SmartStream Technologies explores how, as new initial margin regulations from the Basel Committee on Banking Supervision and the International Organisation of Securities Commissions become a pressing concern for more firms, technology service providers are highlighting the opportunity this brings to entirely rethink their processes – and leverage their unique data insights and capabilities
As uncleared margin reform imposes requirements on 1,100 new counterparties, 9,500 new trading relationships and 19,000 new custody account agreements, the industry is being forced to innovate, according to tech vendors.
Faced with the choice of hiring more staff or enhancing automation, collateral management experts at SmartStream say firms are being forced to move away from manual processes. “With the vast number of new counterparties coming into scope in September next year, there’s going to be a huge increase in volumes – and therefore a requirement to automate processes as much as possible,” says Trevor Negus, product manager for transaction lifecycle management (TLM) collateral management at SmartStream.
Fortunately, the by-product of increasing volumes is the ability to harvest much more data. In the near future, this data can be applied to use cases for AI and machine learning. “We’re now looking at how we can implement artificial intelligence (AI) across our products,” Negus says.
Working smarter, not harder
Looking to the future, SmartStream believes AI and machine learning will become essential to staying competitive. In today’s highly regulated and competitive environment, margins are under attack, and cost-reduction levers are all but exhausted. Among global banks, the income-to-cost ratio hasn’t grown in years, while customers are becoming increasingly demanding and discerning. “It is also becoming more difficult to attract talent, which is driving businesses towards new technology and automation,” says Negus.
Automation can also give firms an edge by allowing them to focus on more intellectually stimulating work, while leaving mundane administrative tasks to machines. SmartStream’s own innovation lab works with partners who are exploring specific applications of AI and machine learning, starting with a proof of concept and, if successful, passing it along to its product centres, which includes TLM solutions. “In terms of collateral management, we’re looking at a number of use cases for machine learning, for example to automate the process for sending out margin.
The ultimate goal, Negus says, is learning what a user does – and creating generalised rules and exception-handling protocols that streamline the process as much as possible. This means looking for data errors or missing data, outstanding disputes and the status of reconciliations. All of this then feeds into a decision tree, and if there is enough data, a process can be automated. Given a shortage of staff, this becomes a fundamental issue, especially for smaller firms coming into scope that are less likely to have big teams for sharing the increased workload. “If we can relieve that strain through automation, machine learning and AI, then that is obviously all for the better,” he says.
Scaling up innovation
Automating workflows brings many side benefits – not least increasing availability of data that can be tracked, processed and analysed. In addition to overall efficiency gains from eliminating the need to manually handle routine processes, automation frees up resources to explore the possibilities of exploiting newly acquired data insights. It is important to bear in mind that machine learning outcomes are only as good as the available data – the classic computer science adage ‘garbage in, garbage out’ still applies. “Data is really the new gold,” says Negus, noting that having the right systems in place to start with is a prerequisite to acquiring the right data. “If we have it, we can use machine learning. Lots of information is already digitised, so it is possible to leverage, but you need a consistent approach.”
Effective implementation of straight-through processing is particularly critical for collateral management – as it connects upstream and downstream interfaces and allows the integration of front-, middle- and back-office systems. This makes processes traceable and explainable – a prerequisite for further efficiency gains through automation.
Hosting strategic systems in the cloud enables firms to scale up their technology infrastructure quickly and cost-effectively without it becoming a drain on manpower. Cloud hosting not only ensures scalability, but effectively mutualises non-proprietary IT costs while minimising the risk of outages and loss of business continuity – avoiding the disastrously expensive potential failures from hosting data processing on local servers. Besides cost savings, outsourcing IT infrastructure frees up internal resources to focus on value-adding activities more specific to a firm’s unique business proposition. Firms need to check if their data is commingled with other firms’ data and determine if they wish to take the risk of a compliance issue, given that not all cloud solutions separate out their client data into distinct, isolated client-specific environments.
By adopting machine learning to collateral management, minimising manual processes and focusing only on exceptions, users are freed up to think proactively and strategically about their decisions in real time – and can focus their attention on activities that offer them a distinct competitive advantage.