CSDs are expected by their uses to explore new technologies as paths to lower costs and reduced risks. They are also in a powerful position to influence the pace at which new technologies are adopted. Dan Thieke, managing director and general manager for settlement and asset services at the DTCC, advises CSDs to work with FinTechs to explore how DLT, AI, RPA and Big Data can be used to cut their own costs and enhance existing services, but to be mindful of their systemically important role in the financial markets they serve.
“Partnerships, investments, acquisitions and start-ups are all viable options for central securities depositories (CSDs) as ways of working with FinTechs,” says Dan Thieke, managing director and general manager for settlement and asset services at the Depository Trust and Clearing Corporation (DTCC) in New York.
He is certainly a man who knows, for the American CSD is not just talking about working with FinTechs but spending time and money looking at how their technologies might transform the infrastructures it runs. Indeed, DTCC is taking its own advice. In a January 2016 white paper on distributed ledger technology, 1 it called for greater collaboration in exploring the potential of new technologies.
Not every proof-of-concept leads to a decision to proceed
By the time that white paper was published, DTCC had already made an investment in Digital Asset Holdings, the DLT technology firm then led by Blythe Masters. It was substantial enough for DTCC CEO Michael Bodson to take a seat on the board. A few months later, in March 2016, DTCC announced a joint proof-of-concept with Digital Asset.
The goal of the exercise was to determine whether distributed ledger technology (DLT) was capable of supporting the clearing and settlement of US Treasury and agency mortgage-backed repo market transactions. As it happens, the conclusion was that traditional technologies were an equally effective way of addressing the pain-points inherent to this use-case. After lengthy discussions with all parts of the industry, DTCC decided repo was not the best opportunity for DLT after all.
A partnership project has proved more successful. In January 2017 DTCC joined forces with IBM, New York-based DLT vendor Axoni and the R3 banking consortium to work out how to shift its credit derivatives Trade Information Warehouse (TIW) transaction recording system on to a combined DLT and Cloud platform.
The timing was perfect. TIW had to refresh its ageing technology platform anyway and DLT could, on the face of it, support the necessary volume of transactions, and the various interactions with investment banks and regulators. Whether or not DLT can support the existing volume of business is turning out to be an important test of its viability in any field.
DLT sets challenges of scalability and inter-operability
“A lot of our focus is on running proofs-of-concept of the scalability of the technology in order to see to what extent it can support infrastructure, even in high volume areas such as equity clearing,” says Thieke. “Given the equity volumes we have in the United States, we had to be sure DLT was able to support that much activity.”
In October 2018, DTCC announced that it could. A benchmarking study – conducted with Digital Asset, R3 and Accenture – concluded that DLT could support an average of 100 million cleared equity trades a day. This is a far greater volume of activity than the refurbished TIW needs, which is why DTCC expects the new platform to go live in the early part of this year. It is currently in its final testing phase with 15 global banks.
The goal of the test is not scalability at all, but simply to validate the inter-operability of the platform with other market infrastructures, including the TradeServ derivative confirmation platform owned by MarkitSERV. If it works, derivatives houses will from 2019 be able to process trades internally and report them to the TIW from the same data record.
Partnerships with FinTechs offer faster time-to-market
If so, the DLT-based TIW will have proved a business case: DLT can increase the efficiency with which day-to-day business is conducted. It will also have proved the business case for forming partnerships, rather than investing in start-ups or making acquisitions. Re-dedicating existing staff or recruiting experts in areas such as Cloud, DLT and artificial intelligence (AI) is bound to take time. It can also be expensive when the supply of labour is short and the market is booming.
Partnerships, by contrast, offer instant access not just to new technologies but to subject matter experts. “Partnerships bring time, skill sets and capabilities that we do not have yet,” explains Thieke. “At the time we embarked on the partnership with IBM, Axoni and R3, things were moving very rapidly and we thought it was an opportunity for us to capitalise on certain situations.”
So it is not surprising that starting or buying a FinTech is a choice DTCC has yet to sample. Neither is natural territory for a CSD, but Thieke does not rule them out. “CSDs do start new services all the time, and they do make acquisitions as well, but they are not well known for either,” he says. “However, it would be interesting if either emerged as a major trend. For the right purpose and the right reason – which means helping the industry solve some pain point – we would definitely consider it.”
AI is useful primarily for combing through large quantities of data
DTCC is also experimenting with artificial intelligence (AI). It is currently being used to identify potential cyber-threats, with the more credible ones passed on to an analyst for evaluation. “It frees up the time of the analysts to focus on the cyber-threats that have risen to the top of the risk agenda,” explains Thieke. A second application is assessing the credit risk DTCC incurs in clearing and settling transactions on behalf of its users. It also shares its most worrying findings with an analyst.
To assess the predictive capabilities of AI, DTCC is using it to comb through the data it receives when processing payments or corporate actions, looking for transactions with characteristics that have led to operational challenges. In the mutual fund industry – which DTCC also services – AI is being used to capture, analyse and distribute critical fund prospectus data faster with fewer resources. Information is parsed through a data-sourcing engine to identify key data points and discrepancies are flagged automatically for review by fund managers, freeing up capacity for data analysis.
RPA means something old-fashioned: higher levels of STP
Robotic process automation (RPA), on the other hand, is being applied in a familiar way: to enhance what is already a high level of automation of repetitive manual tasks. RPA is being used, for example, to automate the on-boarding of clients of the DTCC Global Trade Repository, a service that OTC derivatives market participants use to report their trades to regulators.
