When digital assets stop being an asset class
September 2026
Institutional adoption is usually measured by how much capital moves into dedicated crypto funds. That may be the wrong measure. David Lloyd, founder and CEO at CV5 Capital, considers how the boundary that created the digital asset category is becoming progressively harder to locate
Image: CV5 Capital
The clearest evidence that digital assets have been absorbed into institutional finance may eventually be that institutions stop discussing digital assets as a separate subject.
For most of the past decade the industry has organised itself around the opposite premise: that digital assets constitute a self-contained category requiring dedicated funds, dedicated venues, dedicated custodians, dedicated managers, and a dedicated conference circuit. That premise has been commercially productive and, for a period, operationally necessary. It may also prove self-liquidating. The more successfully digital assets are institutionalised, the harder it becomes to identify where a crypto portfolio ends and an ordinary institutional portfolio begins.
The question worth putting to allocators is therefore not whether traditional institutions will become crypto investors. Many already are, in modest size. It is whether the classification itself survives contact with mature market infrastructure. In 10 years, ‘digital asset manager’ may read as oddly as ‘electronic securities manager’ reads today. Electronic execution was once a strategy. Now it is simply the market.
Exposure without participation
The mechanics of institutional exposure have changed considerably more than the underlying assets have, and three changes did most of the work.
The US Securities and Exchange Commission (SEC) permitted in-kind creations and redemptions for crypto exchange traded products (ETPs) in July 2025 and approved generic listing standards for commodity-based trust shares that September, converting product listing from bespoke approval into a standards-based process. The accounting and supervisory obstacles to US bank custody were removed through the rescission of Staff Accounting Bulletin 121 in January 2025 and Office of the Comptroller of the Currency interpretive letters in March and May 2025 confirming that national banks may custody and execute crypto assets and may use sub-custodians. And CME Group shifted its cryptocurrency futures and options to continuous trading on 29 May 2026, matching the calendar of the underlying market.
The distinction that matters here is between economic exposure and technological participation. An allocator buying a regulated ETP acquires the first without the second. There are no wallets, no private keys and no chain: the exposure is sized inside an existing risk budget, held with an existing custodian, and reported through existing systems.
The 2026 EY-Parthenon and Coinbase survey of more than 350 institutional investors found 81 per cent preferring spot exposure through a registered vehicle and 66 per cent already holding it that way, against 36 per cent holding spot crypto directly. The survey is co-published by a commercial beneficiary of adoption and samples institutions already engaged with the asset, so it describes participants rather than the market.
Read that way it is still informative: participants strongly prefer conventional wrappers.
The wrapper standardises the plumbing. It does not standardise the risk. BlackRock’s iShares Bitcoin Trust reported a one-year net asset value return of minus 45.62 per cent for the 12 months to 30 June 2026, against a since-inception return of 12.18 per cent from January 2024. Conventional access, unconventional volatility. Institutionalisation has not domesticated the asset.
Stablecoins are settlement technology, not an asset class
The institutional stablecoin conversation has little to do with retail payments. It concerns settlement, collateral movement, treasury management, out-of-hours liquidity, movement of value between custodians, venues and counterparties, and increasingly fund subscriptions and redemptions.
Scale should be stated carefully. The Bank for International Settlements put stablecoin market capitalisation at approximately US$320 billion as at end-May 2026, and estimated annual transaction volume of US$28 trillion in 2025, which it characterised as less than three business weeks of settlement volumes in the largest US wholesale payment systems.
That is the correct frame: a fast-growing settlement technology still operating at a small fraction of the scale of existing wholesale infrastructure.
Four forms of tokenised money now compete for the same institutional workflows: fiat-backed stablecoins; bank-issued tokens and tokenised commercial bank deposits; tokenised money market and Treasury products; and, in wholesale contexts, tokenised central bank money. The BIS Annual Economic Report of June 2026 concluded that current stablecoin designs “fall short on foundational properties of money and threaten financial integrity”, and argued for bringing tokenisation into the existing two-tier system rather than around it.
The Bank of England (BoE), in its June 2026 policy statement on sterling-denominated systemic stablecoins, took a more plural view of what may count as money, but still capped per-stablecoin issuance at £40 billion on a temporary basis.
