New York-based Depository Trust and Clearing Corporation, which processes hundreds of millions of financial transactions every day, is considering a dozen more pilot programs to use blockchain, in addition to ongoing projects for putting repurchase agreements and credit-default swaps on the new technology.
Robert Palatnick, chief technology architect and managing director at the DTCC’s office in nearby Jersey City, New Jersey, says the firm’s efforts in developing blockchain solutions is as much about trying to ensure continuity in how the technology develops as much as solving particular operational problems.
He sat down with DigFin to explain the firm’s implementation of distributed-ledger technology (DLT), in which blockchain is separated from its roots as the infrastructure for bitcoin, and made into a tool to serve a variety of enterprise needs. [And click here to listen to our podcast about DTCC and blockchain.]
Even within DTCC, DLT is being put to different uses, with different technology partners. A variety of specialists are trying to ensure DLT works among a diverse set of counterparties. Just as DLT promises to clear out a lot of steps and back-office providers, new ones are popping up to make the new technology work.
“There’s a whole new ecosystem of intermediaries emerging,” Palatnick said, including exchanges, digital wallets, vendors providing integration services, and other specialists promising means of getting different ledgers to speak to one another.
Many distributed ledgers
The prospect of DLT becoming fragmented may be unstoppable, but DTCC sees part of its job to mitigate the trend. “Our strategy is to be a leader in driving blockchain technology interoperability because it’s consensus-based, and so it doesn’t make sense to have silos,” Palatnick said.
To that end, DTCC is both a pioneer in implementing DLT, as well as a contributor of code to both solutions under open-sourced Hyperledger programs, and to Enterprise Ethereum Alliance, two of the major umbrellas under which DLT is being developed for commercial use.
“We’re likely going to be in a world of many distributed ledgers,” Palatnick said. “What’s going to be the standard? How do we ensure interoperability among them? And where is there going to be a need for governance? The challenge of an ungoverned, decentralized ecosystem is that it still requires governance. Smart contracts are still only as good as the people who program them.”
The DTCC’s membership of banks and other financial institutions has tasked it with trying to play a role as a responsible governor even as DLT solutions proliferate, and this is one reason why DTCC is implementing its own DLT projects.
Governance includes protecting the system, from both mundane problems and from crises. For example, DTCC has to process securities and insurance contracts that were written decades ago. If a settlement date turns out to hit a new public holiday, or some other detail goes awry because no one thought of it, someone has to be prepared to intervene to ensure the transaction is processed correctly. The same goes for a flash crash that paralyzes markets. “There needs to be a person in charge in case there’s a surprise,” Palatnick said.
One way a responsible entity carries out such duties is by having a role in the permissioning process – agreeing who can be on a decentralized ledger, overseeing the KYC process, validating the nodes of technology vendors, and ensuring the system is free of malware. DTCC already does this in its traditional networks, but it’s harder to confirm a given DLT is compliant and safe.
He is, however, upbeat about DTCC’s implementing its projects for repo and CDS, noting the firm is on track to have the CDS project go live in the second half of 2018. As a technologist, the advent of blockchain is an exciting development. “It’s cool that all of this is happening now,” he said.
The firm is in discussions with members about a variety of use cases for DLT, but adopting the technology is not straightforward. DLT is inspired by the crypto-currencies that needed blockchain to be mined, validated and traded. But transacting bitcoin is not the same as processing assets – securities, swaps, loans, insurance contracts – that involve corporate actions and other ongoing events that must be acted on ‘off the chain’.
On the other hand, DLT has the potential to enhance DTCC’s mission as operating a central ledger, as depository and register of securities and swaps. The decentralized nature of blockchain means the current infrastructure required to communicate and interpret messages among each user’s database can be merged into a common ledger. Users would no longer maintain their own database, but would operate a node on the shared one, provided they meet DTCC’s infrastructure, compliance and safety standards. So far, DLT hasn’t operated at industrial scale, but that’s what DTCC’s pilot programs are testing.
Palatnick says there are other problems that DLT cannot address, especially reporting. Distributed ledgers are a technology to record transactions, immutably and securely. But DLT can’t provide analytics, search, or reporting, which is a necessary part of processing trades. So shifting operations to DLT creates a new problem, one that requires advances in artificial intelligence and big data, which DTCC is also pursuing with third-party vendors.
CDS warehouse: full replacement
In the case of credit default swaps, DTCC is completely replacing its CDS warehouse (a repository and reporting service) with DLT, without disrupting day-to-day operations. With distributed ledger, DTCC is also moving processing to the cloud, working with a third-party vendor, in order to have the computing power necessary to handle huge volumes.
The challenge is to replace the DTCC’s system, but not how it connects with counterparties (banks). So it is adding what Palatnick terms ‘translation layers’ so that banks can continue to submit trades while developing their own technology stack, enabling them to adopt nodes on the CDS distributed ledger. Banks will also need to migrate CDS processing to the cloud, although they can use any vendor that can handle scale, or use on-premise cloud computing.
This is not straightforward, however. “We’re learning what’s important to put on the chain in smart contracts, and what shouldn’t go there,” Palatnick said.
Once the CDS warehouse is fully shifted to DLT, however, DTCC and its members can add asset classes and functionality. One reason why DTCC is on track to go live in 2018 is because credit default swaps are well-defined, standardized instruments. The data quality is high. The challenge is therefore about the migration to DLT and the cloud.
Repos: changing business models
The repo project is a different matter. The initial business case for DLT was to take the starting leg of repo transactions and net it. Banks use repo on an overnight basis, swapping cash for securities, to ensure their daily liquidity needs. Banks calculate their overnight funding requirements and execute the start leg each morning.
Palatnick says this is not about smoother processing for its own sake. In fact, DLT is slower than how US Treasuries are settled now, which is instantaneous. But because DLT allows all parties to see the information at the same time, it means trades can be netted.
“The long-term goal of this project is to take new repo in the morning and net if off the previous night’s transactions,” Palatnick said. “There could be multiple nets throughout the day.” In this way, if all relevant parties know a trade is on, they could execute it later in the day, meaning they don’t have to do everything in the morning. Settlement times and processes could be changed to respond to new trading patterns, as banks use this mutual real-time information to be more precise in their liquidity management. “Banks will be able to manage their daily liquidity more efficiently. That is very compelling to big firms,” Palatnick said.
But whereas with CDS, the DTCC is replacing its current mainframes to process trades, with repo it is adding DLT on top of its existing platform. The extent to which it replaces the systems banks use to connect and enter trades to DTCC will depend on how much benefit they think DLT will give them in liquidity management. So this project is likely to evolve as banks test the limits of DLT’s ability to help them change their entire daily operations.
These two projects, along with the others on the whiteboard, show why DTCC is working with different vendors in DLT. “The challenge is that there is no single ledger,” Palatnick said. “Each vendor is different.”
For example, DTCC works with Axioni, a blockchain developer, on its CDS project because the vendor’s service includes integration. But other vendors just provide the software, and leave integration to their clients.
“Open-source software is important because decentralizing the ownership of databases doesn’t make sense if you then put all the risk on the viability of a single vendor,” Palatnick said. Each vendor has its own technology stack to run and license software, and even in an open-source environment, there will be certain vendors that provide the majority of code.
“The vendors are the risk,” Palatnick said.
That is why DTCC is putting such store in standards and interoperability as it pursues various DLT projects. Financial institutions prefer to work with different vendors using different programming languages and business models, and this helps foster competition and innovation. “We’re trying to find the model in which ledgers can talk to one another,” Palatnick said.