Qualifying Income from Bank Statements: How to Calculate It for Loan Underwriting

Calculating qualifying income from bank statements is one of the most common - and most error-prone - steps in lending underwriting, especially for self-employed, 1099, and bank-statement-loan borrowers. The math is simple. The judgment calls about which deposits to include and which to exclude are not.

This is the four-step workflow lenders use, the inclusion and exclusion rules that distinguish a qualifying deposit from a non-qualifying one, and how the result feeds DTI. It covers the most common bank-statement income programs: Fannie Mae self-employed (Form 1084), bank-statement non-QM, and Freddie Mac's analogous program.

Want to use this guide on your blog? Copy this embed code:

Step 1: Gather the Right Number of Months

The number of statements you need depends on the program:

Pull statements directly from the bank when possible (downloaded PDFs straight from the bank's portal). If the borrower emails statements, verify the file is from the bank by checking PDF metadata - see our fraud red flags guide for what to look for.

To get the data into a workable format (Excel, CSV, or QBO for your underwriting system), use PDFSub's Bank Statement Converter. It handles 20,000+ bank templates and preserves transaction-level detail you'll need for the categorize/exclude step.

Step 2: Sum All Deposits, Then Exclude

This is where most calculation errors happen. The raw sum of every deposit is not qualifying income. You have to back out the non-income deposits first.

Always include

• W-2 payroll direct deposits
• 1099 contractor payments
• Rental income (collected directly, not through a property manager who issues a 1099)
• Recurring gig-economy payouts (Uber, Lyft, DoorDash, Airbnb, etc.) with at least 12 months of history
• Pension and Social Security direct deposits
• Court-ordered alimony or child support with 12+ months of history
• Self-employed business deposits (subject to expense factor on some programs)

Always exclude

• Inter-account transfers (between your own accounts)
• Loan disbursements (personal loan deposit, HELOC draw, credit-line advance)
• Refunds and reversals (a deposit that immediately offsets a prior withdrawal)
• Tax refunds (one-time, non-recurring)
• Insurance settlements (one-time)
• Gifts (one-time deposits from family without a pattern of recurrence)
• Cash deposits over a threshold without source documentation (typically $1,000+ on bank-statement non-QM programs)
• Reimbursed expenses (employer reimbursements that show as deposits)
• Returns of capital (sale of an asset)

Use judgment

• Windfalls: bonus payouts, large quarterly commissions. Some programs allow these if they're recurring year-over-year; many require averaging them in differently.
• Large round-number deposits from unknown sources: flag for documentation. See fraud red flags.
• Foreign-currency deposits: convert at the date-of-transaction exchange rate if applicable.

The categorize step is yours

PDFSub extracts every deposit (date, description, amount, source bank if available) into a structured format. The categorize/exclude judgment is the underwriter's - no tool can reliably distinguish "loan advance" from "income" just from a transaction description, because banks describe them inconsistently. What you get is clean data ready for you to apply the rules.

In Excel, the typical workflow:

1. Filter the converted statement to deposit rows only
2. Add a Category column with values: INCOME, TRANSFER, LOAN, REFUND, GIFT, WINDFALL, OTHER
3. Tag each deposit
4. Use =SUMIF(Category, "INCOME", Amount) to get qualifying total

Step 3: Divide by Months for Average Monthly Income

Once you have the qualifying total, the math is simple:

Qualifying total / Number of months = Average monthly income

For example: $72,000 qualifying total / 24 months = $3,000/month.

The 50-75% factor on bank-statement non-QM programs

Some bank-statement non-QM programs apply an expense factor on top of the deposit average to account for unreported business expenses. Common factors:

• Personal bank statements: typically no factor (you're already excluding non-income)
• Business bank statements: 50-75% factor (so $10,000/mo average deposits = $5,000-7,500/mo qualifying income)
• CPA-prepared P&L cross-reference: lender may use the lower of (a) bank statement average or (b) P&L net income

Confirm the factor with your AUS or the specific program guidelines. Fannie Mae self-employed (Form 1084) uses a completely different methodology based on tax returns, not deposits.

