A three-statement model links the income statement, the balance sheet and the cash flow statement so that they move together. Change a sales assumption and the model shows not just the new profit, but the new debtor balance, the cash that is tied up as a result, the effect on the overdraft and whether the bank covenant still holds. That is the whole point of building one. A profit forecast tells you whether a plan makes money; an integrated model tells you whether the business can afford to carry it out.
This is a practical how-to for FP&A analysts, management accountants and Financial Controllers: structure, build order, the links, working capital, debt and covenants, scenarios, checks, common errors and model hygiene. If you are an employer looking for someone to build and own the model, our page on FP&A recruitment covers permanent, interim and fractional hires.
What a three-statement model is and why integration matters
The three primary statements are three views of the same financial position. Profit for the period flows into retained earnings on the balance sheet. Movements in the balance sheet, together with profit, explain the change in cash. The cash flow statement then closes to the cash line on the balance sheet. When those links are built in, a change anywhere flows correctly through all three statements and the balance sheet balances without being forced.
Statements built separately can produce impossible results, such as a balance sheet that only balances because of a plug. Integration makes the balance check a genuine test of the model, and it shows consequences a profit-only forecast hides. Strong sales growth on sixty-day terms improves profit and consumes cash; only the integrated model shows both at once.
The structure of a sound model
Good models separate three things: inputs (assumptions and historical data), calculations (the supporting schedules) and outputs (the three statements, a summary and the checks). It is the most important design principle: every assumption can be changed in one place, and someone other than the builder can follow the logic.
A typical workbook layout, left to right:
| Sheet | What it holds |
|---|---|
| Cover and contents | Purpose, version, owner, change log, colour key |
| Inputs | Timeline, scenario selector, growth, margins, working capital days, capex, financing, tax rates |
| Historicals | Two or three years of actual P&L and balance sheet, mapped to the model’s line structure |
| Revenue and costs | Volume and price drivers, cost of sales, overheads, headcount |
| Schedules | Fixed assets, working capital, debt, tax, equity and dividends |
| Statements | Income statement, balance sheet, cash flow (indirect method) |
| Checks and summary | Balance check, cash check, covenant tests, KPIs, one master error flag |
Match the detail to the purpose: a refinancing model needs a careful debt schedule; a planning model needs richer revenue and headcount drivers. Our guide to driver-based planning covers how to choose the operational drivers that belong on the inputs sheet, and building driver trees shows how to map them back to the P&L.
Step 1: Set the timeline and inputs
Decide the periodicity first, because it is painful to change later. Monthly for the first twelve to eighteen months and annual thereafter is common for operating businesses; quarterly suits longer-range strategic models. Build a single timeline row (period dates, days in period, financial year flag, actual or forecast flag) and reference it everywhere.
Put every assumption on the inputs sheet in a consistent colour, with units stated. For scenarios, lay out base, downside and upside columns side by side with one selector cell choosing which feeds the model, rather than keeping copies of the workbook named “v7 downside FINAL”.
Step 2: Load and map the historicals
Bring in two or three years of actual results and the latest balance sheet, mapped to the line structure the model will forecast. The opening balance sheet matters most: every forecast balance rolls forward from it, so if it does not balance, nothing downstream will. Reconcile it to the trial balance or the signed statutory accounts before going further. Historicals also supply the ratios that anchor the assumptions (gross margin, debtor days, capex to revenue), so a forecast that departs sharply from history has to be explained. Keep the mapping to the dozen or so lines that matter for forecasting.
Step 3: Build the income statement to EBITDA
Forecast revenue from drivers rather than a single growth rate where possible: customers × average order value, units × price, or hours × rate × utilisation. Cost of sales follows from the same drivers or a margin assumption. Overheads are best built from headcount (with employer National Insurance, which rose to 15% in April 2025, and pension costs) plus the main non-pay cost lines. Stop at EBITDA for now: depreciation, interest and tax depend on schedules not yet built.
Step 4: Build the supporting schedules
Each schedule follows the same corkscrew pattern: opening balance, plus additions, less reductions, equals closing balance, with the closing balance becoming the next period’s opening. Build them one at a time and check each before moving on.
Fixed assets
Opening net book value, plus capital expenditure, less depreciation, less disposals, equals closing net book value. Depreciation feeds the P&L; capex feeds investing cash flows (see capex appraisal and investment cases). With significant leases, decide early how they will be shown; lease accounting under IFRS 16 explains the mechanics for IFRS reporters.
Equity and dividends
Opening retained earnings, plus profit after tax, less dividends, equals closing retained earnings. Share capital moves only when shares are issued or bought back. This is the first link between the income statement and the balance sheet, and it should be a direct reference to the P&L, never a separately typed number.
