How to convert a bank statement PDF to Excel without breaking the columns

4 August 2026

Short answer: use a converter that knows what a bank statement is, then verify the result against the statement’s own running balance. The conversion is the easy half. Knowing that the Excel file contains every transaction — and only the real ones — is the half that matters, and it is the half almost every tool skips.

Why bank statements break generic PDF converters

A PDF does not contain a table. It contains a set of text fragments, each placed at an x/y coordinate on a page. The neat columns you see are visual alignment, nothing more. Any converter has to reconstruct the table by inferring which fragments belong to the same row and the same column — and Indian bank statements are unusually hostile to that inference:

• Narrations wrap onto multiple lines. One transaction can occupy three or four visual lines. A naive parser emits three rows, two of which have no amount. Axis statements make this worse by wrapping upward, so a continuation line sits above the row it belongs to.
• Amount columns are positional, not labelled. Debit and credit are frequently distinguished by horizontal position alone. Shift by a few points — a longer narration, a wider font — and a payment becomes a receipt.
• Page furniture looks like data. Headers repeat every page, footers carry running totals, and corporate formats such as HDFC ENet append payment-advice tables that look exactly like statement rows.
• Numbers are Indian-formatted. 1,23,456.78 is lakh-grouped, and a converter that assumes thousands grouping will silently misread it. Trailing Cr/Dr markers carry the sign in some formats.

This is why the output of a general PDF-to-Excel tool tends to look 90% right and take an hour to fix — and why the 10% that is wrong is invisible.

The four options, honestly

1. Copy and paste. Free, immediate, and reliably the worst. Copy order follows the PDF’s internal text order, not the visual layout, so rows interleave and multi-line narrations collapse. Fine for eyeballing five transactions, unusable for a year.

2. Excel’s built-in PDF import. In Microsoft 365: Data → Get Data → From File → From PDF. Power Query detects tables page by page. It is genuinely useful and it is free, but on most Indian statement layouts you will be appending 40 page-tables by hand, re-promoting headers, and repairing wrapped narrations. Worth learning if you convert the same format every month; painful otherwise.

3. Generic online PDF-to-Excel converters. They handle simple grid tables well. They know nothing about debits and credits, so they cannot tell you when they have got it wrong — and they upload your client’s statement to their server to do it. See is it safe to upload bank statements.

4. A statement-aware parser. Something that knows a statement has a date, a narration, a debit or credit, and a running balance can use that structure to check its own work. This is the only option that can tell you the output is complete rather than hoping so.

The check that makes the output trustworthy

Every bank statement is self-verifying, and almost nobody uses this:

previous balance + credit − debit = current balance

Apply it to every row. If the chain holds from the opening balance to the closing balance, and the final figure matches the closing balance printed on the statement, then no transaction was dropped, duplicated or misread. If a row breaks the chain, you have found the exact row to inspect — instead of discovering the problem three weeks later during a reconciliation.

TieOut runs this check on every conversion and flags failures rather than hiding them. A fully green table is not a claim that the extraction worked; it is arithmetic proof.

Doing it with TieOut

1. Download the native PDF e-statement from netbanking — not a scan or a phone photo. The text is already inside a native PDF, so nothing has to be guessed from pixels.
2. Open tieout.in and drop the file in. If it is password-protected, enter the password when prompted; decryption happens locally in your browser.
3. Check the validation summary. Green means every row ties out.
4. Export Excel or CSV — or Tally XML if the destination is Tally Prime.

Nothing is uploaded at any step. Switch off your Wi-Fi before step 2 and the conversion still works, which is the simplest way to confirm that for yourself.

One habit worth keeping

Before you use a converted file for anything that matters, compare three numbers against the PDF: the opening balance, the closing balance, and the transaction count. It takes fifteen seconds and catches the entire class of errors where a page silently failed to parse.