Loading component...
Tokenized Finance and Money
May 22, 2026
Loading component...
IMF Communications Department
MEDIA RELATIONS
PRESS OFFICER:
Phone: +1 202 623-7100Email: MEDIA@IMF.org
As prepared for delivery
Thank you for the opportunity to speak today. Today, I want to discuss how the architecture of finance is changing, and what that change means for policy. The institutions that support the monetary system, its infrastructure, and the choices that will shape where money and finance go in the years ahead.
I will cover four areas: what tokenization may imply for the financial sector, the question of what settles tokenized transactions and at what risk, the macro-financial implications of stablecoins for both issuing and recipient jurisdictions, and the regulatory landscape as it stands today. Let me take these in turn.
Tokenization and the Financial System
Previous waves of digitization in finance improved efficiency within existing institutional structures. Electronic trading, online banking, and mobile payments made the plumbing faster, but they did not change the plumbing itself. Tokenization, however, has the potential to reconfigure how trust, settlement, and risk management are organized. The boundaries of the financial system shift, not just the speed of operations within them. In a recent IMF paper, I have argued that this constitutes a structural reallocation of trust within the financial system.
Three features distinguish tokenization from earlier innovations. First, programmability: smart contracts can execute financial logic automatically, triggering margin calls, coupon payments, or collateral transfers through code without human intervention. Second, shared ledgers: a single synchronized source of truth replaces bilateral reconciliation across institutions, so that each counterparty no longer maintains its own record. Third, atomic settlement: delivery versus payment can occur simultaneously in near real time, collapsing multiple stages of the traditional value chain into one synchronized process.
The implications can be significant. When financial contracts are governed by code, the locus of risk shifts from institutions to infrastructure. Balance sheets and legal processes give way to data feeds, algorithms, and the governance arrangements that surround them. Supervisory frameworks designed around capital adequacy and conduct risk are insufficient when failures can originate in a faulty price oracle or a poorly audited smart contract. And because of this, the most consequential transformations are not happening in permissionless crypto but within regulated finance, in banks, asset managers, and financial market infrastructures.
Consider what this means in practice. In banking, tokenized deposits can unify payments, settlement, and liquidity management on a single infrastructure, enabling atomic execution and reducing reconciliation costs. But around-the-clock settlement also reduces banks’ ability to smooth liquidity through end-of-day netting cycles. Real-time liquidity management becomes essential. On the asset side, tokenized lending can embed interest accrual, collateral triggers, and covenant enforcement directly in code. The risk is that automated enforcement amplifies procyclicality if margin calls respond mechanically to market prices. Governance frameworks must allow for override and discretion.
In capital markets, tokenized securities enable atomic delivery versus payment, reducing counterparty risk and operational friction. But eliminating settlement lags shifts liquidity demands from discrete points to continuous real time. Credit exposures decline, while intraday liquidity needs increase. Near-real-time collateral mobilization improves efficiency in normal times but can accelerate collateral withdrawals in stress, transmitting shocks more rapidly across institutions.
Financial market infrastructures are also evolving. Central counterparties (CCPs), central securities depositories (CSDs), and payment systems are moving toward permissioned shared ledgers. This is a deliberate shift away from permissionless architectures and toward institutionally governed infrastructure, with identifiable participants, governance structures, and compliance with anti-money laundering and counter-terrorism financing requirements. Consolidation on shared ledgers improves liquidity and reduces duplication, but it also creates critical infrastructure nodes whose failure could disrupt the entire market. Supervision must extend to the design, governance, and resilience of the infrastructures themselves.
Settlement, Speed, and Stability
Settlement lags in traditional finance are costly. They tie up capital and generate counterparty exposure. But they also serve as buffers. Those buffers give institutions time to net claims, mobilize funding, and respond to shocks. They also give authorities time to intervene before settlement becomes final. Tokenized systems eliminate these buffers by design.
The risks that follow are distinctive. Automated margin calls triggered by price movements can force rapid asset sales, reinforcing procyclical dynamics. Algorithmic risk differs from traditional operational risk in that errors propagate instantaneously and autonomously, without human intervention. A faulty price feed or a coding error can trigger cascading liquidations before any authority has time to respond.
