Reconciliation Automation: Shaping a Fit-for-Purpose, High-Performing Finance Function

Achieving balance in life is essential for well-being, and it takes constant close monitoring to get right. The same goes for the financial health of a business. Account reconciliation – comparing internal financial data against external data to ensure the figures are fully balanced – helps detect errors, missing payments, or fraud. It’s vital to check everything is aligned and fix any discrepancies. Letting it slide can put a business’s finances, integrity and regulatory requirements at risk.

While reconciliation might look straightforward on the surface, it is an increasingly complex process in the fast-evolving world of digital payments where you must match core system data against multiple suppliers, partners, purchasers, and others. 

This article looks at how Reconciliation Automation can iron out the complexities to power faster, more frictionless cross-border payments, ensuring businesses stay in the absolute best shape to grow.

Let’s use the example of Jane, who manages the finance department for a major e-wallet provider. Jane’s customers consistently move funds into and out of their wallets to pay for goods, “cash out”, and transact with friends. And her customers are doing this rapidly throughout the day. 

Suppose Jane’s core system cannot actively match the flow of information about the balance of her customer’s wallets to the sources of record. In that case, there are one of three possible implications:

1. Jane could have miscalculations in forecasted cash flows, causing transaction failures in her customer traffic and transaction costs (FX, if cross-border).
2. Core accounting could be flawed, and Jane’s team will be left with a problematic audit of their company financials, with possible penalties and write-offs.
3. In the case of cross-border, Jane could be left holding additional “emerging” or “non-operational” currencies that will significantly impact her company’s balance sheet. Her team is not equipped nor interested in taking a view on any currency.

Difficulties in Conventional Reconciliation 

• Real-time data: Payments organisations used to do a “reconciliation exercise”, a financial accounting mechanism with monthly statements that cross-referenced core data. With the expansion of real-time processing and interoperability across systems, payment methods, and regions, data management has become imperative in payments. Reconciliation of accounts is no longer an exercise at the end of the month but a core anchor in the daily functions of payment processing finance teams.

• Cross-currency: When multiple currencies exist in the payments flow, reconciliation goes beyond the value of the transaction. What benchmark do you and your partner use for your conversion? What time did you calculate the buy rate for another currency? Are your commissions calculated at the same time? Are the rates different for principal value and commission? The questions go on and on. Cross-currency payments layer multiple elements within the reconciliation process that must be recognised to fully “match” the source of record to core systems. Treasury Management Systems must incorporate this component into the matching logic itself and workflows for finance teams to adjust reconciliation reports for these nuances. 

• Partners: Yes, it’s true, APIs have provided a solution for the real-time management of transaction receipt and reconciliation, and we’ll discuss this further below. However, it’s important to recognise, at this stage, that not all institutions have API integrations available, especially in emerging markets. Yet, real-time transaction demand remains constant, so solutions must be flexible to receive data in various ways. Whatever data source and information (for example, Excel, .txt file, SFTP, or email) a partner shares about transactions must be consistent on a period-by-period basis to systematise the reconciliation process.

• Transaction life cycle: Payments take on multiple statuses over their processing, and the result of that payment (for example, submitted, pending completed, declined, or reversed) is a consistent point of mismatch and discrepancy between internal ledgers and a partner record. Again, APIs solve many of these problems, but any number of reasons could lead to a “false completion”. And think about the impact: if the account has a completed transaction, and internally it was noted as declined, that means your balance is lower than you think, which can lead to dangerous scenarios such as overfunding or depleted balances.

Concepts to consider in Reconciliation Automation

• APIs where possible: API connectivity with a transaction partner, supplier or payments provider is preferable. But you must still ask additional questions: Will the API only include the partner’s balance? Or will it allow me to receive all transactions? If the API can receive all transactions, what will my internal system do with these transactions? Is there a mechanism to check every transaction for a core system mismatch? What will we do after we spot a mismatch? The questions are extensive because APIs, in themselves, are not solving reconciliations. It takes other operational processes and technical systems surrounding this API to solve mismatches and complete the reconciliation process.

• “Normalisation”: Before we start checking data of one place versus another, especially when multiple partners are involved, data must be presented in an apples-to-apples format. Systems are necessary to structure multiple forms of data rapidly in a manner that can be reconciled against a core platform. Systems require common rulesets to create “normalised” data from partner reporting, combined with machine learning to identify common attributes and extract information from complex and unique sources. 

• Pattern recognition: Are there commonalities with a partner or provider when attempting to reconcile transactions? Reconciliation engines should be flexible enough to spot systemic mismatches in data. Implementing a simple counting mechanism against a known ruleset is a good initial step (for example, there were 20 transactions with a mismatched commission value). Still, the services should evolve to capture new trends and make decisions about why transactions will be mismatched in the future. 

• Workflow tools: However automated the reconciliation systems can be, there will still be a need for human interaction at the point of resolution. Jane’s finance team will have to review the mismatch patterns across partners to determine a resolution (for example, reach out to a partner, make tactical adjustments, or fix potential API issues). Communication between finance, treasury, and support teams, who all have a crucial role in reconciliation, requires dashboard quality and other streamlined workflows. Operational procedures surrounding reconciliation require analytics, workflow management tools, and flexible alerting structures, just to name a fews.

Reconciliation may seem elementary on the surface, but there are many hidden challenges, especially in cross-border payments, where FX and a lack of API integration can add complexities. The team managing the operational process must address these intricacies systematically. To deliver a smooth and efficient service, Jane and her finance team have to consider reconciliation as a core process in their day-to-day job, not as an activity that happens at the end of each month. 

Like to learn more about how Thunes drives faster, more frictionless cross-border payments with Automated Reconciliation? Chat with our friendly team today.