European Central Bank, Director General, Market Infrastructure and Payments
Financial Conduct Authority, Head of Department for Project Innovate
Eric J. Pan
U.S. Commodity Futures Trading Commission, Director, Office of International Affairs
European Securities and Markets Authority, Executive Director
European Commission, Head of Unit, Startups and Innovation, Co-Chair, Task Force on Financial Technology, Directorate General Communications Networks, Content and Technology
BNY Mellon, Head of Product Management for Custody, Cash and Foreign Exchange
The Depository Trust & Clearing Corporation, Vice Chairman
Detailed SummaryThe Chair indicated that the aim of the session was to look into the prospects of new fintech technologies, such as blockchain or Distributed Ledger Technology (DLT)1 , in the capital markets area: from issuance to the trading and post-trading of securities and derivatives.
1. Short term opportunities for DLT applications and related benefits
The industry representatives on the panel believed in the potential of blockchain and DLT technology and described different use cases and the related benefits of these technologies.
1.1. New automation and standardisation opportunities
An industry representative emphasized that their institution (a major financial market infrastructure), which sees itself as one of the first fintech companies, considers DLT as transformative, and is excited by its potential and its application to the post-trade environment. The speaker’s institution’s primary focus is risk management; all other considerations are secondary. Therefore, they are not designed to merely seize the latest or the flashiest new tech that comes to the market. Their approach is instead very deliberate and methodical because they are responsible for protecting the safety and the stability of markets.
The speaker’s institution is looking for ‘white space areas’ in the market where DLT technology can be useful. The best short term opportunities are in certain areas where automation and standardisation are limited; for example, DLT can help address an informational gap that has led to inefficiencies in the repo market. It could also assist in addressing costs created by market challenges in the credit derivatives space.
Regarding the repo market use case, the speaker’s institution has partnered with a developer of DLT solutions to potentially add a DLT layer on top of their existing technology. They are exploring together how this technology could enable them to become the settlement counterparty to the start leg of the repo transactions2 in real time. By leveraging DLT, the speaker’s institution and its customers could share standardised information about positions and receive updates on changes to them as they happen. Most importantly, if they become the settling counterparty to the start leg, it would create additional netting opportunities and offsets for customers, which would reduce the risk and capital requirements they face today. The industry representative was pleased to announce that the first phase of the proof of concept was successful. Phase two is now underway, and a working group of industry participants will dig deeper into the solution and hopefully align that with industry needs. At the conclusion of phase two, they will determine whether or not to move ahead with the development of this effort. For those involved in the repo market, the cost savings and the risk minimisation would be substantial if this project proves successful.
In its credit derivatives initiative, the speaker’s institution is completely re-platforming its trade information warehouse (TIW) with blockchain and smart contract technology. This is the first large-scale project of its kind, and they are proud to be leading the industry in this important initiative. TIW is a post-trade platform for approximately 98% to 99% of all credit derivatives globally. It also provides reporting and execution for various CDS post-trade events as well. They are pursuing the re-platform with customers because the CDS market has changed dramatically since the financial crisis, due to market structure changes brought about by policy reforms as well as new capital requirements. As a result, they were forced to consider how to reduce the cost in that service. For this project, they have brought multiple firms to collaborate together, and are working in partnership with a consortium led by a technology firm. The vendor partners are providing the smart contract software that will run on a permission-based blockchain. This software will facilitate greater automation of credit event processing, and the blockchain will provide the shared, standardised ledger that eliminates many of the internal reconciliation costs of customers today.
These developments notably in the credit derivatives space should lead to significant cost savings of at least 20% and hopefully substantially more.
Another industry representative from a leading custodian bank agreed that blockchain offers many benefits provided by the main components of the technology i.e. decentralised trust, distributed consensus, transparency and immutability.
DLT is however not a ‘silver bullet’ that is going to solve all the problems within capital markets. Many current use cases of blockchain are not using all the attributes of the DLT technology but actually mainly provide standardisation and harmonisation of existing processes.
There are nevertheless a number of very interesting use cases in addition to those previously mentioned that are situated mainly in the ring-fenced areas of capital markets that tend to be still more opaque and quite manual, and involve a large amount of paper-based procedures. Two examples of interesting initiatives in that space going into production are in the area of trade finance and gold trading and settlement. The business case of these initiatives is attractive because they are not replacing an existing legacy infrastructure that already works, but are putting technology where there was nothing before.
