AI and the future of global trade
Artificial intelligence (AI) is becoming entrenched in our daily lives, but the technology is still surrounded by misconceptions and skepticism. Ask the public and they may jump to dystopian scenarios where robots have taken over the world.
While this makes for a good sci-fi blockbuster plot, the reality is different and more benign. Those products that Amazon suggested you buy? AI. That TV series you were recommended to watch on Netflix? AI. That self-driving Tesla car you crave to take for a spin? You guessed it: AI.
There is no single industry that is not being re-shaped by technology. Until recently, however, there was one noteworthy exception: global trade. Fortunately, that is slowly changing.
The mechanism that underpins global trade – trade finance – is an industry that remains largely paper-based and reliant on manual processes. This US$18tn a year industry is now being influenced by a new wave of technological innovation, including AI.
Exploring the potential of AI in Trade Finance
AI refers to the use of computer-aided systems to help people make decisions or make decisions for them. It relies on large volumes of data and models to make sense of information and draw intelligence.
In trade finance, AI is helpful in analysing quantitative data, and the repetitive nature of trade finance means that there is a lot of non-traditional data at our disposal.
This means that when trade finance providers need to assess the risks of funding a transaction, AI models can be a very efficient tool for data analysis and reveal intelligence and risks relating to small companies.
AI helps the industry move beyond traditional credit scoring processes, which are often outdated and remain reliant on historical accounting entries – a barrier that prevents small companies from accessing trade finance and has resulted in a $1.5tn global shortfall.
Overcoming the barriers
AI can tackle this shortfall by creating accurate credit scoring models. This can include a company’s payment history, measure the risks of funding a transaction, identify supply chain risks, and benchmark them against their peer group.
Trade finance providers can use this information to communicate effectively with their SME clients, ultimately helping establish better business relationships.
Towards a technological utopia?
The adoption of AI has the potential to do a lot of good in the industry, and the industry is in the early stages of radical transformation.
Advances are driven by fintechs as well as a willingness to change. The industry is working together to create new infrastructure for distributing trade finance assets to other investors in a transparent, standardised format.
The creation of infrastructure is possible due to improvements in technology and integrated across the trade ecosystem in cooperation with banks, insurers, and other industry participants.
It’s collaboration at its best: together, the industry is using technology to re-shape global trade as we know it.
Reimagining operational risk management for business value
The events of 2020 and 2021 have fundamentally changed how we do business, upending every industry, including investment banking. Once bustling trading floors went silent as the switch to work from home led traders to disperse locations – and gave rise to new operational risk challenges.
Today’s dynamic regulatory landscape coupled with ongoing technological innovations have made legacy approaches to operational risk management ill-suited to tackle current challenges and complexity. And while many financial institutions have turned to digital automation and transformation projects to adapt traditional ‘revenue generating’ functions to meet their challenges and help drive growth, they must now do the same with their Operational Risk Management (ORM) functions - or risk being left out in the cold.
The Basel Committee defines operational risk as the “risk of loss resulting from inadequate or failed internal processes, people and systems or from external events.” Unfortunately, many financial institutions still view ORM as a regulatory and compliance necessity rather than a business function that delivers real value. That means executives and risk management departments must now change their risk approach to ensure they are dynamic and flexible, can guide their organizations through complex situations, and can readily meet the evolving expectations of regulators and their clients.
Operational Risk Management is still a young field compared to other risk sectors in the financial markets, but it has always been viewed under a broad umbrella that encompasses risks and uncertainties difficult to quantify and manage in traditional manners. ORM has also been the convergence point where corporate governance issues overlap with revenue-generating business activities, causing potential confusion between departments.
Investment banks have too often placed undue emphasis on creating governance frameworks designed to ensure they meet Basel Committee on Banking Supervision (BCBS) standards instead of recognizing that a sophisticated ORM function can bring quantifiable value. Their desire to merely meet BCBS standards and avoid historic risks has in effect led to an outdated, analogue approach in an increasingly digital world. Savvy investment banks have grasped the value potential of ORM and begun to drive a shift in awareness about the importance of a comprehensive risk identification, measurement, and mitigation program.