Tax
Separate the tax charge in the P&L from the tax paid in cash. The corporation tax creditor rolls forward as opening balance plus charge less payments. Timing matters: smaller companies typically pay nine months and one day after the year end, while large companies pay by quarterly instalments, so the cash outflow can fall in a different year from the charge. Current corporation tax rates and quarterly instalment rules are set out on GOV.UK. Most operating models ignore deferred tax unless it is material; where it is, our note on deferred tax fundamentals sets out what to model.
Working capital drivers
Working capital is where much of the insight sits, and where many models are weakest. Forecast each balance from a days assumption tied to the relevant P&L line:
- Trade debtors = revenue in the period ÷ days in period × debtor days.
- Stock = cost of sales ÷ days in period × stock days.
- Trade creditors = purchases (or cost of sales plus relevant overheads) ÷ days in period × creditor days.
- Accruals, prepayments and deferred income as a share of the related cost or revenue line, or held flat where small.
Three practical points. Trade debtors and creditors include VAT while the P&L does not, so gross up standard-rated sales and purchases. Model the VAT creditor and its quarterly payments rather than burying it in “other creditors”. And in a monthly model, use actual days per period from the timeline row.
The change in each working capital balance from one period to the next feeds the cash flow statement. An increase in debtors or stock is a cash outflow; an increase in creditors is an inflow. That is how growth consumes cash. For the short-range, week-by-week equivalent built from ledgers rather than days assumptions, see our practical guide to the 13-week rolling forecast.
The debt schedule and covenants
Each facility gets its own corkscrew: opening balance, plus drawdowns, less repayments, equals closing balance, with interest calculated alongside. Term loans follow their amortisation profile; a revolving credit facility or overdraft is usually modelled as the balancing item that draws when cash would fall below a minimum and repays when there is surplus.
Interest and circularity
Interest on the average balance is more accurate but circular: interest affects profit, profit affects cash, cash affects the revolver, and the revolver affects interest. Either calculate interest on the opening balance (simpler, no circularity) or enable iterative calculation with a labelled circuit-breaker switch that sets interest to zero if the model blows up. For floating-rate debt, link the rate to a base-rate assumption plus margin; the Bank of England publishes Bank Rate decisions.
Covenant tests
Build the covenants exactly as the facility agreement defines them, not as a generic textbook ratio. Common tests include leverage (net debt to EBITDA), interest cover (EBITDA to interest) and cash flow or debt service cover. Facility definitions of EBITDA and net debt often include specific adjustments. Show each covenant as actual ratio, threshold and headroom, with a red flag if headroom falls below a level the business is comfortable with. Monitoring this is routine work that suits automation; our guide to AI for covenant and KPI tracking covers the practical options.
Step 5: Complete the statements and link them
With the schedules built, finish the income statement: EBITDA, less depreciation, equals operating profit; less interest from the debt schedule, equals profit before tax; less the tax charge, equals profit after tax.
Build the balance sheet entirely from references to schedule closing balances; no line should contain a calculation of its own.
Build the cash flow statement using the indirect method:
- Operating: profit after tax, add back depreciation and interest, add back the tax charge, less tax paid, less the increase in working capital.
- Investing: capital expenditure, disposal proceeds.
- Financing: drawdowns and repayments, interest paid, dividends, share issues.
Opening cash plus net cash flow gives closing cash, and that closing cash is what the balance sheet cash line should reference. This final link closes the loop: if every other link is right, the balance sheet now balances.
Need Someone to Build and Own the Model? — 0204 553 8893
We recruit FP&A analysts, FP&A managers and Financial Controllers who build integrated models, permanently or on an interim or fractional basis. Email recruitment@accountancycapital.co.uk; we respond the same working day.
Checks that catch errors
A robust model tells you when it is wrong. Build these checks on a dedicated sheet and roll them up into a single master flag shown at the top of every output sheet:
- Balance check: total assets less total liabilities less equity equals zero, every period.
- Cash check: cash flow closing cash equals balance sheet cash.
- Roll-forward checks: each schedule’s opening balance equals the prior period’s closing balance.
- Historical check: the model’s historical balance sheet agrees to the source accounts.
- Sense checks: no negative stock or debtors, cash never below the minimum without the revolver drawing, facility never above its limit.
Test against a small tolerance rather than exactly zero, to allow for rounding residues. A difference that grows each period usually points to a cash flow sign error.
Using the model for forecasting and scenarios
Because every assumption flows through all three statements, a scenario shows the full consequence: profit, cash, balance sheet and financing. A plan that looks profitable may reveal that its working capital takes the business through its facility limit in month nine.
