Reference data is the unlikely rocket fuel propelling us into a tokenised future

Download the book

As long as tokenised assets are issued, traded and purchased on separate blockchains, the security and fund token markets will struggle to develop scale and liquidity. It is multi-lateral and not bi-lateral links that are needed to unblock growth, and solutions are now being developed, but blockchain-based markets also need full interoperability with traditional financial markets. Here, the inefficiency of traditional post-trade processing is having paradoxical effects. It is an outcome rather than a cause of the high levels of intermediation in traditional markets and has become a barrier to tokenisation rather than an incentive to adopt it. The immediate result is clumsy workarounds to enable investors to switch between blockchain-based asset classes and traditional asset classes. But SSImple, an 18-month-old blockchain-based database for the storage, sharing and enrichment of Standing Settlement Instructions (SSIs), might just offer a way accelerate the pace of the transition of the capital markets from legacy systems to a tokenised future.

The potential value to be created by issuing and trading tokenised assets on blockchain networks is vast. In terms of equities, bonds, funds and real estate alone – to say nothing of commodities such as oil, grain and precious metals and minerals - tokenisers are contemplating a US$650 trillion global opportunity. But this lavish opportunity will remain out of reach as long as tokenised assets are issued on to blockchains that cannot communicate with each other.

Tokenised assets and tokenised asset services trapped in siloes will not grow

Multiple blockchain protocols are currently contesting for mastery of the cryptocurrency and digital asset markets. If one party to a transaction holds their asset in a wallet on Ethereum and the other holds their asset in a wallet on Hyperledger, it is impossible to exchange the assets unless some sort of interface is built between the two blockchain networks. This is not accidental. Different blockchain protocols want to retain business on their own protocol.

But if a tokenised asset, or even data about a tokenised asset, cannot be transferred from a wallet on one blockchain network to a wallet on another, issuers and investors will limit their involvement. Why? Because they will know that blockchain technology cannot fulfil the basic function of a financial market: to establish the price at which the supply of financial assets is in balance with demand for them by enabling investors to buy assets where they are cheapest and sell them where they are dearest.

Point-to-point links between blockchains are not an adequate answer

At present, this lack of interoperability is being tackled via bi-lateral links between blockchains. The commonest links are token bridges. These work by locking a digital asset on the source blockchain and then minting the same digital asset on the other blockchain, or by “burning” the digital asset on the source blockchain and minting it on the destination blockchain, or by locking tokens in a pool on the source chain and unlocking fungible tokens held in a separate pool on the destination chain.

There are many problems with one-to-one – or, as they are sometimes called, point-to-point - links of this kind. They are expensive and time-consuming to build, maintain and develop, because the costs are not shared. They create link-specific risks of their own. They rest on different perceptions of security threats and operate to different timetables. They are not, in a word, standardised, so cannot facilitate multi-lateral links.

Unfortunately, the growth of tokenised markets through network effects hinges on the development of standardised many-to-many links. Bi-lateral links mean not all sellers and buyers can be present when a transaction is agreed, so the price cannot be efficient, because buyers and sellers that may have different perceptions of the value of the asset are absent. Fewer buyers and sellers also make it harder to buy and sell an asset, so the asset is less liquid, further distorting its price, and deterring even more market participants.

Automation by smart contracts is inhibited by the lack of cross-chain links

If the price of a tokenised asset is unreliable, this has further consequences. “Smart contracts,” which are the means by which blockchain technology automates functions presently carried out by expensive intermediaries, rely on accurate and up-to-date price information from data “oracles” to initiate an action. If “oracles” are consuming inadequate price information, this leads to further market distortions.

Furthermore, since actions initiated by “smart contracts” are implemented by transfers of tokens, the actions themselves are inhibited if the transfer requires movement between blockchains. The lack of interoperability between blockchains means that the transmission of value and data through token transfers initiated by “smart contracts” embedded in other blockchain networks or embedded in assets issued on to different blockchain networks, cannot take place if a link between them is absent.

Issuers and investors need to be confident that tokenised assets can be transferred across blockchains

Up to now, this lack of inter-operability between blockchain networks has not mattered much. It is a tiresome problem but not a critical one, because of the trivial size of the cryptocurrency and security token markets relative to the traditional financial markets. Even at their peak in 2021, the market value of cryptocurrencies did not clear US$3 trillion. (1) At the time, the global equity, bond and funds markets were worth more than 100 times as much. (2)

These disparities in size are a reminder that

ultimately the valuable token opportunity lies not in the cryptocurrency markets but

in the securities and fund markets (to say nothing of “real-world” assets such as real estate and commodities). But it is unrealistic to expect participants in these large, long-established markets to adopt a tokenised alternative on a large scale quickly if there is no mechanism for sparking the network effects needed to bring issuers and investors to market.

