## ftnea2025004

## Source details

**Canonical URL:** [ftnea2025004](https://www.imf.org/-/media/files/publications/ftn063/2025/english/ftnea2025004.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/ftn063/2025/english/ftnea2025004.pdf.md)
- [Structured JSON version](/-/media/files/publications/ftn063/2025/english/ftnea2025004.pdf.json)

---

### Purpose and potential benefits
- Policymakers aim to increase adoption and usage of retail digital payment systems to:
  - Reduce transaction costs and information asymmetries (Dubey and Purnanandam 2023).
  - Expand access to credit by creating electronic records of payments (Ouyang 2021; Berg, Fuster, and Puri 2022; Alok and others 2024).
  - Permit new means of debt enforcement that unlock further lending opportunities (Rishabh and Schäublin 2021; Brunnermeier and Payne 2023).
  - Encourage formalization (Klapper, Miller, and Hess 2019) and lower crime by reducing cash carriage (Wright and others 2017).

### Interoperability versus closed-loop systems
- Definitions and implications:
  - Closed-loop payment system: requires payer and payee to use the same payment provider; network effects can onboard users but may lead to a small number of dominant providers.
  - Interoperability: technical, semantic, and business compatibility that allows end users to transact seamlessly across systems (Boar and others 2021); network effects operate primarily at the platform level, reducing concentration on individual providers and preserving user choice.
- Examples of interoperable retail fast payment systems: Unified Payments Interface (UPI) in India, Pix in Brazil, TIPS in the euro area, CoDi in Mexico, and FedNow in the United States.

### Context: India’s Unified Payments Interface (UPI)
- Structure and operation:
  - UPI is an instant payments platform built over Immediate Payment Service (IMPS), India’s real-time interbank electronic fund transfer service.
  - Banks and third-party application providers (TPAPs) that participate can exchange payment messages through UPI apps.
  - End users execute free push (pay) and pull (receive) transactions identified by a unique virtual ID.
  - Fintech firms can provide user-facing apps while funds remain in users’ bank accounts.
  - NPCI runs UPI and ensures payer’s and payee’s banks debit and credit accordingly.
  - Up to four participant organizations can be involved in a single transaction: the payer’s provider, the payee’s provider, the payer’s bank, and the payee’s bank.
- Settlement:
  - Settlement for end users is immediate.
  - Settlement among financial institutions is managed through deferred net settlement with ten daily cycles.
- Scale and timing:
  - Since its launch in 2016, UPI has grown quickly.
  - UPI now processes more than 18 billion transactions per month (UPI volume as of March 2025).
  - UPI reached more than 600 banks and more than 200 apps.
  - India now makes faster payments than any other country (Appendix Figure 1.1; ACI Worldwide (2023)).

### Mechanisms: how interoperability improves user experience and adoption
- Two main channels:
  1. Direct freedom of choice: users can pick their preferred app (brand trust, reliability), and switch to better apps when available.
  2. Indirect competitive pressure: interoperability increases incentives for providers to improve quality to retain users and lowers barriers for innovative new providers to enter.
- Expected user-level improvements:
  - Higher quality apps and more varied app choices.
  - Increased likelihood of using the digital payment system rather than cash.
- Conceptual framework highlights:
  - Variety case: interoperability prevents coordination failure where users must pick the same app to transact.
  - Quality case: interoperability allows users to adopt higher-quality apps without losing access to counterparties.

