Abstract
This paper examines how the updated macroeconomic statistical standards reflect the ongoing financial innovations with regards to new financial instruments, service providers, and activities. It discusses the updated statistical methodologies (recently updated 2025 SNA/BPM7) to capture activities related to fintech, crypto-assets, and other emerging phenomena in digital finance, as well as challenges and aspects still to be addressed. Moreover, the article highlights the significance of data in understanding the evolving financial landscape and assessing macroeconomic implications of fintech and crypto assets.
Keywords
Introduction
What happens when credit is crowdfunded without banks? When do savings sit in mobile money accounts rather than bank deposits? When payment platforms and decentralized networks perform functions once reserved for banks—but outside the regulated perimeter?
Financial technology (fintech) is rapidly integrating into the global financial landscape, with central bank digital currencies (CBDCs) and crypto assets playing pivotal roles in this evolution. Crypto assets have spawned new asset classes, from Bitcoin to stablecoins like USDT, while CBDCs have the potential to extend access to central bank reserves beyond traditional banking institutions. Distributed Ledger Technology, in particular blockchain which is the technology behind crypto assets, enables market participants to directly transact with each other in the digital space (remotely) without the need for trusted third party such as for example a bank. Fintech lenders and payment platforms are reshaping how credit is created, liquidity circulates, and value is stored and transferred across the financial system.
The question is whether our statistics are keeping up. As Carstens 1 notes, central banks and statistical agencies must adapt their frameworks to maintain the reliability of monetary and financial data in an era of rapid technological change. If statistical frameworks fail to capture these trends in the digital age, important activities could go unmeasured or misclassified and—as a consequence—policymakers might not get accurate measures and aggregates. Most importantly, without up-to-date standardized statistical guidance, emerging digital financial activities may remain hidden or inconsistently measured, ultimately undermining the effectiveness of monetary policy and financial stability monitoring.
International statistical standards (ISS) are actively addressing this challenge. Acknowledging the need for statistical frameworks to incorporate these new innovations to better support both policy effectiveness and macroeconomic analyses, the international statistical community has embarked on updating the methodological guidance. The System of National Accounts 2025 (2025 SNA) and the Integrated Balance of Payments and International Investment Manual, seventh edition (BPM7)—the two foundational frameworks to measure domestic and international economic activity—released in March 2025, have introduced new guidance to better integrate financial digitalization and emerging instruments into countries’ macroeconomic statistics. In parallel, the International Monetary Fund's 2016 Monetary and Financial Statistics Manual and Compilation Guide (MFSMCG) is being updated to reflect changes in financial intermediation and ensure that monetary statistics evolve alongside the changing nature of money and credit creation.
This paper examines how recent updates to ISS tackle the measurement challenges posed by digital finance and highlights areas where additional operational guidance is required. It is structured in two parts. First, it reviews how the recently updated 2025 SNA and BPM7 along with the 2016 MFSMCG in some cases, define, classify, and sectorize key fintech products like electronic-money, peer-to-peer lending, CBDCs, and crypto assets. Second, it identifies remaining conceptual gaps—particularly in capturing digital finance within monetary, liquidity, and credit aggregates and monetary and financial statistics more broadly. Through this discussion, the paper connects the measurement of financial innovation to its broader macroeconomic and financial stability implications.
The rest of the paper is organized as follows: Section 2 demonstrates the scale and economic significance of selected digital finance innovations and identifies the specific measurement questions each poses for traditional statistical frameworks. Section 3, the core part of this piece, examines the 2025 SNA/BPM7's treatment of fintech and crypto-assets—including how the new assets and instruments are classified, and how fintech companies are sectored. Section 4 outlines the way forward, focusing on how financial innovations affect measures of money and liquidity, the operational challenges of data collection and valuation, and ongoing international efforts under the G20 Data Gaps Initiative 3 (DGI 3) to enhance the coverage of digital finance statistics. Section 5 concludes.
Digital finance ecosystem: Innovations and measurement gaps
The innovations reshaping finance are not abstract possibilities—they are measurable economic phenomena at significant scale. This section documents five key developments: mobile money, fintech lending, CBDCs, crypto assets including stablecoins, and tokenization, identifying the measurement questions each poses for traditional statistical frameworks.