If issuers are to issue tokens, they need to be confident that institutional investors will buy their tokens. And if institutional investors are to buy tokenised securities and funds and real-world assets, they will need reassurance not only that they can switch assets between blockchain networks but that they can, during the transition, switch their exposures seamlessly between the digital asset markets and the traditional equity, fixed income and funds markets.

Legacy markets erect legacy barriers to inter-operability between blockchains and traditional markets

So there needs to be interoperability not just between blockchain networks, but between blockchain networks and traditional markets. However, at the interface between blockchain markets and traditional markets, interoperability confronts a fresh set of obstacles: legacy processes and systems and the obstructions and inefficiencies they create.

Unlike blockchains, which create a single version of every transaction and share it with all parties to that transaction, traditional securities market transactions create a record for every party to a trade: the global brokers, the local brokers, the exchange, the asset managers, the global custodians, the sub-custodians, the central counterparty clearing houses (CCPs) and the central securities depositories (CSDs).

These separate records of the same event mean data about the same transaction is processed - which can mean manual re-entry into proprietary systems - multiple times by different intermediaries as it is captured, matched, valued, margined, aggregated, netted, settled, allocated and custodied. At every stage in the repeated exchanges of data there is ample scope for errors and omissions, and consequent breaks in the reconciliation of the records.

Yet until the records are reconciled, a transaction cannot be settled, or its proceeds allocated to investors, or the assets placed in custody for safekeeping and servicing. No wonder employees working in post-trade operations are, according to R3, spending somewhere between 50 and 80 per cent of their time reconciling different records of the same transaction.

Inadequate reference data is the root cause of many legacy system settlement failures

A major source of breaks in those reconciliation processes is poor quality “reference” data. This is the semi-static data - market, customer, bank, instrument, currency and country identification codes and account details - that enable a transaction to be processed. A Bank Identification Code (BIC) or an International Securities Identification Number (ISIN) are classic examples of reference data in the traditional securities markets.

Reference data does not change much or change often but it does need to be kept up-to-date (counterparties do change their banks and account numbers), interpreted intelligently (different terms are used to describe the same things) and used correctly (it is easy, when keying reference data into a system, to transpose letters or numbers) or it is useless in automating a process.

Crucially, reference data is what feeds the Standing Settlement Instructions (SSIs) that counterparties publish to get paid or take delivery of securities efficiently. SSIs record the key information that remains the same from one transaction settlement to another – essentially, the name of the custodian or cash correspondent bank, the account number and the account name – even though the amount and the value date have changed.

The Depository Trust and Clearing Corporation (DTCC) owns and operates a centralised database of SSIs called ALERT that endeavours to keep SSIs up to date by sourcing reference data from asset managers, prime brokers and custodian banks. SWIFT runs a second SSI database (SWIFTRef) whose contents are sourced from correspondent banks.

Both these databases aim to validate and cross-check reference data all the time, but the problem of maintaining accuracy has proved impossible to solve. In the traditional financial markets, the work of maintaining and managing the reference data that underpins SSIs has become an interminable task, frustrating market participants and adding time and cost to post-trade processing.

Blockchain solves the legacy system settlement failure problem but adoption will not be rapid

In theory, wholesale adoption of blockchain technology throughout the financial markets would solve a large part of the reconciliation problem by creating a single record of every transaction at the outset, eliminating the need to reconcile different accounts of the same transaction altogether.

Wholesale adoption of blockchain would also enable transactions to settle “atomically.” In other words, if there is not a simultaneous and irreversible delivery of one asset (say, a security token) against another (say, a cash token) the transaction dies. It does not fail or partially succeed, to be repaired later by operations staff finding missing cash or assets or correcting mistakes in the data about the transaction. The transaction simply never happens at all.

These are appealing prospects. However, wholesale adoption of blockchain is not a realistic possibility without decisive steps to deliver the openness and network effects on which rapid growth in the equity and bond markets depends (3) – and these will take time to be put in place. Instead, blockchain is likely to continue to be applied initially to asset classes that have never developed a comprehensive infrastructure, such as real estate and privately managed assets.

This does not mean the interoperability problem goes away. Asset managers and trading firms will still want to switch between these newly tokenised asset classes and traditional asset classes that do have a comprehensive infrastructure in place already. So there will remain a need for seamless interfaces between the new blockchain infrastructures and the legacy infrastructures.

