Primary bond markets are ripe for renewal

Problems with the present system

But just because the system in operation works does not mean there is no capacity for improvement.

Many of the processes involved in executing deals in today’s capital markets are extremely manual and labour-intensive. It has the hallmarks of an organically developed system — no one would design a system from the ground up that requires so much duplication of effort when there are more efficient technological systems available.

For example, the issuance of a normal bond is a process that, at present, involves the exchange of unstructured data between many participants.

Most bonds are issued from programmes, meaning that the same legal documentation can be used for multiple transactions. But even so, each individual bond requires the agreement of sizes, coupons, prices and maturities between the issuer, a group of several banks and all their accompanying legal teams.

Each exchange of information takes place in emails and instant messaging services. Term sheets are typically sent as word attachments. Once priced, the information must be sent to a central securities depository to be officially registered. The term sheet’s details need to be manually transposed into each participant’s internal systems, often at several different stages. ‘When we speak with issuers, the first they want to see improved is that process of sending term sheets back and forth multiple times,’ said Benjamin de Forton, bond origination and digital assets expert at BNP Paribas.

Each iteration of this process increases the risk of errors being introduced as well as occupying the valuable time of finance professionals who might otherwise be advising clients.

Similar tasks have to be performed throughout the lifecycle of a bond. Even something as commonplace as coupon calculation introduces problems. With floating rate notes, the daily rate must be compounded to produce the appropriate rate for quarterly or biannual interest rate payments. ‘The issuer, paying agent, investors and lead managers all do the calculation and then reconcile it to get the right number, which doesn’t always match,’ said Richard Teichmeister, senior funding officer at EIB. ‘The golden source of a blockchain based instrument removes the need for that.’

All these stages cost resources. Participating in capital markets is an expensive business. This makes access to markets the preserve of larger institutions, while smaller institutions have to rely on bank loans. ‘If we can reduce the costs of issuing bonds, we can broaden access to smaller issuers, particularly in the small- and medium-sized enterprise sector,’ said de Forton.

Solving problems with structured data

There is a great deal about this process that can be improved by technology. ‘A lot of the inefficiencies in primary markets revolve around the need for repetitive key entry of the same terms,’ said Vic Arulchandran, co-founder of Nivaura, a London-based fintech focusing on primary capital markets technology. ‘We’re working to automate that process and create a means of sharing data in a structured way that can flow straight to downstream processes and market infrastructure.’

Blockchain is one means of improving this process. The information need only be entered once, creating a ‘golden source’, which all participants (with permission) are able to agree on and refer to.

‘Even when a new infrastructure seems technically reliable and offers them sufficient powers of oversight, regulators still have to be satisfied that it represents a meaningful improvement to the status quo.’

But the real issue is properly structured data. Creating a platform that can pull the appropriate data from word documents or emails into a system that can interact with each participant’s internal systems does not, in fact, need to take place on a distributed ledger. Reducing the resources required to process each deal should bring down costs and open the opportunity for capital markets participation to a wider audience.

That data can be exchanged efficiently between the issuer and their banks, but to achieve faster settlement and creation of the security, it must be passed to infrastructure providers like central securities depositories.

Origin Markets, a capital markets technology provider, received investment from Clearstream, a European CSD and post-trade services provider. By extending its platform to include Clearstream, Origin is able to ensure that securities can be processed quickly and easily at the CSD level.

Raja Palaniappan, founder of Origin Markets, said: ‘There are two key issues in digitalisation. First, there is the structure of the data. You have to fix that issue before you can look at the second question, which is whether you process that data in a centralised or a decentralised fashion.’

Blockchain bonds and new opportunities

Beyond harmonised data and the middle-office efficiencies it promises for traditional settlement infrastructure, creating new blockchain-based, digitally native assets opens up a range of new possibilities.

Charlie Berman, founder of agora digital capital markets, said: ‘We’re all going to work in a world of native digital assets. At present, we’re still charting the course to get there.’