### Empirical evidence and natural experiments
- Data sources and coverage:
  - Granular data covering all UPI transactions, combined with data covering all transactions facilitated by a major fintech firm that initially operated a closed-loop wallet.
- First natural experiment (post-demonetization adoption choices):
  - Demonization announced on November 8, 2016.
  - After demonetization, many users tried digital payment methods; sampled users increasingly chose interoperable UPI over the closed-loop alternative.
  - Transactions that require interoperability (sender and recipient using different apps) constituted a substantial part of this growth.
  - By March 2017, usage of the non-interoperable closed-loop incumbent plateaued while UPI adoption continued to grow through 2017.
  - UPI grew roughly three-fold from March to October 2017 while the non-interoperable alternative was flat.
- Second natural experiment (integration of closed-loop provider into UPI):
  - A major incumbent closed-loop provider joined UPI after a Reserve Bank of India directive mandating interoperability.
  - Over the first year after integration, total peer-to-merchant (P2M) transaction value per person increased by 8 Rupees per month more in an average district with an above-median incumbent presence than in an average district with a below-median presence.
  - That difference is persistent and substantial, equal to more than 80% of average P2M digital payments per person in above-median districts in the month prior to integration.
  - Footnote evidence: “This difference amounts to 118% of average monthly P2M digital payments per person in below-median districts in the year after interoperability.”
  - The increase came from all three transaction components: both-counterparty-incumbent, one-counterparty-incumbent, and neither-counterparty-incumbent, benefiting both the incumbent and apps already on UPI.
  - Measuring digital payments relative to cash (using ATM withdrawals as a proxy for cash usage) shows a similar substantial and persistent rise in digital payments after integration in districts with larger increases in de facto interoperability.
- Additional granular findings:
  - Many users first joined through trusted and familiar brands, then switched to more innovative and reliable apps once able.
  - Incumbent providers were incentivized to upgrade services to avoid user loss.
  - Innovative app providers could enter and attract users despite initial incumbent dominance.
  - Early adoption: more than half of users accessed UPI through their existing bank’s app during the early launch period.
  - App-bank affinity: over the whole sample period, for a given app-bank pair, transaction values are ten times higher if both are from the same entity.
  - Convergence to higher-quality apps: share of users choosing their bank’s own app fell as platform matured; users migrated to top three fintech apps.
  - Decline-rate evidence: if an app has a one percentage point lower decline rate in a given quarter, then the total value of transactions it processes in the next quarter is 4 percent higher (controlling for app-wise differences and national trends).

### UPI ecosystem growth, enablers, and innovations
- Ecosystem and entrants:
  - UPI reached more than 600 banks and more than 200 apps.
  - August 2017: 45 percent of payers used one app (“App 1”), which accounted for 29 percent of overall transaction value.
  - Multiple new UPI apps entered the market and two later achieved at least a 20 percent share of overall payer-side transaction value.
  - September 2017: major new provider (“App 2”) entered and built substantial market share during 2018 by servicing primarily customers while merchants continued to use incumbent apps; App 2 began establishing a substantial merchant-side presence from mid-2019 and by 2023 it had gained a substantial share of the merchant market.
- Technological and public enablers:
  - Pradhan Mantri Jan Dhan Yojana opened hundreds of millions of new bank accounts.
  - Aadhaar biometric ID provided a unique verifiable digital identity to speed transaction authentication and Know Your Customer checks.
  - The cost of mobile data fell by roughly 96 percent during the mid-2010s.
  - Technological innovations introduced by app providers include quick-response (QR) codes and “soundboxes.”
  - Pilot projects in 2020 and 2021 led to new offline payment functionalities for small-value payments in limited connectivity areas.

### Key quantitative highlights
- UPI volume as of March 2025: more than 18 billion transactions per month.
- UPI ecosystem size: more than 600 banks; more than 200 apps.
- Mobile data cost decline: roughly 96 percent during the mid-2010s.
- Early concentration: August 2017 — 45 percent of payers on one app; that app accounted for 29 percent of transaction value.
- Entry timing and market evolution:
  - App 2 entered in September 2017, expanded customer share in 2018, merchant expansion from mid-2019, substantial merchant share by 2023.
- Quality impact: a one percentage point lower app decline rate → 4 percent higher subsequent transaction value.
- App-bank affinity: transaction values are ten times higher when app and bank belong to the same entity (over the sample period).
- Integration shock effect: 8 Rupees per month increase in P2M transaction value per person in above-median districts relative to below-median districts in the first year after integration; this difference equals more than 80% of average P2M digital payments per person in above-median districts in the month prior to integration.

### Risks to long-term choice and competition
- Potential erosions of interoperability benefits if providers become dominant:
  - Proprietary QR codes initially created steering effects by being readable only by a provider’s own app; RBI prohibited non-interoperable QR codes to preserve openness (Reserve Bank of India, 2020).
  - Even with interoperable QR codes, nonstandardized branding can create impressions that codes are app-specific, generating network effects.
  - Economics of app provision (zero-fee UPI core with complementary revenue from lending, insurance) can create market power if secondary markets exhibit scale or scope advantages; bundling may reduce switching even if better UPI services exist elsewhere.
  - Market concentration has increased: several aggregate metrics show rising concentration among UPI ecosystem participants; in half of UPI transactions, the payer and payee use the same app (stable share).
  - Regional specialization of the two largest apps’ user bases suggests providers may steer users toward co-adoption, constraining freedom to choose.

