Open an OFX statement in Excel, correctly

Excel will open a CSV happily and still get it wrong in four specific ways. Each has a setting here that prevents it.

1. Every row lands in one cell

The most common complaint, and it is not a broken file. Excel splits a CSV using your machine's list separator, which is a semicolon across most of Europe and a comma in the UK, US and India. Hand a comma-delimited file to an Excel that expects semicolons and every row arrives as a single cell.

Choose Semicolon under column separator before downloading and it opens correctly on a double-click. Nothing else needs to change.

2. Accented names arrive as mojibake

CAFÉ becomes CAFÃ‰. Excel does not assume UTF-8; without a marker it falls back to the system code page. The marker is a byte-order mark, and the Excel-safe accents option writes one. It is on by default.

3. Dates come out as the wrong month

03/04/2026 is 3 April in most of the world and 4 March in the United States, and Excel decides which using the locale of whoever opens the file. A statement that crosses a month boundary can therefore be silently rewritten.

Setting	Writes	Use when
ISO (default)	`2026-03-04`	Always, unless something downstream insists otherwise. It cannot be misread and it sorts correctly as text.
Day first	`04/03/2026`	UK, EU, India — software that demands a local format
Month first	`03/04/2026`	US software that demands it

If Excel shows ISO dates left-aligned, it has read them as text. Select the column, Data → Text to Columns → Finish, and it will convert them to real dates in one step.

4. The amount column will not sum

This one is worth understanding because it is invisible. A CSV field beginning =, +, - or @ is treated by Excel as a formula, which is a genuine security problem: anyone who can send you money can choose the description on the transaction, and therefore the contents of that cell.

The usual defence is to prefix such fields with a tab — but applied carelessly that also catches every negative amount, turning -42.75 into text. The column then looks perfectly normal and refuses to add up. Here the guard is applied only to fields that are not plain numbers, so payees are defused and debits stay numeric.

A quicker route than Text to Columns

Convert and download the CSV.
In Excel: Data → From Text/CSV rather than double-clicking the file.
Set the delimiter and, under Data Type Detection, check the preview before loading.

This route shows you the parse before it commits, which is the difference between noticing a wrong column today and finding it in a reconciliation next quarter.