In general, efforts are still focused on the first of the two issues outlined by Palaniappan: structuring the data in a standardised and usable way. But once that is achieved, companies like Berman’s agora, Archulandran’s Nivaura and Société Générale’s Forge, among others, are working on new, decentralised ways of transferring purely digital versions of financial assets.

The World Bank sold a two-year bond in August 2018, working with the Commercial Bank of Australia. The A$110m blockchain operated new debt instrument (bond-i) was the first global bond to be issued on the blockchain. The issuer added another A$50m to the bond via CBA, RBC Capital Markets and TD Securities, as well as bringing in an offshore investor.

‘The experiment promises some unique benefits on the operation side,’ said Patrick Cheng,  lead financial officer at the World Bank. ‘Often when we ask a simple question about a conventional security, regarding coupon payments or positions, we end up with different answers to reconcile. With bond-i, we would have much more ability to glimpse into the product and see positions.’

The World Bank’s bond was on a private blockchain. The EIB elected to do its own blockchain bond on Ethereum, a public blockchain, via Société Générale’s Forge platform. Goldman Sachs, Société Générale and Santander ran the €100m one-year trade. ‘Of course, our first transaction took 18 months of work, which is obviously not cost effective for a €100m deal,’ said Teichmeister. ‘But as it becomes mainstream, we believe we will start to realise cost savings, and improvements to transparency and settlement speed.’

Realising this requires an additional step of complexity since it involves not just structuring the underlying data conveniently but actually changing the nature of the securities themselves. Simply structuring the data in a more efficient and automated way then settling it through the traditional centralised systems will cause investors little alarm. ‘We’re taking a transitional approach,’ said Berman. ‘The beauty of that is that we can achieve a lot now in terms of efficiency savings by structuring the data without requiring investors to change their behaviour, but gradually, we’ll move towards digital securities.’

For a system of blockchain bonds to be effective, there has to be an architecture of digital wallets capable of holding new types of securities. If the system is not widely available and adopted swiftly, liquidity in the market would be compromised. However, while some investors may wish to open their own wallets and hold securities directly, many will be likely to prefer to rely on custodian services as they do now.

Cash settlement — a must

If the new capital markets infrastructure is to be built on distributed ledger technology, then a precondition for achieving atomic settlement and the benefits that come with it is a means of settling the cash leg of securities transactions on-chain.

‘For payment versus payment, If you’re going from a tokenised version of cash to fiat cash in order to settle via the real-time gross settlement system, that’s less efficient than if you have a traditional fiat digital currency for both sides of the payment ,’ said Mark Williamson, managing director at HSBC.

Wholesale central bank digital currencies are frequently touted as the obvious solution.

In the absence of blockchain-based securities networks, the value proposition of wCBDCs is not easy to uncover since modern RTGS systems provide an adequate means of digitally settling central bank reserves for inter-bank transfers.

But the key to atomic settlement is digital currency — whether it is provided by a central bank or private institution is a secondary consideration. EIB’s Teichmeister said digital coins are a ‘must’ to realise the full benefits of blockchain based settlement. ‘If it’s provided by a central bank, that might be a plus, but it’s not a necessity,’ he added.

One possible alternative to the wholesale CBDC is the regulated liability network — a concept touted by Citi’s Tony McLaughlin, managing director of emerging payments and business development. The RLN is a means of creating a digitalised representation of commercial bank money, rather than introducing a digital representation of central bank cash.

Tokenisation

One possible means of introducing blockchain bonds is via tokenisation. Rather than creating a digital security from scratch, one might instead be able to create the asset in the traditional way, then lock it with a custodian who creates a digital token that represents legal ownership. Williamson of HSBC said: ‘When you tokenise a bond and the associated cash used to settle the bond, then using DLT to track the transfer of ownership, you get the benefits of a more efficient network with full traceability and auditability.’

