Start with a reviewed CSV for each statement. Keep copies with matching headers, retain the source of every row, and append them in Excel. Check overlapping statement periods before using the combined totals. BankStatementMate processes one file at a time; the combining work described here happens in your spreadsheet.
1. Prepare one account and currency at a time
Use a suitable CSV download from your bank when available. For supported PDFs or images, follow the statement extraction guide, check each result against its source and export CSV separately. Do not combine unreadable or incomplete exports and hope their totals will reveal every error.
Keep originals untouched. Make a working folder containing only the reviewed files for this task. Record each file's account label, currency, statement period, transaction count and opening/closing balances in a separate review sheet. Use a non-sensitive account alias instead of putting full account numbers into filenames.
Begin with one account and currency. Combining different accounts needs explicit Account and Currency fields on every row and separate checks for each group. Do not sum amounts across currencies.
2. Make the column meanings consistent
Power Query's Append operation stacks tables and matches columns by header name. A missing header produces null values in that column. Resolve differences such as “Withdrawal” versus “Debit” in the working copies only after confirming their meanings.
| Field | Review rule |
|---|---|
| Date and Description | Keep the original transaction meaning and source posting order. |
| Debit and Credit | Confirm money direction and numeric values; keep missing values distinct from zero. |
| Balance | Retain as a source checkpoint. Never sum running balances as transaction amounts. |
| Source file and source row | Preserve a route back to each original record before sorting or excluding anything. |
For dates and amounts, inspect the actual values before loading. Microsoft's data-type guidance explains using a locale when converting text into dates or numbers. A display format cannot fix a misread day/month order. Keep identifiers as text, and resolve missing years from the statement before conversion.
3. Combine the files in desktop Excel
Microsoft's folder import instructions require consistent file structure and warn that subfolder files can be included. In an Excel environment with this connector, choose Data → Get Data → From File → From Folder, select your working folder, and inspect the file list.
- Select Transform Data to filter the list to the intended filenames, extensions and folder paths.
- Combine the selected files. Check the sample file's delimiter, encoding and column interpretation.
- Leave “Skip files with errors” off while reviewing. Investigate failures rather than silently losing a statement.
- Keep the source filename column if present. Verify that you can still identify each row's source; stop and add that information if it is missing.
- Check the resulting types and row counts, then use Close & Load to load the review table.
Menu availability varies by Excel environment; this desktop folder workflow is not a promise of phone support. If that connector is unavailable, open each reviewed export in your spreadsheet and copy its data rows beneath one shared header, adding source labels as you go. Never paste repeated header rows into the transactions.
4. Check a fictional overlap example
Download our fictional statement A CSV and fictional statement B CSV. Both describe one invented deposit account in one currency, with positive Debit as money out and positive Credit as money in. A starts at 1000.00. These teaching files are not customer statements or evidence of an Excel app test.
| Source / data row | Date | Description | Debit | Credit | Balance | Decision in this example |
|---|---|---|---|---|---|---|
| A / 1 | 2026-09-01 | SYNTHETIC Salary | 2500.00 | 3500.00 | Keep | |
| A / 2 | 2026-09-02 | SYNTHETIC Rent | 1200.00 | 2300.00 | Keep | |
| A / 3 | 2026-09-03 | SYNTHETIC Coffee | 4.50 | 2295.50 | Keep | |
| B / 1 | 2026-09-03 | SYNTHETIC Coffee | 4.50 | 2295.50 | Confirmed repeated A / 3 | |
| B / 2 | 2026-09-03 | SYNTHETIC Coffee | 4.50 | 2291.00 | Keep: separate purchase | |
| B / 3 | 2026-09-04 | SYNTHETIC Refund | 10.00 | 2301.00 | Keep |
We deliberately authored B / 1 as the same event as A / 3, while B / 2 is another purchase. Equal dates, descriptions and amounts do not establish duplication in a real file. Compare bank references, source pages and posting sequence; if identity is uncertain, flag the pair for review rather than deleting it.
Appending these three A rows and three B rows gives six input rows. Excluding the one confirmed repeat leaves five. Credits total 2510.00; debits total 1209.00. Net movement is 1301.00, and 1000.00 + 1301.00 = 2301.00. The sample's row counts and arithmetic were independently checked; Power Query itself was not executed for this example.
5. Preserve the review trail
Keep a raw combined table and a separate reviewed result. Record which source row was excluded, which retained row it repeats, and the reason. Avoid a blanket “Remove Duplicates” operation on only date, description and amount: it would also remove the separate coffee purchase in this example.
Check row counts per source before exclusions, then reconcile the retained count to documented exclusions. For the same account, inspect the boundary between statement periods as well as dates within an overlap. A chronological sort alone may reorder same-day entries, so retain source-row order for balance checks.
6. Save and decide the next step
Save the working review as XLSX so the workbook can retain your sheets and review information. Keep it outside the input folder to avoid collecting derived copies later. Before refreshing, inspect the input list again: replacement exports and extra copies can change the result.
Combined totals and a matching closing balance do not prove completeness or agreement with independent books. Use the Excel reconciliation guide for that separate task. If an accounting application is the next destination, prepare a separate import copy to its requirements rather than uploading this review workbook blindly.
BankStatementMate offers supported statement extraction and Excel/CSV export, one file at a time. It does not merge separate statements, certify duplicates or perform reconciliation. Use the CSV extraction workflow only when you still need reviewed rows. This guide supplies instructions and fictional samples; it does not supply an Excel license or run the merge for you.
Exports contain sensitive financial information. Keep files in an authorized location. Uploading a workbook to a cloud service shares that copy with the chosen service under its settings and terms; browser-local extraction does not make later sharing private automatically.