Legacy settlement processes are embarrassingly inefficient and getting worse in some places

In creating those interfaces, the difficulties are exacerbated by the persistent inefficiency of legacy post-trade processes, at least in Europe. According to data published by the European Central Bank (ECB), the proportion of transactions failing to settle on time in the TARGET2-Securities settlement system rose from 2.37 per cent by value in 2019 to 6.71 per cent in 2022, and from 3.07 per cent to 6.26 per cent by volume in the same period. In 2022, over one million cash penalties were issued for failed settlement instructions per month. At an average value of €145, that amounts to €1.74 billion over the year. (4)

According to a review of the causes of settlement failure published in November 2023 by the European Central Securities Depositories Association (ECSDA), the worsening performance reflects basic operational shortcomings. Counterparties cannot find the securities they are meant to deliver. Transactions are matched too late to proceed to settlement on time. And, of course, poor-quality SSIs, thanks to inadequate or out-of-date reference data, mean crucial information is missing. (5)

In other words, post-trade operations in the traditional securities markets still rely on multiple parties reconciling their different records of the same transaction by rootling around in in-house and third-party databases for missing pieces of information before a transaction can be settled. Even if it can be found, the chances of it being out-of-date or re-keying it incorrectly are high.

Traditional settlement procedures are an archaic set of processes ill-suited to the impending truncation of settlement timetables to trade date plus one day (T+1) from the current trade date plus two days (T+2). The United States, Canada and Mexico will switch to T+1 on 27-28 May 2024, and Europe can be expected to follow suit. Without drastic improvements in settlement inefficiency, settlement failure in Europe is going to increase.

Inter-operability between blockchain networks and traditional financial networks depends on work-arounds

A further challenge is to work out how the ailing settlement process in the traditional markets can be connected “seamlessly” to blockchain networks that generate and share a single record of a transaction and settle transactions atomically. It is necessarily (if not absurdly) complicated to connect the one form of settlement to the other. Efficient automation, in other words, is unattainable.

So it is not surprising to find proponents of blockchain-traditional market interoperability not even trying to build “seamless” interfaces between such fundamentally incompatible arrangements. This has the further disadvantage of focusing the search for automatic interoperability solutions on cross-blockchain protocols only, fostering an acceptance that links between blockchains and traditional markets must be intermediated by Heath Robinson contraptions.

Linking blockchains and traditional markets with Heath Robinson contraptions encourages traditional market participants and their service providers to believe they can take part in tokenisation initiatives without effort.

The purpose of these contraptions is to encourage traditional market issuers and investors to take part in tokenisation initiatives by minimising the changes they and – importantly – their service providers must make. The contraptions are workarounds that enable traditional post-trade service providers such as custodian banks and CSDs to provide tokenised asset settlement and custody services to their buy-side clients without having to make any changes to their existing systems and processes.

For example, in August 2023 SWIFT, the Brussels-based cooperative whose financial messages are at the heart of the traditional post-trade reconciliation and settlement processes, announced the results of experiments conducted with a group of mostly traditional financial institutions that included ANZ, BNP Paribas, BNY Mellon, Citi, Clearstream, Euroclear, Lloyds Banking Group, SIX Digital Exchange (SDX) and The Depository Trust & Clearing Corporation (DTCC).

The experiments proved to the satisfaction of SWIFT and its collaborators that a traditional SWIFT interface can provide a single point of access to multiple blockchain networks. In accomplishing this, the crucial ingredient was a Cross-Chain Interoperability Protocol (CCIP), a cross-chain “router” devised by Chainlink, a blockchain “oracle” network built on Ethereum that ensures smart contracts are fed with secure and reliable data from the traditional world.

CCIP was used already by Decentralised Finance (DeFi) protocols such as Aave (a cryptocurrency lending and borrowing service) and Synthetix (a cryptocurrency derivatives trading platform that enables users to obtain synthetic exposures to cryptocurrencies) to secure tokens against being misappropriated as smart contracts transferred them from the source blockchain to the destination blockchain using the usual burn-and-mint process.

The first traditional financial institution to make use of CCIP independently is ANZ, a participant in the SWIFT experiment. ANZ wanted to test the ability of CCIP to settle the purchase of tokenised assets denominated in an Australian dollar Stablecoin (A$DC) with a Stablecoin denominated in New Zealand dollars (NZ$DC). An experiment was conducted in which CCIP duly enabled a buyer to purchase a tokenised asset with NZ$DC and the seller to be paid in A$DC. (6) In other words, ANZ proved that one form of tokenised cash on one blockchain could be exchanged “atomically” for another form of tokenised cash on another blockchain.

In the SWIFT experiment, Chainlink connected the SWIFT network to the Ethereum mainnet (or at least to its Sepolia testnet). The CCIP was then deployed on the testnet to enable firms participating in the experiment to transfer data from a source blockchain to a destination blockchain. In simulated form, tokenised assets were transferred successfully between wallets on the same blockchain, between wallets on different public blockchains and between a wallet on a public blockchain and a wallet on a private blockchain.

More prosaically, the experiment worked because CCIP was able to absorb the data in traditional SWIFT messages, re-format the content into tokens, perform “atomic” exchanges between tokens on different blockchains, and then translate what happened back into SWIFT message formats.