The useful question is not whether stablecoins become an asset class; most institutional users do not want them to be. They want an instrument that holds par, settles at weekends, and carries no duration and minimal credit risk. That is a description of infrastructure. The contest is over which form of tokenised money institutions settle in, not which token appreciates.
Tokenisation runs in the opposite direction
The original proposition was that investors would move from traditional finance into crypto. Tokenisation increasingly involves the reverse: conventional financial assets moving onto distributed ledger infrastructure.
J.P. Morgan Asset Management launched a tokenised money market fund (MMF) on public Ethereum in December 2025, restricted to qualified purchasers and accredited investors and investing only in US Treasuries and Treasury-backed repo. The BNY and Goldman Sachs tokenised money market fund solution, launched in July 2025, is architecturally instructive for a different reason: BNY retains the official books and records, and the tokens mirror the fund shares. On the collateral side, the US Commodity Futures Trading Commission (CFTC) issued an advisory in December 2025 encouraging tokenised collateral in liquid underlyings with established haircuts, naming Treasuries and MMF shares, subject to legal enforceability and a perfected security interest.
The conceptual point is routinely lost. A tokenised Treasury bill is a Treasury bill. The economic asset, its issuer, its credit and its duration are unchanged.
What has changed is issuance, the ownership record, transfer, settlement, and collateral mobility. Tokenisation is a change to market infrastructure, not the creation of an asset class. Treating a tokenised money market fund as a ‘digital asset’ for allocation purposes confuses the register with the instrument and that is a confusion the existing category actively encourages.
Scale again warrants discipline. The European Central Bank (ECB) estimated tokenised assets on public blockchains at approximately €38 billion in February 2026, up from €7.4 billion at the start of 2024, against global assets estimated at approximately €241 trillion at end-2025. The direction is clear; the penetration is not yet material.
Institutional engagement with DeFi, without the ideology
Institutions are not going to migrate portfolios wholesale into permissionless protocols. What is emerging is narrower and more interesting: decentralised protocols used as an infrastructure layer, accessed through controlled wallets, whitelisted counterparties and regulated entities.
Societe Generale’s digital asset subsidiary deployed its euro and dollar stablecoins onto public lending and exchange protocols in September 2025, with transfers restricted to permitted transferees. That is the shape of it: public infrastructure, permissioning applied at the token and counterparty level, a regulated issuer at the centre.
The obstacles are operational rather than conceptual. Wallet governance and transaction authorisation, smart contract and protocol risk, counterparty identification, sanctions screening against pseudonymous addresses, independent valuation, administrator and depositary visibility, segregation, and the location of the regulatory perimeter all have to be resolved before a fiduciary can act. The Bybit compromise of February 2025, the largest theft on record, is instructive: the multi-signature quorum did not fail, the transaction interface presented to the signers did. Chainalysis put total stolen crypto at US$3.4 billion in 2025. Institutional engagement with these protocols is not primarily a strategy question. It is an operating architecture question.
The hybrid portfolio, and the specialist manager
Consider a single alternative investment fund that holds listed equities through a conventional prime broker, trades regulated futures, holds bitcoin with an institutional custodian, settles certain flows in stablecoins, owns tokenised Treasury or money market instruments, takes exposure to tokenised credit, and moves collateral onchain.
The point is not that every fund will look like this, but that asset classification, custody technology and settlement infrastructure are separating from one another. Traditional economic exposures can be delivered through digital infrastructure, and digital exposures through entirely conventional financial infrastructure. Once those three variables move independently, a binary traditional-versus-digital classification stops carrying useful information.
Specialist managers are not displaced by this. Token selection, basis and market-neutral strategies, digital asset derivatives, onchain strategies, and specialised execution require expertise generalist managers have no reason to build. But there is a difference between a specialist strategy and a specialist asset-class silo. A convertible arbitrage manager is a strategy specialist; convertibles are not a separate financial sector. Much of what is currently bespoke digital asset infrastructure will become standard infrastructure available to mainstream funds, at which point the technology stops defining the category.