Another example is Know Your Customer (KYC), which is a significant opportunity for DLT, because it involves an exchange of a great deal of similar information between many different people who want to know the same thing. A shared, distributed database is a good way forward in that case.
1.2. Improved resilience and efficiency of existing systems
An industry representative emphasised that a completely different area of DLT application is using DLT to add additional resiliency to existing systems. The speaker’s institution is a very important settlement agent in the US Government bond market, and has a platform that settles about 85% of all US Treasury transactions. In addition to the normal system for backups, they have built a distributed ledger database that backs up all transactions that are executed on that settlement platform in real time, every 10 minutes.
They are not running DLT instead of their normal backup: they have their primary system, the backups, all of the contingencies, and then also the DLT database. DLT also runs completely within the data centre, so all the firewalls around the data centres that existed before are still in place. Thirdly, everything done on DLT is encrypted, whereas historical databases are not. It is too early to draw conclusions as to whether DLT is better than traditional databases, although some people might have opinions on that, but there is certainly quite a significant amount of intrinsic security in DLT, and possibly better security than at the present time.
This provides a variety of benefits. Firstly, it is a real-life implementation of blockchain, and one that does not require any regulatory approval because it is run completely in-house. The speaker’s institution is the only one that has access to it, but it is spinning multiple nodes in the bank’s own technology environment. The distributed ledger also helps to improve transparency providing access to data in ways that could not happen before. For instance, a CUSIP number3 can be tracked all the way through the chain to see who had it before. It is also possible to run predictive analytics on some settlement ratios, and the banks are also able to predict with machine learning whether failures are going to occur and how likely they will be.
Finally, something that can also be done is to examine throughput volumes during the day and see, for instance, whether particular inter-dealer brokers are late in sending their instructions and their deliveries. This all helps to create a more efficient market and more client benefits, and this is the impact of DLT: it is not the impact that was initially imagined, with many people doing transactions on a common platform, but it is a real-life application that is very powerful.
Another industry representative concurred that there is a place for using DLT to help in business continuity planning in particular, as well as in reconciliation efforts.
A regulator agreed that DLT has many benefits (efficiency, enhanced reporting capacity, reduced costs, and reliability) and that early applications are focused on the optimisation of processes that already exist and using DLT within the current market structure, rather than in a completely new or separate way.
A policymaker added that blockchain and DLT have been identified as one of the main work streams of the fintech taskforce of the EU Commission (EC), but they are ‘not naïve’ when approaching these technologies and have not bought into the idea that blockchain can do, or replace, everything.
2. Feasibility of wide-scale applications of DLT and key issues to be addressed
2.1. Feasibility of wide-scale applications of DLT
A regulator stressed that so far, as previously mentioned, the goal of blockchain initiatives has often been optimising less automated processes in certain niche markets or in certain market segments where there is lower volume, or else supporting existing processes with new technology. For large-scale application, there is clearly some challenges that remain to be addressed before making it workable.
An industry representative said that their institution does not believe that DLT is going to replace traditional systems, or at least, not any time soon. A couple of years ago, when people started getting excited about DLT, they predicted that traditional market infrastructures such as the speaker’s institution would be soon out of business, and that T+0 would quickly become the norm using DLT technology. However, they did not realise that the technology used by existing market infrastructures could already do same day clearing and settlement; the issue is not the technology per se, but the overall environment in the financial industry. There are customers who are in different time zones; there are still customers in the US who use cash and have physical securities; and there are asset managers and very small broker-dealers. This is why T+3 (and in some cases T+2) is still the norm, rather than because of the technology used by clearing houses.
The Chair agreed with these remarks about systems already settling in real time; RTGS, by definition, is real time gross settlement for payment. Existing security settlement engines are also settling more or less in real time, and the issue is more about the organisation of the market and how much DLT can bring additional opportunities and comfort to it.
Another industry representative stressed that if the industry wishes to see wide-scale adaptation of DLT for Delivery-versus-Payment (DVP) settlement of classical instruments, it is essential to find a way to get cash on the DLT. This is why the speaker’s institution is working with a consortium of banks, and also technology companies, on developing what they call the “Utility Settlement Coin”. This is meant to be a stopgap for the period between what exists now and when central banks will be issuing directly on the DLT, which is a state of affairs that is still some way in the future.
