It takes a brave CSD to adopt blockchain as the foundation of its core systems, yet the technology is nevertheless rich in new opportunities for CSDs, notably as data managers. But the least risky way for CSDs to experiment with blockchain is by growing in tandem with the crypto- asset markets, says Alexander Chekanov, chief architect at NSD in Russia. Far from disintermediating CSDs, he says, crypto-assets issued, traded and settled on blockchain networks have the potential to put them in touch directly with end-investors.
“I want to see this technology succeed,” says Alexander Chekanov, Chief Architect at National Settlement Depository (NSD), the Russian central securities depository (CSD), in Moscow. “But I cannot endorse it for core systems or for important functions just yet.” It is a characteristically thoughtful and nuanced assessment of blockchain or Distributed Ledger Technology (DLT) by a technologist charged with striking the right balance between innovation and the stability of the technology platform of a systemically important financial market infrastructure.
“To a large extent I focus on innovation – on how technology can bring new business models to the organisation, how we can leverage technology to change the way we work and to be better positioned in a changing market,” explains Chekanov, who joined NSD two years ago from the MFEX mutual fund platform in Stockholm. “But my main responsibility is the overall enterprise architecture of NSD and, while blockchain can be made to work, it goes against some key architectural principles that we normally use.”
Blockchain not mature enough or cheap enough for core systems
In practice, those architectural principles reduce to two that matter. First, any technology needs to be flexible enough to meet new needs, as well as existing ones, since no business remains static for long. And the trouble with blockchain is that use-cases tend to be particular rather than general, and internal rather than external, making it hard to argue that it can be adapted to future needs.
Secondly, the cost needs to be controlled – and blockchain technology turns out to be relatively expensive. “Implementations of distributed ledger technology (DLT) are not yet mainstream, and that means they do not have enough tools to be efficient,” explains Chekanov. “You cannot hire cheap developers and systems analysts to work on it. Things you get easily with conventional technology are always subject to some sort of R&D with DLT. That has its own value, but it is not what you want when adapting you most stable systems of record. Blockchain can provide value in innovative applications that establish new business processes but, as an architect, I would not build a long-lasting core system on top of it just yet. I could make it work, probably, but cost- and risk-wise it is not efficient.”
In fact, he thinks it is a particular mistake to treat blockchain as a solution to ageing core systems. “Ageing can be used an as excuse but it is never an argument to re-write something in a new and fashionable technology,” says Chekanov. “No organisation will ever spend millions of dollars to replace a system just because of age. But competitive pressures are another story. If your current system needs a huge re-work, a huge re-write, to make you competitive on the market, enough to make you throw it away, you will be considering using the latest and not fully proven technologies that best fit your re-designed business processes. That is definitely an opportunity for DLT to appear in the enterprise landscape, perhaps even among core systems.”
Blockchain ideally suited to the creation of a single golden source of data
And the most tempting competitive advantage offered by DLT, he adds, lies in data management. Chekanov argues that DLT has the potential to provide the securities industry with a single source of reliable data – a single register, if you like – which can be accessed and used by multiple entities. This can of course be provided by conventional technology, but existing solutions lack the principal attraction of the blockchain model: its ability to disintermediate existing links in the value chain without making existing operational processes unworkable.
“At present, we have a chain of intermediaries, and displacing one intermediary from the chain is very difficult,” explains Chekanov. “Each intermediary owns some part of the process, so removing any part means re-shaping the interactions between the intermediaries on either side of the one that you want to remove. This is obviously difficult to do, for legal, technical and business relationship reasons. It is why value chains tend to grow, rather than to shrink.”
DLT threatens CSDs with disintermediation as well as the opportunity of reintermediation
In a blockchain, by contrast, the interactions take place not between links in a chain but between nodes on a network. “With services attached to a single golden source of data rather than adding data to a transaction as a step in a chain, you can easily exclude intermediaries that have become less useful and the process will not lose from it,” says Chekanov. “It becomes much easier to work around that problem. The opportunity for CSDs is to provide and control the core infrastructure.”