The allocation data is consistent with a market at exactly this stage. State Street’s 2025 study of 324 asset managers and asset owners found more than half with less than 1 per cent exposure to digital assets, while 60 per cent planned to increase beyond 2 per cent within a year. The AIMA and PwC report of November 2025, covering 122 institutional investors and managers with combined assets of US$982 billion, found 55 per cent of traditional hedge funds with some exposure, up from 47 per cent in 2024, and most allocations below 2 per cent of assets. Participation is broad, allocation is small, and the infrastructure question is arriving faster than the money.
Where this thesis could fail
Convergence is not inevitable, and the case against it is stronger than its advocates usually concede.
Regulation remains fragmented. The Financial Stability Board’s thematic review of October 2025 found implementation of its global crypto framework “incomplete, uneven, and inconsistent”, creating, in the words of the review’s chair, “opportunities for regulatory arbitrage”. In the US, most of the 2025 and 2026 liberalisation arrived through staff guidance, interpretive letters, and supervisory policy rather than adopted rules, and market structure legislation was still not enacted as at early September 2026. Guidance can be withdrawn considerably faster than it was issued.
The infrastructure case is also unproven at scale. IOSCO’s final report on tokenisation in November 2025 found the promised benefits, particularly secondary market liquidity, not clearly evidenced in the use cases examined, and identified the absence of cross-chain interoperability and of credible settlement assets as binding constraints. The European Union’s DLT Pilot Regime, the flagship attempt to create regulated tokenised market infrastructure, had authorised three infrastructures between March 2023 and May 2025, recorded a debt issuance of €401 at one venue, and seen no secondary market transactions at all.
There is a structural argument too. The BIS Head of Research has argued that the economics of public permissionless blockchains work against consolidation: capacity constraints are necessary to compensate validators, congestion prices users onto alternative chains, and “instead of convergence toward a single medium of exchange, the result is fragmentation”. If that is right, the destination is not one integrated market but several partially connected ones, which for liquidity and collateral mobility is worse than the status quo.
Legal certainty lags further still. The Financial Markets Law Committee observed in July 2026 that the UK settlement finality regime derives from a directive not drafted with distributed ledger technology (DLT) in mind and not technology neutral. Insolvency treatment turns on old-fashioned questions: the Celsius ruling of January 2023 held that customer assets in the Earn programme formed part of the estate, leaving accountholders as unsecured creditors, on the basis of the contractual title language and the absence of segregation. The technology was not the determinant. Prudential treatment is unresolved, with the Basel Committee running an expedited targeted review of its cryptoasset standard as at February 2026. And fiduciary duties are unchanged by any of it.
Convergence could therefore stall, or proceed unevenly by asset class and by jurisdiction, for a considerable period. One point digital asset advocacy tends to avoid: for many processes, existing infrastructure remains superior. Central bank money is already the safest settlement asset, and established depositories already deliver finality with a legal certainty that tokenised alternatives cannot yet match.
Infrastructure becomes the competitive question
If the thesis holds even partially, the institutional question ceases to be ‘what digital assets can this fund buy?’ and becomes ‘can the operating architecture support these assets to institutional standards?’
That is a question about governance, custody arrangements, counterparty controls, wallet permissions and transaction authorisation, segregation, valuation policy and net asset value production, anti-money laundering (AML) and sanctions controls, independent oversight, liquidity management, operational resilience, reconciliation, auditability, and reporting. Convergence does not reduce the importance of any of these. It increases it.
A portfolio spanning conventional and blockchain infrastructure has more counterparties, more settlement conventions, more reconciliation points, and more failure modes than one confined to either, and the burden rises fastest during the transition, when onchain and offchain versions of the same exposure coexist.
The long-term measure of institutional adoption is therefore not the capital sitting in dedicated crypto funds. It is the point at which a portfolio can move without friction between conventional securities, digital assets, and tokenised instruments, using whichever custody, execution, and settlement architecture suits the transaction, under a single control framework.
None of this implies the disappearance of bitcoin, of blockchain infrastructure or of specialist digital asset managers. It concerns the convergence of market architecture, and the declining usefulness of a classification built on a distinction the infrastructure is quietly erasing. The winners may not be those who persuade institutions to become crypto investors. They may be those who make the distinction irrelevant.
