HSBC Redefining Treasury Forum

As we head into the fourth quarter of a turbulent year, adapting to the fast-changing environment has become business as usual for corporate treasurers in the Middle East. Their top priorities are knowing where their cash is, guaranteeing liquidity when and where it’s needed, maximising returns on excess cash, and payment resiliency.

Grappling with increased economic volatility, many treasurers are seeking to improve visibility, control and resiliency, as well as lower costs, through centralising cash, risk, and banking operations. They are also investigating use cases for artificial intelligence (AI) deployment, such as cash flow forecasting, to free up their teams from repetitive manual processes, thereby improving workforce productivity and enabling strategic decision-making.

The Forum, held in September in Dubai, brought together more than 50 treasurers from diverse sectors in the region to share knowledge, experience and best practice in improving treasury operations.

Across industry sectors, participants identified several common challenges: forecasting collections and payouts due to a reliance on business units, spreadsheets and inconsistent data; managing multiple bank accounts and bank relationships; complex foreign exchange (FX) workflows to deal with suppliers and customers across different countries; payment delays; and arduous Know Your Customer requirements.

Bank connectivity remains a regional constraint, with limited interoperability in some markets, which impedes progress on automation and standardisation. In addition, many lean treasury teams struggle with multi-bank, multi-process administration.

Mobilising liquidity is also a challenge, as cash may be available at subsidiary level but difficult or slow to move because of regulatory, tax, documentation or banking restrictions in different jurisdictions.

To address these issues, many treasurers across the region are embarking on treasury modernisation programmes, moving from transactional to strategic treasury by automating, centralising and adopting AI.

The road to centralised operations

While some treasurers are satisfied with a decentralised treasury set-up, which allows for greater local flexibility, the majority are looking to increase centralisation to achieve better cash visibility, cost efficiency and standardised risk management. Many view treasury centralisation as essential for strategic decision-making.

But despite the desire to consolidate operations, a straw poll of the forum participants found that most are still in the early stages of their centralisation journey.

For treasurers currently in exploratory mode, a major hurdle is identifying where to start. For example, a treasurer from a large fitness company, which is expanding rapidly in Asia, questioned whether they should first look to centralise their operations in a free zone, such as ADGM, or automate their processes, which are currently managed by Excel spreadsheets in each country.

The company has an in-house – but limited – enterprise resource planning (ERP) system, so peers advised taking a gradual step-by-step approach by implementing one banking platform to improve visibility and follow on with host-to-host (H2H) connectivity to help with automation.

Consolidating cash and FX management is a common first phase of centralisation, followed by standardised transaction processing in payment factories, shared service centres and netting centres. The third phase involves centralised transaction processing, such as full in-house banking and regional treasury centres.

Several other participants also expressed their intention to begin with cash pooling. As a next step, they were considering payments-on-behalf-of, or POBO, arrangements.

But while cash pooling and liquidity concentration are a priority for many treasurers, execution is constrained by currency, tax and regulatory limitations. For example, POBO and collections-on-behalf-of arrangements are not permitted in some contexts. In addition, fragmentation across markets necessitate multiple local banking relationships and numerous bank accounts.

Several multi-country operating realities were highlighted during the forum, such as the need to concentrate collections received in local currencies into AED and/or USD, with the discussion focusing on the timing of transfers, buffering/minimum balances locally, and where to perform FX conversion (onshore versus offshore).

Undoubtedly, designing regional pooling, intercompany funding, governance and appropriate treasury structures requires guidance on local law, tax and regulations. To support our clients in making the right decisions, in partnership with KPMG, we have launched the UAE Treasury Centralisation Guide. The guide aims to help treasurers understand the regulations, account management, tax issues, and operating models within the UAE to ensure success for their business.

Supporting technology and systems

In general, topics around tech-enabled efficiency within treasury resonated with Forum participants. There was much debate about technology and systems used for treasury centralisation, such as ERPs, treasury management systems (TMS), cloud and bank AI automation. But as with cash pooling, the Middle East has distinct local challenges that need to be factored into any treasury technology roadmap.

Forum conversations focused on connectivity for payments and trade. There was a wide variety of set-ups among the corporate treasuries represented in the room. For example, a treasurer from a large agricultural machines manufacturer, which is in the early stages of treasury transformation with limited automation to date, said that they rely on HSBCnet for most of their activities.

Other participants reported using a variety of connection channels including Swift, H2H and application programming interfaces (APIs). One argued that the ideal set-up is using an e-channel connection with banks via their ERP as a first option, allowing them to use HSBCnet as a backup option.

Some highlighted the usefulness of API connectivity between ERPs and banking systems, which improves data quality and leads to better decision-making.

While implementing a TMS is usually reserved for more mature treasury operations that are further down the automation and centralisation path, many acknowledged the important role a TMS can play in helping them manage their business more efficiently. A treasurer from an energy services company outlined their journey, which began with automating payments in more than 15 markets and then focused on cleaning up their database in the second phase, as well as looking to adopt AI within their TMS.

There was general consensus that there remains a place for manual paper-based payments as an emergency payment option in light of recent escalations in regional tensions, which have resulted in system outages and remote working challenges.

Having a tried and tested Plan B is essential for ensuring resilience in critical activities, such as payments. Another straw poll of forum participants confirmed that the majority have a tested recovery plan in place in case of a payment outage.

AI applications

There were noticeably different levels of AI maturity among the forum participants, from the basic fear of missing out on AI/being in the exploratory phase to more structured and elaborate AI deployments depending on business requirements.

On the mature end, a representative from a consumer group reported seeing benefits with regards to improved efficiency in their treasury decisions and visibility over cash at click of a button. Interestingly, another talked about using AI to create a customised ERP, as well as working on other treasury projects.

Embedded AI functionality in ERP and TMS platforms are already bringing better visibility, reconciliation, fraud detection and sanctions compliance. Undoubtedly, treasury staff are using enterprise-wide available tools on their desktop to automate routine tasks, such as preparing reports, analysing data, providing commentaries, translating documents, and comparing bank proposals.

However, AI remains at an early stage of adoption for the majority, even among treasuries that have worked towards TMS implementation.

For example, despite being touted as a killer treasury use case, AI in cash flow forecasting continues to be challenging, as many remain unconvinced of the accuracy of the output. Most agreed that the top priority is to get the data right. It is commonly understood that if the starting data quality is poor, more advanced tooling won’t deliver value.

One participant openly discussed their unsuccessful experiments with large language models: the outputs were incorrect because the underlying data was faulty. In another reference to taking small steps when adopting new things, one participant reported focusing on MT940 statement messages last year, as an initial step to improve data consistency and downstream processing.

Some participants are introducing automation using AI tools such as Anthropic’s Claude and Microsoft Copilot to streamline processes. A few have already embedded Claude into their ERP systems, and reported improvements in reporting and visibility across global operations. They explained how they integrated with various systems, including emails, Excel and other raw data, to automate reporting and streamline workflows.

For one retail company, embedding Claude in their platform improved cash flow forecasting and vendor management. Another company reported using AI to display a real-time view of similar information on their internal dashboard, while another mentioned that they used AI to support provisions for vendors.

While some have identified and built strong use cases for AI agents, they faced challenges when trying to get sign off from CFOs, CROs or internal audit/compliance teams due to the perceived risks involved.

Continual transformation

Treasury modernisation should not be seen as an end state, but as an evolving journey. Moving from transactional treasury, automation and process efficiency should be seen as the first step. The next level is centralisation, which can be seen as a means to deliver measurable value to the business through flexibility and efficiency. In the third step, AI adoption can contribute to strategic decision-making and provide financial leadership to the organisation.