SETL, a DLT provider working with Citi’s RLN framework, has been developing an infrastructure for applying this concept to securities settlement. In essence, the system would involve those participating in the RLN becoming members of a single network, which would be partitioned for each institution. Each institution creates tokens that represents their customers’ holdings of securities or cash. The security itself would still be held at the CSD.

If an investor wishes to buy a bond from someone at their own institution, the ownership of the tokens for securities and cash is updated. If buying from a customer of another institution, the institutions send each other messages through the network and, when the transaction is agreed, extinguish the cash tokens of the buyer and the security token of the seller and create equivalent tokens within their own partition. Settlement would take place in the partitions of the CSD and the central bank. ‘The key difference with this means of tokenising is that everyone deals only with their own customers — no Citi customers end up holding Bank of America tokens, for example,’ said Anthony Culligan, chief engineer at SETL.

Atomic settlement and the end of counterparty risk

Atomic settlement means the exchange of delivery and payment simultaneously, which removes counterparty risk. With digital currency and digital securities operating on the same or mutually interoperable chains, it is possible to ensure that an exchange takes place instantly and with finality (meaning it cannot be unwound afterwards).

Settlement can be instant or scheduled for hours or days in the future when funds or securities have arrived. In a post-bitcoin world, where peer-to-peer transfer of value is possible, it is perhaps difficult to justify the function of multiple external counterparties within the exchange.

‘Reducing settlement time is extremely valuable, because it means you can reduce the capital that needs to be posted for settlement periods,’ said Chris Agathangelou, NatWest Capital Markets’ head of digital and flow credit.

It is important to note that blockchain bonds are not the only means of reducing settlement time. With well-structured data, much more automation can take place within the present centralised system than is currently the case. ‘We can achieve T+0 settlement via straight-through-processing at the CSD level,’ said Palaniappan of Origin Markets.

Disintermediation: worth the effort?

Since blockchain’s invention, peer-to-peer transactions have been a huge part of the prospective value it offers. While in the case of bitcoin, this is to escape oversight, the principle can be extended to transaction chains in financial markets.

The history of the attempts to integrate blockchain into finance has been full of promises of disintermediation. First, technologists thought they might disintermediate banks and allow issuers to sell bonds directly to investors. They quickly realised that banks add value as intermediaries, generating deal ideas, advising issuers, pitching and connecting them with investors and more.

Now, the promise of moving to digital or blockchain bonds for some means disintermediating certain parties in the exchange, which could, in theory, reduce costs. It is certainly technically possible for one investor to send a digital asset from their wallet directly to the wallet of another investor, cutting out exchanges, clearing houses, central securities depositories and custodians. However, it is important to examine what costs those intermediaries bring (in terms of fees, time or resources) and what services they provide. It is impossible to do away with some intermediaries for legal reasons but others provide services that market participants may not wish to take on themselves.

Consensus has yet to develop on these questions and market participants’ opinions differ widely.

Many market participants acknowledge that intermediaries like central securities depositories are necessary from a regulatory perspective, rather than a technical one. John Whelan, managing director of crypto and digital assets at Santander said: ‘If you have programmable digital cash, you can do all kinds of things in reducing intermediaries. We know, for example, that if you have programmable digital securities on one side and programmable cash on the other you don’t actually need a central clearing counterparty because you can enforce atomic settlement via a cryptographic cross-ledger.’

Others feel that custodians are likely to struggle. ‘It is very easy for investors to create a wallet on a public blockchain like Ethereum,’ said Jean-Marc Stenger, chief executive of Société Générale’s Forge. ‘Digital securities can be transferred directly, wallet to wallet, in 10 to 15 minutes. Custodians take two to three days.’

But not everyone takes this view. Berman of agora said: ‘We reject this idea of representing this as a battle between the old and the new worlds. We need interoperability and it makes sense to work with existing players who are themselves digitalising. They have relationships forged over decades, capital strength and experience which you can’t just dismiss. Bonds are defined in statutes and rule books which require regulatory licenses and governance appropriate for these huge markets measured in the tens of trillions of dollars.’