### Policy implications and recommendations
- Early-stage acceleration:
  - Providing a public app can kick-start network effects (BHIM case: BHIM initially accounted for more than half of payer-side total transaction value prior to major fintech firm take-off).
  - Encourage existing payments platforms to connect to the interoperable system to integrate trusted financial institutions (banks, incumbents) and bridge legacy users.
  - Ensure complementary enabling investments are in place (e.g., reduction in mobile data costs, expansion of bank accounts and digital identity).
- Ongoing regulatory vigilance:
  - Regulators should monitor app dominance and be prepared to act to maintain a level playing field among apps.
  - Use granular usage data and qualitative user research to identify mechanisms generating market power for dominant apps and tailor interventions to those mechanisms.
  - System operators should consult with private sector participants to ensure design choices support ecosystem health.
- Design and operational safeguards:
  - Preserve fully open, interoperable, and competitive system design choices (e.g., prohibitions on non-interoperable QR codes).
  - Watch for bundling and secondary-market advantages that could entrench market power and constrain switching.

### Appendix Figure note
- Appendix Figure 1.1. Volume of Fast Payment Transactions (Millions) — chart not reproduced here.
- Fast payments defined as: real-time or near real-time transfers of funds between accounts of end users as close to a 24/7 basis as possible (Frost and others 2024).
- US data comprise Zelle from 2017 and RTP from 2020.
- Data sources include: BIS; Statista; TCH; IMF staff analysis.

*Fintech Notes: Growing Retail Digital Payments — Introduction (excerpt); IMF staff; ftnea2025004.*

### Introduction ...........................................................................................................

### Introduction

### Purpose and potential benefits
- Policymakers aim to increase adoption and usage of retail digital payment systems to:
  - Reduce transaction costs and information asymmetries (Dubey and Purnanandam 2023).
  - Expand access to credit by creating electronic records of payments (Ouyang 2021; Berg, Fuster, and Puri 2022; Alok and others 2024).
  - Permit new means of debt enforcement that unlock further lending opportunities (Rishabh and Schäublin 2021; Brunnermeier and Payne 2023).
  - Encourage formalization (Klapper, Miller, and Hess 2019) and lower crime by reducing cash carriage (Wright and others 2017).

### Interoperability versus closed-loop systems
- Closed-loop payment system: requires payer and payee to use the same payment provider; network effects can onboard users but may lead to a small number of dominant providers.
- Interoperability: technical, semantic, and business compatibility that allows end users to transact seamlessly across systems (Boar and others 2021); in an interoperable system, network effects operate primarily at the platform level, reducing concentration on individual providers and preserving user choice.
- Examples of interoperable retail fast payment systems: Unified Payments Interface (UPI) in India, Pix in Brazil, TIPS in the euro area, CoDi in Mexico, and FedNow in the United States.

### Context: India’s Unified Payments Interface (UPI)
- UPI is an instant payments platform built over Immediate Payment Service (IMPS), India’s real-time interbank electronic fund transfer service.
- Banks and third-party application providers (TPAPs) that participate can exchange payment messages through UPI apps.
- UPI design:
  - End users execute free push (pay) and pull (receive) transactions identified by a unique virtual ID.
  - Fintech firms can provide user-facing apps while funds remain in users’ bank accounts.
  - NPCI runs UPI and ensures payer’s and payee’s banks debit and credit accordingly.
  - Up to four participant organizations can be involved in a single transaction: the payer’s provider, the payee’s provider, the payer’s bank, and the payee’s bank.
- Settlement:
  - Settlement for end users is immediate.
  - Settlement among financial institutions is managed through deferred net settlement with ten daily cycles.
- Scale and timing:
  - Since its launch in 2016, UPI has grown quickly.
  - UPI now processes more than 18 billion transactions per month (UPI volume as of March 2025).
  - India now makes faster payments than any other country (Appendix Figure 1.1; ACI Worldwide (2023)).