Trending and consistency checks

Lenders look at month-over-month consistency. A borrower with $3,000/mo average split as $1,000, $1,000, $1,000, $9,000, $1,000, ... is treated differently than one with $3,000, $3,100, $2,900, $3,200, $2,800, .... The former gets a lower qualifying income (or full rejection on some programs) due to inconsistency.

In Excel:

Average:  =AVERAGE(monthly_income_range)
StDev:    =STDEV.P(monthly_income_range)
CV%:      =StDev / Average    (coefficient of variation)

CV under 25% is usually acceptable. CV over 50% triggers additional documentation requirements.

Step 4: Calculate DTI

DTI (debt-to-income ratio) is the standard underwriting affordability metric:

DTI % = (Monthly debt obligations / Monthly qualifying income) * 100

Monthly debt obligations include

• Proposed new mortgage principal + interest + taxes + insurance (PITI)
• HOA dues, if applicable
• Existing mortgage payments on other properties
• Auto loan payments
• Student loan payments (specific rules vary by program for income-driven repayment plans)
• Credit card minimum payments (from credit report)
• Personal loan payments
• Alimony and child support obligations
• Co-signed loan payments (under specific conditions)

Monthly debt obligations exclude

• Utilities (electric, gas, water, internet)
• Insurance premiums other than PITI
• Cell phone bills
• Daycare and tuition
• 401(k) loans (in most cases)
• Voluntary deductions

DTI thresholds

Front-end DTI is housing-only (PITI / income). Back-end DTI is total debt / income.

Common Errors That Send Files Back to the Borrower

Counting transfers as income. When a borrower has multiple accounts and you only review one, transfers from the other account look like deposits. Always confirm with the borrower whether large unidentified deposits are transfers from their own accounts.

Counting loan disbursements as income. Personal loan disbursements, HELOC draws, and credit-line advances show as deposits. The borrower's credit report will show the corresponding loan - cross-reference to catch these.

Counting refunds. A $2,000 deposit on 02/15 that follows a $2,000 withdrawal on 02/13 is almost certainly a refund or reversed transaction, not income.

Including non-recurring windfalls. A $15,000 deposit labeled "TAX REFUND" or "INSURANCE SETTLEMENT" is not qualifying income unless it repeats annually with documentation.

Forgetting to apply the 50-75% factor on business bank statement programs. Easy to miss on personal-account programs that look similar. Confirm whether the program treats deposits as net income or gross.

Using the wrong number of months. Fannie Mae self-employed wants 24 months and Form 1084 (tax-return-based). Bank-statement non-QM wants 12 or 24 months of statements. Confirm program before you start.

Mixing business and personal accounts. Some programs require business statements only; some require personal; some allow either. If the borrower deposited business income into a personal account, the program may not allow you to count it. Confirm program rules.

Where PDFSub Fits in the Workflow

For automated bank-statement income calculation at scale, lenders typically use a stack of:

• Direct bank connection (Plaid, Finicity, MX) for borrowers who can authenticate
• Statement upload + parsing + fraud detection (Ocrolus, Inscribe) for everyone else
• Manual review and Excel modeling for edge cases, international banks, and pre-approval / education

PDFSub fits the third bucket and partially the second. It converts the PDF statements to structured Excel/CSV/QBO data quickly, including international and non-US banks where Ocrolus and Plaid may have gaps (130+ languages, 20,000+ banks). The categorize/exclude step stays manual - that's the lender's judgment call - but you start with complete, clean transaction data instead of typing it in row by row.

For format choice when feeding into your AUS or underwriting system, see QBO vs CSV vs OFX. For the broader fraud-detection picture, see 8 Red Flags on a Bank Statement. For the basic conversion workflow, see How to convert bank statements for mortgage applications.

Summary

1. Gather the right number of months (typically 12 or 24 depending on program)
2. Sum and exclude non-income deposits (transfers, loans, refunds, one-time windfalls)
3. Average monthly by dividing the qualifying total by the number of months (apply 50-75% expense factor if the program requires it for business accounts)
4. DTI = monthly debt / monthly qualifying income

The math is simple. The judgment calls about what to include and exclude are the underwriter's. PDFSub gets you to clean transaction-level data fast, so the time you spend is on the judgment, not on data entry.