Extract ALTA, HUD-1, and Closing Disclosure to Excel

To extract an ALTA settlement statement to Excel, build the output as one row per settlement line item, not one row per document. Each row should identify the document type, buyer or seller side, section or line reference, description, debit amount, credit amount, whether the amount was paid before or at closing, a classification hint, the source file, and the source page.

That same structure works for HUD-1 data extraction to spreadsheet and for closing disclosure line items to Excel because it captures the accounting meaning of each line instead of copying the visual layout of the form. A bookkeeper or CPA can then review prorations, deposits, title and lender fees, recording charges, commissions, credits, and basis candidates before the data moves into an accounting system or tax workpaper.

The key is to treat the PDF as source evidence, not as the final format. A closing statement often has the information needed for several downstream jobs at once: a basis schedule for a rental purchase, a commission check against a CDA, a buyer or seller credit reconciliation, a fund closing file, or an import staging sheet for QuickBooks, Xero, NetSuite, or a tax workbook.

ALTA Settlement Statements, HUD-1s, and Closing Disclosures all describe a real estate closing, but they do not arrange the same information in the same way. The spreadsheet needs to survive those format differences. A row that says "Recording fee, buyer side, debit, paid at closing, source page 3" is more useful than a copied block of OCR text because it gives the reviewer a place to test accounting treatment.

The extraction file should not decide tax treatment by itself. It should make review easier by separating potential basis items, deductible expense candidates, refundable credits, loan charges, prorations, deposits, and proceeds into rows that preserve the original line reference and page source.

Identify whether the file is ALTA, HUD-1, or Closing Disclosure

Start by naming the form in a dedicated column. That small step prevents a mixed batch from turning into a spreadsheet where line references mean different things from one row to the next.

• ALTA Settlement Statement: usually found in title or escrow closing packages; keep buyer, seller, and shared description fields separated.
• HUD-1 Settlement Statement: common in legacy or non-TRID closing files; preserve numbered line references, not just buyer and seller totals.
• Closing Disclosure: used in consumer mortgage closings under TRID; capture section and page context for loan costs, prepaid items, escrow items, and summaries.

An ALTA Settlement Statement is common in title and escrow files. It usually presents buyer, seller, and description information around settlement charges, credits, and disbursements. ALTA says its standardized settlement statements itemize fees and charges paid by the homebuyer and seller, are used with the federal HUD-1, do not replace the CFPB Closing Disclosure, and are available in four versions through ALTA's standardized settlement statement forms. For extraction, that means the buyer and seller sides should stay separate even when the description looks shared.

A HUD-1 Settlement Statement is the legacy two-page settlement form still encountered in older files and certain non-TRID transactions. It has familiar numbered lines and summaries, which makes the line reference valuable in the spreadsheet. Do not flatten a HUD-1 into only buyer totals and seller totals. The accounting value is in the individual rows that explain how those totals were built.

A Closing Disclosure is the current TRID form used for most consumer mortgage closings. It is longer, with loan terms, projected payments, closing costs, cash-to-close calculations, and summaries of transactions. The extraction job should identify the page and section because a charge that looks like a simple fee may need to be traced back to a loan-cost section, prepaid item, escrow item, or transaction summary.

Recognition matters because the same economic event can appear under different labels. A tax proration, seller credit, owner title policy, lender title policy, recording fee, or escrow deposit may sit in a different part of the form depending on the document type and closing software. If your team also handles UK property files, the parallel workflow is UK completion statement extraction to Excel, but US ALTA, HUD-1, and Closing Disclosure work needs its own schema because the form families and settlement terminology are different.

Build the Excel schema around accounting review

The spreadsheet should be designed for review and import, not for visual resemblance to the PDF. A practical schema for real estate closing statement to Excel work usually needs these columns:

• Document type
• Closing date
• Property or file reference
• Buyer or seller side
• Section or line reference
• Original description
• Debit amount
• Credit amount
• Paid before closing or paid at closing
• Classification hint
• Source file
• Source page

For example, a row might read: ALTA Settlement Statement, buyer side, recording fee, debit amount 185.00, credit amount blank, paid at closing, classification hint "recording fee", source file "123-Main-Closing.pdf", source page 2. That row is ready for review because the amount, party side, classification hint, and page evidence are separated.

One row per settlement line item is the rule that makes the file usable. If a PDF has separate lines for transfer tax, recording fees, lender title insurance, owner title insurance, prepaid interest, escrow deposits, and seller credits, each line should become its own row. A single "closing costs" total is too blunt for accounting review.

Keep debit and credit amounts in separate numeric columns instead of mixing signs into one amount column. That makes pivot tables, checks by party side, and import mapping cleaner. Text fields should stay text: document type, line reference, description, classification hint, source file, and source page. Dates should be consistently formatted. Currency columns should be numeric, with cents preserved.

Classification hints are not final accounting decisions. They are review labels that help route the row: basis candidate, deductible expense candidate, refundable credit, loan cost, title or recording fee, commission, tax proration, HOA proration, prepaid item, escrow deposit, or net proceeds. The original description and source page still matter because the reviewer may need to override the hint.