Mobile money and electronic payments: When phones serve as banks
One of technology's most visible impact on finance has been on payments, particularly through mobile money platforms in developing economies. The scale is substantial: in several countries—including Ghana, Senegal, and Cambodia—mobile money transaction values in 2024 exceeded annual GDP (Figure 1), demonstrating the extent to which these platforms permeate daily economic activity. Moreover, mobile money accounts have evolved beyond their initial role as payment instruments, increasingly functioning as small-scale savings vehicles. 2 In leading Sub-Saharan markets, average balances maintained per mobile-money account more than doubled over the past decade—from USD 9.7 to 25.4 in Ghana and USD 0.7 to 17.9 in Zambia (Figure 2).
Value of mobile money transactions (during the reference year 2024) as percentage of GDP in selected countries. Source: Authors’ analysis using the IMF's Financial Access Survey.
Average balance per mobile money account in USD in selected countries. Source: Authors’ analysis using the IMF's Financial Access Survey.
Beyond Africa, use of mobile payments—including India's UPI, China's Alipay and WeChat Pay, and advanced-economy services such as Venmo and Cash App—has become widely used, fundamentally reshaping how value moves through economies.
While the 2016 MFSMCG provided initial guidance on the treatment of e-money as part of broad money, the evolving nature of these platforms raises new questions. Should institutional classification depend on whether the provider is bank-affiliated or independent? How should statistics account for layered structures where mobile money liabilities of the service providers are backed by deposits at commercial banks—does this create a risk of double counting?
Fintech lending is a growing segment of direct lending or financial intermediation that remains relatively modest in overall size. Fintech lending encompasses all lending activities facilitated by electronic platforms (This broad definition may include activities undertaken both by banks and non-banks. Current work under the DGI 3 Recommendation 12 on Fintech-enabled financial inclusion and Financial Access Survey 2025 Pilot are heading into this direction.). Such fintech platforms facilitate a range of lending activities, including secured or unsecured lending, funding through debt securities (a bond, debenture, or subordinated debt), or funding through the purchase of invoices or receivables from a business. Fintech lending platforms may operate in two broad modes: As financial auxiliaries, they can be in the form of an online “marketplace platform”, which allows lenders to transact directly with borrowers (peer-to-peer lending and crowdfunding platforms), As financial intermediaries, when they use their balance sheets to originate the lending, i.e., they extend loans on their own account bearing the associated risks and rewards.
Fintech lending is expanding rapidly and has already begun to surpass traditional lending activities in specific niches. According to the FSB data survey focused on non-bank fintech (For the purpose of FSB coordinated data exercise non-bank fintech lending was defined as “lending activity facilitated by electronic platforms that are not operated by commercial banks”. 3 ) lending, ten jurisdictions reported about USD 42 billion in aggregate outstanding fintech credit for 2023, equivalent to roughly one percent of their combined other financial intermediaries loan assets. While modest overall, this share varies considerably across countries: in some markets, fintech platforms account for over 30 percent of nonbank loan portfolios. 4 Balance-sheet intermediaries such as structured finance vehicles and finance companies, which together comprise about 73 percent of total fintech lending and auxiliary marketplace platforms that simply broker loans, facilitating an additional USD 3.7 billion in 2023. 4
From a statistical standpoint, fintech lending presents both classification and sectoring challenges. How should statisticians distinguish between loans originating on balance sheets versus those merely facilitated? Should peer-to-peer loans between nonbank entities be included in aggregate credit measures? And can platform facilitators provide the detailed data necessary for consistent coverage in financial accounts and credit aggregates? Whether platforms are classified as financial intermediaries or auxiliaries determines both their sectoral placement and the treatment of their lending in credit aggregates.
Central banks worldwide are exploring digital versions of their currencies. According to BIS study, 5 85 out of 93 central banks surveyed in 2024 were exploring CBDCs. Among active CBDC projects, advanced research and preparations or pilots, notable examples include Bahamas (Sand Dollar), Eastern Caribbean (D-Cash), Nigeria (eNaira), Jamaica (Jam-Dex), China (e-CNY), India (Digital Rupee), Brazil (Drex), and euro area (digital euro).
Adoption rates vary significantly, particularly for retail CBDCs, with some achieving broader usage while others remain in early stages. India's retail digital rupee (e₹-R) illustrates the growth potential of retail CBDCs. Launched in December 2022, circulation grew from ₹57.2 million in March 2023 to ₹10,165 million (USD 120 million) by March 2025—a 178-fold increase in just two years (Figure 3).
India's retail CBDC: transaction volume and value, 2023–2025. Sources: Reserve Bank of India and IMF staff calculations.
