Accuracy report · November 2026

How Accurate Is Aloan?

A measurement of AI spreading on tax returns and financial statements, on 36 borrowers whose every printed number we know.

99.98% of figures on clean files read correctly 99.87% on phone photos. A figure Aloan misses shows up as an empty cell, so it's counted separately.
254 of 279 mistakes and typos in the borrower's documents flagged or auto-corrected by Aloan Of the other 25, 10 got an advisory note for the analyst to check and 15 were not caught.
0 of 1,728 credit outcomes changed by a re-read or a damaged copy DSCR, leverage and global cash flow, each against its line.
9,377 of 9,430 (99.4%) tax-return figures covered by a validation check in Aloan that identifies inconsistencies A total, a K-1 against its Schedule K, an owner's return, or last year's balance.

Of the 288 tax returns, 286 came back with no wrong figure on any of three reads. Of the 168 financial statements, 166 did.

One number isn't enough

A credit officer asking "how accurate?" is asking four questions

  1. Did it read the number? The figure on the page is the figure in the spread.
  2. Will I know when it's wrong? A wrong figure the software can't see is worse than one it flags.
  3. Is it the same every time? The same file read twice gives the same answer.
  4. Does it change the decision? If something was misread, would the credit outcome have been different?

A single accuracy percentage, usually measured on clean documents, answers only the first and easiest of these. We measured all four and report them separately.

We also split the first question in two. A wrong figure is a plausible number that isn't what the page printed, so the underwriter has no reason to doubt it. A blank is a printed number that didn't come back, and the underwriter can see the empty cell. The two carry very different risk, so we never add them together. The headline is wrong figures.

We count only what Aloan reads. Anything we deliberately don't extract isn't counted for or against us.

How we measured

Synthetic borrowers, so the answer to every printed number is known

A generator writes three years of books for an invented company and its owners, files their taxes on the IRS form edition for each year, and prints financial statements from the same books. It records every printed number as it goes, and each return passes the generator's own consistency checks before it is printed.

The test has 36 companies: 12 kinds of business, each generated three times with different names, figures, owners, tax software and statement styles. Because every company is new, Aloan can't have learned the answers in advance.

What each test borrower comes with

Click any page to see it larger

12 kinds of businesseach generated three times with new names and figures

  • ManufacturerS corp
  • DistributorC corp
  • ContractorC corp
  • Services firmPartnership
  • Restaurant and its landlordS corp and partnership
  • Two-property real estatePartnership
  • Medical practiceS corp, owner's Schedule C
  • Distributor with a holding company and trustPartnerships and trust
  • NonprofitForm 990
  • FarmS corp
  • Retailer, owner has a rental houseS corp, owner's Schedule E
  • HotelPartnership

3 years of tax returnseach on that year's IRS form

  • Business return
  • Supporting statements
  • K-1s to each owner
  • Owners' personal returns

Financial statementsin six different forms

  • CPA reviewed
  • CPA compiled
  • QuickBooks export
  • Company-prepared
  • Spreadsheet
  • Interim, year to date

4 kinds of papereach still readable by a person

  • Clean PDF
  • Photocopy
  • Phone photo, light
  • Phone photo, heavy

Why the test files are hard to read

Real borrower files are messy, so we made the test files messy too. They include the preparer's full tax return with all its attachments, financial statements that don't always match the return, and owners who are paid through other companies. We know the correct value of every number on every page. Click any page to see it larger.

Line items are often on a separate page

Page 1 of a Form 1120-S, and the statement at the back of the same return

Form 1120-S, page 1

Statement page, page 13

Line 20 shows one total, $303,500, and says to see Statement 1. The details are on page 13. Aloan has to find that page, read each item, and check that the items add up to the total. Other assets and other liabilities on the balance sheet work the same way. Meals are listed at $6,500 here because only half is deductible, while the company's books show $13,000.

Each tax program prints statements differently

The same statement from two other returns in the test. The return above says “See Statement 1.”

Return that says “STMT. 1”

Return that says “See Stmt. 1”

We vary the tax software from borrower to borrower. Each program has its own headers, fonts, layout and way of pointing to a statement. Software that only handles one of them would miss the others.

Prior-year numbers sit right next to current ones

A two-year comparison from a tax return, and a QuickBooks profit and loss

Tax return, two-year comparison

QuickBooks profit and loss

Both pages show 2024 next to 2025, plus a column of changes. A number taken from the wrong column looks perfectly reasonable, so we count it as wrong. We know what every cell on the page holds, so these mistakes don't slip past the grading.

Expenses and depreciation aren't always where you expect

Financial statements from two companies in the test

Income statement with one line for all expenses

Expense schedule, four pages later

Income statement with no depreciation line

Cash flow statement, the only place depreciation appears

On the left, all operating expenses are one line on the income statement. Depreciation of $126,000 is included in that line, and only the schedule at the back lists it separately. On the right, depreciation is folded into other operating expenses and appears only on the cash flow statement. The same company's tax return shows $414,000 of depreciation for that year, so the two documents disagree, as they often do in real files.

One borrower, five tax returns

A borrower owned through a holding company and a family trust. Each arrow is a K‑1.

Harbor WholesaleForm 1065 · the borrower Harbor Holdings, LLCForm 1065 · holding company Oakhurst Family TrustForm 1041 Cornelius OakhurstForm 1040 Seraphine TarrowForm 1040 60% 40% Part II Part III Part III
The borrower's income passes to its owners through a holding company and a trust. Each K-1 should match the return it came from and the owner's return it goes to, in all three years. “Part II” and “Part III” are the sections of the owner's Schedule E where it's reported. Aloan also has to keep track of whose return is whose. The names are made up and change every time the test is generated.

