QIF to CSV Converter

A QIF file looks deceptively simple when you open it: line after line of a single letter followed by a value, with a caret every so often. That structure is the reason a spreadsheet cannot read it. There is no delimiter, no header row, and no fixed number of lines per transaction — a record might be four lines or fourteen depending on whether it has a memo, a check number, a cleared flag, or a split. Drop a .qif file into Excel and you get one column of text where each cell begins with a letter, which is faithful to the file and useless as data.

Converting inverts the shape. The parser walks the file line by line, collecting fields into a transaction until it hits a caret, then flushes that transaction as a row and starts a new one. The date on the D line is normalised to YYYY-MM-DD, the amount on the T or U line is stripped of currency symbols and separators, and the payee and memo lines are carried through with proper CSV quoting so that a description containing a comma survives the trip intact. Header blocks such as !Type:Bank contribute no transaction and therefore no row.

The people who need this generally need it once and urgently. A Quicken subscription lapses and years of history are stuck in a format nothing else opens. An accountant asks for a client's transactions and gets a QIF they cannot use. Someone leaves Microsoft Money — discontinued for well over a decade — and needs the data somewhere modern. Or a spending analysis needs a pivot table, and the finance app's built-in reports refuse to cut the numbers the required way. In each case CSV is not the destination so much as the universal way station.

Dates came through unchanged instead of as YYYY-MM-DD. Older Quicken builds sometimes write dates with an apostrophe before a two-digit year, in the shape D 3/ 4' 5, padding with spaces rather than zeros. That is not a slash-separated date and it passes through as-is. A find-and-replace in a text editor before converting is the quickest fix; alternatively re-export from a newer Quicken build, which writes four-digit years.

A split transaction shows only its total. Splits are stored as S, E, and $ lines nested inside a single record, so one transaction may cover five categories. The CSV gives one row per record with the total on it. If the category-level detail is what you need, it has to be pulled out of the original file — the CSV cannot represent it.

A file with several accounts merged into one undifferentiated list. QIF marks account boundaries with !Account header blocks that carry no transaction data. Since the CSV has no account column, transactions from a checking account and a credit card end up interleaved in one sheet. Split the QIF at the !Account lines and convert each part separately if you need them kept apart.

Nothing was found in the file. Every exported row needs a D or a T line, so a file yields no output if it contains only list data — a QIF of categories, accounts, or a memorised payee list rather than transactions. Opening it in a text editor and looking for lines beginning with D confirms this in seconds.