Nearly three out of four treasury practitioners cite cash management and forecasting as their top priority, as per the AFP 2025 Treasury Benchmarking Survey. Meanwhile, according to the PWC 2025 Global Treasury Survey, only 57 % use a Treasury Management System (TMS) for exposure management, with 36% still incorporating manual processes.
Modern treasury teams are no longer settling for "good enough". They are leveraging hybrid connectivity, automated structures, integrated forecasting, and robust governance in an effort to turn treasury into a strategic growth driver. Here are five best practices to make this a reality:
The first and most critical step is gaining real-time visibility into your treasury operations.
A lot still runs on MT940 in Europe; yet manual MT940 downloads are batch processing in a world that runs in real time. With MT940 messages, you typically only see end-of-day positions, which can create reconciliation backlogs and may increase operational risk through manual file handling.
Bank APIs for real-time treasury can provide up-to-the-minute account data for cash positioning and payments. In contrast to SFTP host-to-host connections delivering MT940 messages, APIs act as a real-time messaging layer, potentially allowing treasurers to refresh positions intraday, trigger alerts, and automate cash sweeping as events occur.
For high-volume environments, a hybrid model often works best. Use APIs for intraday balances and host-to-host SFTP for bulk statements and payment files. This combination can help give you comprehensive cash visibility while maintaining the robustness needed for audit trails and large transaction volumes.
Centralising all bank data into a single dashboard can help eliminate the need to log into multiple different portals. A treasury management system links to each bank via the available protocol and pulls statements automatically, aggregating balances by entity, region, bank, and currency.
Scheduled pulls combined with intraday APIs can update balances throughout the day, potentially removing manual downloads and spreadsheet consolidation. The system converts all balances into a reporting currency using live FX rates, aiming to give a near-instant group-level position across all currencies.
After real-time visibility is enabled, liquidity can be managed more proactively by:
Optimising cash flow management through automated pooling can let the group fund internal deficits before drawing on external credit lines, potentially lowering net interest expense.
In a physical pool, a company typically automatically sweeps end-of-day surpluses from subsidiary accounts into a header account at the holding or group level.
Notional pooling, on the other hand, “virtually” offsets credit and debit balances across entities, so interest is calculated on the net position without physically moving funds.
Both structures can help reduce idle balances and may allow a surplus entity to fund deficit entities inside the group, potentially cutting reliance on costly overdrafts.
It is generally considered best practice to involve automating pooling with clearly laid out tax and transfer-pricing policies, monitored through a TMS that shows group-wide surplus and deficit.
For multinational groups, the goal is seeing not only where cash sits, but which currencies create risk so hedges can be executed on time. Modern treasury teams often use a TMS that automatically consolidates foreign exchange (FX) risk from various sources such as the company’s ERP, forecasted cash flows, bank balances, and intercompany loans.
Automated rules can then flag when exposures breach defined limits or when market rates hit trigger levels, potentially enabling timely execution of hedges and other mitigation strategies. For a deeper dive into managing cross-border payments, read our article on international payment management.
The AI revolution now enables improved forecasting accuracy in Treasury. According to industry vendors and research, machine learning models may improve short-term cash forecast accuracy by up to 50 % compared to manual spreadsheet methods, though results can vary significantly depending on implementation and data quality.
AI can therefore improve forecasting accuracy, both directly and indirectly, as well as scenario planning overall:
Scenario planning adds a "what if" layer on top of your baseline forecast so you can see how liquidity might hold up when markets move against you. Treasury teams can model severe but plausible shocks like revenue drops, payment delays, FX swings, rate spikes and simulate their potential impact on cash flows and covenant compliance.
The real value often comes when treasury uses scenario results to pre-agree contingency actions and embed this into regular governance. AI-powered tools, such as Embat's TellMe, can accelerate this process by surfacing insights from your data.
The financial close is a dreaded exercise in most finance departments, as every system must perfectly reconcile to produce accurate financial statements. For more on how automation can improve decision-making, see our guide to automated banking reconciliation. Treasury best practices here include:
Modern treasury teams often receive statements in multiple formats, while the ERP needs a single, consistent structure. A suitable approach is to route all statement files through a central normalisation layer that maps each bank's format into one internally accepted format.
This normalisation thereby harmonises key fields—such us value date, booking date, amount, currency, and transaction codes—so matching rules can be applied consistently. Richer data from ISO 20022 can mean better information about the payment for reconciliation, potentially enabling more efficient matching and fewer manual processes.
