Is innovation a double-edged sword for capital markets teams?

Morgan Stanley unveiled its refurbished Canary Wharf office in London at the start of 2019 to great acclaim, due to its open-plan offices and bright furniture. A visitor could be forgiven for thinking they had walked into a tech company rather than an investment bank. Rob Rooney, who decided to forego his position as head of Europe, the Middle East and Africa at the bank and focus solely on technology, has reportedly been driving the revamp.

But the change is not merely aesthetic. The banking sector has historically been seen as rigidly resistant to rapid change. In the decade since the financial crisis, however, banks have put themselves in the vanguard of innovation. Not only are they pushing technological boundaries internally, they are pushing tech further through investment in fintech start-ups.

Capital markets 2.0

Even highly regulated capital markets are not immune to the tremendous pace of innovation. “The challenge isn’t ‘are there enough ideas in this space?’, but ‘how workable are they?’,” says Lee Braine from the chief technology office of Barclays’ corporate and investment bank. The challenge is to move from ideas to real-life applications, while dealing with creaking and inflexible core systems and market infrastructure as well as a full regulatory agenda.

“In terms of capital markets, we are looking for more agile ways of working, making processes more efficient and transparent,” says Ivar Wiersma, global head of ING Labs Wholesale Banking. These are the main areas of overlap in all the applications under development, whether they are using distributed ledger technology (DLT), artificial intelligence (AI), cloud technology or even quantum computing.

Most of these emerging technologies have tremendous cost- and time-saving potential, but to replace old systems, no matter how unwieldy, there needs to be agreement on the way forward. “Under the hood, a lot of work is being done on this in capital markets,” says Mr Wiersma. “This area of the financial markets is already very interconnected and, particularly with DLT advancements, will require working together as an industry.”

Banks have traditionally worked independently, using their own systems and even their own jargon, but capital markets transactions have always required teamwork. “The industry is changing – we are adapting to a new way of working, towards a more collaborative, curious and action-oriented culture,” says Philip Smith, a partner at Allen & Overy’s capital markets business. 

Settling with DLT

The main target for this transformation is the labour-intensive and inefficient parts of the market. The biggest inefficiencies in capital markets are at the clearing and settlement stage. When any transaction is signed, it has not technically completed until final settlement has taken place, which takes time and is closely overseen by the regulators.

One way to simplify the process would be to use tokens instead of currency. The utility settlement coin, created by UBS, is a new digital instrument similar to cash and asset backed by central bank funds. It can be used both internally by banks to move liquidity around more easily and externally to achieve legal settlement finality instantly on a peer-to-peer basis for transactions, according to Mr Braine. Five banks joined the network alongside Barclays in September 2017: Credit Suisse, Canadian Imperial Bank of Commerce, HSBC, MUFG and State Street.

Investors could settle through DLT with instant legal finality, but the challenges in moving from the old architecture to a new system are enormous. Every bank has technical debt and business process debt, so the costs of moving to a new system are significant, according to Mr Braine. There are also questions to be resolved. How long should the old system be run in parallel? Are the counterparties interested in the same technology? These are questions fintech companies have rarely considered when pitching ideas. However, for large-scale changes, Mr Braine believes costs could be shared by participants, making the proposition more palatable.

This theory will sound attractive to banks, moving money around faster and cost-efficiently, but what are its wider implications? If this type of trading had been available to investors during the financial crisis, the run on the banks may have been even more devastating. The friction in the market kept is widely considered to have upright in this case. “There is an important debate about the consequences of the tokenisation of currency,” says Mr Smith. Moving away from a well-understood and highly regulated process into a new system carries risk.

After the sector as a whole agrees on a unified system and builds it, the law will need to catch up. “A major shift in financial architecture in a decade or so is inevitable, but it won’t happen tomorrow,” says Mr Smith. “DLT is a typical case where I think people overestimate the short-term effects, but underestimate what it means in the longer term,” adds Mr Wiersma.

Regulator heaven

A great advantage of DLT is that trading tokenised assets is more transparent, but it is not always possible to achieve settlement finality under the current applicable regulations. Mr Wiersma therefore advocates a high level of engagement and proactive dialogue with the regulator.

For this to be effective, regulators not only need to understand the technology, but must be well-equipped to deal with its applications. Banks work closely with partner fintech companies to make sure their technology can be applied within the legal and regulatory framework of capital markets, but when projects are taken to the regulator, sometimes it is unable to give an opinion. “Initiatives sometimes have to accept the risk that you may carry something forward quite far before you know whether it will get the regulatory green light,” says Mr Braine. 

Nevertheless, regulators are delighted at the prospect of DLT – instant, transparent, immutable – being widely adopted, according to Mr Smith. Indeed, the attractions of moving to a single shared platform as the point of reference where everything to do with a deal is recorded is clearly something that is desirable for a regulator. “The regulator could get access to reportable events in real time, via standardised and consistent reporting structures,” says Mr Braine. Mr Smith notes that this would allow the regulator to directly access the information it needs in case of a problem, rather than having to ask parties for information.

As regards security, in DLT the points of entry and exit are considered to be the most vulnerable. Public DLTs are therefore deemed to be more secure: the more users put links in the chain, the more copies there are of the chain and the harder it becomes to tamper with it. Ironically, public DLT is not currently regulated, though the smaller and hence less secure private DLTs are compliant with the rules.