In the meantime, commercial bank cash can be used on the DLT, which is sufficient to a certain extent; however, it traps liquidity and requires pre-funding, so there is a credit risk involved. The Utility Settlement Coin wants to bridge the gap and find a way of putting central bank money on a DLT but in a tokenised way: i.e. completely backed one-by-one with real multi-currency digital assets that can then be freely transferred and used in solutions for capital markets. That is an essential building block that will unlock the possibilities represented by DLT. However, this requires good regulatory scrutiny and the cooperation of all the central banks. The consortium is working on that with the different parties to move it forward, but the speaker’s institution is very positive about the early proof of concept, and is also quite encouraged by some of the interactions it has had with regulators and central banks regarding this.
Answering a question from the audience regarding what is currently preventing central banks from moving into a cryptocurrency, an official stated that nothing is preventing this. The question is why, and whether there is any need for it. There are market infrastructures which work properly and allow digital electronic money, settled through accounts from different services. There are mechanisms that allow for the transfer of money without having recourse to cryptocurrency today. If tomorrow, a central bank considers this to be a solution for their markets, there is nothing impeding them.
However, it has to be made sure that confidence in the use of a certain currency is maintained. This is a very important point for the central bank; this has to be seen as a method of transmission of monetary policy to the economy, and that this is under control and is done properly. This needs to be defined. Provided that these caveats are present, there is nothing in principle preventing this, but it is simply not considered necessary today.
A regulator mentioned that the Bank of England is doing some work and some thinking on this question of central bank issued digital currency; the potential implications are ‘absolutely enormous’. With a central bank issued currency on a smartphone, people would not bother to have a current account with a commercial bank. This has massive implications for the relationship between the central bank and the commercial banks, and it also raises huge questions in relation to privacy, because presumably the central bank would then be able to see what someone was doing with their money. People who are trying to tackle anti money laundering and terrorist finance might think that this is a good development, but there are other values at stake as well. If one or two States opted for central bank issued digital currencies, citizens could choose between keeping the pound sterling in their pocket or the US dollar on their mobile phone. This is a huge question, and the ramifications are significant.
An industry representative added that having a central bank issuing cash directly on a blockchain is not essential, but ideally there should be a better way than tokenising the cash that sits on a balance sheet of a commercial bank, which is creating new risks in blockchain that ideally should be avoided.
2.2. Key issues and challenges to be addressed regarding blockchain and DLT
2.2.1. Main challenges
A policymaker emphasized that the EC is considering blockchain technology and DLT in a positive way, in the context of a pro-innovation framework, but is also taking the risks and the challenges that they represent very seriously. Some of the identified challenges being discussed and addressed concern the entry and exit points of blockchain, where the technology comes into contact with existing regulatory and legal frameworks.
The fintech taskforce of the EC has identified issues such as identity verification, anti-money laundering and KYC and EU data protection regulation compliance, as well as the capacity to handle a large number of operations, scalability, legal integration into existing frameworks, legal recognition of immutability and proof of ownership, territoriality, and liability issues. There are also the interoperability and standardisation challenges that have been addressed. In order to deal with all of this, the EC is intending to actively engage in blockchain developments. They are aiming to set up an open and agile European centre of research and expertise on blockchain and its challenges and uses. This is the European Parliament pilot project that was foreseen, which is also part of the inspiration for the EC fintech taskforce.
A regulator mentioned the main challenges that ESMA has identified in their work regarding DLT: these are interoperability, privacy issues, legal certainty, scalability, and standardisation. Standardised data is required to be able to use some of this technology, so there are preconditions in some areas that need to be looked at.
An industry representative stressed that standardisation and collaboration in the industry are essential. One of the reasons why the speaker’s institution used the Trade Information Warehouse is that it uses very standardised information. This was information that the dealers and the buy side put into the trade information warehouse to solve current operational problems. That data was already in place, and was very standardised. In addition, they needed to cooperate and work together to solve those operational risk problems. There is an absolute necessity to use this data. The speaker’s institution thinks that standardisation, as well as cooperation for the industry, is going to be critical for moving forward.