‘There’s a host of service providers — CSDs, paying agents — that service securities for banks, simply because banks don’t want to do them,’ said Palaniappan. ‘I’ve never heard anyone complain about paying agent fees or clearing fees, so I struggle to see the value in disintermediating them.’

‘Important work is being done to shift the market’s operations from the present unstructured approach that rests on emails and attachments towards a more efficient approach of exchanging structured data.’

A more realistic solution will be that these intermediaries will find their roles changing because of digitally native assets. This might mean fewer staff, perhaps lower fees or higher margins, but most will simply perform similar services within the new infrastructure. Paul Snaith, the World Bank’s director of treasury operations said: ‘We all have an interest in seeing these intermediaries operate on better technology.’

Smart securities

With a means of settling cash digitally, the possibilities expand dramatically. A blockchain-based security can provide a universally agreeable source for the calculation of a coupon, but what is even more powerful would be a means of automatically paying those coupons.

Financial instruments are, in essence, legal rights and obligations to cashflows. Accordingly, digital assets can, via smart contracts, perform the cashflow operations themselves, rather than simply contain the instructions to do so. This requires a wholesale digital currency that can be used across a bank’s systems for other liabilities since it would be inefficient.

But beyond the automation of lifecycle payments, smart securities promise a much wider range of functionality.

Smart contracts can take in external information and modify their behaviour accordingly. Some investors have specific mandates preventing them from holding instruments below a certain credit rating. Digital financial instruments could contain software allowing them to notify investors of changes in their issuer’s rating status, or indeed information pertaining to sanctions placed upon the issuer.

This can also have consequences for know-your-customer checks. ‘We can embed KYC features within the instrument itself,’ said Jean-Marc Stenger. ‘The network can contain a list of whitelisted investors, which means KYC can be enforced. A smart bond can prevent itself being sent to a sanctioned entity or someone who has not been whitelisted by a KYC process.’

He added that Forge is working on structured products. ‘Instruments with more complex pay-offs could also be automated,’ he said. ‘The dream is to be able to execute hedging formulas in derivatives and interesting products like that automatically.’

Interoperability

Rolling out the infrastructure required for investors to shift to digital securities will prove challenging. One possible risk is that multiple different and competing infrastructures emerge, splitting the market and fragmenting liquidity. Investors will either have to maintain multiple different wallets or be shut out of certain markets. Liquidity will be compromised if investors are unable to trade assets interchangeably.

The present rate of development suggests that there will be an explosion of DLT initiatives at all levels of capital markets infrastructure, but the speed of progress carries with it the risk of the arrival of different and therefore competing standards.

Fortunately, this is an issue that issuers, banks and fintechs are all keenly aware of and anxious to avoid. ‘We want to avoid fragmentation at any cost,’ said BNP Paribas’ de Forton. ‘At this stage, we’re just building and experimenting, but before things can be rolled out to market, we need to converge on common standards.’

Soren Mortensen, director global financial markets at IBM said: ‘The fact is the market is not going to operate on one blockchain that rules all. Interoperability across different networks and different protocols is going to be absolutely key.’

He added that ensuring this is managed effectively will mean that the market cannot be regulated by a single jurisdiction. ‘If these networks are to operate globally, then we need more regulation at a supranational level. Global oversight will be hugely important.’

A key part of the process is ensuring that digital bonds and digital cash can be exchanged between different blockchains. An experiment in February by NatWest Capital Markets, Santander, Fnality and Nivaura demonstrated exactly this capacity.

Banque de France completed an experiment in March 2020, settling blockchain bonds with CBDC. This experiment, led by HSBC, involved connecting existing technologies with multiple different blockchains, making use of tokenisation.

