HM Treasury Cryptoasset and Stablecoin Consultation Response
21 March 2021
Dear Sir/Madam,
HM Treasury UK regulatory approach to cryptoassets and stablecoins: Consultation and call for Evidence
Please find attached responses to your questions on the UK regulatory approach to cryptoassets and stablecoins.
I am a long-standing, UK-based payments professional - I led Accenture’s European payments practice for many years, and more recently had a senior role at Ripple, a Silicon Valley payments Fintech and I am currently setting up a payments Fintech with a business partner. I am providing these responses in a personal capacity, as a stakeholder with a keen interest in the success of UK payments and adoption of cryptoassets.
The key themes in my responses are:
- Regulation for some stablecoins is a good idea, but the proposed token classification should distinguish between unregulated stable tokens and regulated stable tokens;
- The government should raise its ambition and re-invent regulation with technology; and have a principle of “better technology, better regulatory outcomes”;
- There should be explicit objectives for privacy and protection from surveillance, and for digital payments inclusion, enabling a digital alternative to coins and paper cash;
- Before drafting regulations, regulators should use their sandboxes to test out regulatory ideas;
- We are entering a new world of co-existing multi-ecosystem interoperable, borderless financial services, with parts outside the full control of any single country, requiring a new regulatory framework that goes far beyond incremental change to the existing one;
- In a shared ledger, there is no concept of clearing and settlement; payment regulations based on traditional clearing and settlement and interbank payments are less relevant to stablecoins;
- Algorithmic platforms and shared ledgers outside anyone’s control call for regulation to distinguish between “regulatable” entities and “unregulatable” open software execution.
Should you wish to discuss any in further detail, I am at your disposal,
Yours faithfully
Jeremy Light
Question 1: Do you have views on continuing to use a classification that is broadly consistent with existing guidance issued by UK authorities, supplemented with new categories where needed?
The token classification used by UK authorities, outlined in paragraph 1.12, is a useful one and should be retained, as its gives clarity on whether, and how, tokens are regulated. This is a pre-requisite for confidence in taking solutions to market developed using tokens.
However, HMT needs a clearer definition of tokens that reflect how they can be used. For example, the word token is often used as an option for a CBDC as a bearer asset, in contrast to using a CBDC in identifiable accounts. This distinction itself is unclear as tokens are not necessarily bearer assets. Thus, HMT should differentiate in its classification between bearer tokens and non-bearer tokens to be clear on any different regulatory treatment of the two types.
HMT should also classify tokens on whether they are cryptoassets issued and tracked on a shared ledger (a digital asset), or tokens that represent assets held off-ledger (a claim on an asset). As an example, there is a big difference between a tokenised security where a token on a shared ledger (e.g. used for trading) represents a security held on a different ledger (company share register, CREST etc), and a security token where a company issues its shares as tokens on its own shared ledger. Here, the ledger acts as its record for ownership of its shares, which in turn can be used for trading its shares, shareholder communication, shareholder voting and initiating dividend payments.
HMT is agnostic on the technology underpinning the use of a token, but the classification system should recognise that technology and tokens can be tightly integrated. For example, on the Ethereum ledger the ERC-20 and ERC-721 tokens are actually smart contracts, or self-executing scripts of code, and smart contracts such as these are driving a proliferation of stablecoins and non-fungible tokens.
HMT should also consider a definition and classification for non-fungible tokens (NFTs). These have hit the headlines recently in the context of digital art and collectibles such as digital gaming cards. The use and demand for NFTs can appear puzzling, but in the context of digital rights and licensing they have huge potential when combined with programmability. For example, NFTs can be used to prove ownership and automatically collect royalties and licence fees on behalf of the owner on demand when copyright digital photos are used, digital music is played, videos are downloaded and so on. For HMT purposes, NFTs probably are best classified as unregulated.
2 Do you have views on the proposed new regulated category of ‘stable tokens’?
It makes sense to introduce the ‘stable token’ classification for stablecoins, but there are different types of stablecoins and different ways of collateralising them – for example, fiat-collateralised, commodity- collateralised, crypto-asset collateralised and algorithmically collateralised (e.g. DAI which is a USD stablecoin, collateralised by other non-fiat, non-commodity cryptoassets, using an algorithm to maintain parity with the US dollar).
Given collateralisation is a frequently used regulatory tool or requirement, it makes sense to consider distinguishing between unregulated stable tokens and regulated stable tokens. HMT would need to define what goes into each category, but for example, stablecoins collateralised with unregulated tokens would logically sit in the unregulated stable token category.
