Direct vs Indirect Cash Flow in Reporting and Forecasting

Direct vs indirect cash flow is two questions under one name. In a statement of cash flows, the two methods reach the same operating cash flow and differ only in what they show. In a forecast, they are different models that produce different numbers, and the evidence favours building the near term directly.

Standard-setters have encouraged the direct method since 1987 and research finds it better at predicting cash, yet few companies report it. The table below sets the two methods side by side in both settings.

The same two methods answer different questions in a report and in a forecast

	Direct method	Indirect method
In the statement of cash flows
Starts from	Gross receipts and payments: from customers, to employees and suppliers, interest, taxes	Profit, adjusted for non-cash items and changes in working capital
Operating cash flow	The same total by construction (Publix, 2025: $5,908M)	The same total by construction (Publix, 2025: $5,908M)
US GAAP (ASC 230)	Encouraged since 1987; the indirect reconciliation must be shown as well	Permitted
IFRS (IAS 7)	Encouraged since 1992; no reconciliation required	Permitted; from 2027 it must start from operating profit
Who uses it	A small minority of companies	97.5% of a 1991 AICPA sample; still the vast majority of US companies in 2025
In a forecast
Built from	Dated invoices, contracts, payroll and supplier terms	Forecast profit and assumed ratios: receivable days, payable days, deferred revenue
Horizon it suits	As far as dated items reach	Months to years
How it fails	Misses what is not yet invoiced or contracted	Each month’s error stays in every later balance; balance-sheet changes miss acquisitions and currency effects

Sources: FASB Statement No. 95 (1987); IAS 7, paragraphs 18 and 19; IFRS 18; Publix Super Markets, 2025 Form 10-K; Bahnson, Miller and Budge (1996) as reported by Abu-Abbas (2014); Zeff (ICAEW, 2025). Forecasting rows: CX Cash synthesis.

Operating cash flow = receipts from customers − payments to employees and suppliers − interest and taxes ± other = net income + non-cash items − increase in operating working capital

Direct vs indirect cash flow: what is the difference?

The direct method lists where operating cash came from and went; the indirect method starts from profit and explains why cash differed from it. The Financial Accounting Standards Board defined the direct method in 1987 as reporting “major classes of gross cash receipts and gross cash payments and their arithmetic sum” (FASB, 1987). The indirect method reaches the same total from the other end. It starts with net income and removes deferrals, accruals and items whose cash effects belong to investing or financing; IAS 7 uses the same definitions (IFRS Foundation). Publix Super Markets, a large US retailer that reports the direct way, prints both versions for 2025.

Publix reports its 2025 operating cash flow both ways, and both reach $5,908M

Direct method (the statement), $M		Indirect method (the reconciliation), $M
Cash received from customers	62,974	Net earnings	4,734
Cash paid to employees and suppliers	(56,421)	Depreciation and amortization	1,091
Income taxes paid	(844)	Retirement contributions paid in stock	561
Self-insured claims paid	(699)	Deferred income taxes	403
Dividends and interest received	486	Gain on investments	(720)
Other operating cash receipts	458	Changes in working capital (6 lines)	(276)
Other operating cash payments	(46)	Other non-cash items (3 lines)	115
Net cash provided by operating activities	5,908	Net cash provided by operating activities	5,908

Source: Publix Super Markets, Form 10-K for the year ended 27 December 2025. The working-capital and other lines are grouped by CX Cash; the groups sum to the filed totals (our arithmetic).

The direct column says where the cash came from and went: $62,974M from customers, $56,421M to employees and suppliers. The indirect column says why cash differed from profit, including $720M of investment gains that sat in net earnings but brought in no operating cash. A lender who saw only the reconciliation would not know what customers paid; a shareholder who saw only the direct statement would miss the gains inside earnings.

Do GAAP and IFRS require one method?

Neither requires companies to use the direct method. Both encourage it, and US GAAP asks direct-method companies for the indirect reconciliation as well. That is why Publix prints both: the 1987 standard requires it “regardless of whether the direct or indirect method” is used. IAS 7 has no such rule.

The rules are moving toward a better indirect bridge. IFRS 18, issued in April 2024, makes the indirect method start from operating profit for periods beginning on or after 1 January 2027 (IFRS Foundation). On 23 September 2026 the IASB tentatively decided to propose that companies disclose, line by line, the non-cash changes that stop an indirect reconciliation from tying to the balance sheet: currency translation, buying or losing control of subsidiaries, and reclassifications. Nine of its 12 members agreed (IASB, 2026).

Why do most companies use the indirect method?

Their ledgers record accruals, and building gross receipts and payments from them costs money. When the FASB proposed a statement of cash flows in 1986, it received more than 450 comment letters. “A majority of respondents to the Exposure Draft asked the Board to require use of the direct method”, most of them commercial lenders, who said receipts and payments were particularly important in judging a borrower’s ability to repay. Preparers replied that they “do not presently collect information in a manner that will allow them to determine amounts such as cash received from customers or cash paid to suppliers directly from their accounting systems” (FASB, 1987).