Williamson’s comments about the solutions used to address the challenges of interoperability echoed those of the Banque de France on the transfer of data and assets, as well as on the atomic exchange of assets across different blockchains. ‘The Banque de France and HSBC have demonstrated the possibility of such interoperability, essential to ensure that the multiple environments, on which the efficient functioning of markets rely, can coexist,’ the Banque reported in December.

Public/private blockchains or a centralised solution

A common debate when integrating blockchain to capital markets is whether the system should be based on a public blockchain, like Ethereum or Tezos, or a private and permissioned blockchain, like R3’s Corda.

A public blockchain allows anyone to set up a node and participate in the verification of the transactions. However, that does not mean that everyone would be able to see the details of every transaction or participate. Access to securities and information can still be controlled via a whitelist of approved investor addresses. ‘KYC features can be embedded into the smart contract structure, which means that if your public address has not been whitelisted by Société Générale Forge or another regulated entity, the security token can’t be transferred,’ said Stenger of Société Générale’s Forge, the platform behind the issuance of the European Investment Bank’s blockchain bond.

A few years ago, banks would generally only consider operating on private blockchains but, as technology has developed and they have learned more about it, there seems to be growing trust in the security and privacy standards that public blockchains can deliver.

Nevertheless, some are still concerned about the openness of public chains. Maud Le Moine, head of sovereigns, supranationals and agencies debt capital markets at Goldman Sachs, which worked on EIB’s Ethereum bond, said:

‘After that transaction, we are exploring the benefits of private networks. It is likely that some borrowers would prefer greater privacy controls and the ability to vet network participants.’ Le Moine points out that the transaction fees are likely to be higher than on private chains and more volatile than centralised systems and it will be difficult for them to compete with private networks on speed.

There is also a risk that the increased transparency of a public blockchain might make it more difficult for investors to make confidential trades. However, much of the same techniques they use at present — different portfolios, brokers — are likely to still work on the blockchain.

But Stenger believes that capital markets will move to public blockchains. ‘It was more difficult to develop on public blockchains, but we think that it’s how the market should develop,’ he said. ‘There will be uses for private blockchains, but they’re more like intranets. The really important developments — the internet — will be public blockchains.’

Public blockchains certainly provide reach and accessibility, and the purity of decentralised verification, since private blockchains require participants to set up their own nodes, which requires much more infrastructure than simply opening a wallet or creating an account to transact on a public chain.

Le Moine points out that the transaction fees are likely to be higher than on private chains and more volatile than centralised systems and it will be difficult for them to compete with either architecture on speed.

It is also worth noting that, while fans of blockchain tout its immutable ledger, fans of centralised systems point out that mistakes can be fixed more easily in a centralised system.

Bitcoin’s consensus mechanism, proof-of-work, relies on vast expenditure of electricity to verify it. Private blockchains can make use of much lighter consensus mechanisms and are therefore more environmentally friendly. There are alternative mechanisms for establishing consensus in public blockchains like proof-of-stake. While this does hugely reduce the energy used by the network, it is likely to still be higher than for a centralised network or private blockchain.

Conclusion

Digitalisation in the primary bond market is in an immensely vibrant and dynamic infancy. Important work is being done to shift the market’s operations from the present unstructured approach that rests on emails and attachments towards a more efficient approach of exchanging structured data.

Those developments will give market participants more freedom to choose what they feel is the best approach. There will likely be no single infrastructure governing all bond market transactions. Traditional centralised settlement rails will coexist alongside new decentralised systems, with issuers and investors able to transact on one or other according to their preferences.

The result should be more liquid, efficient capital markets, where transaction costs are brought down, giving participants more opportunity to spend their time on the parts of their jobs that add value and cannot be effectively automated. Lower costs should mean broader participation from the SME sector where access to capital markets, rather than bank loans, should improve growth.

The public sector institutions that have led the way so far will play an important role in helping to establish the standards that must become ubiquitous in order for a healthy digital capital markets ecosystem to develop.