3. Do you have views on the government’s proposed objectives and principles for cryptoassets regulation? Do you have views on which should be prioritised, or where there may be tension between them?
The government should be more ambitious in its objectives. Cryptoassets and stablecoins will transform financial services over the next decade, and the government should use this as an opportunity to reinvent regulation, with the objective of “better technology, better regulatory outcomes”. In particular, technology such as stablecoins and cryptoassets should be harnessed to ensure better financial crime outcomes, rather than simply applying existing rules such as AML.
In applying this principle of “better technology, better regulatory outcomes”, HMT should consider putting in place benchmarks of existing regulatory outcomes against which better ones can be targeted. For example, benchmark the % of fraud in total payment flows under today’s systems and set this as a ceiling on fraud using cryptoassets and stablecoins. Other benchmarks could be on the prevalence of money laundering today, system resilience, speed of settlement and so on – these will allow innovators to demonstrate the benefits of their solutions and their contribution to better regulatory outcomes.
Principles and objectives for privacy and surveillance are missing. These are critical to consumer protection as technology becomes more embedded and integrated in society. There is always a suspicion that surveillance is being introduced and privacy compromised with new technology, and payment systems are open to abuse in this area. Ignoring them reinforces this suspicion, and explicit principles and objectives are required.
Given that stablecoins allow payments to be accessible anywhere on the internet, the government should also have an objective for digital payments inclusion and a digital alternative to cash which retains the properties of cash such as inclusion, ubiquity and privacy.
Lastly, there also needs to be a principle to avoid regulatory overreach such as the recent FCA directive to ban retail consumers from buying derivatives in cryptoassets, including exchange traded products (ETPs). The FCA appears to be trying to protect consumers from the entities it regulates, which is contradictory - the objective should be to protect consumers from the actions of unregulated entities. ETPs make it easier and safer for consumers to buy cryptoassets, and potentially at better prices, while derivatives allow them to manage their risks. Instead, the British consumer is unable to manage their risks with derivatives, while they still can buy ETPs in Europe from entities outside the FCA’s reach. Through this ban, the FCA has reduced consumer protection and increased their risks.
This point highlights an obvious tension between the objectives of consumer protection and competition and innovation (and UK competitiveness). As this example shows, competition, innovation and UK competitiveness should take precedence over consumer protection, particularly when it is based on perceived protection and perceived risks rather than actual ones.
The issue is that with new innovation and new entrants, the risks, probabilities and downsides are unknown, so regulating for these is impossible to get right. One solution is for the regulators to use their sandboxes to test out their own regulatory ideas to calibrate risks, probabilities and downsides to inform subsequent regulation. A further solution is to introduce time-based regulation for testing out new proposed regulations, and also to tier regulation based on the value of stablecoins issued by a stablecoin provider. Such regulations make it easier for new entrants to establish themselves, promoting competition, and it provides visibility to uncover bad actors from those who are genuinely trying honest ground breaking ideas.
4 Do you agree with the approach outlined, in which the regulatory perimeter, objectives and principles are set by government and HMT, with detailed rules to follow set by the UK’s independent regulators?
Yes – there needs to be a clear 1:1 relationship between a cryptoasset classification and the regulator with oversight for it.
It is important for clarity, hence for successful innovation, that the government determines which regulator is responsible for each class of cryptoasset, or at least which regulator has primary oversight for each class. It is critical to avoid conflicting, or ambiguous regulation emanating from different regulators.
However, the regulatory perimeters are likely to change dramatically over the next decade and the objectives and principles will need to keep up. Distributed finance (DeFi) embracing public blockchains and digital assets is a growing force that will be felt soon in mainstream financial services, and soon there will be a further ecosystem of CBDCs. Thus, there will be three financial services ecosystems co-existing, the two new and the current. The new ecosystems are borderless with some parts inside a domestic regulatory perimeter, and some outside, with components such as bridge currencies for cross-border payments outside all domestic perimeters. The regulatory environment and proposed incremental changes to it are rooted in today, but a very new and very different regulatory environment is needed to work with the new ones. In particular, a regulatory framework that supports interoperability between the old and the new is needed, and one that works with borderless payments and finance. For example, based on evidence from stablecoins, it is very likely that we will see fiat-collateralised stablecoins using CBDCs, or wrapped-CBDC, used in programmable payments. How will regulators deal with that?
In my view, new types of regulators will be needed for the new ecosystems of DeFi and public blockchains to manage this interoperability. So, although it is fine for now develop crypto-regulation into existing frameworks, HMT must start preparing for this new world of co-existing multi-ecosystem interoperable financial services.