This has direct policy implications. Governance of code must be mandatory for systemically important contracts: formal verification, independent audits, transparent change management, and the ability to pause execution under predefined emergency conditions. Central bank backstops designed around business-day cycles are insufficient in a 24/7 environment. Effective backstops may need to operate directly within tokenized systems, at machine speed. And legal clarity is required on whether tokenized records constitute proof of ownership and whether on-ledger settlement carries finality, an area where a dual-layer approach is emerging that combines smart contracts for operational rules with traditional legal agreements for rights and dispute resolution.
Concrete pilots have been already testing these boundaries. In the DTCC initiative, banks, custodians, CSDs, and CCPs used tokenized US Treasuries as collateral, mobilizing high-quality liquid assets in real time to meet margin calls while maintaining legally enforceable control. Eurex Clearing obtained regulatory non-objection for DLT-supported collateral mobilization, embedding the process within the existing CCP risk framework rather than building a parallel system. These pilots demonstrate that tokenization can automate and accelerate core CCP functions. But they also illustrate that when margining and collateral substitution are governed by smart contracts, errors in code or data inputs could trigger automated, procyclical responses. Governance of the algorithm and data inputs becomes systemic risk management.
The Settlement Asset Question
In traditional markets, central bank money serves as the ultimate settlement asset. It anchors convertibility and preserves what we call the singleness of money: the principle that different forms of money trade at par. Tokenization reopens this question by enabling a broader range of digital forms of money to circulate on shared ledgers.
Tokenized deposits are a digital extension of existing bank liabilities. Prudential regulation and deposit insurance apply. They retain the two-tier monetary system but remain exposed to bank credit risk where deposits are uninsured.
Wholesale central bank digital currencies (CBDCs) offer a direct digital claim on the central bank, eliminating settlement asset credit risk. Project Jura demonstrated cross-border atomic settlement using wholesale CBDC, and Project Agorá is exploring a unified ledger that integrates tokenized deposits with central bank reserves.
Regulated stablecoins are privately issued, reserve-backed tokens. Par convertibility depends on reserve quality, the issuer's operational capacity, and the liquidity of underlying funding and government securities markets.
Each of these forms embodies a different allocation of risk between the public and private sectors. Where the ultimate backstop sits is fundamentally a policy choice, not a purely technical question. That choice shapes the structure of the monetary system: who bears the risk, who provides the liquidity, and who governs the infrastructure.
Par Exchange and the Fragility of Stablecoin Pegs
Four pillars sustain par exchange in the traditional monetary system: (1) settlement finality, (2) the central bank as ultimate settlement asset, (3) standing liquidity facilities, and (4) deposit insurance. Resolution regimes complete the framework by ensuring continuity of the payment function even through institutional failure. Tokenized deposits and tokenized central bank reserves inherit this architecture. Par exchange is sustained by the same machinery as their non-tokenized equivalents.
Stablecoins still sit outside this architecture. They have no access to central bank settlement, no deposit insurance, and no resolution framework. The binding constraint for stablecoins is liquidity, not solvency: whether reserves can be liquidated fast enough to meet redemptions at par, even when they are sufficient in value. Major issuers currently impose some limits on access to the redemption window. Only a small number of authorized participants can redeem directly, while retail holders must sell on secondary markets even where prices can deviate from par.
This is economically analogous to a fixed exchange rate with limited convertibility. The peg is sustained by rationed access to reserves, not by unconditional redemption. The dynamics are the same as those observed in money market funds, but without the regulatory scaffolding: no gates, no swing pricing, no central bank backstop.
In the absence of the institutional architecture that sustains par, trade-offs and liquidity pressures may relax during calm periods, but in stress episodes the constraints can become binding. Let me borrow from the international macro literature and present a trilemma that illustrates the structural trade-off. Without an external backstop, a stablecoin issuer cannot simultaneously maintain a par peg, open convertibility, and choose its reserve backing freely. At most two of the three can be achieved; the third must be sacrificed.
