A finance systems administrator at a mid-size manufacturer got an email last month from the company’s German house bank. Starting in the first week of the new quarter, the daily MT940 statement feed that had fed Business Central’s bank reconciliation for six years would stop. In its place: a CAMT.053 XML file, delivered on a different schedule, with a different structure. The email asked, almost as an afterthought, whether the ERP would “just work” with the new format. It was not a rhetorical question the bank expected an answer to. It was a notice.
That conversation is playing out across Business Central shops right now, and it is worth understanding why the Business Central CAMT.053 migration is arriving in 2026 specifically, what Business Central’s native tooling actually does with a CAMT.053 file, and where that tooling quietly runs out before a consultant finds out the hard way during a month-end close.
Why the Business Central CAMT.053 Migration Is Landing Now
The ISO 20022 migration has two distinct halves, and they are not moving on the same clock. The first half concluded in November 2025, when SWIFT ended coexistence for cross-border payment instructions. MT103, MT202, and the rest of the legacy payment message types stopped flowing in their original form; everything now moves as ISO 20022 MX, primarily pacs.008 and pacs.009. Institutions that kept sending legacy MT payment formats after that date found them auto-converted by SWIFT, with additional charges applied starting January 1, 2026.
The second half, covering statement and reporting formats rather than payment instructions, was deliberately left on a slower and less coordinated timeline. MT940 and MT950 (account statements), MT900 and MT910 (debit and credit notifications), and MT941 and MT942 (intraday reporting) are being replaced by camt.053, camt.054, and camt.052 respectively, but SWIFT is not converting MT9xx traffic to camt format automatically. Each bank negotiates its own cutover, account by account, on whatever internal schedule its own core banking migration allows. That is precisely why one Business Central customer converted a year ago without much drama, another is getting the notice this quarter, and a third will not hear anything until 2027. There is no single migration date to plan around. There are dozens of them, one per banking relationship, arriving on nobody’s schedule but the bank’s.
The Business Central Side of the CAMT.053 Migration
Business Central handles SEPA CAMT import through a Data Exchange Definition, a mapping layer that translates XML element paths in the bank’s file into fields across three tables: the General Journal table, the Bank Acc. Reconciliation header, and the Bank Acc. Reconciliation Line table. The mapping itself is fairly mechanical once you see it laid out. The transaction amount and its sign come from Stmt/Ntry/Amt combined with the credit or debit indicator at Stmt/Ntry/CdtDbtInd, with a configurable negative-sign identifier (typically set so that DBIT entries post as negative). Posting and transaction dates come from Stmt/Ntry/BookgDt, which can arrive as either a plain date or a full timestamp; Business Central accepts both. Payer information for matching comes from Stmt/Ntry/NtryDtls/TxDtls/RltdPties/Dbtr/Nm, and the free-text remittance information that drives most of the automatic matching logic comes from Stmt/Ntry/NtryDtls/TxDtls/RmtInf/Ustrd.
The field that actually determines whether any of this works is easy to miss: Stmt/Ntry/NtryDtls/TxDtls/Refs/EndToEndId. This is the identifier Business Central relies on to treat each line as unique within an open reconciliation. If a bank’s first-generation camt.053 export leaves this element empty, or populates it with a placeholder value repeated across an entire batch, the failure mode is not an import error. The file parses without complaint. What actually happens is worse in a subtler way: lines with a missing identifier get silently dropped, and lines that share a duplicated identifier with something already posted will import but fail automatic matching one by one, with no message pointing back to the cause. A team that has not specifically gone looking for this will spend a reconciliation cycle assuming the new statement format is simply less complete than the old one, when the real problem is a configuration gap that a five-minute look at the raw XML would have caught before go-live.
Where the Generic Definition Runs Out
Business Central ships a generic SEPA CAMT data exchange definition, and it is a reasonable starting point, not a finished solution. CAMT.053.001.08 is a schema, not a single fixed file layout, and individual banks populate it with their own dialect: some banks include optional elements Business Central’s default mapping ignores, some omit fields the default mapping expects, and namespace versions drift between institutions even within the same country. In practice, most multinational Business Central deployments end up cloning the generic definition per banking relationship rather than trying to maintain one shared version that quietly breaks every time a new bank is onboarded with slightly different habits.
