CBDCs and tokenised money
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
A CBDC is a digital payment instrument, denominated in the national unit of account,
which is a direct liability of the central bank.28 Much attention has recently focused on retail
CBDCs that are accessible by households and businesses (discussed below). Yet wholesale
CBDCs also offer new functions for payment and settlement, and to a much wider range of
intermediaries than domestic commercial banks. They could unlock significant private sector
innovation across a range of financial services.
Wholesale CBDCs can allow intermediaries to access new capabilities that are not
provided by the reserves held by commercial banks with the central bank. These are particularly
relevant in permissioned DLT networks, where a decentralised network of trusted participants
accesses a shared ledger. As discussed below, decentralised governance is a useful feature of
multi-CBDC systems involving multiple central banks and currencies. Yet the functions could in
principle be offered in more centralised payment systems. Key are self-executing smart contracts
that let participants make their transactions programmable. Transactions thus settle only when
certain pre-specified conditions are met. In security trading, such automation can allow payment
vs payment (PvP) and delivery vs payment (DvP) mechanisms, meaning that payments and
delivery of a security are made only all together or not at all. Such atomic settlement can
significantly speed up settlement and mitigate counterparty risk.29
One benefit of wholesale CBDCs is that they could be available to a much wider range of
intermediaries than just domestic commercial banks. Allowing non-bank PSPs to transact in
CBDC could make for much greater competition and vibrancy. New protocols built on wholesale
CBDCs could be open source, making the source code freely available for a community of
developers to develop and scrutinise. This feature would allow for libraries of protocols that can
be used to combine functions, thus facilitating the composability of different functions and
enabling new services to be built on top of the programmability function of CBDCs.
By construction, wholesale CBDCs would allow for finality in payments. The mechanics
of how finality is attained in permissioned DLT platforms are described in more detail in Box C,
but their essence can be explained through the simple analogy with a physical banknote. The
recipient of a physical banknote wants to be assured that the note is genuine, not counterfeit.
Ensuring that payment is in genuine money in a digital system is accomplished by proving the
origin or "provenance" of the money transferred. Crypto proves its provenance by publicly
posting the full history of all transactions by everyone. When real names are used, such public
posting would violate privacy and would be unsuitable as a payment system. This is where
cryptographic techniques such as zero-knowledge proofs (ZKPs) provide a solution. As the name
signifies, "proof" denotes that a statement is true, and "zero-knowledge" means that no additional
information is exposed beyond the validity of the assertion. Cryptographic techniques allow the
payer to prove that the money was obtained from valid past transactions without having to post
the full history of all transactions. Depending on the detailed implementation, a "notary" may be
needed to prevent the same digital token being spent twice; in many cases, the central bank can
play this role. The common theme is that decentralisation can be achieved without the structural
flaws of crypto.
As issuers of the settlement currency, central banks can support the tokenisation of
regulated financial instruments such as retail deposits.30 Tokenised deposits are a digital
representation of commercial bank deposits on a DLT platform. They would represent a claim on
the depositor's commercial bank, just as a regular deposit does, and be convertible into central
bank money (either cash or retail CBDC) at par value. Depositors would be able to convert their
deposits into and out of tokens, and to exchange them for goods, services or other assets.
Tokenised deposits would also be protected by deposit insurance but, unlike traditional deposits,
they would also be programmable and "always on" (24/7), thus lending themselves to broader
uses in retail payments – eg in autonomous ecosystems. This way, they could facilitate
tokenisation of other financial assets, such as stocks or bonds. This functionality could allow for
fractional ownership of assets and for the ability to exchange these on a 24/7 basis. Crucially,
this could be done in a regulated system, with settlements in wholesale CBDC.