Crypto assets market capitalization in USD billion 2016–2025. Sources: Haver and IMF staff calculations.
Beyond retail CBDSCs, countries are experimenting with wholesale CBDCs that provide digital central bank money exclusively to financial institutions for interbank settlements and securities transactions which are now increasingly being termed “tokenized reserves”. 6
These developments raise important statistical questions. Should retail and wholesale CBDCs receive identical statistical treatment, given that both represent central bank money? How should retail CBDCs be classified—as currency, deposits, or a distinct category? Does the programable, token-based, or distributed ledger nature of wholesale CBDCs change their economic function sufficiently to warrant different treatment from conventional reserves?
Crypto assets including stablecoins have expanded rapidly alongside CBDC innovations (Figure 4). Initially introduced to enable peer-to-peer transactions without intermediaries and to serve as global payment instruments, they have since diverged from this original purpose. As of mid-December 2025, their total market capitalization was around USD 3 trillion, dominated by Bitcoin (BTC) and Ether (ETH). In practice, these assets are now held primarily for investment and speculative purposes rather than as means of payment.
Unlike conventional instruments, unbacked crypto assets have no issuer or corresponding liability, raising a core statistical question: should they be treated as commodities, financial assets without counterpart liabilities as an exception, or a new hybrid class? Their prices reflect only supply and demand, yet they function as investment and collateral assets in decentralized finance, blurring traditional boundaries of financial activity.
A related but distinct development is the rise of stablecoins which are crypto assets designed to maintain stable value against reference assets. Stablecoins, by contrast, remain closer to their intended purpose. Their relative price stability and asset backing give them greater potential to serve as effective payment instruments, particularly in cross-border and digital-market contexts. This notwithstanding and although smaller in scale (about USD 314 billion as of December 2025, their reserve holdings across banks and securities create new transmission channels for financial instability 7 . The March 2023 collapse of Silicon Valley Bank, which held USDC reserves, briefly de-pegged the coin and exposed USD 3.3 billion in at-risk funds, while the TerraUSD failure in 2022 erased USD 40 billion in value. These events revealed how such cross-sector exposures remain largely unmapped in sectoral balance sheets.
These developments raise an important question: how are stablecoins different from unbacked crypto assets in statistical sense. At the same time, their growing linkages to the banking sector and securities markets create new channels for risk transmission and expose fundamental gaps in macroeconomic measurement, which raises important questions. Should they be considered as financial assets and included in existing instrument categories or treated separately as investment instruments? A separate question is whether they should be included in measure of money or liquidity?
Other tokenized assets
Building on the developments in payment technologies, blockchain technology is now being applied beyond the realm of payments to represent conventional financial instruments through tokenization that as trend may have potentially significant macrofinancial implications. 8 Tokenization refers to creating digital representations of traditional financial or physical assets—collectively known as real-world assets (RWAs)—on distributed ledgers. It takes two main forms: (i) direct issuance of assets natively on-chain, such as tokenized government bonds or money market funds; and (ii) tokenization of already existing off-chain assets, such as real estate, commodities, or private credit.
As shown in Figure 5, the value of distributed tokenized RWAs exceeds USD 18 billion, spanning public and private equity, commodities, private credit and government debt, which represents the largest segment as of December 2025. Recent large-scale tokenized bond issuances by banks, corporates, and even governments (Table 1), demonstrate how tokenization of assets is now mainstream and being adopted across stakeholders.
Total distributed real world assets value as of December 2025 (in USD billion). Sources: RWA.xyz and staff calculations.
Selected recent tokenized bond issuance.
Source: Authors’ compilation.
From a statistical perspective, tokenization raises key classification and recording questions. Should tokenized assets share the same classification as their underlying reference assets, or qualify as another asset category with characteristics distinct from the underlying assets? How should custodians, tokenization platforms, and smart-contract entities be sectored within existing frameworks? If both the token and its underlying asset are recorded, does this lead to double counting in balance-sheet aggregates?
These innovations, as discussed above, pose specific measurement challenges that existing statistical frameworks must address. Section 3 examines how the 2025 SNA and BPM7 address these challenges through updated guidance on the classification of financial instruments and sectoring of institutional units involved.
Technology neutrality: The cornerstone principle
The 2025 SNA and BPM7 address digital finance innovations through a foundational principle: technology neutrality. Under this principle, the economic substance of an instrument—not the technology that enables it—determines its statistical classification (2025 SNA §22.78-22.79, BPM7§ 16.74–16.75). This ensures that the digitalization of finance does not fragment statistical comparability or render historical time series obsolete.