When it comes to AI, many start-ups have grand ideas about ways to apply neural network technologies to specific capital markets issues. But when even developers themselves are not entirely sure how their own technology works, it will quickly run afoul of regulators, warns Mr Braine. Banks cannot simply apply ‘black box’ technology in such a highly regulated environment, even if doing so solves an issue. It is safer to explore the data science aspects of AI: as an analytics tool it will be filtered through a human rather than being exposed directly to clients, according to Mr Braine.

Augmenting or limiting?

A major limitation of DLT in banking is its capacity, according to Mr Smith. How many transactions can be put through at a time, and how long does it take? What improvements must be made before the technology can be rolled out on a significant scale? This includes updating the systems within banks, which still have back offices plugged into the old system.

“In capital markets, notional margins are still very high and access requires a lot of manual work,” says Mr Wiersma. Most of the technology in development aims to make the market run more smoothly, but this raises the question of whether progress will come at the cost of a large number of human jobs.  

Mr Smith dismisses the notion that optimisation of workflow, such as automising the drafting of contracts, will materially affect the number of people working in capital markets. For example, in areas where contracts are highly commoditised, such as bond trading, the uses for a program that could accurately machine-read contracts are huge. All the contractual terms remain the same, leaving only the commercial terms needing to be adjusted. If this could be done automatically throughout the document, it would drastically reduce the lead time for bond issuance. “The speed of execution would be vastly increased and the transaction would be less exposed to market volatility,” adds Mr Smith.

Rather than replacing labour, some argue that technology could augment the existing workflow. In 2018, ING adopted an AI-driven tool called Katana, which uses predictive analytics to help bond traders. “A lot of analytics go on inside a trader’s head. This algorithm helps traders to get better quotes and pricing for clients more efficiently,” says Mr Wiersma. The analytics are an asset, but traders are still in control. “Tech like this could have a tremendous effect on positions if used widely,” he adds.

Mr Smith adds: “The theory is that if the system is more efficient, it will be accessed more often. It would likely reduce lumpiness in the market, where issuers may choose to do more smaller on-demand deals.” Technology can smooth out communication between the various parties, but many believe it is unlikely an AI program will ever replace legal counsel. Allen & Overy already operates virtual ‘deal rooms’, where documents are shared between relevant lawyers. When a change is made to documentation an alert is sent out, eliminating the need for sending it back and forth.

From incubation to execution

Many of the proposed changes are sweeping and radical and the road ahead is littered with challenges. The best way to engage with disruptive fintech is balancing these with quick wins: minimum disruption for maximum benefit, according to Mr Braine. A problem here is that banks are not the best equipped for the nimble processes and business arcs of fintech start-ups.

“Within a bank, the vendor approval process can be lengthy for good reasons related to performing necessary due diligence,” says Mr Braine. But if a bank just wants to test a project, which is not yet going live, some flexibility on the risks is required. “The fintech space needs a controlled way to conduct proofs of concepts without having to go through a full version. A slim-lined environment capability is useful for testing,” he adds.

It seems that it is better to start small, with applications that may not in themselves be game changing, but have cleverly built in the optionality for further innovation. In the DLT field, bilateral payment netting business CLSNet embodies this by standardising calculations. “It is only a stepping stone as it doesn’t perform the settlement yet, but it opens the door for a broader suite of services off the back of it,” says Mr Braine.

A real-life case study of this type of approach is the Bookbuilding app, launched by developer SIA in mid-2017 in collaboration with Refinitiv’s Eikon platform, which provides a fully digitised platform for equity capital markets transactions. “The SIA Bookbuilding app services all participants – issuers, investors and syndicate banks – to offer full transparency throughout the bookbuilding process in real time,” says SIA capital markets director Deborah Traversa.

It has “a native compliance approach in terms of the Markets in Financial Instruments Directive II and market abuse regulation with full digital traceability including allocation justification, record keeping and market sounding”, she adds. The next step for SIA would be to deepen data integration by including real-time feedback from investors and using analytics to provide actionable insight. In the medium to long term, the app has its eyes on DLT developments. “We already have a private blockchain infrastructure called SIAchain that is fully compliant with the rules, secure and dedicated to the financial world,” says Ms Traversa.

The new frontiers

Although most of these innovations are still some way from full integration into the DNA of the capital markets, many applications are already live, with most institutions hard at work on further developments. A longer term and perhaps more remote area of technological advancement is quantum computing.

This has numerous possible applications in investment banking; for example, Barclays is experimenting with portfolio optimisation by mapping volatility, according to Mr Braine. However, quantum computing is such a difficult subject that the thresholds for development are high. The biggest concern is that a sufficiently powerful quantum computer could potentially crack existing cryptography in future. This is a serious threat for the banking world and it is a matter of when, not if, the hardware can be built, Mr Braine adds.

Cybersecurity is at the top of the list for many clients, agrees Mr Wiersma. “Yes, quantum computing will in the future be instrumental, but at the moment it is more important to change the behaviour of employees,” he says. While the whole industry is concerned about the threat of tech innovation, it could be that human error is the greater threat.