### Mechanisms: how interoperability improves users’ experience and adoption
- Two main channels:
  1. Direct freedom of choice: users can pick their preferred app (brand trust, reliability), and switch to better apps when available.
  2. Indirect competitive pressure: interoperability increases incentives for providers to improve quality to retain users and lowers barriers for innovative new providers to enter.
- Expected user-level improvements:
  - Higher quality apps and more varied app choices.
  - Increased likelihood of using the digital payment system rather than cash.

### Empirical evidence from UPI and natural experiments
- Evidence sources: granular data covering all UPI transactions, combined with data covering all transactions facilitated by a major fintech firm that initially operated a closed-loop wallet.
- First natural experiment (post-demonetization adoption choices):
  - After demonetization, many users tried digital payment methods; sampled users increasingly chose interoperable UPI over the closed-loop alternative.
  - Transactions that require interoperability (sender and recipient using different apps) constituted a substantial part of this growth.
  - Relevant date: demonetization announced on November 8, 2016.
- Second natural experiment (integration of closed-loop provider into UPI):
  - The fintech provider later joined UPI, expanding possible transactions for both its users and UPI users.
  - Regions with higher pre-integration presence of the closed-loop provider experienced larger increases in de facto interoperability and saw significant increases in adoption of digital payments, both in absolute terms and relative to cash.
- Additional empirical findings (summary of granular results reported in the Note):
  - Many users first joined through trusted and familiar brands, then switched to more innovative and reliable apps once able.
  - Incumbent providers were incentivized to upgrade services to avoid user loss.
  - Innovative app providers could enter and attract users despite initial incumbent dominance.

### Challenges in identification and interpretation
- Isolating the role of interoperability versus closed-loop systems is difficult because India might have seen rapid growth of retail digital payments even without interoperability given the lack of convenient alternatives to cash.
- The Note uses complementary natural experiments and rich transaction-level data to disentangle these effects.

### Conclusions and policy implications
- Interoperability appears to be a promising policy tool to transition away from cash by:
  - Expanding users’ freedom to choose apps.
  - Increasing competition among providers and encouraging quality improvements and innovation.
  - Increasing the attractiveness of digital payments relative to cash and pulling more users into the system.
- Risks and recommended vigilance:
  - As overall adoption grows and more providers join, policymakers should remain vigilant to the emergence of market power.
  - Authorities should be prepared to act to maintain a fully open, interoperable, and competitive system.

*Fintech Notes: Growing Retail Digital Payments — Introduction (excerpt).*

### 1. Value (Percent of GDP) 2. Volume (Transactions per Capita)

### ftnea2025004 - 1. Value (Percent of GDP) 2. Volume (Transactions per Capita)

### Overview and Data Definitions
- Sources: RBI; Haver Analytics; BIS; WDI; and IMF staff analysis.
- The graph shows the value and volume of UPI and other electronic retail payment methods in India.
- Prepaid payment instruments include smart cards and mobile wallets that are preloaded with value using cash, card, or other methods.
- UPI includes both peer-to-merchant and peer-to-peer transactions as many retail payments to small merchants are recorded as peer-to-peer.
- UPI = Unified Payments Interface.

### Interoperability versus Closed-Loop Systems
- Interoperable systems (example: UPI) allow end users to transact across systems without both counterparties using the same provider.
- Closed-loop systems require both payer and payee to use the same provider; funds are loaded into provider-hosted wallets and transfers occur within that network.
- Closed-loop systems generate network effects because the more users a provider has, the more attractive it is to new users; this network effect is not present for app providers under UPI.

### UPI Ecosystem Growth
- UPI reached more than 600 banks and more than 200 apps.
- Many users initially joined through bank-provided apps and subsequently migrated to fintech apps offering added features (automatic bill payments, rewards and credits, investment or insurance products).
- Technological innovations introduced by app providers include quick-response (QR) codes and “soundboxes.”

### Enablers of Rapid Adoption
- Pradhan Mantri Jan Dhan Yojana opened hundreds of millions of new bank accounts.
- Aadhaar biometric ID provided a unique verifiable digital identity to speed transaction authentication and Know Your Customer checks.
- The cost of mobile data fell by roughly 96 percent during the mid-2010s.
- Pilot projects in 2020 and 2021 led to new offline payment functionalities for small-value payments in limited connectivity areas.