As with AI, RPA frees up the time of employees to focus on more challenging tasks than entering client information manually or reconciling breaks in invoices and payments. “We view both AI and RPA as a way to enhance the employee experience, within the organisation, not as a replacement for human beings,” explains Thieke. He adds that, by reducing mistakes, machines mitigate risk. “RPA is a risk play, not just a cost play,” he says.
What matters most to a CSD is managing risk
This repeated emphasis on internal risk management is not surprising. DTCC, like every CSD, is a highly regulated, systemically important financial market infrastructure. It has to take its responsibility to avoid outages seriously. In fact, in October 2017 DTCC published a white paper urging policymakers to monitor FinTech developments for any signs that they are changing the nature of risk or increasing systemic risk overall.
CSDs must always be mindful of the contribution they make to financial stability, or instability. “Resilience is top-of-mind for us as we do anything,” says Thieke. “In adopting or working with new technologies, we want to make sure we are not introducing any new risks.” This is why DTCC has developed its own framework for assessing FinTech risks.
“FinTech can be leveraged to help manage risk but it does present a series of new challenges that also need to be considered,” explains Thieke. “Interconnectedness risk, for example, increases with new service providers that are different from traditional service providers. So does concentration risk, if everybody is rushing to use the same service provider. The key Cloud providers, for example, are becoming critically important market infrastructures. There is a fragmentation issue too, where market infrastructures are forced to interact with new entrants.”
Big Data can help CSDs manage risk more effectively
This helps to explain why, insofar as DTCC has a Big Data project in hand at all, it is driven almost entirely by risk management considerations. The risk management group at DTCC is now using information from a data warehouse in its day-to-day work. The data warehouse has for the last two years been devouring a growing proportion of the information the CSD collects across all of its activities in the normal course of business.
“We have been able to run scenario analyses and simulations against data that in the past took us a couple of months to get our hands on, and another couple of months to refresh,” says Thieke. “When the United States moved to settlement on T+2 in 2017, for example, we were able to run multiple impact scenarios from a margin and a settlement perspective pretty much in real-time.”
Internal efficiencies are easier to capture than new revenue streams
It is yet another example of what DTCC has found in using new technologies in general: the internal efficiencies have proved easier to realise than the revenues from new business opportunities. However, this does have the advantage of giving the organisation experience of new technologies before exposing them to users of its services, with its concomitant risks of financial and reputational damage.
And it does not preclude making services based on tested technology available to customers later. “We have seen more of the benefits internally so far, but we are certainly considering ways in which these technologies can be deployed externally and create benefits for our clients as well as DTCC,” says Thieke.
The best technology cannot fix broken processes
But there is a major constraint on realising those benefits for either party. This is the fact that so many post-trade processes and procedures are not standardised. Instead, they have evolved through operational work-arounds and innovations that satisfied immediate needs, without the guidance of an overarching strategy or even a clear sense of direction.
“Even the best technology cannot solve the problem if you do not have participants willing to get together to improve the process,” says Thieke. “In fact, it is better to improve the process and then evaluate the technology solutions.” He warns that such get-togethers must encompass the entire eco-system, or there is a risk of fragmentation as some market participants adopt a revised process based on the new technology but others do not.
Thieke argues that CSDs are well-positioned to lead discussions about how to improve operational processes, but accepts consensus has to be close to universal before they can introduce a new technology, so changes can never be introduced quickly. “At DTCC we have had a lot of successes doing that,” he adds. “CSDs are in a great position to drive those conversations and those potential process changes forward.”
CSDs have something FinTechs need: trust
There is another conversation which CSDs could have – and that is with FinTechs. They are characterised by innovative technologies, but usually lack customers, and tend to be are short of the one substance that would enable them to get some: trust. In theory, CSDs can solve the problem of trust and empower FinTechs by allowing them to clear and settle on their platforms.
In practice, says Thieke, risk management will always trump the appetite for innovation. “As a crucial market infrastructure, we cannot choose who we interact with, because we must remain open to all market participants,” he says. “But clearly our third-party risk analysis, evaluation and monitoring activities are going to be even more important in the future than they are already.”
Thieke understands what that entails. He has spent most of his career at DTCC on the CSD side of the business, looking at how new standards (DTCC was the first to apply ISO 20022 to corporate actions) and emergent technologies can make settlement and asset servicing more efficient. It is work which has obliged him to interact with the technologists at DTCC as well as FinTechs.
“Newer technologies always provide an opportunity to re-evaluate a process or its underlying technology,” explains Thieke. “It is an opportunity to improve settlement cycles or asset servicing, and your interaction with other providers. But ultimately it comes down to a cost/benefit analysis. If the benefits do not outweigh the costs, that argues for maintaining the status quo.”
But Thieke is eager to emphasise that a thorough cost/benefit analysis must always be open to learning from others. “What is most valuable about the World Forum of CSDs (WFC) is that it is an opportunity to learn from our counterparts in different parts of the world,” he says. “Others may have made advances we can deploy in our own markets. We can also learn how to overcome the challenges to implementation.”
Wednesday 10 April 2019
FinTech technologies and how to embrace them
Moderator: Tom Zeeb, head of securities and exchanges, SIX Group AG
Pawel A. Stefanski, Middle East and Africa banking leader, IBM
Abhijit Akhawe, Accenture
Dan Thieke, Managing director and general manager, settlement and asset services, DTCC
Sherman Lin, Chairman and CEO, TDCC Taiwan