Technology neutrality allows statisticians to integrate innovations into existing categories of financial assets and liabilities and to classify their service providers within the existing financial sector framework avoiding the need to create new taxonomies for each technological advancement. The 2025 SNA and BPM7 do not establish a distinct fintech sector because fintech activities span multiple financial services already captured within the sector classification framework. Instead, fintech service providers are allocated to established sector categories based on their economic function, and fintech products are classified within existing instrument categories. However, when innovations have sufficiently distinct economic characteristics, new subcategories may be introduced. For instance, the 2025 SNA/BPM7 introduce a subcategory for crypto assets designed to act as a general medium of exchange with corresponding liabilities—capturing backed stablecoins within the currency and deposits category.
Yet technology neutrality does not mean statistical invisibility. Recognizing the analytical importance of monitoring digital finance, the 2025 SNA and BPM7 recommend supplementary “of which” disaggregation (2025 SNA §22.78; BPM7 §16.74). For example, loans could include “of which: extended by fintech lenders,” with further granularity distinguishing balance-sheet from peer-to-peer lending where data permits. Similarly, sectoral breakdowns may identify “of which: financial service providers predominantly operating digitally.” These are not separate SNA categories but analytical supplements that increase granularity without altering core aggregates.
Classification of digital financial ASSETs
Having established the principle of technology neutrality, this subsection examines how the 2025 SNA and BPM7 classify specific digital financial instruments. The classification of each instrument type addresses the practical measurement questions raised in Section 2 regarding classification boundaries and sectoring. Where relevant, this discussion also references the 2016 MFSMCG, which—as the last manual updated in the previous round of revisions—already provided initial guidance on certain digital finance issues, particularly electronic money.
E-money
The 2016 MFSMCG defines electronic money (e-money) as a payment instrument whereby monetary value is electronically stored—either on a physical device or remotely at a server—representing a claim on the issuer. Examples include prepaid cards, mobile wallets, and mobile money. The 2025 SNA and BPM7 adopt consistent definitions, characterizing e-money as stored monetary value that can be used for third-party payments and functions as a close substitute for transferable deposits.
Importantly, e-money refers to stored value like balances held in mobile-money accounts, prepaid cards, or digital wallets that represent the issuer's liability. Using a debit card or bank account through a mobile app does not create e-money; these are merely digital channels to access deposits already on bank balance sheets. Similarly, store-specific gift cards or limited-use prepaid instruments do not qualify as e-money.
All three frameworks—the 2016 MFSMCG, 2025 SNA, and BPM7—classify e-money as deposits rather than currency, and specifically as transferable deposits when usable for direct third-party payments (2025 SNA §5.152 and §5.154; SNA § 22.83; BPM7 §4.130 and §4.132; BPM7 §16.79 MFSMCG §4.38).
Fintech lending
Digital platforms may transform how credit is delivered, but not what it fundamentally represents. Fintech lending, whether through balance-sheet lenders or peer-to-peer platforms, still produces loans as the underlying financial instrument. While the platform's business model affects the institutional classification of the parties involved (discussed in Section 3.2), the underlying financial instrument is classified consistently: loans originated by balance-sheet lenders and loans facilitated through peer-to-peer platforms are both recorded as loans (2025 SNA §22.80 -22.81 BPM7 §16.76–16.77).
Together, these models demonstrate that digital platforms change who provides credit and how it is delivered, but not the fundamental statistical treatment of loans as financial assets.
CBDCs
The 2025SNA (§22.86) and BPM7 (§16.82) classify CBDCs as digital financial assets with corresponding liabilities under the currency and deposits category.
Retail CBDCs issued to households and nonfinancial corporations are specifically recorded as currency (with an “of which” subcategory to distinguish them). Wholesale CBDCs, however, present classification ambiguity. While retail CBDCs mirror currency in function and accessibility, wholesale CBDCs, which are like tokenized reserves used exclusively by financial institutions for interbank settlements, are not yet explicitly addressed in the updated standards but most likely are to be classified as deposits rather than currency. The programable, token-based, or distributed-ledger technology underlying wholesale CBDCs does not alter their economic function as reserves held by financial institutions at the central bank. Under the principle of technology neutrality, classification follows economic substance, not technological form.
Crypto assets and stablecoins
Crypto assets can be classified as fungible (divisible and non-unique, such as Bitcoin) or non-fungible (indivisible and unique, commonly known as non-fungible tokens or NFTs).