5 What are your views on the extent to which the UK’s approach should align to those in other jurisdictions?
Alignment with other jurisdictions is sound in theory, but probably impractical. Every jurisdiction is going up a steep learning curve, and has different agendas and financial services. Given the UK’s pre-eminence in financial services globally, the UK’s best approach is to take the lead in crypto-regulation and alignment will come from other jurisdictions following. Provided the UK takes a global, or borderless view of crypto-regulation which caters for global networks such as for cross-border payments, then by taking the lead, this alignment should follow naturally (as it has done with English law used as the basis for much international trade and transactions).
6 Do you agree with the government’s assessment of risks and opportunities?
The risks stated need evidence – they appear to be perceived risks rather than actual risks.
What are the specific risks on financial stability and market integrity using crypto-assets and stablecoins? These should be stated and compared to existing risks in the financial system. Such a comparison is likely to show that in many areas crypto-assets and stablecoins actually reduce risks. For example, DeFi lending is built on collateralised loans, algorithmically controlled and much less risky than fractional reserve lending using money created by commercial banks; real-time cross-border payments using cryptoassets as liquidity bridges reduce liquidity, settlement and credit risks; and shared ledger transactions, under the control of an immutable protocol are much less risky than the ledger-hopping transactions in today’s payment systems.
Consumers losing money through volatility is different to losing money being unaware of the risks. Regulators should focus on the latter, rather than the former, and let consumers decide for themselves. It is preferable to tell consumers if they buy and hold a cryptoasset it could fall dramatically in value in days, rather than to ban them from doing so.
The competition risk cited is again a perceived risk. As stated in para 2.1, a government objective is to promote competition, therefore concerns that competition will challenge the dominance of market incumbents is misplaced, and contradictory. Such competition can only be good for society. History has shown that transformational change such as contactless cards, real-time payments, mobile wallets (Apple Pay etc) and challenger bank adoption still take a decade or more to become fully established, as consumer inertia is the limiting factor – Open Banking is another example in its early stages of adoption. Provided regulators keep up with the market, rather than slow it down with bans, restrictions or hesitancy then regulators should be unconcerned by market forces chipping away at incumbent financial institutions.
Instead, the government should be concerned about the cost of transitioning to stablecoin payments and transaction costs. If the government protects incumbents through restricting competition, how do they protect consumers from high costs of a dominant player?
The government should also be concerned by the risk of capital and intellectual flight to jurisdictions that support and encourage crypto services. The UK has fallen behind in the digital world evidenced by the lack of global brands emerging from the UK during the era of rapid internet growth. The government should avoid repeating this with cryptoassets.
7 Do you have views on the proposed initial scope of UK cryptoasset regulation as summarised above?
Limiting UK cryptoasset regulation initially to stablecoins is sensible. As pointed out in the consultation, fiat-collateralised stablecoins are very similar to e-money. e-Money regulations are established and proven and can be expanded to include stablecoins.
As per my answer to question 2, each stablecoin should be classified as a regulated stable token or an unregulated stable token. The government will need to decide which stablecoins to regulate initially and which to leave unregulated (for example, those collateralised against unregulated tokens).
8 Do you agree that this approach best balances the government’s stated objectives and principles?
Yes.
Stablecoins are growing in use and are driving innovation, thus meet the competition and innovation objective.
e-Money regulations have largely been successful for many years in protecting consumers, so regulating stablecoins as a form of e-money will meet the consumer protection objective.
The recent Wirecard scandal highlighted risks in the e-money value chain. However, these risks were inherent to the (lack of) resilience of the players in the value chain and this is a key area where stablecoin regulation needs to be defined to meet stability and integrity objective. Issuers of stablecoins should have contingency plans and resilience built into their operations so that if they fail, or have an outage, or a supplier fails (e.g. a custodian) the end users of the stablecoins are unaffected. Use of shared ledger technology and consensus protocols will mitigate much of the risks in this area, and any regulation in this area should recognise and accommodate this (“better technology, better regulatory outcomes”).
9 Do you agree that the activities and functions outlined above are sufficient to capture the activities that should fall within the scope of regulation?
Wallets should be differentiated between custodial and non-custodial. While it makes sense to regulate custodial wallets, where the service provider has custody of customer funds, it makes no sense to regulate non-custodial wallets as with these consumers have custody of their own funds, so there is nothing to regulate. Non-custodial wallets simply provide access to an owner’s funds on a shared ledger and are just software. Regulating non-custodial wallets would be equivalent to regulating a web browser.