The Board chose encouragement, and Statement No. 95 passed four votes to three. Two of the three dissenters, Lauver and Swieringa, objected to the indirect method itself: by permitting it, they wrote, “the Board has foregone the opportunity to make a significant contribution to the quality of financial reporting”. The majority wrote that further movement toward the direct method “should be permitted to develop” as preparers and users gained experience (FASB, 1987).

It did not develop. Bahnson, Miller and Budge found that 97.5% of the companies in the AICPA’s 1991 survey sample used the indirect method alone, as a 2014 review reports (Abu-Abbas, 2014). In a 2025 ICAEW article the accounting historian Stephen Zeff summed it up (Zeff, ICAEW, 2025): “the vast majority of US companies, and apparently a majority of entities using IFRS Accounting Standards, have opted to use the indirect method”. The figure of 98% that circulates online, credited to the AICPA without a date, matches samples from the 1990s and should be read as a figure from that decade.

The cost argument has not moved either. At an EFRAG roundtable reported in 2024, one preparer said her company “had attempted to draw up a draft version of the direct method but in the process failed to come up with a solution”, citing multiple currency exposures (EFRAG, 2024). In April 2025 a symposium co-hosted by the FASB and the University of Chicago heard from its panel of standard-setters that “The indirect method is widely used and there have not been calls to fully transition to a direct method statement” (FASB and IASB, 2025).

Which method is better for forecasting cash flow?

The direct method, for as far as dated receipts and payments reach. Both standards justify their encouragement in forecasting terms. IAS 7, for one, says the direct method “provides information which may be useful in estimating future cash flows and which is not available under the indirect method” (IAS 7, paragraph 19).

The research supports them. Gopal Krishnan and James Largay found that past direct-method data “predict future operating cash flow better than indirect method cash flow data” (Krishnan & Largay, 2000). Steven Orpurt and Yoonseok Zang showed why outsiders cannot simply rebuild the missing lines: items such as cash received from customers “are not reliably estimable using income statements and either balance sheets or indirect method statements of cash flows” (Orpurt & Zang, 2009). Jeffrey Hales and Orpurt’s 2013 review found evidence that, for many firms, the recurring benefits of direct-method information likely exceed its recurring costs (Hales & Orpurt, 2013). These studies test outsiders forecasting listed companies from published statements; none tests a company’s own forecast. The transfer rests on the mechanism they identify, that gross receipts carry information which profit and balance-sheet ratios lose, and a company holds its own receipts in full.

Balance-sheet ratios are the usual substitute, and they carry errors of their own. Paul Hribar and Daniel Collins showed in 2002 that measuring accruals from successive balance sheets, instead of from the cash flow statement, introduces errors that bias results when mergers, acquisitions or discontinued operations are involved (Hribar & Collins, 2002).

Atlassian’s latest annual report shows the gap in a real company. In the year to June 2026, its balance sheet shows receivables rising by $491.6M, while its cash flow statement, which reports working-capital changes net of business combinations, shows $484.5M. Deferred revenue rose $180.4M on the balance sheet and $153.3M in the cash flow statement. Rebuilt from the balance sheet, those two lines alone would overstate operating cash flow by about $20M, 1.5% of the $1,353M reported (our arithmetic). That year Atlassian also paid $1,229M for acquisitions net of cash acquired (Atlassian, 2026).

How does each kind of forecast work, and how does it fail?

In a forecast, direct vs indirect cash flow is a choice between two models: one estimates profit and converts it into cash, the other places dated cash items where they fall. The FASB sketched the indirect one in 1987: estimate future income first, then convert it into cash “by allowing for leads and lags between cash flows and income”. Wikipedia calls the direct route, a schedule of receipts and disbursements, “best suited to the short-term forecasting horizon of 30 days”, the period with actual data (Wikipedia, Cash flow forecasting). Among the indirect variants it lists the pro-forma balance sheet method, and observes with some understatement that “if all the other balance sheet accounts have been correctly forecast, cash will be correct, also”.

That condition carries the whole method. An indirect forecast starts from a known position, the opening bank balance, and advances it month by month on assumed ratios: growth, margin, receivable days, the share of revenue billed in advance. Each month’s closing balance opens the next, so an error stays in every later balance, and a wrong margin or growth rate adds to it every month. A direct forecast places dated items (invoices already issued, contracts with billing dates, the payroll calendar) where they fall, and it stops where the dated items stop. Its items are forecasts too, since customers can pay late, but each one can be checked against a document. The free 13-week cash flow model and forecast template below are a starting point for a direct forecast.

What does one quarter look like forecast both ways?

Different by $370k in the example below, and every dollar of the difference points to a balance-sheet line. The company is hypothetical: a B2B software business with $800k of monthly revenue, a net loss of $150k a month and $60k a month of non-cash costs, mostly stock compensation and depreciation. Its indirect forecast holds receivables, deferred revenue and payables flat, so operating cash flow is −$90k every month. In the direct forecast, monthly-billed customers owe $480k a month on 30-day terms, and one of them, worth $60k a month, has started paying at 60 days. Annual contracts worth $3.84M a year follow a renewal calendar that brings in $410k, $150k and $90k over the quarter. Payments are the same in both forecasts.