In principle, the experiment proved that any financial institution with a SWIFT connection – and there are more than 11,500 of them around the world - can transfer data from any blockchain to any another blockchain, in much the same way that they exchange messages about conventional financial transactions. A SWIFT interface with a suitable contraption attached can, in principle, become an interface to blockchain markets as well as traditional markets.

SSImple could accelerate the growth of tokenised capital markets

It sounds like a temporary workaround, and it is. Preserving SWIFT interfaces and the downstream post-trade process they feed and the traditional financial institutions that operate them is not the best way to secure the benefits of tokenisation. The best way to secure those benefits is obviously to shift all issuance and investing activity on to (inter-operable) blockchains.

To encourage that to happen, it would be helpful to make the missing SSIs that account for so many settlement failures in the traditional markets available on blockchains. The SSIs are still needed to settle transactions on-chain, and if they need to be retrieved from databases that are off-chain, it will be impossible for blockchain technologies to effect much improvement. SSIs need to be on-chain.

To make that happen, the SSImple start-up, founded by former DTCC executive director for ALERT Bill Meenaghan in 2022, is already hosting securities and foreign exchange market SSIs on a Corda blockchain. This is a significant step by a business whose origins lie not in inter-operability between blockchain protocols or the urge to accelerate the transition to a tokenised future, but in the realisation that blockchain technology could improve on existing SSI database services in the traditional markets.

SSImple enables custodian banks to publish their SSIs on a blockchain. They then give their buy-side clients permission to access and share them – also on the blockchain - with counterparties (such as brokers) that they work with. By this means, any change in an SSI by a custodian bank cascades automatically through the shared database, ensuring the SSIs used by asset managers and brokers are always up to date. Access is via Application Programme Interfaces (APIs).

The service is getting traction. In September 2023, SSImple announced its first contract, with Equilend, the securities lending trading platform. Equilend sell-side customers - custodian banks and prime brokers - and their buy-side counterparties are now accessing the SSIs they need to settle securities lending transactions directly from the SSImple database via an API.

Inadvertently almost, SSImple finds itself prefiguring a fully tokenised future for the capital markets. True, more accurate, up-to-date and timely SSIs will reduce settlement failure rates in traditional markets, which will reduce the risk of transfers between blockchain markets and traditional markets failing as well. But Meenaghan recognises that the larger opportunity lies in accelerating the transition from traditional markets to tokenised blockchain markets.

At the moment, SSImple is making up-to-date SSIs available to both sides of traditional securities and FX transactions. It is already extending this service to transactions agreed on or between blockchains. It is well-placed to do so, because it is putting SSIs on a blockchain already, and so has only to enlarge the data sets available.

Adding data currently off-chain to the data sets on-chain is not technically difficult but will obviously take time because it entails working with the incumbent intermediaries. “If everything was on-chain already there would be no need for SSImple, because the reference data would be there and moving as part of the process of swapping the assets,” explains Meenaghan. “For blockchains to be truly effective you have got to get to that point, but it is a long, long way off.”

The main reason it is a long way off is that the custodian banks do not want to give up intermediating transactions, or change their SWIFT-based systems, and nor do their sell-side and buy-side clients want to accommodate or pay for change either. However, what those same clients do want is custodians that can help them trade and invest in tokenised assets. It follows that a period of coexistence in which data is both on and off blockchains is inevitable - but not unending.

Indeed, Meenaghan believes the period of coexistence that is now beginning is the necessary prelude to explosive growth in tokenised markets.

“There will never be a single blockchain with everything on it, and there will always be islands of activity where firms are protective of what is being transacted but if you are able to connect them together, so you can safely move assets from one blockchain to another, or represent changes of ownership on the same blockchain, people will have sufficient confidence in the process to increase transaction volumes from low single digits to the tens and the hundreds and eventually the millions of transactions a day.”

Bill Meenaghan, CEO and Founder, SSImple

---

(1) Coinmarketcap records a peak of US$2.86 trillion in November 2021. See https://coinmarketcap.com/charts/

(2) Sifma, Capital Markets Fact Book 2023, 23 July 2023; Investment Company Institute, Worldwide Regulated Open-End Fund Assets and Flows Third Quarter 2023, 9 January 2024.

(3) See “For true tokenisation to triumph a public initiative is needed,” pages 14 to 40.

(4) See https://futureoffinance.biz/european-capital-markets-are-inefficient-so-why-arent-european-csds-doing-more-about-it/

(5) European Central Securities Depositories Association (ECSDA), Settlement efficiency considerations, November 2023.

(6) See Chainlink, Cross-Chain Settlement of Tokenised Assets Using Chainlink CCIP, September 2023.