### Conceptual Framework: How Interoperability Boosts Adoption
- Interoperability improves users’ experience directly and indirectly:
  - Direct benefits: freedom to choose preferred app (variety) and ability to select higher-quality apps (quality).
  - Indirect benefits: lowers entry costs for new app providers and incentivizes incumbents to upgrade app quality.
- Variety case: interoperability prevents coordination failure where users must pick the same app to transact.
- Quality case: interoperability allows users to adopt higher-quality apps without losing access to counterparties.

### Evidence: Direct Benefits for Users
- Early adoption: more than half of users accessed UPI through their existing bank’s app during the early launch period.
- App-bank affinity: over the whole sample period, for a given app-bank pair, transaction values are ten times higher if both are from the same entity (Appendix Table 1.1).
- Convergence to higher-quality apps: share of users choosing their bank’s own app fell as platform matured; users migrated to top three fintech apps.
- Decline-rate evidence: if an app has a one percentage point lower decline rate in a given quarter, then the total value of transactions it processes in the next quarter is 4 percent higher (controlling for app-wise differences and national trends).

### Evidence: Indirect Benefits through App Providers
- Concentration and competition:
  - August 2017: 45 percent of payers used one app (“App 1”), which accounted for 29 percent of overall transaction value.
  - Despite this concentration, multiple new UPI apps entered the market and two later achieved at least a 20 percent share of overall payer-side transaction value.
- Facilitation of entry:
  - September 2017: major new provider (“App 2”) entered and built substantial market share during 2018 by servicing primarily customers while merchants continued to use incumbent apps.
  - Only from mid-2019 did App 2 begin establishing a substantial merchant-side presence; by 2023 it had gained a substantial share of the merchant market.
- Interoperability enabled new providers to target specific subgroups (e.g., customers first), lowering the user threshold needed for viability and increasing the variety and quality of apps available.

### Key Findings and Quantitative Highlights
- UPI ecosystem: more than 600 banks; more than 200 apps.
- Mobile data cost decline: roughly 96 percent during the mid-2010s.
- Early concentration: August 2017 — 45 percent of payers on one app; that app accounted for 29 percent of transaction value.
- Entry timing: App 2 entered in September 2017, expanded customer share in 2018, merchant expansion from mid-2019, substantial merchant share by 2023.
- Quality impact: a one percentage point lower app decline rate → 4 percent higher subsequent transaction value.
- Transaction values are ten times higher when app and bank belong to the same entity (over the sample period).

### Policy-Relevant Implications (inferred from analysis and evidence)
- Interoperability can increase digital-payment adoption by:
  - Allowing users to choose apps that match personal preferences and higher-quality providers.
  - Reducing app-level network effects that deter entry, thereby fostering competition and innovation among app providers.
  - Creating incentives for incumbent providers to improve quality to retain users.
- Public investments in complementary digital public infrastructure (financial inclusion programs, digital identity, affordable mobile data) facilitate rapid scale-up of interoperable payment systems.
- Offline payment functionalities increase accessibility for small-value transactions in areas with limited connectivity.

*Sources: RBI; Haver Analytics; BIS; WDI; and IMF staff analysis.*

### 1. August 2017

### 1. August 2017

### Interoperability and provider quality
- Interoperability incentivized existing providers to upgrade service quality, evidenced by users moving to higher-quality apps as measured by transaction decline rates.
- The largest three apps, which now facilitate more than 95 percent of payer-side transactions, exhibited especially large improvements in decline rates even as their total transaction volumes grew exponentially.
- Distributional evidence (Figure 10) shows a downward trend in decline rates across apps, with percentile bands highlighting improvements across the cross-app distribution.

### Public provision of payments apps (BHIM case)
- The National Payments Corporation of India launched the Bharat Interface for Money (BHIM) app in late 2016 when total UPI usage was small and there were fewer other providers.
- BHIM initially accounted for more than half of payer-side total transaction value prior to the take-off of major fintech firm apps, highlighting a potential catalytic role for direct public provision in overcoming coordination failures (low user adoption ↔ lack of high-quality apps).
- Box Figure 2.1 reports BHIM district-wise shares of UPI transaction value over time, with payer-side and payee-side median shares and 25th–75th and 10th–90th percentile bands.

### Evidence on adoption of digital payments
- Conceptual framing: interoperability can increase adoption both directly (greater freedom of choice for users) and indirectly (encouraging provider entry and quality upgrades).
- Two natural experiments are used to benchmark adoption under interoperability versus closed-loop systems.