The 2025 SNA and BPM7 confirm that all fungible crypto assets fall within the asset boundary (SNA §4.98; BPM7 §3.79). For macroeconomic and financial statistics purposes, fungible crypto assets are the primary focus, as they function as tradable financial or investment instruments with implications for monetary aggregates, sectoral balance sheets, and financial stability.
The 2025 SNA and BPM7 distinguishes fungible crypto assets based on whether they have a corresponding liability—that is, whether an identifiable issuer promises redemption or provides an underlying claim.
Unbacked crypto assets (Bitcoin, Ether, and approximately 18,000 similar tokens) have no corresponding liability. The 2025 SNA/BPM7 classify these as nonproduced, nonfinancial assets. They are not treated as currency, since they are not liabilities of any central bank or any other issuer. Even when used in transactions, payments are statistically recorded as bartering a nonfinancial asset for goods or services. However, the 2025 SNA and BPM7 acknowledge that the classification of unbacked crypto assets remains an open issue, as their potentially widespread use as a general medium of exchange could warrant future reevaluation.
Crypto assets with corresponding liabilities encompass several types: Stablecoins maintain stable value relative to specified assets through reserve backing. Fiat-pegged stablecoins like USDT and USDC—backed by bank deposits, government bonds, and money market instruments—are classified under currency and deposits in a new subcategory for crypto assets designed to act as a general medium of exchange with underlying liabilities (2025SNA§12.56; SNA§22.86-22.87; BPM7 §5.43; §16.82–16.83). Note however, that algorithmic stablecoins, which maintain price stability through protocols that adjust token supply rather than reserve backing, have no corresponding liability and would be classified as nonfinancial assets similar to unbacked crypto assets. Security tokens (tokenized shares or bonds) are classified as equity or debt securities, identical to their traditional counterparts—consistent with technology neutrality (BPM7 §16.81 and SNA §22.85). A tokenized government bond is recorded as a debt security, no different than if the bond were issued in paper form or as a traditional electronic book entry. The use of distributed ledger technology to record ownership does not change the economic nature of the claim. Utility tokens that grant access to platform services and can be traded are generally classified as debt securities when the issuer has an obligation to provide services (BPM7 §16.81; SNA §22.85). If a utility token is tradable, has monetary value, and represents an issuer's obligation to provide future services, it functions analogously to a debt instrument. The principle here, again, is that the economic nature of the claim determines the classification, not the label attached to the instrument.
Sectoring the new intermediaries: Where do fintech firms belong?
Under the technology-neutral approach of the 2025 SNA and BPM7, institutional units are classified according to their economic functions. Where digital delivery is material, compilers may use “of which: predominantly digital providers” to retain analytical visibility in sectoral tables.
E-money issuers
Electronic money institutions with liabilities that form part of broad money should be classified as deposit-taking corporations (S.122), regardless of whether they are bank-affiliated or independent mobile network operators (MNOs). The 2025SNA §5.154 /BPM7 §4.132 explicitly states that e-money institutions should be classified as deposit-taking corporations if they are financial corporations and if the e-money issued is included in broad money.
Fintech lending platforms
The BPM7 recognizes that digital platforms have transformed the delivery of credit, but not the nature of the underlying financial instrument. According to 2025SNA (§22.80–22.81) / BPM7 (§16.76–16.77), fintech lending platforms fall broadly into two categories—balance-sheet lenders and peer-to-peer (P2P) intermediaries—depending on whether they assume credit risk on their own account or merely facilitate transactions between others. Balance-sheet fintech lenders that extend loans using their own funds and hold the loans as assets on their balance sheet are classified as other financial intermediaries (S.125). As noted in BPM7 §16.76, these entities operate similarly to traditional finance companies, incurring liabilities (from investors, equity, or other funding sources) and acquiring financial assets (loans) on their own account, unlike financial digital platforms that provide rather matching services. Activity of balance sheet fintech lenders constitutes financial intermediation, as they take on risk and earn income through interest rate spreads or fees. Peer-to-peer (P2P) and marketplace lending platforms, by contrast, are classified as financial auxiliaries (S.126) under BPM7 §16.77 (2025SNA §22.81), since they match lenders and borrowers without taking ownership of loans. The underlying loan exists directly between the ultimate lender and borrower, and the platform's role is limited to facilitation, earning a service fee.