The list of activities in para 3.21 contains activities that are found in traditional payment systems, but which are very different in shared ledgers and should be excluded to avoid confusion and avoid restricting innovations. The consultation refers frequently to “payment chains”, which although missing a definition, I infer mean the sequence of crediting and debiting accounts in different ledgers held by financial institutions in the process of executing, clearing and settling a payment. A process managed by the institutions and other third parties such as central clearing mechanisms. On a shared ledger, there is no concept of clearing and settlement and there are no entities managing the ledger. It runs itself using its protocol (which is binary, it either works flawlessly, or the ledger is useless). Even where transactions cross shared ledgers (say from USDC to ETH) this can be handled by smart contracts executing atomic swaps, without intervention or under the control of any entity.
Therefore, transmission of funds on a shared ledger should be removed from the list in 3.21, it makes no sense to include it.
Similarly, executing transactions in stable tokens should be removed, as transactions are initiated by a wallet and executed by a smart contract, without intervention or under the control of any entity. It should also be noted that if executing transactions is regulated this would put all entities that execute smart contracts and programmable payments under regulatory supervision, stifling innovation and favouring large organisations.
Transaction validation on a shared ledger as part of a consensus mechanism should also fall outside regulation. All a validator does is participate in a consensus mechanism that maintains the integrity of the shared ledger. Regulating validation would drive validators off the shared ledger, leading to centralisation and the risk of bad actors gaining control of the ledger. This includes staking and mining which are missing from the list in 3.21, and which should be explicitly highlighted as outside the regulatory perimeter.
In Table 3.A which lists the scope of stablecoins to regulate, the regulation should reflect existing regulation of the underlying asset e.g. a gold stablecoin should be treated in a similar way to a gold ETF and to its custody arrangements.
10 Do you agree that the government should primarily use existing payments regulations as the basis of the requirements for a new stable token regime, applying enhanced requirements where appropriate on the basis of mitigating relevant risks? What other existing legislation and specific requirements should also be considered?
Existing regulation is a good starting point, in particular e-money regulation. However, payments between ledgers are very different to payments on shared ledgers (no traditional clearing and settlement for example), therefore great care must be taken to avoid imposing regulation on stablecoins that is unsuitable to how they operate.
A detailed risk analysis of stablecoins is required listing all the risks to consumer protection, privacy, competition and market stability, and comparing these to the risks mitigated by existing regulation. Where these risks match, the relevant regulation may be suitable, where they differ, new regulation is needed. Crucially, these risks need first to be observed and calibrated, rather than be perceived, and my suggestion in Q3 for regulators to first test regulation in their sandboxes should be considered.
The consultation is unclear on the planned reserve requirements for stablecoins. I assume that the government intends that stablecoins are backed £ for £ by reserves, without allowing any form of fractional reserves, even if issued by a commercial bank. Given this, the risk profile of stablecoins can be different to that of commercial bank deposits if they are 100% collateralised by say, government bonds, and this needs to be reflected in the regulation of stablecoins – and may be different to existing payment regulations.
11 Do you agree with the high-level requirements outlined? Do you consider that any additional requirements are needed?
The requirements listed are for a traditional authorisation regime based on physical organisations to regulate in the traditional way. Stablecoins are used programmatically by smart contracts, protocols, distributed apps (dapps), machine devices and other constructs where there may be no clear organisation to regulate.
This distinction between “regulatable” entities and “unregulatable” open software execution needs to be reflected in the requirements to avoid regulatory cul-de-sacs which could inhibit innovation or force it underground.
12 Do you have views on whether single-fiat tokens should be required to meet the requirements of e-money under the EMRs, with possible adaptation and additional requirements where needed?
Stablecoins held in a custodial wallet on behalf of a consumer are similar to e-money and the requirements of e-money are relevant. Stablecoins held in a smart contract or in a non-custodial wallet are different, and the e-money requirements may be less relevant.
Clarity is also needed on any different requirements between sterling-denominated stablecoins and stablecoins denominated in other fiat currencies.
13 Do you have views on whether exclusions to the authorisation regime are needed in relation to the stable tokens regime, in light of the government’s objectives? If so, which activities do you think should be excluded?
- Fund transmission
- Transaction execution
- Transaction validation
See answer to Q9.
14 What are your views on the appropriate classification and treatment of (unbacked) tokens that seek to maintain a stable value through the use of algorithms?
Logically these should be unregulated if they are backed, or collateralised by unregulated tokens. If they are backed by other tokens including regulated tokens, then their classification and treatment will depend on the collateral. Given these tokens are very new, and their use is still nascent, it is best to keep a watching brief on how they develop before deciding on a classification and regulatory approach.