Waiting until month-end to reconcile bank accounts can turn routine control into a stressful bottleneck. Daily automated reconciliation aims to ensure new bank transactions are matched against ERP entries continuously, so exceptions can be identified and resolved while fresh.
This may reduce suspense items, help uncover process issues earlier, and improve mid-month reporting accuracy. For month-end close, many accounts may already be reconciled, so finance typically only deals with residual exceptions, potentially shortening close timelines and freeing capacity for analysis rather than firefighting.
The governance layer is one of the most important aspects of a Treasury function. Learn more about how digitisation supports fraud prevention in our article on treasury digitisation and fraud prevention. Its most critical aspects are:
Segregation of duties, often referred to as the four-eyes principle, aims to ensure no single person can create, approve, and release a payment end-to-end. In a robust setup, a TMS can enforce role-based access so users only perform specific tasks aligned to their job function.
Configuring tiered approval workflows based on rules like amount, entity, or payment type is critical.
For example:
Robust segregation of duties and procedures help reduce internal fraud risk, may prevent unauthorised payments, and can provide clear evidence for auditors.
Every hop between your TMS and ERP and any external system such as bank APIs should use secure, encrypted channels to protect payment data in transit. Avoiding the manual manipulation of payment files is critical, as every manual intervention introduces the risk of human error or fraud. Payment files should flow straight from the TMS to the bank through controlled, automated interfaces, with integrity checks confirming files have not been tampered with. Together, automated secure transmission and strong governance are important to protect the integrity of the data.
Adopting these tools is not just about efficiency; it is about freeing up the Treasurer to become a strategic partner to the board. The role of treasury is continuing to evolve into a more strategic, data-driven partner critical to enterprise value creation.
The combination of hybrid connectivity, automated cash pooling, integrated ERP forecasting, continuous reconciliation, and embedded payment governance can help create a modern treasury stack that can scale with growth. When treasury moves from reactive firefighting to proactive planning, the CFO will gain a trusted partner who can model stress scenarios, optimise funding costs, and provide real-time liquidity insights when needed.
That shift from cost centre to value creator is widely regarded as the ultimate best practice — and it's what helps professionals like Tom finally leave the office on time.
At Embat, we understand the pain points that can keep treasury teams working late. Our all-in-one treasury management platform is designed to deliver real-time bank connectivity, automated cash forecasting, AI-powered reconciliation, and centralised payments—all integrated with your ERP and designed for teams across the EU and the UK.
Whether you are managing multi-entity cash pools, automating your month-end close, or building board-ready liquidity forecasts, Embat's platform is designed to help you save up to 75 % of your team's time, based on customer feedback.
Book a demo to see how leading finance teams aim to turn treasury best practices into competitive advantage.
Corporate treasury best practices generally refer to processes, policies, and technologies that aim to optimise cash management, liquidity, risk mitigation, and compliance. This typically means real-time cash visibility via hybrid connectivity, automated cash pooling, integrated ERP forecasting, continuous reconciliation, and robust payment governance.
Connect all bank accounts to a TMS using a hybrid approach: that combines real-time APIs for intraday balances with host-to-host SFTP for high-volume statements. This can help eliminate manual portal logins and aims to give you a single, consolidated view of global cash positions.
Physical cash pooling involves actual movement of funds, sweeping surplus balances into a central header account. Notional pooling leaves funds in local accounts but offsets them virtually so the bank calculates interest on the net balance. Both can help reduce idle cash and external borrowing.
ERP integration feeds live transactional data directly into your forecast model, replacing manual spreadsheet updates. Automated data flow may improve forecast accuracy by approximately 30% in some implementations, can reduce errors, and may enable both direct and indirect forecasting from a single source of truth.
Consider implementing segregation of duties, tiered approval workflows, role-based access controls, automated alerts for vendor bank-detail changes, and end-to-end encrypted transmission between the TMS and banks. Best practice generally includes eliminating manual file editing and maintaining full audit trails.
This article has been prepared using sources from the AFP 2025 Treasury Benchmarking Survey, PwC 2025 Global Treasury Survey, SWIFT ISO 20022 documentation and other sources. All statistics and figures were current as of the date of publication and may have since changed. The strategies and approaches discussed are for informational purposes only and may not be suitable for all businesses. Results from implementing any strategy will vary by organisation.