In the case of mixed-model platforms combining both activities should, where possible, be split by function based on separate accounts. If that is not feasible, the dominant activity principle applies. Example of challenging case might be bigtech firms engaged in lending- these will require case-by-case analysis depending on organizational structure. When large technology or e-commerce platforms extend credit to sellers or customers and this activity is separated from the nonfinancial activities, classifying would be as follows: Other financial intermediaries (S.125), if it runs balance-sheet lending to independent entities Captive financial institutions (S.127), if the borrower is capital-related to the platform Financial auxiliaries (S.126), if the platform merely matches independent lenders and borrowers.
Crypto exchanges: Financial or nonfinancial?
The 2025SNA § 5.169 / BPM7 § 4.147 provides specific guidance on crypto asset exchanges and platforms, with classification depending on the types of crypto assets primarily traded: Exchanges whose activity centers on crypto assets with corresponding liabilities—for instance, backed stablecoins, tokenized securities, or security tokens—are classified as financial corporations (financial auxiliaries, S.126), since they facilitate transactions in financial assets in a manner similar to conventional securities exchanges. In contrast, exchanges that mainly trade unbacked crypto assets (such as Bitcoin or Ether) are classified as nonfinancial corporations, as unbacked crypto assets are treated as nonfinancial assets, analogous to commodities.
The same principle applies to custodial and wallet services provided by crypto platforms: when they hold crypto assets with corresponding liabilities, they fall within the financial sector (S.126), whereas services related to unbacked crypto holdings are considered nonfinancial.
When the institution disappears: Measuring decentralized finance (DeFi)
For many digital-finance providers, sectoring under the 2025 SNA/BPM7 is conceptually straightforward. Central banks issuing CBDCs remain classified in the central bank subsector as S.121, and neobanks or e-money institutions that accept deposits continue to fall under deposit taking corporations except central bank.
The real challenge arises with entities that lack a clear institutional perimeter like Decentralized-Finance (DeFi) protocols. These systems perform financial intermediation functions through self-executing smart contracts on permissionless blockchains, often without any identifiable legal owner, management, or jurisdiction. 9 The activities are real, including lending, liquidity provision, and trading, but the institutional perimeter is diffuse, and counterparty relationships are often opaque.
While the 2025 SNA (22.80–22.82) and BPM7 (16.76–16.78) do not address the treatment of such arrangements directly, the principles outlined in the standards suggest a pragmatic approach: when such activities exhibit governance, continuity, and an ability to generate income (e.g., through transaction fees), they can be treated as quasi-corporations, with residence proxied by the location of the core development team or foundation. In such cases, DeFi protocols could be sectored as financial auxiliaries (S.126), with transactions recorded directly between users of the protocols.
Yet many protocols remain borderless and pseudonymous, making both sectoring and residency assignment problematic. Without identifiable institutional units, statisticians face a challenge: financial activities exist, but no entity can be identified and located. As regulatory frameworks evolve and large protocols gain legal status, future practice will likely move toward recognizing major decentralized networks as quasi-corporate entities, bridging this measurement void. While the 2025 SNA and BPM7 provide a conceptual basis—emphasizing the recognition of quasi-corporations, technology neutrality, and functional classification—the practical sectoring of decentralized protocols remains unresolved. These issues underscore the need for further operational guidance.
The next phase of digital finance in macroeconomic statistics
The 2025 SNA and BPM7 mark a major conceptual leap in integrating digital finance into the macroeconomic measurement framework. Looking ahead, the challenge lies in translating these conceptual advances into practice, by preparing topical compilation guidance. The work on compilation guidance focused on crypto assets in the context of 2025 SNA and BPM7 has already advanced.
The next phase of conceptual work involves scoping money, liquidity, and credit for the digital era, followed by developing detailed compilation guidance on data sources and valuation methods and advancing coordinated international efforts under the G20 Data Gaps Initiative 3 (DGI-3) and beyond to strengthen cross-country coverage of digital finance statistics.
Redefining money in the digital age: The next frontier for statistical standards
While the 2025 SNA and BPM7 lay the conceptual groundwork for classifying digital financial instruments and institutional units involved, the measurement of money, liquidity, and credit in this digital age—particularly in light of emerging payment methods such as stablecoins and retail CBDCs—remains a focal point for refinement. The forthcoming update to the 2016 MFSMCG will play a crucial role in clarifying how these new instruments should be incorporated into measures of money and other aggregates, guiding macroeconomic statisticians on which digital assets qualify as “money” within macroeconomic statistics and ensuring consistency in measuring the evolving money, liquidity, and credit supply.