2.2.2. Legal aspects
A policymaker mentioned that there are some jurisdictional issues to be tackled, such as the law applicable to distributed ledgers and the liability issues as regards the ultimate responsibility for events taking place on the ledger. Also, there is the question of legal recognition: that distributed ledger data is true and accurate, and has legal value. The same applies to the legal validity of documents produced and stored on the ledger, including smart contracts, and all of these issues have yet to be resolved and tested in real-life legal systems.
An industry representative added that blockchain and some of the other technologies that have been mentioned, such as artificial intelligence and big data, have to be addressed in the context of the present legal system. The technology must be used to further what is already being done. In the use cases that the speaker’s institution in particular is conducting, they are not looking at challenging any existing legal aspects or doing anything that is going to change the current legal relationship with their clientele, but are only aiming to enhance operation efficiency. The issue is really about how to make it more reliable or faster, more secure, and how to look at the white spaces and begin to make the marketplaces much more efficient.
A regulator agreed that there are some legal issues that might need to be considered. For instance, there might potentially be a need for some clarification about the legal concepts on settlement finality and similar aspects, but certainly in terms of the overall regulatory regime, there is no particular need to have a completely new set of regulations to cater for this.
2.2.3. Hacking risks
The Chair queried whether the assertion by blockchain supporters that it is ‘unhackable’ is true.
An industry representative answered that for an example such as Bitcoin, the consensus seems to be that hacking the Bitcoin DLT itself is today technically impossible, in the sense that the computer power to run through all of the possible combinations in order to decrypt the key seems not to be available. However, that does not mean that Bitcoins cannot be stolen; with the key, anything can be done. The key on the DLT acts almost as the proof of ownership, and with the key, the asset can be reached. The asset does not have a tag attached, stating who it belongs to. It belongs to the person or to the computer that has the key.
The issue of ‘hackable or not hackable’ is not necessarily a problem. It is clear that it is a safer technology than some of the technologies that exist today, but that does not mean that it is fool proof, or that there is no possibility that assets might get lost or stolen. There needs to be a full framework around what happens ‘if the unthinkable happens’, as exists today.
Another industry representative considered that these technologies will exist alongside the legacy technologies that everyone has. The market structure is very complex; there are going to be small players who will not have the resources to be able to adapt to the new technology, and so those services will still need to be provided. They will be hackable, and wherever the weakest point is, this is where the vulnerability will be. The entire market structure needs to be looked at, as well as the vulnerability of that market, in order to protect it. People therefore have to be very careful when stating that a technology is not hackable.
The other issue is that DLT works on the basis of smart contracts, and smart contracts are not documents; they are code. Code can be written in a way in which things might happen within the code that people may not know about. There was a very famous case like this, the Dow case, where at least $60 million were effectively taken away because of the way the contract was written. Care is required when discussing hackability and other risks. The associated risks need to be understood, and the issue of cyber security needs to be considered in particular when looking at any new technology; this is an ‘absolute must’.
3. Regulatory and supervisory approach regarding fintech and blockchain developments
3.1. A technology-neutral and pro-innovation regulatory approach
A regulator stressed that the DLT technology is still evolving, it is probably too early to say at the present time what the regulatory response might need to be long term. ESMA will need to keep on monitoring what is happening, and how the different technologies are being used. But the conclusion at the moment is that the current regulatory framework is workable, does not pose any particular obstacles to the use of DLT and therefore needs to be respected.
It is also very important to look at how DLT is being adopted, not only in Europe but globally, because these technological developments do not respect borders. It is important that there should be close cooperation at the international level to make sure that views are exchanged, and that it is understood where the new challenges might be coming from. The industry should come forward and identify where they see clashes with the current regulatory regime, so that ESMA can look at that early, rather than being behind the curve.
Another official speaking in a personal capacity agreed with the views of the previous speaker. Fintech is also a key topic for US regulators and the CFTC in particular. Regarding the big picture of what they are trying to achieve, there are five areas in which the CFTC can ensure that entities based in the United States, or entities that are located outside the United States that may have an impact on the US market will not be deterred from pursuing innovation.
