This page is about:

    Machine Learning - Artificial Intelligence

    • Elias A. Baltassis, Director Europe, Data & Analytics, Boston Consulting Group (BCG)
    • Dr. Ash BOOTH, Head of Artificial Intelligence, Corporate & Institutional Digital, HSBC

    The ubiquity, depth and granularity of data in the world today is unprecedented, and it has been facilitated by ordinary consumers, who are creating, sharing and accessing more information than ever before through online channels or via mobile applications, and it shows no sign yet of decelerating. Such is the scale of the sprawling data ecosystem that experts predict the world will be generating in excess of 163 zettabytes (ZB) of data per year by 2025, an increase from the estimated 16.3 ZB that was produced in 2017.1

    Experts – speaking at HSBC's International Day 2019 in London on March 7 – offered their insights into how corporates can leverage big data and augment their business propositions.

    The data tsunami: Too big to manage?

    Of the top ten corporates by market capitalisation size a decade ago, only one company (Microsoft) had a data-driven business model, in contrast to today, where big technology companies have become increasingly dominant in public markets. With the spectacular growth of digitalisation, online connectivity and the Internet of Things (IOT), vast tranches of data are waiting to be tapped. This, could help companies better understand their clients' needs, enabling them to optimise services by tailoring products for customers or utilising advanced analytics to identify future supply and demand trends, in addition to streamlining operations and their supply chain management processes.

    Despite the advances in data analytics, Elias A. Baltassis, Director Europe, Data & Analytics at Boston Consulting Group (BCG), said the full impact of the data boom will become more visible over the next 10 years as technology evolves and penetrates even more aspects of economic life. Acquiring a vast repository of data or information on its own is of limited value, but those dynamics change significantly when artificial intelligence (AI) is thrown into the mix. Through the proliferation of open source software, AI solutions have become far more commoditised, enabling organisations to sift through data on a scalable basis, and perform comprehensive analytics off the back of it.

    Untangling AI: Untangling fact from fiction

    As a concept, the term "AI" is routinely misapplied and misused, acknowledged Baltassis, who added there were a number of different variants of the technology. Robotic Process Automation (RPA) is often referred to as AI when it is not anything of the kind. Applications such as RPA are serving a number of useful purposes and are being integrated across many leading companies, principally to expedite repetitive, rules-based activities such as regulatory or shareholder reporting. As RPA can digest data very quickly and accurately and input it into reports in a fraction of the time that it takes a human, the technology can shave a number of basis points off operational costs and dramatically improve client experiences.

    Sophisticated AI models are also coming to market, most notably machine learning (ML), an algorithm-powered technology which provides users with, predictive and prescriptive analytics by assimilating nascent concepts such as neural networks and deep learning resulting in new applications such as computer vision or natural language processing.

    Baltassis highlighted that more banks are now drawing on AI and ML to monitor counterparty credit risk and operational risks, enabling firms to adopt a more predictive – and less reactive – approach towards risk management.2 Beyond financial services, the audience heard that pharmaceutical companies are trialling AI in drug discovery, by using the technology to interrogate huge data sets allowing for quicker product development. Baltassis added these innovations have the potential to revolutionise the pharmaceutical industry, by dramatically reducing the time involved in drug testing and experimentation.

    Corporate treasury and AI

    While the short-term direct impact of AI on corporate treasury activities could be less profound than – say – in the pharmaceuticals sector, it will have positive implications nonetheless, principally by increasing the quality of forecasts and removing some of the costly manual processes. Potential use cases may include the application of ML in optimising cash management and hedging strategies; and machine executable regulatory compliance.3 Baltassis suggested predictive analytics performed by AI could also help treasurers improve their procurement processes, supply chain management activities and sales forecasts.

    "Bank service providers such as HSBC Global Banking & Markets are also adopting AI to improve corporate user experiences, by rolling out automated virtual assistants and intelligent chatbots," said Ash Booth, Head of AI at HSBC. These solutions can communicate with clients about queries which are fairly rudimentary and that require an immediate/time-sensitive response. However, Booth said problems could potentially arise were the bank to fully systematise the entire client relationship life-cycle, as there will always be a need for human intervention in scenarios where AI is incompatible or inappropriate. While AI is a facilitator for client relationship management, it will not disintermediate humans anytime soon.

    Finding the AI equilibrium

    Many corporates accept that AI will have a material impact on their enterprises. Through its phenomenal ability to deconstruct large volumes of data, AI will improve a number of operational activities, enabling cost synergies to be obtained and new sources of revenue to be exploited. While these technologies can streamline a lot of business processes, human intermediation in areas such as client engagement and query resolution is not going to be usurped by chatbots, suggesting that a balance will need to be maintained by service providers if they are to keep clients happy.

    1Forbes (April 13, 2017) What will we do when the world's data hits 163 zettabytes in 2025?

    2Finextra (January 10, 2018) How AI is changing risk management and compliance

    3Eurofinance – The 7 signs of intelligent treasury

    This page is about:

    South-South trade
    Trade between developing countries is evolving
    Join the conversation?

    Join our Linkedin group to get an unparalleled view of macro and microeconomic events and trends from a bank that is a leader in both developed and emerging markets.