A base case, a realistic downside and sometimes an upside are usually enough. Sensitivity tables (for example, revenue growth against gross margin, showing minimum cash headroom) are a quick way to show which assumptions actually matter. Our guide to scenario and sensitivity analysis goes further, and the model also becomes the engine of a rolling forecast once actuals are loaded each month and the forecast periods roll forward.
The common errors to avoid
Three-statement models go wrong in recognisable ways:
- Failed integration. The balance sheet needs a plug. Build every line from a schedule and treat the balance check as non-negotiable.
- Hard-coded numbers inside formulas. Invisible, and they will not change when the inputs do.
- Sign errors in the cash flow. Working capital increases shown as inflows, or repayments added rather than deducted. Adopt one sign convention and state it on the cover sheet.
- Inconsistent formulas across a row. One period overwritten with a different formula is one of the hardest errors to spot. Every row should carry the same formula from the first forecast period to the last.
- Unmanaged circularity. Iterative calculation switched on silently, so the model stops calculating properly and nobody notices.
- Too much complexity. Detail nobody uses and macros only the builder understands.
Building a robust and maintainable model: model hygiene
A robust model does not break when inputs change; a maintainable one can be updated by someone other than its builder. Both come from habits:
- Consistent colour coding: inputs in blue, formulas in black, links to other sheets in green.
- One formula per row, copied across; columns for the same period in the same position on every sheet.
- Short, readable formulas. Break a long calculation into helper rows rather than nesting six IF statements.
- A change log on the cover sheet and version control by date, not by “final” and “final final”.
- Peer review before any model is used for a board paper, lender submission or investment decision.
The UK government’s Aqua Book, published on GOV.UK, is a useful reference on quality assurance for analytical models, and the ICAEW Excel Community publishes practical material on spreadsheet good practice. Our comparison of FP&A tooling covers when it is worth moving off Excel, and using AI inside Excel covers where AI assistants help with formula review and where they need close supervision.
Who should build and own the model
In most mid-sized businesses the three-statement model sits with FP&A where that function exists, and with the Financial Controller or Finance Manager where it does not. An FP&A analyst often does the build and monthly update; an FP&A manager owns the assumptions, challenges the business on them and presents the outputs. The Financial Controller cares most about the opening balance sheet, the actuals and the covenant tests.
Modelling is one of the most testable skills in finance hiring: a short exercise, such as fixing a deliberately broken model, tells an employer more than “advanced Excel” on a CV. Our FP&A interview questions include modelling questions worth using. When a business needs a model built quickly for a refinancing, fundraise or budget season, an interim is often the practical answer; see interim FP&A budget-season cover. For the Financial Controller role in a transaction, our guide to the FC’s role in fundraising and due diligence explains where the model fits. Where the work sits at finance director level, our sister firm FD Capital handles CFO and FD searches.
Frequently asked questions
Should I use the direct or indirect method for the cash flow?
The indirect method is standard in a three-statement model because it derives cash from profit and balance sheet movements, which is what makes the integration work. Short-range cash forecasts, such as a 13-week cash flow forecast, use the direct method instead.
Why doesn’t my balance sheet balance?
Check, in order: the opening balance sheet; that profit flows to retained earnings; that every balance sheet movement has a matching cash flow line with the right sign; and that closing cash on the cash flow is the figure the balance sheet uses. If the difference first appears in a particular period, start there.
What should a modelling test at interview look like?
A proportionate timed exercise: build or fix a small integrated model from clear assumptions, then walk through the checks used.
Adrian Lawrence FCA — Founder, Accountancy Capital
Fellow of the ICAEW | Chartered Accountant holding an ICAEW practising certificate | Part of FD Capital Group
Adrian is a former listed-company Finance Director with a BSc from Queen Mary College, University of London. His view is that a good three-statement model is not clever: it balances without a plug, every assumption sits in one place, and someone else can pick it up next year. Asking a candidate how they check a model is one of the quickest ways to tell a capable analyst from a basic one.
Related Pages and Resources
Forecasting Guides
The forecasting work the model supports.
Cash and Drivers
Short-range cash and operational drivers.
FP&A Hiring
Analysts and managers who build models.
Financial Controller Hiring
Permanent, interim and fractional FCs.
→ Financial Controller recruitment
Hiring an FP&A Professional Who Can Model? — 0204 553 8893
Call us or email recruitment@accountancycapital.co.uk. We respond the same working day, with permanent shortlists typically within one to two weeks and interim or fractional candidates within days. Fees are agreed in writing before a search starts.