For most of the past decade the industry has organised itself around the opposite premise: that digital assets constitute a self-contained category requiring dedicated funds, dedicated venues, dedicated custodians, dedicated managers, and a dedicated conference circuit. That premise has been commercially productive and, for a period, operationally necessary. It may also prove self-liquidating. The more successfully digital assets are institutionalised, the harder it becomes to identify where a crypto portfolio ends and an ordinary institutional portfolio begins.
The question worth putting to allocators is therefore not whether traditional institutions will become crypto investors. Many already are, in modest size. It is whether the classification itself survives contact with mature market infrastructure. In 10 years, ‘digital asset manager’ may read as oddly as ‘electronic securities manager’ reads today. Electronic execution was once a strategy. Now it is simply the market.
Exposure without participation
The mechanics of institutional exposure have changed considerably more than the underlying assets have, and three changes did most of the work.
The US Securities and Exchange Commission (SEC) permitted in-kind creations and redemptions for crypto exchange traded products (ETPs) in July 2025 and approved generic listing standards for commodity-based trust shares that September, converting product listing from bespoke approval into a standards-based process. The accounting and supervisory obstacles to US bank custody were removed through the rescission of Staff Accounting Bulletin 121 in January 2025 and Office of the Comptroller of the Currency interpretive letters in March and May 2025 confirming that national banks may custody and execute crypto assets and may use sub-custodians. And CME Group shifted its cryptocurrency futures and options to continuous trading on 29 May 2026, matching the calendar of the underlying market.
The distinction that matters here is between economic exposure and technological participation. An allocator buying a regulated ETP acquires the first without the second. There are no wallets, no private keys and no chain: the exposure is sized inside an existing risk budget, held with an existing custodian, and reported through existing systems.
The 2026 EY-Parthenon and Coinbase survey of more than 350 institutional investors found 81 per cent preferring spot exposure through a registered vehicle and 66 per cent already holding it that way, against 36 per cent holding spot crypto directly. The survey is co-published by a commercial beneficiary of adoption and samples institutions already engaged with the asset, so it describes participants rather than the market.
Read that way it is still informative: participants strongly prefer conventional wrappers.
The wrapper standardises the plumbing. It does not standardise the risk. BlackRock’s iShares Bitcoin Trust reported a one-year net asset value return of minus 45.62 per cent for the 12 months to 30 June 2026, against a since-inception return of 12.18 per cent from January 2024. Conventional access, unconventional volatility. Institutionalisation has not domesticated the asset.
Stablecoins are settlement technology, not an asset class
The institutional stablecoin conversation has little to do with retail payments. It concerns settlement, collateral movement, treasury management, out-of-hours liquidity, movement of value between custodians, venues and counterparties, and increasingly fund subscriptions and redemptions.
Scale should be stated carefully. The Bank for International Settlements put stablecoin market capitalisation at approximately US$320 billion as at end-May 2026, and estimated annual transaction volume of US$28 trillion in 2025, which it characterised as less than three business weeks of settlement volumes in the largest US wholesale payment systems.
That is the correct frame: a fast-growing settlement technology still operating at a small fraction of the scale of existing wholesale infrastructure.
Four forms of tokenised money now compete for the same institutional workflows: fiat-backed stablecoins; bank-issued tokens and tokenised commercial bank deposits; tokenised money market and Treasury products; and, in wholesale contexts, tokenised central bank money. The BIS Annual Economic Report of June 2026 concluded that current stablecoin designs “fall short on foundational properties of money and threaten financial integrity”, and argued for bringing tokenisation into the existing two-tier system rather than around it.
The Bank of England (BoE), in its June 2026 policy statement on sterling-denominated systemic stablecoins, took a more plural view of what may count as money, but still capped per-stablecoin issuance at £40 billion on a temporary basis.
The useful question is not whether stablecoins become an asset class; most institutional users do not want them to be. They want an instrument that holds par, settles at weekends, and carries no duration and minimal credit risk. That is a description of infrastructure. The contest is over which form of tokenised money institutions settle in, not which token appreciates.