Embracing a data-driven approach
Market players now recognize that adopting a digital strategy will allow them to deploy diverse and agile risk management mechanisms. It will also empower them to develop a strong and dynamic understanding of risks while adding real value to the business. This value goes beyond meeting regulatory and compliance mandates introduced as part of the Standardized Measurement Approach developed under Basel 3. A robust approach to risk allows the ORM functions to provide actionable intelligence to support business decision-making and assume a more commercial role that supports the various business units’ day-to-day activities. And that requires an intelligent, data-driven approach with a mandate to match, one that is championed at all levels of the organization.
This type of aggressive approach and embrace of digital transformation can also strengthen how ORM functions handle ambiguous and/or improbable events, especially as traditional methods of risk analysis prove unable to manage the ever-increasing volume of data. In 2010, the total amount of data created, captured, copied and consumed equaled about two zettabytes, compared to 2018 when volumes reached about 33 zettabytes. This 26% compounded annual growth rate means that if the rate of growth steadily continues by 2024, we can expect 149 zettabytes of data created per annum.
Available data levels will make it difficult for analogue ORM functions to successfully meet the executive expectations, however organizations that adopt a data-driven approach will find increased data volumes provide them the insights to gain a competitive advantage and ability to proactively manage their risk.
Leveraging AI and advanced analytics for high impact
Cognitive computing technologies like artificial intelligence (AI), data mining and natural language processing (NLP) can supplement a data-driven approach and help financial institutions confidently automate decisions, optimize processes and provide a deeper insight into available data. These cognitive computing technologies can help reduce or eliminate time-intensive and repetitive tasks, often related to data collection, handling and analysis which are better suited to automation. That in turn can free up critical employees to deploy their experience, knowledge of policies, and powers of assessment to support ORM functions and achieve their goals and focus on high-impact, high-value deliverables.
Cognitive computing can teach computers to recognise and identify risk, which is especially useful to handle and evaluate unstructured data – the kind of data that doesn’t fit neatly into structured rows and columns on a spreadsheet. Natural language processing (NLP) can analyze text to derive insights and sentiments from unstructured data, which a 2015 study by the International Data Group estimates accounts for 90% of all data generated daily. When combined with the estimated future data volumes, cognitive computing functionality presents an immense opportunity for ORM functions to add additional business value in ways previously impossible. A detection model built on cognitive analytics can manage risk on a near real-time basis and can also unlock organizations’ historic datasets that have been compiled for internal, regulatory, or compliance purposes. These datasets often contain free text descriptions that contain a potential wealth of untapped, institution-specific information and could provide valuable insight into historic operational risk losses, providing data to augment employee’s qualitative experiences.
Teaching an old dog new tricks
There are certainly challenges to launching digital transformation projects, implementing new data-driven approaches, and introducing cognitive computing technologies, including employee uncertainty and ethical considerations. That means financial institutions must preemptively address and prepare for potential challenges before they adopt a technology-enabled approach to Operational Risk Management. They must also secure employee buy-in to ensure stakeholders use these new technologies to their full potential and to assuage any concerns that technology diminishes employees’ important role in the organization.
It’s critical that investment banks now shift their Operational Risk Management functions and focus on becoming more adaptive and agile in an increasingly volatile, complex, and uncertain world. Over 66% of banking executives report that adopting new technologies like AI and NLP will be a key driver in IBs development through to 2025. Yet for many investment banks, their ORM functions do not leverage the powerful new tools available to them – including increased computing power, digitization, advanced analytics, and data visualization techniques – much less harness the power of cognitive computing technologies. Until ORM functions leverage these tools, executive leadership cannot allocate resources and solidify ORM’s role in business strategy, performance, and decision-making processes.
Old habits die hard, but it’s time for ORM functions to keep pace with these new technologies, methodologies, and approaches to position themselves and their organizations for success in today’s ever-changing world. If they do not adapt, there is a real risk they may stifle the wider organization, impede new opportunities and inhibit paths to valuable business growth.