That opportunity is also, of course, a threat. “As CSDs, we are the core infrastructure that provides the centralised set of data,” he adds. “The promise of blockchain corresponds to what we do today. So for us to survive and thrive in the new market it will be important to think about how we will play this role of the central infrastructure in future. Because if we cannot re-think it and change ourselves according to how the market wants to evolve we are threatened with being replaced by something else.”
In responding to the threat, warns Chekanov, CSDs will be constrained by incumbent service providers such as custodians, transfer agents and data vendors. They will resist disintermediation, making it difficult for CSDs to act pre-emptively. “Implementation will not be easy,” he says. “Any new technology is difficult to implement in the market, but if it also threatens to take out a lot of intermediaries, it will not go smoothly. Indeed, it does not go smoothly today.”
Crypto-asset markets present CSDs with the ideal starting point for DLT
Chekanov thinks the best chance for CSDs to make rapid progress is through the safekeeping, servicing and settlement of the new class of so-called crypto-assets. “It is an emerging market that has not established itself, so it can write its own rules without worrying about its legacy of political and business interests,” he says. “It does not have this baggage, so it will shape itself in a more efficient manner. If crypto-assets become mainstream over time, the legacy has no cards to play. I think this is the biggest opportunity to re-shape the current system.”
Chekanov speculates that settlement of crypto-assets will be performed by combinations of smart contracts, denting the settlement revenues of CSDs from the new market. But he accepts that this is a bold claim, since smart contracts are even less mature than blockchain technologies. “DLT overall is not mature enough,” he says. “But it is obviously good enough for crypto-assets, since they are issued and traded on DLT. And I do not think smart contracts are any more vulnerable than what we have today – in fact, they may even be less vulnerable.”
Chekanov characterises smart contracts as a form of “distributed straight through processing” – automated contracts rather than “smart” contracts. Instead of automating a business process within a single organisation that is trusted to do it well, smart contracts make the automation of a process visible to the whole market. “Today, we have limited experience of how to handle a smart contract,” he adds. “As experience is gained and best practices are built, I am certain they will be an excellent and secure way of automating routine and costly business processes that span securities markets.”
CSDs should seize opportunities to own, operate and connect infrastructural platforms
Reality for CSDs is of course bound to be more complex than this steady transition to a crypto- assets-plus-smart-contracts future implies. “I believe there will be some kind of platform operator who will provide the basic infrastructure on which various services, including settlement, take place,” he says. “It could be CSDs. They are obviously well-positioned to own infrastructure. But at this point I am certain only that there will be infrastructure roles, and that they will be more digitalised and less centralised than today.”
Chekanov thinks these platform operators of the future will not remain purely local. “My favourite parallel here is with the Internet,” he says. “That is a global network for sure but, as we have seen, each country is trying to regulate its own segment of the Internet. In financial services, countries will definitely want to control their own platforms too, so we will have local operators. But these could develop into inter-operable local networks. We could also have a single global network with minimal functions, like SWIFT today, with local platform operators built on top of it. Both ways could work but I think the second way, though more efficient, will be harder to set up.”
The decision by the Australian Stock Exchange (ASX) to replace the ageing technology platform of its CHESS CSD with DLT is a case in point. ASX is aiming initially to create a local network of users on DLT, but naturally hopes that it can construct links with overseas markets more efficiently as they also transition to the new technology. Chekanov is fascinated by the decision (“very interesting, daring and ambitious”), which he sees as an attempt to “future-proof” the CSD.
“I do not believe that they will get a dividend as soon as the system is implemented,” he explains. “But it has allowed them to agitate the whole market to adopt this disruptive technology, which makes their system future-proof, in the sense that what they build today will be very different from how they will adapt it five years down the road. It is much easier to adapt from one DLT system to another than from a non-DLT legacy solution to a DLT one.”
Rate of adoption of DLT by CSDs will vary according to their circumstances
Chekanov does not expect ASX to inspire imitators at other CSDs until the Australian DLT system has started to prove its value. As he points out, many CSDs have invested in applications of DLT to limited areas of their business, or run proofs-of-concept, without deciding to go ahead with full implementations of any of them. NSD itself built a proxy voting solution on DLT, but users found it offered no meaningful improvement on the existing electronic proxy voting service.
“It was cool, and it was certainly worth a press release to buy a bit of media fame, but it did not offer much in terms of actual cost savings or improved user experience or user value,” says Chekanov. “We found that it did not happen because, well, we had a solution that was a direct competitor and was more evolved and complete than the DLT solution. To spend another $2 million to get it to the same level was not cost-efficient for us.”
Other organisations without an incumbent system doing as good a job have found the case for a blockchain-based proxy voting system more compelling. The first project undertaken by the CSD Working Group on DLT convened by the International Securities Service Association (ISSA) – whose membership includes 15 CSDs – was actually to draft a set of requirements for a proxy voting product based on DLT. NASDAQ, which pioneered a blockchain-based proxy voting service in Estonia, has now sold the system to STRATE in South Africa.
It is a reminder that the adoption of blockchain, like the adoption of any new service, depends on the coincidence of a viable technology with buyers that are ready to change and incentivised to do so. “Adoption of blockchain depends on your situation, it depends on what you already have,” explains Chekanov. “If you already have a solution which can do mostly the same thing, and you cannot get away with selling just automation of a business process, the value of your new blockchain solution will depend greatly on a favourable eco-system being in place to offer synergy.”
Blockchain cannot replace well-functioning systems without network effects
What he means is that no blockchain network can thrive in isolation: it must become part of a wider network or network of networks. “If blockchain was completely native to everyone, and people were using tokens on a blockchain to pay for their coffee every day, using blockchain to vote their stock as well would be far more intuitive,” says Chekanov. “Until the eco-system is there, I do not think doing the same logic on blockchain provides enough value. With DLT you need to engage the whole marketplace, not just one horizontal part of it.”
This is something he believes ASX got right. “Other markets which have developed blockchain-based services have not tried to push it on their market like ASX did,” notes Chekanov. “I think ASX are absolutely to be commended for managing to get this going, for finding the right arguments, and for actually convincing their market to go ahead with blockchain.” But even ASX will eventually face the challenge of how to link its domestic DLT network with networks abroad.
Many think CSDs are natural exponents of exactly that art. After all, they have a history of linking up with other CSDs, which makes them an obvious choice as the drivers of inter-operability agreements and the consequent acceleration of network effects. Chekanov is not so sure. “The problem is that CSDs are intermediaries for enterprises communicating with other enterprises,” he says. “They do not engage with end-customers – and network effects are highly propagated by end-customers, because end-customers do not need to sign complicated legal contracts when they sign up to a new service. They have a far lower cost of switching.”
As long as CSDs work with intermediaries which have a limited interest in re-shaping the way in which they work, they are powerless to influence the behaviour of the end-investors that drive the network effects that accelerate and entrench new ways of doing things. Again, however, Chekanov sees hope in the development of the crypto-asset markets, precisely because they could disintermediate the multiple intermediaries that keep CSDs apart from end-investors.
Crypto-asset markets could put CSDs in touch with end-investors
“With crypto-assets issued and traded on DLT networks, CSDs might get access to end-investors,” he says. “Their role can be as network orchestrators, providing the core infrastructure. Then, achieving inter-operability and promoting network effects will be in the hands of CSDs. If this is a role which CSDs can evolve into, I think it will be a huge success for them in the future. And the role is consistent with the direction crypto-asset markets are taking.”
In other words, the opportunity which blockchain technology has created – the safekeeping and servicing of crypto-assets, and the settlement of transactions in them – is likely to prove more important for CSDs than the application of the technology itself to their existing activities.
“Blockchain is just a technology – a means to an end,” concludes Chekanov. “Building a DLT solution gets you nowhere. To succeed, you need to understand what the new opportunities are and how your business model needs to change to take advantage of them. What do you want to do? What kind of markets do you want to be in and what role do you want to play in them? And how can you use technology to achieve those goals? A strategy structured in this way will be practical and useful.”
Wednesday 10 April 2019
Blockchain: is it an opportunity or threat and what should CSDs do to embrace the future business model?
Moderator: Vipin Y S Mahabirsingh
Andre Eduardo Demarco, Director of depository operations, B3, Brazil
Lina Hediah, Executive director, Consensys, Dubai
Woochol Shin, Head of innovative technology team, KSD, Korea
Alexander Chekanov, Chief architect, NSD, Russia
Cindy Chen, Managing director for post-trade, HKEK, Hong Kong