The test also includes:

In our QA files from real banks, about 11% of scanned pages are phone photos. We use that share to weight a typical bank file.

Did it read the number?

On clean files, 11 of 54,744 figures were wrong

Aloan read 32,433 figures from tax returns and 22,311 from financial statements. 11 were wrong and 17 came back blank. 3 of the 11 wrong figures failed a check, and the other 8 raised nothing.

Wrong figures and blanks, per 10,000 figures read

Tax returns and financial statements together, by kind of paper

On clean files, photocopies and heavy phone photos, blanks outnumber wrong figures, so most misses there are ones the underwriter can see. Light phone photos are the exception: 26 wrong against 13 blank. The chart doesn't show whether a wrong figure was flagged; on clean files 3 of 11 were.
As Aloan received itPaperFigures readWrong1 inBlank
Clean54,744115,00017
Photocopy17,23453,40013
Phone photo, light17,2342666013
Phone photo, heavy17,2342086027
A typical bank file (89% clean, 11% phone photo)3,000

Each thumbnail is the same patch of one borrower's 1120, page 1, lines 1c to 5, from the file Aloan was given. The damaged copies are made from the clean PDF, and each level was tuned so a person can still read every figure. Select one to see it full size.

By document, 99.81% of tax-return figures and 99.91% of financial-statement figures were exact on clean files. 99.1% of statement lines landed in the same spread category as the answer key. Most of the rest are judgment calls an analyst makes either way, such as where a nonprofit's restricted net assets sit within equity.

Will I know when it's wrong?

Of 279 mistakes planted in the borrower's paper, 15 got through

Aloan ties every figure it can to another figure on the paper: totals against their lines, a K-1 against the Schedule K it came from, an owner's return against their K-1, this year's opening balance against last year's closing one, and a statement's subtotals against its printed totals. 9,377 of the 9,430 tax-return figures Aloan keeps (99.4%) take part in at least one such tie.

To test the checks, we changed one printed number in one document, the way people slip: two digits transposed, a digit dropped or doubled, a neighbouring figure copied. Every planted error was drawn fresh for this run, so none of the checks was tuned on it.

What happened to each planted mistake

One changed figure per document

Financial statements give fewer independent ways to prove a figure than a tax return, so more slips get through: 13 of 144, against 2 of 135 on returns. 7 of the 13 were a transposed total assets or net income.
Show the numbers
PlantedFlaggedCorrectedNotedNot flagged
Tax returns135126342
Financial statements14411015613
All279236181015

Aloan doesn't compare a borrower's financial statements with their tax return. Credit policy decides which source governs, and the analyst reconciles the two. When we lowered revenue on a statement against the return, it was flagged in 1 of 18 cases.

Is it the same every time?

99.95% of figures came back the same way in all three reads

Each file was read three times, independently. Of the 18,248 figures on the clean files, 18,238 came back the same way every time.

Every DSCR, leverage and global cash flow figure landed on the same side of its line in all three reads.

Does it change the decision?

No re-read and no damaged copy moved a borrower across its line

A misread matters when it changes the credit outcome. We rebuilt three measures from every read, with the same calculations the underwriting worksheets use:

Each read is compared with the majority of the three clean reads. A flip is a read that lands on the other side of the line.

MeasureBorrower-yearsClean re-reads that flippedDamaged reads that flipped
DSCR990 of 2970 of 297
Leverage900 of 2700 of 270
Global cash flow990 of 2970 of 297

The generator draws its deals close to their lines on purpose: clean DSCR runs from 0.93x to 1.71x, and 6 of the 99 years fail. A small misread is more likely to flip one of these deals than a typical file. DSCR from the financial statements was within 0.02x of the answer key's in all 99 years.

Compared with what?

On clean files, Aloan's error rate is a seventh of double data entry's

People make mistakes too. No study has published a per-figure error rate for bank analysts spreading tax returns, so the closest evidence comes from other fields where people copy numbers from paper.

Error rate per figure, published human studies and Aloan

Log scale: each gridline is ten times the one to its left

Human fields include dates, codes and text, while ours are numbers graded to the dollar. Spreading means finding the right line across many forms, which is closer to abstraction than to copying a form, but nobody has measured analysts doing it.
Show the numbers and sources
TaskSourceError rate
Abstraction from records: finding and interpreting each valueGarza et al., Int J Med Inform 20256.57%
Clinic staff typing meter readings into a recordMays & Mathias, JAMIA 20193.7%
Single entry from a structured formGarza et al. 20250.29%
Specialist keying operatorsKlemmer & Lockhead, J Appl Psychol 19620.14–0.37%
Double entry: two people keying and reconcilingGarza et al. 20250.14%
Aloan, phone photosThis report0.13%
Aloan, clean filesThis report0.02%

People and software make different kinds of mistakes, which is why we report decisions as well as figures.

How a bank can measure its own rate. Pick about 40 recent deals. Have two analysts spread each one independently, and a third check every disagreement against the page. That gives a per-figure error rate to within about a quarter of a percentage point. Run the same deals through Aloan and score it against the same answers.

What to ask any vendor

These are the questions we'd ask of anyone's AI spreading, ours included:

  1. Accuracy of what, and out of how many? Figures, fields or documents? On which forms?
  2. Wrong or blank? How often is a figure wrong without anyone being able to see it?
  3. On what paper? Clean PDFs, or the photocopies and phone photos your bank actually receives?
  4. Will I know when it's wrong? Of the errors in the borrower's own paper, how many does it flag?
  5. Does it change the decision? How often would a misread have moved a DSCR across your line?

To see Aloan on your bank's files, book a 30-minute demo. How we built the test: the blog post.