One quarter forecast both ways for a hypothetical software company, $ thousands

	Month 1	Month 2	Month 3	Quarter
Indirect forecast
Net loss	(150)	(150)	(150)	(450)
Non-cash costs added back	60	60	60	180
Change in working capital (held flat)	0	0	0	0
Operating cash flow	(90)	(90)	(90)	(270)
Direct forecast
Receipts, monthly-billed customers	480	420	480	1,380
Receipts, annual contracts	410	150	90	650
Payments	(890)	(890)	(890)	(2,670)
Operating cash flow	0	(320)	(320)	(640)
Gap (direct minus indirect)	90	(230)	(230)	(370)
of which renewal calendar (deferred revenue)	90	(170)	(230)	(310)
of which slower payer (receivables)	0	(60)	0	(60)

Hypothetical company; figures follow from the assumptions above. Because the indirect model holds balances flat, its implied receipts equal revenue ($800k a month) and its payments equal revenue less the net loss and the non-cash costs.

The gap, direct minus indirect, is −$370k for the quarter, about four months of the burn the indirect model reports (our arithmetic). On $2.4M of opening cash, the indirect forecast ends the quarter at $2.13M and the direct one at $1.76M. Extrapolated, the first reads as about 27 months of runway and the second as about 11. Neither extrapolation is right, because renewals are seasonal. The $310k shortfall on annual contracts should come back when the larger renewals fall due, and if it does not, the deferred-revenue assumption is wrong. The gap does not say which forecast is wrong. Every dollar of it sits in a working-capital line, though, so it says where to look.

Do investors want direct-method statements?

Mostly not, on the evidence of recent consultations. At the EFRAG roundtable, preparers and users noted “that the information provided under the direct method would not help analysts to improve their forecasting”, and the IASB’s observer, Nick Anderson, remarked that most investors prefer the indirect method (EFRAG, 2024).

EFRAG’s participants have a point about published statements. A reported direct statement is coarse: Publix puts $56,421M of payments, about 90% of its receipts from customers, into one line. The indirect reconciliation is where the quality of earnings shows. Atlassian’s 2026 net loss of $53.8M becomes $1,353M of operating cash flow mainly because $1,607M of stock-based compensation, about 30 times its net loss (our arithmetic), is added back. An analyst learns more from that line than from any gross receipt.

Inside a company, most of the cost they describe falls away. It is the cost of rebuilding gross flows from accrual ledgers, across subsidiaries and currencies, for an audited statement. A company forecasting its own cash already holds the invoices, the contracts and the payroll calendar.

Both positions can hold at once. A published statement serves outsiders best with an indirect bridge from profit. A company’s own forecast does better built from dated receipts and payments as far as they reach. Beyond them it can run on the profit and loss and stated ratios, restarted each month from the actual bank balance. How far the dated items reach varies by company. Wikipedia puts the direct route’s natural horizon at 30 days, while the hypothetical software company above could place a full quarter of renewals by date.

Sources and notes

• Abu-Abbas, B. M. (2014). Direct, indirect, or both methods of reporting operating statement of cash flows. International Journal of Finance and Accounting 3(6), 335–340.
• Atlassian Corporation (2026). Form 10-K for the fiscal year ended 30 June 2026.
• EFRAG (2024). Issues with how the IAS 7 Statement of Cash Flows is prepared, contractor’s draft report by C. Zielke and S. Hassanali, proactive research project on the statement of cash flows, phase 2.
• FASB (1987). Statement of Financial Accounting Standards No. 95: Statement of Cash Flows.
• FASB and IASB (2025). Statement of Cash Flows, agenda paper AP20A for the joint education meeting of 3 October 2025.
• Hales, J. and Orpurt, S. F. (2013). A review of academic research on the reporting of cash flows from operations. Accounting Horizons 27(3), 539–578.
• Hribar, P. and Collins, D. W. (2002). Errors in estimating accruals: implications for empirical research. Journal of Accounting Research 40(1), 105–134.
• IASB (2026). Statement of Cash Flows and Related Matters, project page and IASB Update, September 2026.
• IFRS Foundation. IAS 7 Statement of Cash Flows and IFRS 18 Presentation and Disclosure in Financial Statements, standard summaries.
• IAS 7, paragraphs 18 and 19, as adopted in Commission Regulation (EC) No 1126/2008 (legislation.gov.uk).
• Krishnan, G. V. and Largay, J. A. (2000). The predictive ability of direct method cash flow information. Journal of Business Finance & Accounting 27(1–2), 215–245.
• Orpurt, S. F. and Zang, Y. (2009). Do direct cash flow disclosures help predict future operating cash flows and earnings? The Accounting Review 84(3), 893–935.
• Publix Super Markets (2026). Form 10-K for the fiscal year ended 27 December 2025.
• Zeff, S. A. (2025). Cash flow statements: the direct or indirect method? ICAEW, By All Accounts.
• Wikipedia. Cash flow forecasting.

Findings from the journal articles come from their published abstracts. Sums, shares and runway figures given without a citation are derived from the filed numbers, as are the grouped lines in the Publix table. The software company forecast both ways is hypothetical, and its figures follow from the assumptions stated above its table.