Firstly, there is a desire for staff to acquire a better understanding of how these new technologies work. Many of these are lawyers, so part of the challenge is trying to get people to be more technologically aware, in order to be able to better assess their implications. Secondly, a principle of ‘do no harm’ has been put forward; this is a precautionary principle, in the sense that as people and firms are trying to experiment and innovate, the CFTC has to be very careful not to scare away or deter this innovation at an early stage. This is very much along the lines of the ‘sandbox’ approach that many other authorities, such as the UK’s FCA have pioneered, and the CFTC are very supportive of that type of process. Thirdly, at some point, there may be some developments in which the CFTC needs to become involved; fintech applications for regulators (Regtech) could be one of them. Fourthly, there is a need to make sure that the rules are updated if necessary to provide flexibility, and have the right concepts that take into account fintech development. The present rule-making process is often very slow and many rules date back several decades. Finally, global collaboration is essential and there is a desire on the part of the CFTC to learn from what regulators are doing in other regions. Many of the markets that the CFTC regulates are already global markets, so by definition, whatever is done in relation to fintech has to be coordinated, or it will not work. This willingness to coordinate on standards, or other types of protocols, must be part of the picture.
A regulator considered that it is essential to identify the risks associated with blockchain and DLT technologies and to assess whether these risks can be contained to a reasonable extent so that the opportunities created by these technologies can be grasped. However a balanced evaluation of risks is needed and it should not prevent the recognition of the opportunities that these technologies represent (e.g. efficiencies, cost savings and greater reliability). This is essential for promoting better competition in the market in the interest of financial customers.
It is nevertheless quite important not to put blockchain technology ‘on a pedestal’, by itself. There are other very interesting, and perhaps equally interesting, technologies developing. These include big data analytics and artificial intelligence, which potentially have applications not only in the context of banking and securities markets, but also of insurance, for instance.
It is also essential, when talking about gaps and white spaces in the market, to be clear that it is not the regulator that designates these. It is for the market to come up with hypotheses and propositions, and if these can be brought into the market, then it will be possible to see what demand there is for them. Some will be commercial successes, and others will be failures; it is not the regulator’s role to pre-empt this process of ‘Darwinian natural selection’ in the marketplace. Regulators should not put unnecessary barriers in the way of that process.
Having a one-to-one customised dialogue between regulators and innovative fintech business, including the DLT ones is essential. This is what the FCA has been doing in its “Project Innovate” over the last couple of years. More recently, they have also been operating the regulatory sandbox, and are trying to close the space between regulation and the applications of innovative technologies. This is very useful, particularly when some of the technology people come from backgrounds that are far removed from financial regulation; there is a need to explore and contain the risks, and to understand the interplay between the regulatory system and the new propositions. This is a valuable means of learning about risks and trends, which is the other aspect of what regulators need to engage in.
The regulator agreed that the principle of technological neutrality can go a very long way, even in relation to blockchain and does not expect it to be the case that every blockchain application will require a rewriting of the rulebook. There are large tracts of the rulebooks that are capable of taking blockchain applications ‘in their stride’, along with new applications emanating from other technologies. This said, there might be specific areas of the regulatory setup that may require adjustment to accommodate DLT in particular4.
A policymaker mentioned that the EC has published a fintech public consultation, wherein they underline that they believe in regulation that is technology neutral, but are also pro-competition and pro-innovation. That is the track they intend to take, along with initiatives on the free flow of data; like much of the Digital Single Market, this would apply to the financial sector, but is not exclusively for it. In the public consultation, the EC also asks in a positive sense about what DLT or blockchain can do to increase access to finance, especially for SMEs and start-ups, and what the main challenges are in the areas of implementation, technology, data standardisation and interoperability. Regarding the latter, the issue is how to make multiple blockchains work together so that they do not get siloed and fragmented again, also taking into account different national efforts, such as in the area of e-government.
Regarding whether there are regulatory or supervisory obstacles from EU regulation or from national laws, the EC does not believe that there are any. It would be good to find that the existing regulatory framework can be adjusted through a sandbox or some minor changes. Regulatory sandboxes can indeed allow the testing of new approaches together with individualized monitoring. There is also the issue of smart contracts, where the question might be more about whether to use these as more than code. For these to have more legal force and to ensure appropriate consumer and investor protection, more thinking needs to be done, although the EC does not believe that the solutions will be very complex.