In particular, the updated MFSM will need to clarify how new digital instruments interact with money and other related aggregates. The proliferation of stablecoins and retail CBDCs raises questions about the scope of “money” in macroeconomic statistics: under what conditions—if any—should widely use fiat-backed stablecoins or short-term tokenized debt securities be incorporated into broad money? 10 Likewise, unbacked crypto assets, despite their sizeable market capitalization, challenge the boundary between financial and nonfinancial assets and their relevance for liquidity indicators.
Further precision is also required for the statistical treatment of e-money issuers. As digital ecosystems evolve, e-money is increasingly issued by entities outside the traditional banking sector—such as fintech firms and telecommunications providers. The updated MFSM will provide detailed guidance on how to record such issuers’ liabilities, particularly when e-money balances are backed by deposits at commercial banks, to avoid double counting across balance sheets.
Although BPM7 specifies that electronic money institutions should be classified as deposit-taking corporations when they qualify as financial corporations and when the e-money they issue is included in broad money, several practical questions remain for compilers. These issues will require further operational guidance in the forthcoming MFSM update. In particular, how should e-money issued by mobile network operators or large platform companies be recorded when the issuing activity is embedded within a nonfinancial corporation and cannot be easily delineated from the other activities or as a separate institutional unit? Additionally, when e-money issuers back customer balances with deposits at commercial banks, how can statisticians ensure that the same monetary value is not recorded both on the issuer's and on the banking sector's balance sheets?
A more recent challenge is tokenization, which has accelerated rapidly since the finalization of the 2025 SNA and BPM7. These developments will require statisticians to consider whether tokenized instruments should retain the same classification as their underlying assets or be treated as a separate category, as well as how to avoid duplication between the recorded value of the token and its reference asset—a matter that the updated MFSM must explicitly address.
In summary, the forthcoming MFSM update will shape how digitalization transforms not only the classification of financial instruments and institutional units but also the composition of key macro-financial aggregates such as the monetary base and money supply.
Operational guidance and data compilation
As much as the 2025 SNA, BPM7, and the forthcoming MFSM provide the conceptual framework for incorporating digital finance within macroeconomic statistics, the effective implementation of these standards will depend on the availability, granularity, valuation, and timeliness of relevant data. Translating these new concepts into measurable statistics requires compilers to identify credible data sources, apply consistent valuation methods, and adopt innovative data collection methods. The first substantial effort in this area has already been undertaken – the compilation guidance focused on crypto assets is being finalized in the IMF and will be published in 2026.
Valuation and price consistency pose the first major challenge. Digital assets, whether crypto assets or tokenized securities, often trade simultaneously on multiple exchanges with varying liquidity and reporting standards. Prices can diverge across platforms, time zones, and trading pairs, complicating the determination of reference prices for balance sheet and flow statistics. The 2025 SNA reiterates general market-price valuation principles but does not specify fully how to apply them in decentralized or fragmented markets. Future operational guidance will need to define clear protocols—such as using volume-weighted averages, identifying reference exchanges, or applying end-of-period composite pricing—to ensure both comparability and consistency across countries.
Further development of compilation guidance will also need to consider complex institutional cases left open by the 2025 SNA and BPM7, particularly decentralized or algorithmic financial arrangements. These include identifying when DeFi protocols qualify as institutional units or quasi-corporations, how to assign residence, and how to measure their positions and flows. Such issues, recognized but not yet operationalized in current standards, will form an important part of the next phase of statistical guidance.
A related issue concerns data sources. Many digital finance activities, particularly in DeFi—occur outside traditional reporting frameworks, limiting the visibility of sectoral positions and counterparty linkages. Even when data are available, such as on blockchain ledgers, the pseudonymous nature of ownership complicates institutional-sector assignment. Accordingly, data needs must be explicitly specified for each asset type, identifying who reports, what is reported, and how it aligns with existing macroeconomic datasets.
Taken together, these operational dimensions—valuation, data sourcing, and frequency—represent the frontier of digital finance measurement. Conceptual frameworks can define what to measure, but without clear protocols on how to value and collect the underlying data, statistical systems risk omitting some of the most dynamic parts of the modern financial system.
