Why Your Bank Statement PDF Won't Convert
Three reasons a converter refuses your file, how to tell which one you have in five seconds, and what actually fixes each.
On this page
The short answer: your PDF is probably a picture. A PDF can hold real, selectable text, or it can hold a photograph of a page with no text in it at all. They look identical on screen. Only the first can be converted without optical character recognition.
There is also a third case, less common and more confusing: a PDF that does contain text, but text that decodes to nothing readable.
You can work out which one you have in about five seconds.
The five-second test
Open your statement in any PDF viewer and try to select a line of text with your mouse.
| What happens | What you have |
|---|---|
| Individual words highlight, and you can copy and paste them | A text PDF. Convertible. |
| You get a blue rectangle over the whole area, and nothing copies | A scanned PDF. Needs OCR first. |
| Text highlights, but pasting gives boxes, question marks or gibberish | A broken text layer. See below. |
Alternatively, press Ctrl+F (or Cmd+F) and search for a word you can see on the page. If the search finds nothing, there is no text to find.
Case 1: a scanned PDF
What it actually is
A scanned PDF is a picture. When a statement is printed and then scanned, or photographed with a phone, what ends up in the file is an image of a page. Your eyes read the words. The file contains no words, only coloured pixels arranged in the shape of words.
This is why a converter cannot read it. There is nothing to read. It is not a limitation of one particular tool; it is a property of your file.
How to be sure
Zoom in hard, to 400% or more. A text PDF stays perfectly sharp at any zoom, because the letters are drawn from mathematical outlines. A scan goes soft and blocky, because you are magnifying pixels.
What to do about it
Best option: download the original from your bank. Log in, go to statements or documents, and download the PDF directly rather than using a scan, a photo, or a print-then-scan copy. Banks generate these files from their own systems rather than by scanning paper, so the text in them is real text. This is the simplest fix.
If the bank’s original is genuinely unavailable (an old paper statement, a closed account), you will need OCR to turn the picture back into text. OCR guesses at characters, so a misread digit in an amount can become a silent, plausible-looking error.
If you go that route, reconcile the result carefully afterwards. Check that each transaction plus its amount lands on the balance printed beside it. Our guide to checking a converted statement walks through that.
Case 2: a text PDF that still won’t convert
If text selects and copies cleanly but a converter still fails, the text layer is not the problem. Common causes include:
The statement is password-protected
Some bank statement PDFs are encrypted. Open the file with the required password, then use Print → Save as PDF (or File → Export) to produce an unencrypted copy, if your PDF viewer allows it.
The layout is unusual
Statements printed in an unusual arrangement (several columns across, transactions inside boxed panels, cheque listings in a grid) are harder to read reliably than a plain table. Some tools will refuse; others may convert the main table while leaving the rest out.
Check that content specifically, because the usual safety net does not cover it. A running-balance check works by confirming each transaction leads to the balance printed beside it. Transactions printed in a grid or a boxed panel usually carry no running balance of their own, so they were never part of that chain, and if they are missed, the chain still adds up perfectly. The arithmetic has nothing to notice. Your eyes are the only check that content gets.
The file is too large
Browser-based converters have a size ceiling. DTLOVA’s is 20 MB. If yours is a very long statement, or contains high-resolution images or an embedded scan, try downloading a shorter date range from your bank.
It is not actually a statement
Screenshots saved as PDF, statement summaries, and annual tax certificates can look statement-like without containing a transaction table. There may be nothing useful to extract.
Case 3: text that decodes to nothing
This is the confusing one, and it is worth knowing about because it looks like a bug rather than a file problem.
The PDF has a text layer. Your viewer will let you select it. But when you copy and paste, you get replacement characters: boxes, question marks, or apparent nonsense.
Why it happens: a PDF embeds the fonts it uses, and it is supposed to also embed a map from each glyph shape back to the letter it represents. When that map is missing or incomplete, the shapes still render correctly on screen. The file knows how to draw an A, but nothing in it says that shape is an A. Software reading the text gets the glyph codes and no way to interpret them.
Why it matters more than it sounds: without a check for this, a converter can extract what looks like text, find what look like rows, and hand you a confident CSV full of nonsense. A tool that detects it and refuses is doing you a favour.
What to do: re-download the statement from your bank first. It costs nothing, and if the original export was a one-off glitch the new file will be fine.
If the new copy has the same problem, the fault is in how your bank generates its PDFs, and downloading it again will not help. At that point ask your bank for a different format. Many offer CSV, Excel, OFX or QIF, and any of those avoids the problem entirely. Converting a PDF is a workaround for a bank that does not offer one.
Before you convert anything: check your bank first
This is the advice nobody selling a converter wants to lead with, so here it is.
Many banks let you download transactions directly as CSV, Excel, OFX or QIF. If yours does, use that. It is the bank’s own data, with no extraction step and nothing to go wrong. Look for “Download”, “Export”, or “Statements” in your online banking, and check the format dropdown before assuming PDF is the only option.
Converting a PDF is the right answer when your bank offers nothing else, when you need an older period than the export covers, or when you have a statement from an account you no longer control. Those are real situations, just not the only one.
Quick reference
| Symptom | Cause | Fix |
|---|---|---|
| Cannot select any text | Scanned image | Re-download from the bank; OCR as a last resort |
| Blurry when zoomed in | Scanned image | As above |
| Asks for a password | Encrypted PDF | Open, then re-save as an unencrypted PDF |
| Copies as boxes or question marks | Broken text layer | Re-download from the bank |
| Converts, but rows are missing | Grid or boxed content | Check the CSV against the PDF page by page |
| Too large to convert | Size limit | Download a shorter date range |
| No transactions in it | Not a statement | Find the real statement document |
How our converter handles this
DTLOVA’s bank statement converter reads text PDFs and does not perform OCR, so a scanned statement is something it tells you about rather than guesses at. It also detects the broken-text-layer case described above and says so, instead of producing a CSV from characters it could not decode.
Where it converts a statement but could not place some of the content in the table, it states that too. A CSV that is knowingly incomplete is more dangerous than one that failed outright, because nothing looks wrong.
A final check before you rely on the CSV
If the converter produces a file, do not assume that a successful download means the data is complete.
For important financial records, compare the converted transactions with the original statement. If the statement provides running balances or totals, use them as additional checks. Pay particular attention to page breaks, grids, boxed sections, and any other content outside the main transaction table.
For a step-by-step method, see How to Check if a Converted Bank Statement Is Correct.