An industry representative supported the ‘do-no-harm’ approach. One of the issues to consider is that harm can happen in different ways. For example, the CFTC had a record-keeping rule that required that records should be kept in a certain fashion, incompatible with a node in a blockchain or DLT. Therefore it is important to consider also what is already on the books and to check that there are no provisions in the rules that may hamper the development of new technologies. That is the way in which the CFTC has been approaching this issue which the speaker considered very constructive.
3.2. Prospects of Regtech applications
The Chair asked how supervisors are considering the use of technology and data solutions to help monitor and supervise capital markets and firms (so called Regtech).
An official stated that there is great potential in a blockchain reporting system that may help improve market surveillance; it might be many years before this is in place, but in order to get it over the ‘finish line’, the regulatory authorities themselves might need to participate. They will have to be willing to participate in something that is still very experimental, and may be untested.
Regulators see the potential for using DLT for supervision; they are not doing it today, but see it as a vision. The dream is that this would be used first for the oversight of the market. With an operable DLT system, they could have real time observations regarding what is happening in the market. They could also potentially use these technologies for their own internal purposes and for improving the efficiency of their own processes. These are however ‘early days’ and this comes back to the notion of proof of concept and observing what the private sector is doing with such technologies. Once applications of DLT have been more firmly established, then regulators can start applying that to their own internal operations.
A regulator agreed that Regtech applications represent a very interesting potential for the future. Generally speaking, there are more and more reporting requirements, and also data is being reported several times in slightly different ways. One thing that needs to be tackled before talking about the technology that might help deal with that reporting and surveillance of the data would be standardisation: making sure that they have common data sources, that there is a common way of reporting, and that the current duplication in the system is avoided. Much of this can be done without new technology, but there are certainly areas of market surveillance or other areas where using these new technologies would be really interesting. DLT applications are being assessed and tested internally by financial institutions but, ultimately, regulators, central banks and other public authorities that are looking at the data will need to participate in this process. This is potentially a very exciting prospect, but is clearly a long-term prospect, and it is early days.
A policymaker noted that the EC is going to pilot some Regtech solutions, including blockchain and artificial intelligence applications to help supervisors, and it also foresees incorporating this in upcoming budgets and work programmes. This can be an opportunity to then have information available to supervisors almost as soon as it is produced. Then there is the question of whether they have the capacity to look at it that often.
The EC also supports the ‘once only’ principle of e-government policy, by which once the data is given for one purpose – with consent of the citizen or enterprise – it could be shared with other governmental users, so there is not repeated reporting. Issues that need to be considered in doing so are, of course, cybersecurity and incident reporting, among other areas.
A regulator emphasized that the potential of using DLT in areas such as regulatory reporting and transaction reporting, is very obvious. This is not something that is going to happen overnight. Like banks, regulators have legacy systems as well, with all the issues that these give rise to in terms of maintenance and investment.
The other issue to consider is what is done with data once it has been received. The FCA has quite a good transaction monitoring system called ‘Zen’, which enables them to spot many issues happening in the markets, including misconduct. The challenge then is whether they have the supervisory capacity and the enforcement capacity to follow up on what is brought to light by this data analysis. It is not a question of what is brought in at the ‘broad end of the funnel’ and how it is analysed; it is a question of how the application of supervisory and enforcement resources are rationalised and prioritised thereafter.
The Chair summed up the discussion by saying that standardisation is a prerequisite for this DLT technology to work, and global coordination is necessary on those topics. CPMI and the FSB are looking at these questions and assessing those technologies, which is very positive.
There is potential for these technologies to help rather than to replace the existing ones, which are already delivering real time, secure and well-functioning processing, to deliver new services, and also to develop opportunities to deal with cyber challenges. It may not be the case that the DLT technology is not hackable, but it could be an alternative to a polluted centralised technology in a certain way. Efficiency gains can also be considered in relation to this, in particular regarding big data management.
1. Blockchain is a type of distributed ledger technology (DLT). There are potentially other types of distributed ledgers, but in this report we will use blockchain and DLT indifferently.
2. The ‘start leg’ is the portion of the repo transaction when the security is sold, while the subsequent repurchase is called the ‘close leg’.
3. CUSIP numbers identify US financial securities for the purpose of facilitating clearing and settlement.
4. The FCA will be publishing a discussion paper on blockchain in April. They will make the point about technological neutrality in that discussion paper, but also ask whether specific areas of the regulatory setup may require adjustment to accommodate DLT in particular.