This is where prompt-driven extraction is useful. With AI financial document extraction to Excel, a closing PDF can be uploaded with instructions to create one row per settlement line item, preserve source file and page references, and return Excel, CSV, or JSON. Invoice Data Extraction supports prompt-based field selection, custom column names, line-item rows, saved prompts, and structured outputs; for specialized closing files, run a small sample first and confirm that the rows match the team's review needs before processing larger batches.

Treat recurring settlement lines by workflow, not by label

Closing files reuse familiar economic categories, but the labels vary by title company, lender, state, and form type. The extraction should preserve the original wording while adding enough structure for review.

Purchase price or sale price anchors the transaction. For a buyer-side file, it is usually the starting point for acquisition review. For a seller-side file, it ties to gross proceeds before payoffs, commissions, credits, and closing charges.

Earnest money, deposits, and seller concessions need their own rows because they affect cash to close and credits between parties. A deposit already paid before closing should not be confused with a charge paid at closing. Seller concessions should be visible as credits, not buried inside a net number.

Loan charges, title charges, recording fees, and transfer taxes are where basis and expense review becomes sensitive. Some acquisition-related charges may feed a basis workpaper, some may be loan costs, some may be current expenses, and some may be excluded from immediate deduction. For rental purchases, those current-expense lines may later connect to a Schedule E expense worksheet; for fix-and-flip work, they may need to roll into property-level receipt tracking for Schedule C COGS. The spreadsheet should flag the likely category without pretending that extraction has made the tax decision.

Property-tax prorations, HOA prorations, prepaid insurance, prepaid rent, and escrow or impound deposits are frequent sources of errors. The same line family may appear as a debit or credit depending on whether the buyer is reimbursing the seller, the seller is crediting the buyer, or the lender is funding an escrow account. If the spreadsheet keeps party side, debit, credit, and paid status separate, those rows are easier to test.

Commissions and net proceeds matter most for brokerages, sellers, and fund accounting teams. A commission line may need to be checked against a commission disbursement authorization, while net proceeds may need to reconcile to the wire and closing ledger. For property-management teams, closing data may later connect with owner reporting, where a separate workflow such as AppFolio, Buildium, and Yardi owner statements to spreadsheet handles the post-closing monthly statements rather than the settlement file itself.

Review the extracted spreadsheet before import

Before importing any closing-statement spreadsheet, treat the Excel file as a staging workpaper. The review should answer three questions: are the rows complete, do the amounts behave correctly, and can each row be traced back to the PDF?

Start with control checks. Compare row counts against the visible settlement sections, then total debit and credit columns by buyer side and seller side. Look for duplicated rows, amounts in the wrong debit or credit column, blank classification hints, empty line references, dates formatted as text when the import expects dates, and currency values that landed as text instead of numbers. Spot-check source pages for high-value items such as purchase price, loan charges, seller credits, tax prorations, title policies, commissions, and escrow deposits.

Do not rely on one grand total as proof that the extraction is right. ALTA statements, HUD-1s, and Closing Disclosures split and summarize charges differently. A line may appear once in a detailed section and again in a transaction summary. A review that traces rows back to source pages catches duplicated summary lines, omitted details, and charges placed on the wrong party side.

For QuickBooks, Xero, NetSuite, tax software workpapers, or a fund accounting workbook, map the spreadsheet to the receiving system's fields only after the review is complete. Some teams need a journal-entry staging file. Others need a fixed-asset support schedule, a closing-cost classification workbook, or a reconciliation tab that never imports directly. The right import layout depends on the accounting workflow, not on the original PDF.

Invoice Data Extraction supports saved prompts, Excel, CSV, and JSON output, native Excel typing, and source file and page references for verification. For recurring closing-statement jobs, save the prompt that produced the accepted column layout, then reuse it for the same client, fund, title office, or property type. Teams that process broader real estate AP files may use the same review discipline for AI invoice extraction for property management, where source-page checks and consistent columns matter across vendor invoices as well as closing statements.

Use a repeatable prompt for every closing file

A repeatable extraction routine starts with a prompt that describes the spreadsheet, not just the document. Ask the extractor to identify whether each file is an ALTA Settlement Statement, HUD-1, or Closing Disclosure, then create one row for each settlement line item.

The prompt should require buyer or seller side, section or line reference, original description, debit amount, credit amount, paid-before or paid-at-closing status, classification hint, source file, and source page. It should also tell the tool how to handle missing data: leave the cell blank when the PDF does not provide the field, keep amount columns numeric with cents, format dates consistently, and add a note when a row is ambiguous.

For a real estate closing statement to Excel workflow, avoid asking for broad categories alone. "Extract closing costs" is weaker than asking for separate rows for taxes, HOA prorations, prepaid items, title charges, lender fees, recording fees, transfer taxes, deposits, credits, commissions, and net proceeds. The second instruction gives the reviewer rows that map to actual accounting questions.

In Invoice Data Extraction, that prompt can be used with uploaded PDF, JPG, or PNG files and returned as Excel, CSV, or JSON. Once a sample batch has been checked against source-page references, save the accepted prompt for recurring client, fund, title-office, or property-management work. The saved prompt becomes the control point: consistent column names, consistent review fields, and fewer one-off spreadsheet layouts every time another closing file arrives.