What cash-flow underwriting is, and what a statement adds

Reading the money, not only the borrowing history

In a joint statement issued in December 2019, five United States agencies, among them the Federal Reserve Board and the Consumer Financial Protection Bureau, described cash flow data derived from consumers' bank account records as one kind of alternative data, meaning data not usually found in a credit report. They noted that some firms are automating the use of cash flow data to better evaluate borrowers' ability to repay loans, and that cash flow analysis generally focuses on whether a borrower is able to meet new or existing recurring obligations, judged from income and expenses over time.

The two records answer different questions. A credit report shows whether past loans were repaid. A statement shows the income and spending from which the next loan would be repaid.

What a statement shows

The same statement notes that cash flow data can cover categories of income and expenses, such as fixed housing costs, and how a consumer or small business has managed an account over time, for example the balances left in it. In practice, a statement can show:

• Money coming in: wages, payments from customers, transfers from family, government benefits.
• Regular spending: rent, utilities, school fees, phone bills.
• Repayments on other loans, including some that may not appear in a credit report.
• The balance over time: how high it goes after payday and how low it falls before the next one.
• Signs of strain: an overdrawn account, a payment that bounced, a run of small top-ups just before bills are due.

The basic measure: what is left each month

The simplest cash-flow measure is net cash flow: money in minus money out over a month. Transfers between the borrower's own accounts and the proceeds of new loans are left out of money in, because they are not income. Lesson 3 shows how transactions are sorted.

Net cash flow for a month = money in during the month − money out during the month; Average net cash flow = sum of the monthly net cash flows ÷ number of months

Worked example, with illustrative numbers. Over six months a borrower receives 18,000, 22,000, 15,000, 20,000, 25,000 and 20,000, a total of 120,000, or 20,000 a month on average. Spending is 17,000 every month, so average net cash flow is 3,000. A new loan with a monthly instalment of 2,500 would take most of that surplus. In the first month only 1,000 was left, and in the third month money in was 15,000 against 17,000 out, a shortfall of 2,000, so in months like these the instalment could not have been paid from that month's income. The average looks adequate; the month-by-month record shows how thin the margin is. Lesson 4 looks at how to measure that variation.

Evidence that account activity predicts default

Banks have long seen their own customers' accounts. Norden and Weber, in research published in the Review of Financial Studies in 2010, studied the accounts of bank borrowers and found that credit line usage, limit violations and cash inflows showed abnormal patterns roughly a year before default. They found that measures of account activity substantially improve default predictions and are especially useful for monitoring small businesses and individuals, and argued that account activity gives a bank a view of the borrower's cash flows as they happen.

What has changed is who can see such data. When a borrower can share statements from a bank or e-wallet with another lender, a lender without the account relationship can look at the same kind of record. Lesson 2 explains the routes and the consent they need.

Who it can help, and where accounts exist

The US agencies noted that alternative data may help firms evaluate the creditworthiness of consumers who currently may not obtain credit in the mainstream credit system, and that improving the measurement of income and expenses through cash flow evaluation may be particularly beneficial for consumers whose income comes reliably over time from a variety of sources rather than a single job. For example, someone paid by many customers rather than by one employer.

A statement exists only where money goes through an account. The World Bank's Global Findex 2025, published in July 2025, reports that nearly 80 per cent of adults worldwide have a financial account. By World Bank region, it reports 83 per cent in East Asia and Pacific, nearly 80 per cent in South Asia, about 70 per cent in Latin America and the Caribbean, 58 per cent in Sub-Saharan Africa and 53 per cent in the Middle East and North Africa. East Asia and Pacific is a World Bank regional aggregate that includes China; it is not a figure for Southeast Asia, and owning an account is not the same as receiving income into it. For how Southeast Asian markets compare, see the course Southeast Asia in the world.

What a statement cannot tell you

A statement shows only the money that passes through the accounts the borrower shares. Income received in cash, a second account or wallet that is not shared, a one-off transfer that makes a month look better, or a loan from a relative may all be missing or misleading. Cash-flow data is therefore used alongside a credit report and a score, not in place of them (credit scoring course, lesson 7), and it feeds the affordability assessment (affordability course).

The rules still apply. The US agencies encouraged responsible use of alternative data. They said a well-designed compliance management programme analyses the relevant consumer protection laws, including fair lending laws, before such data are used, and that data presenting greater consumer protection risks warrant more robust compliance management. They also observed that cash flow data directly tied to how consumers manage their finances may present lower risks than other kinds of alternative data, and that a creditor's use of it can generally be explained and disclosed to the borrower. For what applies in each market, see the rules comparison.