Regulatory frameworks will place issuers at different points on this triangle. Each jurisdiction's choices about reserves, redemption rights, and central bank access produce a different combination. The trilemma clarifies why regulatory divergence matters: prioritizing convertibility and par stability inherently constrains issuer autonomy, while preserving autonomy means accepting either restricted redemption access or greater risk of depegging. There is no cost-free position. Every model involves a trade-off, and the question for policymakers is which trade-off best serves financial stability.
Risks in the Issuing Jurisdiction
Let me now turn to the macro-financial risks that stablecoins pose. I want to draw a clear distinction between the risks that arise in the jurisdictions where stablecoins are issued and the risks that emerge in the jurisdictions where they circulate. I will address the issuing side first, and then turn to recipient countries.
For the stablecoin issuers themselves, the risks include credit, market, liquidity, and operational risk in the management of reserves. Even when reserves consist entirely of high-quality liquid assets, par convertibility still depends on the operational capacity to meet redemptions and on the liquidity of underlying markets at the moment of need. As I mentioned, most retail holders currently cannot redeem directly. Holders who fall below the minimum threshold must sell on secondary markets, where prices can deviate from par, creating first-mover advantages during stress.
As stablecoins grow, so do the risks in the originating jurisdiction. Concentration of stablecoin reserves in a small number of banks creates two-way contagion: a run on the stablecoin stresses the banks, and trouble at the banks stresses the stablecoin. Silvergate Bank's crypto deposits reached 98 percent of total deposits at the end of 2021, which is the kind of concentration that creates fragility on both sides. If stablecoins grow to the range of 2 to 3.7 trillion US dollars that some project, fire sales during runs could have spillovers to the functioning of Treasury bill and repo markets, both critical for monetary policy transmission. Stablecoins currently hold approximately 2 percent of outstanding US Treasury bills. Ahmed and Aldasoro (2025) suggest that a 3.5 billion US dollar change in stablecoin issuance could move T-bill yields by 2 to 8 basis points. At larger scale, the structure and risk profile of sovereign debt itself would change as stablecoins become major holders of government securities.
Regulation: Progress, Divergence, and Design Choices
The regulatory response is advancing on multiple fronts. Following the Financial Stability Board (FSB) and International Organization of Securities Commissions (IOSCO) finalization of global recommendations in 2023, three major frameworks are now taking shape: the European Union's Markets in Crypto-Assets Regulation, in force since June 2024 with transitional provisions through July 2026; the UK's draft stablecoin regulation, currently in public consultation; and the US GENIUS Act, which is in the rulemaking phase. Thematic reviews by the FSB and IOSCO suggest that critical gaps remain in risk management, capital buffers, and recovery and resolution planning.
The three frameworks share common elements. Direct remuneration by the issuer to holders is prohibited in all three andeach token must be backed one-to-one by corresponding reserves. But significant divergences remain, creating clear arbitrage risks.
On reserve composition, the European Union requires a minimum of 30 percent in credit institution deposits, rising to 60 percent for significant stablecoins. The United Kingdom proposes 5 percent in commercial bank deposits and 40 percent at the central bank for systemic issuers. The United States applies concentration limits but does not mandate a specific deposit minimum.
Differences also emerge in redemption rules The EU prohibits fees except under a recovery plan and requires redemption at par with no delay. The UK permits cost-linked fees and proposes next-business-day settlement, with same-day settlement for systemic issuers. The US permits fees with disclosure, and early OCC proposals suggest settlement at T+2, extending to T+7 under stress.
These divergences become particularly consequential for stablecoins issued in multiple jurisdictions by the same or affiliated entities. Where one jurisdiction requires faster, cheaper, or easier redemption, holders in that jurisdiction have an incentive to redeem first during stress. Reserve rebalancing across co-issuers may not keep pace with redemption pressures. This multi-issuance risk is a concrete financial stability concern that international coordination has not yet resolved.
