How does browser OFX, QFX, and QBO parsing work without uploads?
The converter probes the first 2,048 bytes of each uploaded file for OFXHEADER, CHARSET, and ENCODING markers before parsing statement data inside a local Web Worker. Both legacy OFX 1.x SGML streams and strict OFX 2.x XML files decode entirely on your device up to 25 MB per file, so zero bank records ever travel over the network.
Banks still distribute many checking and credit card downloads as OFX 1.x files structured under W3C Overview of SGML Resources rules, where leaf elements such as DTPOSTED and TRNAMT omit closing tags. Our parser inserts safe leaf closures and decodes XML and HTML character entities before extracting each transaction row.
- Bank statements: Reads checking and savings ledgers from
<STMTRS>containers alongsideBANKID,BRANCHID,ACCTID, andACCTTYPEmetadata. - Credit cards: Extracts card activity inside
<CCSTMTRS>blocks and assigns a unifiedCREDITCARDaccount type when card tags differ from checking downloads. - Investment blocks: Captures cash activity inside
<INVSTMTRS>containers or bare<STMTTRN>elements placed directly under<OFX>.
For every block, the reader also extracts statement bounds (DTSTART, DTEND), ending LEDGERBAL and AVAILBAL balances, and ISO 4217 CURDEF currency codes, defaulting to USD if omitted. Posting timestamps normalize cleanly to ISO 8601 dates (YYYY-MM-DD) regardless of bank timezone offsets appended to DTPOSTED.
When CHARSET specifies 1252, WINDOWS-1252, or ISO-8859-1 instead of UTF-8, the byte decoder switches character tables automatically so European payee characters stay readable. If your archive comes from older desktop Quicken backups rather than Open Financial Exchange exports, you can convert legacy Quicken QIF exports directly.
Comparing standard CSV, Excel, QuickBooks Online, and Xero exports
Different bookkeeping platforms expect incompatible CSV headers, date orders, and byte prefixes even when reading the exact same bank feed. Choosing a dedicated preset under Export format avoids manual column renaming before ledger import and keeps numeric signs consistent across every statement batch.
| Export format | Date format | BOM & structure | Best accounting target |
|---|---|---|---|
| Standard CSV (.csv) | ISO YYYY-MM-DD | UTF-8 BOM, 11 or 12 columns | Excel desktop and archival audits |
| Excel workbook (.xlsx) | Native ISO text | Two sheets with numeric amounts | Multi-account balance review |
| QuickBooks Online CSV | US MM/DD/YYYY | No BOM, 3 or 4 columns | QuickBooks Online bank upload |
| Xero Bank Statement CSV | ISO YYYY-MM-DD | No BOM, 6 mapped columns | Xero manual bank statement import |
Standard CSV and Excel multi-sheet columns
Standard CSV follows IETF RFC 4180 CSV specification quoting rules so commas inside payee strings or memos never shift columns, writing Date, Type, Payee / Name, Memo, Check #, Amount, Currency, Account ID, Account Type, FITID, and Duplicate FITID. Missing payee names fall back cleanly to MEMO or TRNTYPE.
Selecting Excel workbook (.xlsx) builds two structured worksheets inside one download so reviewers avoid opening separate summary files. Transactions stores formatted numeric amounts ready for spreadsheet formulas, while Accounts & Balances records institution IDs, period start and end dates, ledger balances, available balances, and transaction counts per account.
QuickBooks Online and Xero CSV import presets
QuickBooks Online CSV exports -quickbooks.csv files formatted with US MM/DD/YYYY dates and clean Date, Description, Amount columns without trailing metadata. Xero Bank Statement CSV generates -xero.csv files using YYYY-MM-DD dates and exact *Date, *Amount, Payee, Description, Reference, and Check Number column headers.
Teams handling corporate treasury reporting alongside retail bank statements frequently encounter mainframe cash-management logs instead of QuickBooks files. When your finance department receives daily bank balance transmissions from treasury portals, switch to our treasury BAI2 to CSV and Excel parser.
Handling FITID deduplication, split columns, and ZIP bundles
Overlapping monthly bank downloads frequently repeat the final three or four transactions from a previous export window. Dropping raw files together into a spreadsheet without unique identifier checks risks double-counting expenses or deposits during month-end bank reconciliation.
Account-scoped FITID duplicate detection
When Deduplicate matching FITID rows is checked, the worker indexes each entry by combining account ID and FITID so identical trace numbers in separate checking and card accounts never collide. Desktop spreadsheet tools rely on the Byte Order Mark described in Unicode FAQ on UTF-8 and Byte Order Marks to render foreign characters accurately.
Duplicates are flagged with a DUP badge in the interactive preview table and counted in the Duplicate FITIDs summary card before export. Clicking Add files lets you append additional monthly OFX, QFX, QBO, or ZIP archives to the active session while keeping deduplication active across every loaded statement.
Tip: Click Load sample OFX (
harbor-checking-sample.ofx) to test FITID deduplication, split columns, and Excel balance sheets locally before batching overlapping quarterly ZIP archives.
Split Debit and Credit columns and masked accounts
Checking Split Debit and Credit columns turns signed TRNAMT values into positive Debit and Credit fields for Standard CSV and Excel, or Credit and Debit headers in QuickBooks CSV exports. Negative card charges or withdrawals populate Debit cleanly without leading minus signs, while deposits populate Credit.
The converter unpacks ZIP bundles up to a 64 MB uncompressed safety cap, automatically ignoring __MACOSX resource directories while parsing nested .ofx, .qfx, .qbo, and .xml files. Teams consolidating global subsidiaries alongside North American accounts can also process European ISO 20022 XML bank feeds ready for Excel audits.
Toggling Mask account numbers (****1234) redacts leading digits across account numbers inside both the preview summary and exported spreadsheets while leaving the last four digits visible for bank reconciliation. This control is off by default so automated accounting imports keep full account references unless redacted intentionally.
Client-side privacy, CSV safety, and OFX round-trip limits
Financial statements expose routing context, account balances, counterparty names, and payroll memos that should remain strictly inside your browser session. Every byte probe, SGML cleanup pass, deduplication check, and workbook build executes 100% locally in a Web Worker with zero external API requests.
Standard CSV exports pass every text cell through formula sanitization aligned with MITRE CWE-1236 formula neutralization standard before saving to disk. Strings beginning with =, +, -, @, tab, or carriage-return characters are safely neutralized so untrusted payee names never execute spreadsheet formulas when opened in Excel.
Each uploaded file supports up to 25 MB to keep browser tab memory responsive, while the interactive table preview displays the first 300 rows for smooth filtering and scrolling. Clicking Export transactions always writes every valid non-duplicate row across all loaded statements rather than truncating at the 300-row display limit.
When you finish auditing or recategorizing rows in Excel and need to generate fresh statement files for desktop Quicken or ledger re-import, use our CSV to OFX converter. Technical file conversion here is provided for informational purposes only and does not constitute accounting, tax, or legal advice; review final ledger entries with a qualified accountant.