1) Post-demonetization comparison (late 2016–2017)
- The November 2016 demonetization shock sharply increased demand for non-cash payment forms; many users tried both UPI and incumbent closed-loop alternatives.
- By March 2017, usage of the non-interoperable closed-loop incumbent plateaued while UPI adoption continued to grow through 2017.
- The difference in trends is large: while the non-interoperable alternative was flat, UPI grew roughly three-fold from March to October 2017.
- Cross-app payments rose even more than within-app payments, indicating interoperability was a valued feature relative to closed-loop incumbents.

2) Integration shock from incumbent joining UPI
- After a Reserve Bank of India directive mandating interoperability, a major incumbent closed-loop provider joined UPI, effectively combining two platforms into a single network.
- The size of the resulting increase in de facto interoperability varied across districts depending on the incumbent’s ex ante presence.
- Over the first year after integration, total peer-to-merchant (P2M) transaction value per person increased by 8 Rupees per month more in an average district with an above-median incumbent presence than in an average district with a below-median presence.
- That difference is persistent and substantial, equal to more than 80% of average P2M digital payments per person in above-median districts in the month prior to integration.
- The increase came from all three transaction components: both-counterparty-incumbent, one-counterparty-incumbent, and neither-counterparty-incumbent, benefiting both the incumbent and apps already on UPI.
- Measuring digital payments relative to cash (using ATM withdrawals as a proxy for cash usage) shows a similar substantial and persistent rise in digital payments after integration in districts with larger increases in de facto interoperability.
- Footnote evidence: “This difference amounts to 118% of average monthly P2M digital payments per person in below-median districts in the year after interoperability.”

### Risks to long-term choice and competition
- Potential erosions of interoperability benefits if providers become dominant:
  - Proprietary QR codes initially created steering effects by being readable only by a provider’s own app; RBI prohibited non-interoperable QR codes to preserve openness (Reserve Bank of India, 2020).
  - Even with interoperable QR codes, nonstandardized branding can create impressions that codes are app-specific, generating network effects.
- Economics of app provision (zero-fee UPI core with complementary revenue from lending, insurance) can create market power if secondary markets exhibit scale or scope advantages; bundling may reduce switching even if better UPI services exist elsewhere.
- Market concentration has increased: several aggregate metrics show rising concentration among UPI ecosystem participants; in half of UPI transactions, the payer and payee use the same app (stable share).
- Regional specialization of the two largest apps’ user bases suggests providers may steer users toward co-adoption, constraining freedom to choose.

### Regulatory implications and policy recommendations
- Regulators should monitor app dominance and be prepared to act to maintain a level playing field among apps.
- Use granular usage data and qualitative user research to identify mechanisms generating market power for dominant apps and tailor interventions to those mechanisms.
- In early stages of an interoperable system, adoption can be accelerated by:
  - Providing a public app to kick-start network effects.
  - Encouraging existing payments platforms to connect to the interoperable system, integrating trusted financial institutions (banks, incumbents) to bridge legacy users.
- Ensure complementary enabling investments are in place (e.g., reduction in mobile data costs, expansion of bank accounts and digital identity).
- As the interoperable platform matures, watch for emergence of dominant private providers and maintain policies to preserve a fully open, interoperable, and competitive system.
- System operators should consult with private sector participants to ensure design choices support ecosystem health.

*International Monetary Fund — FINTECH NOTES Growing Retail Digital Payments (selected chapter content)*

### Appendix Figure 1.1. Volume of Fast Payment Transactions (Millions)

### Appendix Figure 1.1. Volume of Fast Payment Transactions (Millions)

### Description
- Chart title: Appendix Figure 1.1. Volume of Fast Payment Transactions (Millions).
- Presents volumes of fast payment transactions measured in millions (figure itself not reproduced here).

### Data sources
- BIS
- Statista
- TCH
- IMF staff analysis

### Definition and scope notes
- Fast payments defined as: real-time or near real-time transfers of funds between accounts of end users as close to a 24/7 basis as possible (Frost and others 2024).
- US data comprise Zelle from 2017 and RTP from 2020.

*Source: IMF staff; Appendix Figure 1.1, ftnea2025004.*

---


_Source: https://www.imf.org/-/media/files/publications/ftn063/2025/english/ftnea2025004.pdf_