Tokenisation runs in the opposite direction
The original proposition was that investors would move from traditional finance into crypto. Tokenisation increasingly involves the reverse: conventional financial assets moving onto distributed ledger infrastructure.
J.P. Morgan Asset Management launched a tokenised money market fund (MMF) on public Ethereum in December 2025, restricted to qualified purchasers and accredited investors and investing only in US Treasuries and Treasury-backed repo. The BNY and Goldman Sachs tokenised money market fund solution, launched in July 2025, is architecturally instructive for a different reason: BNY retains the official books and records, and the tokens mirror the fund shares. On the collateral side, the US Commodity Futures Trading Commission (CFTC) issued an advisory in December 2025 encouraging tokenised collateral in liquid underlyings with established haircuts, naming Treasuries and MMF shares, subject to legal enforceability and a perfected security interest.
The conceptual point is routinely lost. A tokenised Treasury bill is a Treasury bill. The economic asset, its issuer, its credit and its duration are unchanged.
What has changed is issuance, the ownership record, transfer, settlement, and collateral mobility. Tokenisation is a change to market infrastructure, not the creation of an asset class. Treating a tokenised money market fund as a ‘digital asset’ for allocation purposes confuses the register with the instrument and that is a confusion the existing category actively encourages.
Scale again warrants discipline. The European Central Bank (ECB) estimated tokenised assets on public blockchains at approximately €38 billion in February 2026, up from €7.4 billion at the start of 2024, against global assets estimated at approximately €241 trillion at end-2025. The direction is clear; the penetration is not yet material.
Institutional engagement with DeFi, without the ideology
Institutions are not going to migrate portfolios wholesale into permissionless protocols. What is emerging is narrower and more interesting: decentralised protocols used as an infrastructure layer, accessed through controlled wallets, whitelisted counterparties and regulated entities.
Societe Generale’s digital asset subsidiary deployed its euro and dollar stablecoins onto public lending and exchange protocols in September 2025, with transfers restricted to permitted transferees. That is the shape of it: public infrastructure, permissioning applied at the token and counterparty level, a regulated issuer at the centre.
The obstacles are operational rather than conceptual. Wallet governance and transaction authorisation, smart contract and protocol risk, counterparty identification, sanctions screening against pseudonymous addresses, independent valuation, administrator and depositary visibility, segregation, and the location of the regulatory perimeter all have to be resolved before a fiduciary can act. The Bybit compromise of February 2025, the largest theft on record, is instructive: the multi-signature quorum did not fail, the transaction interface presented to the signers did. Chainalysis put total stolen crypto at US$3.4 billion in 2025. Institutional engagement with these protocols is not primarily a strategy question. It is an operating architecture question.
The hybrid portfolio, and the specialist manager
Consider a single alternative investment fund that holds listed equities through a conventional prime broker, trades regulated futures, holds bitcoin with an institutional custodian, settles certain flows in stablecoins, owns tokenised Treasury or money market instruments, takes exposure to tokenised credit, and moves collateral onchain.
The point is not that every fund will look like this, but that asset classification, custody technology and settlement infrastructure are separating from one another. Traditional economic exposures can be delivered through digital infrastructure, and digital exposures through entirely conventional financial infrastructure. Once those three variables move independently, a binary traditional-versus-digital classification stops carrying useful information.
Specialist managers are not displaced by this. Token selection, basis and market-neutral strategies, digital asset derivatives, onchain strategies, and specialised execution require expertise generalist managers have no reason to build. But there is a difference between a specialist strategy and a specialist asset-class silo. A convertible arbitrage manager is a strategy specialist; convertibles are not a separate financial sector. Much of what is currently bespoke digital asset infrastructure will become standard infrastructure available to mainstream funds, at which point the technology stops defining the category.
The allocation data is consistent with a market at exactly this stage. State Street’s 2025 study of 324 asset managers and asset owners found more than half with less than 1 per cent exposure to digital assets, while 60 per cent planned to increase beyond 2 per cent within a year. The AIMA and PwC report of November 2025, covering 122 institutional investors and managers with combined assets of US$982 billion, found 55 per cent of traditional hedge funds with some exposure, up from 47 per cent in 2024, and most allocations below 2 per cent of assets. Participation is broad, allocation is small, and the infrastructure question is arriving faster than the money.