The impact of stablecoins on banking systems is similarly not predetermined. It depends on design characteristics, and the choices are best understood as a continuum rather than discrete categories. Five characteristics are decisive. First, the composition of reserve assets determines the credit and liquidity risk profile. Second, whether holders receive remuneration, currently prohibited in all three major jurisdictions but increasingly blurred by indirect incentives through intermediaries, affects whether the instrument functions as a payment device or an investment product. Third, whether the issuer holds reserves directly at the central bank determines the quality of the backstop and the proximity to public money. Fourth, whether the issuer has access to central bank liquidity facilities in stress, an area where the UK is the most advanced in its thinking, determines the system's resilience under pressure. And the fifth, redemption rules - including timelines, fees, minimums, and suspension conditions -determine the holders' practical ability to convert at par.
These choices map onto a broader spectrum between CBDC and privately issued stablecoins. At one end, a full-service model in which all assets are held as central bank reserves, closest to a synthetic CBDC, with the ultimate backstop entirely on the public balance sheet. At the other, a self-service model with no central bank reserve access, where par convertibility depends entirely on private reserve management and market confidence. The Bank of England's proposal for systemic stablecoins sits in between, a partial-service model with at least 40 percent of reserves at the central bank and access to a standing backstop lending facility. The US is closer to the self-service end, with some elements of the light-service model as the Federal Reserve explores payment accounts for eligible institutions.
Moving along this spectrum changes who bears the risk, who provides the liquidity backstop, and how par convertibility is assured. Preserving clarity about the nature of the settlement asset and the location of the backstop is essential for financial stability, regardless of where on the spectrum a jurisdiction lands.
The View from Recipient Countries
I want to turn now to a dimension of the stablecoin discussion that receives less attention in the advanced economy policy debate but is of first-order importance for many of the IMF member countries. What I have discussed so far concerns the design, regulation, and risks of stablecoins in the jurisdictions where they are issued. But stablecoins circulate globally, and their macro-financial impact is borneon the economies where they are used, not where they are created.
The benefits for recipient countries could be material. Stablecoins may reduce the cost and increase the speed of cross-border payments and remittances. They can increase competition, support financial inclusion, and, as integration with artificial intelligence advances, enable new forms of programmable payment. These benefits are being reinforced by the nascent regulatory frameworks I described. But issuer-level regulation, however well designed, does not address the macro-financial risks that fall on recipient economies. Three channels are particularly important.
The first is fragmentation. Stablecoins do not interoperate the way bank deposits do within a national payment system. Different reserve assets, different regulatory standards, and different technical infrastructure create a fragmented landscape in which gaps and the potential for arbitrage are structural, not transitional.
The second is contagion. Runs on stablecoins remain possible even when reserves consist of high-quality assets. Treasury bill prices fluctuate, and forced liquidation during stress could transmit price and capital flow volatility to economies that had no role in creating the risk.
The third, and most consequential, is currency substitution and more broadly, erosion of the macroeconomic toolset. Where inflation is high, the exchange rate is volatile, or institutional credibility is weak, foreign-currency stablecoins can displace local currency in both transactions and savings. This undermines monetary policy transmission, weakens financial stability, and erodes seigniorage income. Capital flow volatility compounds all three channels. Stablecoins increase de facto capital mobility, making capital flow management measures harder to enforce. nyPeer-to-peer transfers through unhosted wallets fall entirely outside the regulatory perimeter.
The empirical evidence, drawn from the most recent Global Financial Stability Report, confirms that these concerns are not hypothetical. Stablecoin flows have surged in recent years, and emerging market economies are increasingly on the receiving end. Stablecoins have overtaken unbacked crypto assets as the dominant vehicle for cross-border crypto activity. This shift is significant: these flows are functionally dollar transfers, not speculative crypto trading. Net stablecoin flows into emerging markets remain small in absolute terms but have been accelerating since 2022. The trajectory matters more than the current level, because the infrastructure for rapid scaling is already in place.
When measured relative to GDP, the picture is more striking. For the most exposed emerging economies, stablecoin inflows are reaching levels that are increasingly significant relative to the domestic financial system. These are economies with thinner foreign exchange markets, less liquid domestic capital markets, and in many cases managed exchange rate regimes, where even modest capital flows can move prices and complicate monetary policy. Stablecoins are originated in advanced economies but the macro-financial impact falls disproportionately on emerging markets that have no voice in the design of these instruments or their regulatory frameworks.