From standards to statistics
Producing relevant, high-quality, and cross-country comparable data on digital finance is a critical element of this evolving discussion. As regulatory frameworks develop to keep pace with rapid innovation in digital finance, there is an increasing need for innovative approaches to data collection and analysis. Accurate and comprehensive data remain essential to provide policymakers with reliable macroeconomic statistics and indicators. While many countries are updating their data-collection frameworks to capture new intermediaries and incorporate more detailed breakdowns, these improved sources will require time to fully mature and deliver consistent global insights.
Among efforts aimed at providing data on digital finance, G20 DGI 3 is a broad and internationally coordinated project. The DGI-3 includes targeted recommendations addressing statistical challenges posed by digital finance innovations: Recommendation 10 on Fintech credit, Recommendation 11 on Digital money, Recommendation 12 on Fintech-enabled financial inclusion.
Recommendation 10 emphasizes the need for improved data on non-bank fintech credit services to strengthen financial stability analysis. As lending migrates from traditional banks to fintech platforms, P2P networks, and other digital credit providers, comprehensive data collection becomes essential for monitoring systemic risks and assessing credit market health. In 2024, FSB members provided fintech credit data for the first time on a best-efforts basis, marking initial progress toward systematic collection of information about fintech lending volumes, borrower profiles, and credit quality across jurisdictions. 3
Recommendation 11 addresses the growing relevance of new instruments that may be used as means of payment, including CBDCs, stablecoins, and unbacked crypto assets. The Task Team for Recommendation 11 led by the IMF prepared test data reporting templates for digital money and crypto assets, built around three key informational dimensions—Who holds What and Where: Who refers to the institutional sector of the holder, such as households, banks, non-bank financial corporations, and nonfinancial corporations; What identifies the type of instrument or asset (CBDCs, stablecoins, other unbacked crypto assets); and Where captures the country of residence of the holder, a critical dimension considering cross-border flows and the potential risks of currency substitution. The data dimensions and structure of this collection exercise are well aligned with the information needs of macroeconomic statistics and consistent with the definitions set out in the 2025 SNA and BPM7.
Recommendation 12 addresses fintech-enabled financial inclusion by advocating granular data on access and usage of digital financial services, disaggregated by income, gender, geography, and education. While focused primarily on sociodemographic factors to evaluate progress toward inclusion goals, the resulting statistics may also inform macroeconomic analysis of monetary policy transmission and consumption patterns.
The essence of these initiatives is to pilot, refine, and scale data-collection frameworks across countries in a resource-sensitive manner while meeting core statistical needs.
Financial landscape has been changing rapidly with new phenomena emerging, such as CBDC, crypto assets including stablecoins, tokenized financial assets, as well as a range of new service providers such as fintech lenders, P2P lending platforms, crypto assets exchanges, and others. Financial innovation is here to stay, and macroeconomic statistics have taken important steps to incorporate it.
The 2025 SNA and BPM7 provide important conceptual foundations and brings digital finance into the statistical fold with clear classifications. Technology neutrality has been indicated as main principle for making statistical classifications in the digital age. These revisions ensure that, at a conceptual level, GDP, sectoral accounts, and external statistics can better reflect the digitalization of finance.
The next critical step lies in the forthcoming update of the 2016 MFSMCG. Building on the 2025 SNA and BPM7, the updated MFSM will define how the digital transformation of finance should be reflected in monetary, credit, and liquidity aggregates.
However, conceptual clarity alone is not sufficient. To translate these standards into practice, further detailed compilation guidance will be needed, covering data sources, valuation methods, and reporting procedures to ensure consistent and practical implementation across economies. Data-collection systems must evolve to keep pace with innovation. The path forward will be iterative: early estimates may be incomplete but will improve over time as methods, regulations, and reporting infrastructures develop.
To contribute to this goal, the IMF Statistics Department aims at releasing in 2026 a compilation guidance focused on crypto assets to provide clarifications on data sources and compilation methods for measuring crypto assets, including stablecoins, in macroeconomic statistics. This guidance will help national compilers by supporting consistent and comparable estimates across countries. It also supports the G20 DGI 3 Recommendation 11.
Going forward, the priority will be to translate conceptual progress into measurable data. This will require continued coordination among data producers, central banks, and international agencies to develop consistent valuation practices, expand reporting coverage, and align digital-finance statistics with existing macroeconomic datasets. As data sources mature and methodologies are refined, the quality and timeliness of digital-finance indicators will improve, ensuring that macroeconomic statistics remain relevant, reliable, and analytically robust in an increasingly digital financial system.
Footnotes
Disclaimer
The views expressed in this paper are those of the authors and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.