Where this thesis could fail
Convergence is not inevitable, and the case against it is stronger than its advocates usually concede.
Regulation remains fragmented. The Financial Stability Board’s thematic review of October 2025 found implementation of its global crypto framework “incomplete, uneven, and inconsistent”, creating, in the words of the review’s chair, “opportunities for regulatory arbitrage”. In the US, most of the 2025 and 2026 liberalisation arrived through staff guidance, interpretive letters, and supervisory policy rather than adopted rules, and market structure legislation was still not enacted as at early September 2026. Guidance can be withdrawn considerably faster than it was issued.
The infrastructure case is also unproven at scale. IOSCO’s final report on tokenisation in November 2025 found the promised benefits, particularly secondary market liquidity, not clearly evidenced in the use cases examined, and identified the absence of cross-chain interoperability and of credible settlement assets as binding constraints. The European Union’s DLT Pilot Regime, the flagship attempt to create regulated tokenised market infrastructure, had authorised three infrastructures between March 2023 and May 2025, recorded a debt issuance of €401 at one venue, and seen no secondary market transactions at all.
There is a structural argument too. The BIS Head of Research has argued that the economics of public permissionless blockchains work against consolidation: capacity constraints are necessary to compensate validators, congestion prices users onto alternative chains, and “instead of convergence toward a single medium of exchange, the result is fragmentation”. If that is right, the destination is not one integrated market but several partially connected ones, which for liquidity and collateral mobility is worse than the status quo.
Legal certainty lags further still. The Financial Markets Law Committee observed in July 2026 that the UK settlement finality regime derives from a directive not drafted with distributed ledger technology (DLT) in mind and not technology neutral. Insolvency treatment turns on old-fashioned questions: the Celsius ruling of January 2023 held that customer assets in the Earn programme formed part of the estate, leaving accountholders as unsecured creditors, on the basis of the contractual title language and the absence of segregation. The technology was not the determinant. Prudential treatment is unresolved, with the Basel Committee running an expedited targeted review of its cryptoasset standard as at February 2026. And fiduciary duties are unchanged by any of it.
Convergence could therefore stall, or proceed unevenly by asset class and by jurisdiction, for a considerable period. One point digital asset advocacy tends to avoid: for many processes, existing infrastructure remains superior. Central bank money is already the safest settlement asset, and established depositories already deliver finality with a legal certainty that tokenised alternatives cannot yet match.
Infrastructure becomes the competitive question
If the thesis holds even partially, the institutional question ceases to be ‘what digital assets can this fund buy?’ and becomes ‘can the operating architecture support these assets to institutional standards?’
That is a question about governance, custody arrangements, counterparty controls, wallet permissions and transaction authorisation, segregation, valuation policy and net asset value production, anti-money laundering (AML) and sanctions controls, independent oversight, liquidity management, operational resilience, reconciliation, auditability, and reporting. Convergence does not reduce the importance of any of these. It increases it.
A portfolio spanning conventional and blockchain infrastructure has more counterparties, more settlement conventions, more reconciliation points, and more failure modes than one confined to either, and the burden rises fastest during the transition, when onchain and offchain versions of the same exposure coexist.
The long-term measure of institutional adoption is therefore not the capital sitting in dedicated crypto funds. It is the point at which a portfolio can move without friction between conventional securities, digital assets, and tokenised instruments, using whichever custody, execution, and settlement architecture suits the transaction, under a single control framework.
None of this implies the disappearance of bitcoin, of blockchain infrastructure or of specialist digital asset managers. It concerns the convergence of market architecture, and the declining usefulness of a classification built on a distinction the infrastructure is quietly erasing. The winners may not be those who persuade institutions to become crypto investors. They may be those who make the distinction irrelevant.
NO FEE, NO RISK
100% ON RETURNS If you invest in only one digital assets news source this year, make sure it is your free subscription to The Digital Assets Edge
100% ON RETURNS If you invest in only one digital assets news source this year, make sure it is your free subscription to The Digital Assets Edge