Quantitative analysis shows that stablecoin adoption in emerging markets is not purely a crypto-native phenomenon. It is, instead, driven by the same macroeconomic and structural factors that have historically driven traditional dollarization: foreign exchange volatility, inflation, weak institutional frameworks, and poor policy credibility. Countries with volatile currencies, high inflation, and fragile policy environments are precisely those most exposed. Dollar stablecoins are, in effect, the digital equivalent of dollarization, but with lower barriers to entry.
The Eurodollar analogy is instructive but limited. Eurodollars circulated among institutions within wholesale infrastructure that was gradually brought under central bank oversight over the course of decades. Stablecoins extend the same logic to retail users on networks with far less supervision. The stablecoin exchange rate in local currency can function as a new type of parallel exchange rate market when there is a large shadow price for accessing dollars outside the regulated market. .
Policy Options for Recipient Countries
Recipient countries face a difficult set of choices. The first line of defense is strengthening macroeconomic fundamentals: credible monetary policy, sustainable fiscal positions, and well-functioning domestic payment systems. Where these are sound, demand for foreign-currency stablecoins as a substitute for local money is lower. But fundamentals alone are not sufficient, because network effects can lock in currency substitution even after the underlying vulnerabilities have been addressed.
Stablecoins increase de facto capital mobility, which intensifies a well-known tension for countries that rely on capital flow management measures to maintain managed exchange rates alongside monetary autonomy. The optionscan be stark: accept greater exchange rate flexibility, accept reduced monetary autonomy, or find new ways to bring stablecoin flows within capital flow management frameworks. None of these is straightforward. Other tools are available, including foreign exchange intervention, providing access to foreign-denominated assets through regulated channels rather than attempting outright bans, and setting limits on stablecoin usage where warranted.
But none is effective without data. Reporting requirements must therefore be embedded in laws and regulation. Without visibility into who holds stablecoins, in what amounts, and for what purposes, the other policy tools cannot be calibrated.
And here authorities face a very relevant constraint. The data infrastructure to support effective surveillance and policy calibration of stablecoin policies is still being constructed. Visibility on the residence and sector of stablecoin holders is limited. Off-chain activity, including transfers within exchanges, over-the-counter desks, and custodial wallets, never appears on-chain.
Policy implementation is further constrained by weak data-reporting mandates in many jurisdictions, limited institutional capacity especially in emerging and developing economies, and the structuraldifficulty of cross-border coordination when issuers, holders, and reserves are located in different jurisdictions.
To help address this, the IMF is working with international partners under the G20 Data Gaps Initiative to establish a test data collection, developing harmonized data templates across multiple data sources to build a more reliable statistical picture.
Concluding Remarks
Let me close with three observations.
First, tokenization is a structural transformation of the financial architecture. It shifts risk from institutions to infrastructure and requires that we govern code, data, and settlement with the same rigor we apply to the institutions that have traditionally been at the center of the financial system. The policy window to shape this architecture is open, but it will not remain so indefinitely.
Second, stablecoins are the most immediate policy frontier. Regulatory divergence across jurisdictions creates arbitrage risks. Policy choices about reserve assets, central bank access, backstops, and redemption rules will determine the shape of the future monetary system.
Third, the macro-financial impact of stablecoins are likely to be felt most acutely in recipient countries. Dollar stablecoins can facilitate the digital equivalent of dollarization, driven by the same macroeconomic vulnerabilities but operating with lower barriers to entry and less supervision. Addressing this requires strengthening fundamentals, but it also requires regulation, data, and international coordination.
The IMF's role is to help member countries navigate these choices: aligning domestic frameworks with global standards, closing data gaps, and supporting emerging and developing economies as they confront stablecoin-driven capital flow pressures. We are committed to this work, both analytically and operationally, and through deep engagement with our membership.
Thank you. I look forward to your questions.