Programmable CBDCs could also support machine-to-machine payments in autonomous
ecosystems.31 Autonomous machines and devices increasingly communicate and execute
processes without human intervention through the Internet of Things, a network of connected
devices. Looking ahead, machines may directly purchase goods and services from each other,
and manage their own budget. Their interconnection will increase the need for smart contracts
and programmable money. For example, they may be equipped with wallets, charged with a
certain budget of digital money. Smart contracts may automatically trigger payments as soon as
certain conditions are met, eg the arrival of the goods. This could lead to significant efficiency
gains, for example in the goods logistics sector, where transactions often take several days and
are still predominantly paper-based. The full potential of these technological developments can
be realised only if machine-to-machine transactions are settled instantly, so that any settlement
risk is removed. Existing private sector cryptocurrency projects for the Internet of Things are still
exploratory and suffer from limits to scalability.32 They also raise concerns about the stability
and convertibility of cryptocurrencies used for payments and would require on- and off-ramp
bridges to connect with traditional payment rails. In this respect, the industry could benefit from
CBDCs, which could underpin a decentralised system, eg by enabling regulated financial
institutions to issue programmable money.33
In short, programmability, composability and tokenisation are not the preserve of crypto.
The benefits of atomic settlement and open-source protocols are fully compatible with central
banks being at the core of the validation process. Yet by relying on central bank money,
wholesale CBDCs would benefit from the stability and singleness of the currency that central
banks provide. They would also draw on the accountability of the central bank and of regulated
intermediaries to society. By supporting innovative private sector services, they would facilitate
adaptability so that the system can meet new needs as they arise.
Retail CBDCs and fast payment systems
Retail CBDCs and retail FPS share many similarities. Retail CBDCs make central bank
money available in digital form to households and businesses. Bank and non-bank PSPs provide
retail-facing payment services. The key difference from retail FPS is that, for CBDCs, the
instrument is a legal claim on the central bank. Retail CBDCs are thus sometimes seen as "digital
cash" – another form of central bank money available to the public.34 In retail FPS, many of
which are operated by the central bank, the instrument being exchanged is a claim on private
intermediaries (eg bank deposits or e-money). Nonetheless, both retail CBDCs and retail FPS
build on public data architecture with APIs that ensure secure data exchange and interoperability
between different bank and non-bank PSPs. Both feature high speeds and availability, as
transfers occur in real time or near real time on a (near) 24/7 basis.
These retail payment infrastructures have already shown their mettle in enhancing
efficiency and inclusion in the monetary system. Unlike crypto, which requires high rents and
suffers from congestion and limited scalability, CBDCs and retail FPS allow for network effects
to lead to a virtuous circle of greater use, lower costs and better services. Because of their
explicit mandates, central banks can design systems to meet these goals from the ground up. An
open payment system resting on the interoperability of services offered by competing private
PSPs can challenge rents in concentrated banking sectors and reduce the payments costs for end
users.
Retail FPS have already made impressive progress in lowering costs and supporting
financial inclusion for the unbanked. For example, in just over a year after its launch, the
Brazilian retail FPS Pix is used by two thirds of the adult population – with 50 million users
making a digital payment for the first time. Powered by innovative products and services offered
by over 770 private PSPs, Pix payments have now surpassed credit and debit card transactions
(Graph 10.A). The costs to merchants of accepting person-to-business (P2B) payments average
one tenth of the cost of credit card payments (Graph 10.B). Equally impressive progress in
inclusive, low-cost payments has been made in other economies.35
Retail CBDCs could play a similarly beneficial role as retail FPS, while offering
additional technological capabilities. For example, Project Hamilton – a joint project by the
Federal Reserve Bank of Boston and the Massachusetts Institute of Technology Digital Currency
Initiative – has shown the technical feasibility of a CBDC architecture that can process 1.7
million transactions per second – far more than major card networks or blockchains.36 The
project uses functions inspired by cryptocurrencies, but it does not use DLT. In its next stage,
Project Hamilton aims to create a foundation for more complex functionalities, such as
cryptographic designs for privacy and auditability, programmability and self-custody. The code
for the project is open-source and can be scrutinised by any developer, to maximise knowledge-
sharing and expand the pool of experts contributing to the code base, including central banks,
academia and the private sector.