15 Do you agree Part 5 of the Banking Act should apply to systems that facilitate the transfer of new types of stable tokens?
No – part 5 of the Banking Act refers to oversight of interbank payment systems. Stablecoins could be used for interbank payments, but they are more likely to be used for payments between wallets, custodial and non-custodial, without any bank in the process. Further, interbank payments are made through centralised clearing systems to transfer money from one bank account to another. Stablecoins are separate to bank accounts, and work on distributed, non-centralised sytems. They also run on protocols which can be self-governing, so unless such protocols are banned, any oversight and enforcement requirements for the Bank of England under the Banking Act would be unclear.
If stablecoin payments become systemically important, then specific regulation may be required for distributed payment networks without central infrastructure.
16 Do you have views on potentially extending Bank of England regulation of wider service providers in the stable token chain, where systemic?
For the reasons given in previous answers, the nature of distributed shared ledger payment systems, with no central control is very different to today’s centralised payments systems, therefore much more analysis and experience of these new systems is required before extending regulation or developing new regulation. Additionally, it will take many years before consumer adoption results in new systemically important payments using stablecoins, so time is on the government’s side to develop appropriate regulation.
17 Do you agree that Part 5 of FSBRA 2013 should apply to payment systems facilitating the transfer of new types of stable tokens?
No – the Payment System Regulator was established to take away control of centralised payment systems from an oligopoly of banks, in order to promote innovation, competition and end user interests.
The e-money regulations are primarily around safeguarding and use of customer funds, rather than payments per-se. The FCA regulates e-money, so if stablecoins are to come under an authorisation regime similar to e-money, then it is logical for the FCA to regulate stablecoins.
Stablecoin payments are independent of banks and bank accounts, and independent of (and a world apart from) the centralised payments systems the PSR regulates. If economic regulation of stablecoin payments is required, then a separate regulator should be established, but this may take years to become clear, so no action is needed now.
18 Do you have views on location and legal entity requirements?
Cryptoassets, including stablecoins are borderless, rather than cross-border, therefore it is difficult to see how the government can regulate anything other than sterling stablecoins issued in the UK and collateralised by assets held in the UK.
The consultation avoids discussion on non-GBP fiat stablecoins, but there does need to be clarity on how these fit into the proposed token classification and authorisation regime.
The recent case of a Thai Baht stablecoin issued on an algorithmically-governed platform in Korea highlights the difficulty of determining the location and legal entity requirements for stablecoins.
19 Are there any areas of existing regulation where clarification or amendments are needed to support the use of security tokens?
As per the answer in question 1, there is a difference between a tokenised security and a security token. A tokenised security on a shared ledger (e.g. used for trading) represents a security held on a different ledger (company share register, CREST etc) – for example, a share of ABC plc represented by an ERC20 token on the Ethereum blockchain. A security token is where a company issues its shares as tokens on a shared ledger, which acts as its ledger of record for ownership of its shares (and which could also be used for trading its shares and initiating dividends) e.g. the ABC plc shared ledger.
A distinction between the two types is needed.
20 What, specifically, are the potential benefits of the adoption of DLT by FMIs? What could be the benefits for trading, clearing and settlement?
n/a
21 What are the potential drawbacks of DLT for wholesale markets and FMIs?
22 Is UK regulation or legislation fit for purpose in terms of the adoption of DLT in wholesale markets and FMIs in the UK? How can FMI regulation/legislation by optimised for DLT?
23 What are the wider industry incentives or obstacles to the adoption of DLT in wholesale markets and FMIs in the UK?
24 If market coordination is required to deliver the benefits of DLT, what form could it take?
25 Would common standards, for example on interoperability, transparency/confidentiality, security or governance, help drive the uptake of DLT/new technology in financial markets? Where would common standards be most beneficial?
26 What should the UK government and regulators be doing to help facilitate the adoption of DLT/new technology across financial markets/FMIs?
27 Do you see value in the government capturing tokens typically used by retail consumers as a form of speculative investment under the regulatory perimeter in the future?
No – it is impractical to do so. The government need ensure only that these tokens are classified (as unregulated) under the FCA’s classification system to provide clarity.
28 Do you have any views on how the government should bring these tokens into the regulatory perimeter in the future?
The government should avoid doing so.
29 What are the risks and opportunities you see in relation to DeFi?
The key immediate opportunity and benefit of DeFi is fully collateralised lending, which is less risky and avoids the systemic risk of traditional commercial bank lending using fractional reserves. There is a scenario that collateralisation against digital assets held on public blockchains may win out against collateralisation against fiat assets in driving DeFi growth.