidence in advance is what makes a verdict:

30. Chapter 25 takes up: