Work with FinTechs, to learn from them and build bridges with them to regulated markets. Use new digital technologies to lower costs for clients rather than for yourselves. Think across the silos that separate trading from clearing from settlement from custody. Remember that CSDs keep assets safe and remind investors of it. Do not fear new technologies but embrace them as the surest defence against obsolescence. These perspectives are among the advice offered by Thomas Zeeb of SIX, moderator of a session at the WFC on this very topic, and a man who is taking his own advice seriously enough to put it into practice with an integrated digital trading, clearing, settlement and custody platform.
“FinTechs are key partners in the development of digital asset eco-systems,” says Zeeb, head of securities and exchanges at SIX Group AG in Zurich. He thinks that central securities depositories (CSDs) should not only partner with FinTechs but invest in them, buy them and even start them. By working closely with FinTechs, CSDs can learn about new technologies, while introducing them to a constraint with which they are barely familiar – namely, regulation. “FinTechs also tend to be somehat rudimentary when it comes to cyber-security,” adds Zeeb. “They trust in the technology they are using to the point where it borders on naivete.”
SIX is putting money behind this approach. It is recruiting a team of people to work with FinTechs, helping them adapt their ideas and technologies to a regulated financial marketplace. Its formation reflects the SIX view that all the most exciting developments in information technology – artificial intelligence (AI), machine learning, Big Data, robotic process automation (RPA) and blockchain, or distributed ledger technology (DLT) – are inter-linked parts of a single digital eco-system that market infrastructures must understand and deploy. As Zeeb notes, “a DLT infrastructure is more valuable if it is also a self-learning one.”
Finding the elusive RoI on digital investments
Adopting any new technology poses a challenge for CSDs that is becoming steadily more obvious, whatever the transformational potential of the new technology. It is that, if an existing process is not broken, automating it with new technology offers no gain, while automating a broken process actually yields a negative return. “If all we are doing is getting a machine to do what a person did before, the cost of the machinery and the maintenance of the machinery makes it hard to build a business case for the investment,” says Zeeb. “If we can take AI to the point where it can predict anomalies, and act accordingly to reduce risk or optimize liquidity, then it starts to get interesting. But we are a long way from that.”
This absence of a clear RoI is the major obstacle to the deployment of RPA, AI and machine learning-based services, and even to the rapid adoption of what is allegedly the greatest technological eliminator of duplication of cost: DLT. In fact, Zeeb says SIX has concluded that there is no business case at all for applying DLT to existing business processes. “We have experimented with blockchain technology in a number of areas internally, including reference data, corporate actions and structured products trading, to see how the technology worked, understand the pros and cons and see where it might be applicable,” he says. “We came very quickly to the conclusion there is not much point in trying to automate an existing process in a different way.”
Instead, SIX has completely reoriented its digital strategy to identify ways in which it can use DLT to help market participants minimize their capital, collateral and funding costs. In early July this year the Group announced it was building SIX Digital Exchange (SDX), an integrated platform on which tokenized equity and fixed income securities and mutual fund units can be traded, cleared and settled. Though it will use components of the SIX CSD to register and custodize digital assets, SDX is run as a stand-alone start-up that is both legally and operationally separate from the existing stock exchange, central counterparty clearing house (CCP) and CSD.
“Being largely independent means SDX can react much more quickly,” says Zeeb. “Decision-making at SDX is a lot more streamlined than it would be if it was developing within the standard Group structure.” Its existence also bears witness to the fact that Zeeb himself now heads not only clearing, settlement, custody and securities finance at SIX but also the stock exchange. It is not a coincidence that an organization led by someone who has oversight of the entire securities value chain is the first to offer an integrated digital asset service.
SDX integrates trading, delivery-versus-payment (DvP) and custody on a DLT platform
“We are able to pull together services that most infrastructure groups do not manage or govern together,” explains Zeeb. “Looking holistically at the entire process is something I am forced to do. It is also easier to do, from my position and absolutely essential. Anybody who looks at this issue seriously has to ask themselves, `What is the point of automating just part of the process if ultimately you can eliminate most of the processing and reconciliation breaks altogether?’ After all, if SDX flies, three legal entities will be replaced by one. That would make a big difference.”
Eliminating the stock exchange, the CCP and the CSD, and replacing them with a single platform, would indeed make a big difference. “Today, if you do a trade, you do it across a minimum of three legal entities – you trade at the stock exchange, clear via a CCP and settle ultimately at a CSD – and you have to deal with a variety of executing brokers and custodians as well,” elaborates Zeeb. “We are taking that structure and collapsing three separate steps into a single, simultaneous process in which you can trade and settle transactions and book them in an e-wallet at a rate of anything up to 500,000 transactions a second.”
Settling that many transactions that fast translates into substantial savings in funding costs, which in turn reduces both collateral and capital costs. But to achieve speeds and volumes of that magnitude, it is not enough to tokenize a security. The cash leg of the transaction also has to be tokenized. The plan for this, subject to regulatory approvals, is for SDX users to place fiat currency cash on deposit in a segregated account at the Swiss National Bank (SNB) and SDX then issues SIX “liquidity coins” – which can be used to trade and settle on SDX – of equivalent value against it.
This does not, contrary to some press reports, make SDX a crypto-currency exchange. But it would enable the SNB to validate that the amount of tokenized currency in issue on the DLT network is matched by an equivalent amount of fiat currency, obviating the risk of money creation. “Because we have responsibility for the payment system, which we operate on behalf of the central bank, as long as I am segregating Swiss francs or euros or other currencies directly at the central bank, they can audit and reconcile the holding at any time,” explains Zeeb.
The ability of SIX to manage the payments system also makes it easy for counterparties to turn digital assets into fiat currency. All they need to do is give SDX details of their bank account. This would mean that users of SDX can ultimately access fiat currency only by coming off the DLT network – the SNB is not yet prepared to issue Swiss francs directly on to the network – but the fact that SIX operates the payments systems makes the process far more efficient than it is in some other DLT networks.
SDX reduces capital, collateral and financing costs
The conspicuous absentee from the DLT-based value chain of SDX is clearing. Zeeb freely accepts that, if SDX takes off and attracts existing equity and fixed income business, it will over time erode the revenues of the SIX x-clear CCP, at least in the cash market. “You do not need a cash market CCP if you are getting settlement finality for 500,000 transactions a second,” says Zeeb. “What do you care if there are 21 fails when number 22 goes through? You are still funding a liquidity requirement measured in fractions of a second. It will reduce funding, collateral and capital costs much more dramatically than clearing via a CCP.”
Those savings are justified because instantaneous trading and settlement of transactions is much less risky than concentrating and netting transactions through a CCP, enabling market participants to skip the initial and variation margin and default fund costs of centralized clearing. That said, CCPs are likely to retain their position in clearing derivatives, since they assume counterparty risks for much longer periods in futures and swaps than they do in the cash markets.
Being prepared to countenance the elimination of clearing revenues is one measure of the boldness of the SDX initiative taken by the board of SIX. SDX also remains the only instance so far of a regulated infrastructure group – and SIX is regulated by both the Swiss Financial Market Supervisory Authority (FINMA) and the Swiss National Bank – choosing to support the emerging market in digital assets – or, more pejoratively, crypto-assets – on any basis, let alone an integrated one.
“The functionality to admit crypto-assets exists, but it has not been turned on yet, and whether it ever will be is still to be decided,” says Zeeb. “There are concerns about the provenance of crypto-assets, so you have to do two Know Your Client (KYC) checks – one on the investor and the second on the investment, to make sure the investor is not funding some form of illicit activity.” The need to manage KYC risk helps to explain why SIX has chosen to run SDX as a private permissioned DLT network rather than a public one. Retaining complete control of admission to membership will also make it easier to link SDX to other blockchain networks in the future – and reassure investors.
The role of the CSD in adding safety and trust to markets is under-valued
“A key part of the value proposition to investors is that they are dealing with a proper market infrastructure, not some guy with two Mac Pros working out of his basement,” says Zeeb. “Technology can always be hacked. People should not put their trust in technology. They should put their trust in processes. A well-structured and well-implemented DLT network may well be more secure than a traditional mainframe system but it would be foolish to abandon all the other cyber-security measures market infrastructures have built into their daily processes, including information-sharing about cyber-threats. There is a value-add there that has nothing to do with the technology.”
Zeeb does not think this value-add is much appreciated yet. “I think there is a huge value in maintaining the `nostro’ structure of CSDs,” he says. “Most investors have no idea that, if their bank goes bankrupt, their assets are still safely ring-fenced at a CSD. If they understood that, they would be less prepared to put their savings into public blockchains that move between 50 jurisdictions. If a bank defaults, you know there is a legal entity that you can approach that has ring-fenced your assets in a client account. It may take some time for you to retrieve your assets, because the receiver has to do his work, but at least you know that everything other than the cash deposited with the bank is safe. That is a fundamental role of a CSD. It has worked well. Watering that function down, or decentralizing it in a DLT network, would be wrong.”
He alludes to the failure earlier this year of London broker Beaufort Securities, after which client assets were used to pay creditors, as a case in point. Zeeb adds that the benefits reach beyond safeguarding investors’ assets. He thinks CSDs can promote financial innovation, by acting as guarantors of asset safety.
“A lot of banks will work against CSDs holding individual end-client accounts even though that model already exists in the Nordic, Indian and Turkish markets,” he explains. “We are a long way from that model now in most markets, but will the industry eventually evolve in that direction?” asks Zeeb. “I believe that it will, but in most markets it is still a long way off. In the meantime, the facility will be there to allow banks to hold sub-accounts on a direct mirroring basis to their client-side, or “loro” accounts.
That said, SDX does represent a step towards such a future. While SDX is not dependent on blockchain technology, building it as a permissioned blockchain network – and especially one protected by all the SIX cyber-security defences – has in principle made it easier for new entrants such as FinTechs and crypto-asset issuers to join conventional banks and brokers as “nodes” on the network. Once investors open accounts on the same network, the “nodes” can issue instructions for the investment and reinvestment of the assets without imperilling their safety.
Offer customers reasons to migrate, not an invoice and a deadline
The SDX model has a wider significance too. Financial market infrastructures are not slow to understand that moving to a new technology platform could enable them to offer more and better and cheaper services. But if existing processes are not broken, users tend to be unwilling to bear the costs even of beneficial changes.
SDX offers a way out of this dilemma: a choice between the existing system and an alternative. “The amount of investment and work required to replace legacy systems is not to be under-estimated,” says Zeeb. “Having said that, market infrastructures can build bridges. If somebody uses our SDX system, there is no reason why they should even care what technology is running behind it. They get all the benefits and none of the costs.”
Of course, building a separate but adjacent platform is not without cost, but the cost and the risk are much lower than attempting a wholesale migration of customers on to a new platform. “If I start playing around and experiment with the existing platform, I automatically have to put more overhead around it as it entails operating on a live patient,” says Zeeb. “Obviously, you will have substantially more scrutiny if the potential exists for a wrongly coded piece of work to bring down the whole CSD system. We always manage enhancements to existing platforms extremely carefully. Running the new technology separately in the form of SDX, with a view to creating a totally new eco-system, offers fantastic upside potential if it takes off and limited downside if it does not.”
SDX is a sign that Zeeb has taken his own advice (“Embrace the change”) on how best to avoid redundancy at the hands of new technology. “Every market infrastructure needs to think about where they are positioned today and where they want to be positioned in the future,” he says. “I believe there are some basic questions that need to be asked and answered by myself and my peers: What is your real position in your own market today? How sustainable is it? How likely is it that you are going to be disintermediated?”
Zeeb believes the threat is real. He warns that the traditional role of a CSD is not guaranteed into the future. “Any one of us, or even all of us, can be disintermediated,” he says. “Developed market stock exchanges once believed that multi-lateral trading facilities could never compete with them – an attitude which eventually cost them half their market share and their status as the reference point for securities prices. As market infrastructures, we need to be very careful not to rest on our laurels, just because we think we have a vault somewhere.”