It also helps to be honest about geographic scope. CAMT adoption is heavily concentrated in the EU and a handful of other markets that have committed to the SEPA-aligned schema. Plenty of banks elsewhere, particularly smaller regional institutions in North America and parts of Asia, are staying on BAI2 or proprietary flat-file formats for the foreseeable future, with no camt.053 timeline at all. A single Business Central tenant with entities in four countries can easily end up running MT940, camt.053, and BAI2 in parallel for years, not as a transitional state but as the normal operating condition. For organizations with a long tail of smaller regional banking relationships, this is where an ISV connector such as AMC Banking 365 Fundamentals earns its licensing cost: it converts against several hundred bank-specific format variants through its own service rather than asking an internal team to build and maintain a bespoke Data Exchange Definition for every regional bank quirk. That trade-off is worth evaluating deliberately, since routing statement data through a third-party conversion service is a data-governance decision as much as a technical one.
A Practical Migration Checklist
The teams that get through a bank-driven format change without a rough reconciliation cycle tend to do a few things in a specific order. They start by inventorying every bank account’s current import format code and confirming the actual cutover date per relationship, since assuming all banking relationships move together is the single most common planning mistake. Before the cutover, they request an actual test camt.053 file from the bank, not a sample pulled from vendor documentation, and run it through the Data Exchange Definition in a sandbox company where a bad mapping does no damage. They open that test file as raw XML and check the EndToEndId population line by line rather than trusting that because a mapping worked for one bank’s camt.053 export, it will behave identically for another bank using the same nominal schema version. Where any deviation shows up, they clone the generic SEPA CAMT definition into a bank-specific version rather than editing the shared default, so a fix for one relationship cannot silently change behavior for every other bank account using the same code. They keep the outgoing MT940 or BAI2 definition active and assigned for as long as the bank is willing to run both feeds in parallel, because a same-day cutover with no fallback window turns any mapping problem into an immediate reconciliation gap. And before retiring the old feed for good, they re-test automatic matching rates against a full month of real transaction volume on the new format, since a quiet drop in match percentage is a far more reliable early warning of a mapping issue than an outright import failure would be.
What Copilot’s Bank Reconciliation Assist Does and Doesn’t Solve Here
Business Central’s Copilot-assisted bank reconciliation feature, currently in preview, is genuinely useful for what it does: it analyzes unmatched transaction descriptions and ledger entries to suggest matches and likely G/L accounts, including scenarios where a single deposit needs to be split across several invoices. It is worth being clear about its boundaries in this specific context, though. It operates entirely downstream of the Data Exchange Definition, on statement lines that have already been successfully parsed into the system. If a mapping problem is dropping or mismatching lines before they ever reach the reconciliation worksheet, no amount of AI-assisted matching afterward will surface or correct that. The feature also performs best when transaction descriptions and G/L account names share a single consistent language, which is worth double-checking specifically during a format changeover, since a bank’s new camt.053 export sometimes reformats descriptive text differently than its old MT940 feed did.
The Deadline Is the Bank’s, Not Yours
None of this is a discretionary upgrade a Business Central team can schedule around a quieter month. It is a mapping and validation project with a deadline set by someone else’s core banking migration, arriving on a per-relationship basis rather than a single company-wide date. The organizations handling it well are the ones treating each bank’s cutover notice as a trigger for testing, not as an announcement to file away until the old feed actually stops. Routeget has walked several clients through exactly this kind of parallel MT940 and camt.053 coexistence during multi-entity rollouts, and the pattern is consistent: the mapping work is small, but only when someone does it before the bank forces the issue rather than after.
#BusinessCentral #CAMT053Migration #BankReconciliation #ISO20022 #DataExchangeDefinitions #ERPIntegrations