Posted On: 2026-08-12
Posted By: Lery

Financial Modelling for Beginners: Building Your First Three-Statement Model

The three-statement model is the foundational skill in financial modelling. Almost everything else, meaning valuations, leveraged buyouts, budgets, and forecasts, is built on top of it or borrows its logic.

It is also the thing most beginners attempt once, fail to balance, and quietly abandon.

That failure is almost never conceptual. People understand what a balance sheet is. They fail because they build the statements in the wrong order, hardcode figures that should be formulas, or do not understand which line is supposed to make the model balance. All three are fixable in an afternoon once someone explains them plainly.

This guide builds the model in the correct sequence, explains every linkage, and includes a diagnostic section for when it does not balance, because the first time you build one, it will not.

What a three-statement model actually is

Three financial statements, linked by formulas, projected forward over time, driven by a set of assumptions.

The word doing the work is linked. If you change one assumption, such as the revenue growth rate, every affected figure across all three statements updates automatically, and the balance sheet still balances. That is the whole test of whether you have built a model or merely typed three tables into a spreadsheet.

The three statements and what each answers:

Income statement. Did we make a profit over this period?
Balance sheet. What do we own and owe at this moment?
Cash flow statement. Where did the cash actually go?

They connect through two figures: net income and cash. Net income flows from the income statement into the cash flow statement and into retained earnings on the balance sheet. Ending cash from the cash flow statement becomes the cash line on the balance sheet. Get those two links right and the structure holds.

If those concepts feel shaky before you start, the money knowledge every professional should have covers the statements and how they relate.

Before you build: set up properly

Structure decides whether your model is auditable or a liability. Do this before typing a single formula.

Separate inputs from calculations. All assumptions live in one clearly marked block. Nothing hardcoded inside a formula, ever. If your growth rate is typed into the middle of a revenue calculation, nobody can find it, and you will not remember it either.

Use the colour convention. Blue for hardcoded inputs, black for formulas. This is the standard in finance, and reviewers check for it immediately.

One formula per row. Column D should contain the same logic as column E and column F. A row where the logic changes halfway across is the single most common source of real modelling errors.

Keep signs consistent. Decide early whether costs are entered as negatives or positives and never mix. Most beginners create errors by adding a negative they meant to subtract.

Label periods clearly across the top, and mark which are historical actuals and which are forecast. Three years of history and three to five years of forecast is a normal starting shape.

Add a check row at the bottom. More on this later, but build the space for it now.

Step 1: Enter the historicals

Take a real company's filings. Do not use a tutorial dataset, because real financials are messier and you learn far more.

Enter three years of actual figures for all three statements. This serves two purposes. It gives you a base to grow from, and it forces you to see how the statements relate in reality before you start projecting.

While entering them, calculate the historical ratios you will use as forecast drivers: revenue growth, gross margin, operating expenses as a percentage of revenue, days sales outstanding, days inventory, days payable, and capital expenditure as a percentage of revenue.

This step is not preparation for the modelling. It is modelling. The assumptions you choose later should come from what the business has actually done, not from a number that felt reasonable.

Step 2: Build the assumptions block

Every forecast driver in one place, clearly labelled, in blue.

At minimum:

Revenue growth rate per year
Gross margin percentage
Operating expenses as a percentage of revenue
Depreciation, either as a percentage of opening fixed assets or from a schedule
Capital expenditure as a percentage of revenue
Working capital days: receivable days, inventory days, payable days
Interest rate on debt
Interest rate earned on cash
Tax rate
Dividend policy, if any

Base each on the historical ratios you calculated, then adjust where you have a reason. Being able to explain why an assumption differs from history is the actual analytical work, and it is what interviewers probe.

Step 3: Project the income statement

Top down, and stop before the bottom.

Revenue. Prior period revenue multiplied by one plus the growth rate. If you can build it from drivers instead, such as units multiplied by price, or customers multiplied by average revenue per customer, do that. Driver-based forecasts are more defensible and far more useful in a real business.

Cost of goods sold. Revenue multiplied by one minus gross margin. Gross profit follows.

Operating expenses. Usually a percentage of revenue, or grown at a separate rate if you have better information. Subtract to reach EBITDA.

Depreciation and amortisation. Pull this from your fixed asset schedule rather than guessing. Subtract to reach EBIT.

Now stop.

You cannot calculate interest yet, because interest depends on debt and cash balances, which depend on the cash flow statement, which depends on the balance sheet. Leave interest, pre-tax income, tax, and net income as blanks or placeholders. You will come back.

This is the step beginners get wrong most often. They calculate interest from the current debt figure, hardcode it, and the model looks complete but is not actually linked.

Step 4: Build the supporting schedules

Two schedules do most of the connective work.

The fixed asset schedule

 Opening fixed assets
+ Capital expenditure
- Depreciation
= Closing fixed assets

Closing fixed assets feeds the balance sheet. Depreciation feeds the income statement. Capital expenditure feeds the investing section of the cash flow statement. One small schedule, three destinations.

The debt schedule

 Opening debt
+ New borrowing
- Repayments
= Closing debt

Interest expense = average or opening debt balance × interest rate

Closing debt feeds the balance sheet. Interest expense feeds the income statement. Borrowing and repayment feed the financing section of the cash flow statement.

A warning about averages. Calculating interest on the average of opening and closing debt is more accurate, but it creates a circular reference, because interest affects net income, which affects cash, which affects debt, which affects interest. Excel will complain. For your first model, calculate interest on the opening balance. It avoids the circularity entirely and the accuracy difference is small. Learn to handle circularity with iterative calculations later, once the fundamentals are solid.

Step 5: Project the balance sheet, except cash

Forecast each line using a driver rather than a growth rate wherever possible.

Accounts receivable. Driven by receivable days:

 = (Receivable days ÷ 365) × Revenue

Inventory. Driven by inventory days:

 = (Inventory days ÷ 365) × Cost of goods sold

Accounts payable. Driven by payable days:

 = (Payable days ÷ 365) × Cost of goods sold

Fixed assets. From the fixed asset schedule.

Debt. From the debt schedule.

Share capital. Usually held flat unless you are modelling an issuance.

Retained earnings. This is a key linkage:

= Opening retained earnings + Net income - Dividends

Cash. Leave it blank. It comes from the cash flow statement, and this is the point where the whole model resolves.

Working capital driven by days is the professional approach because it ties the balance sheet to operating activity. If revenue grows 20%, receivables should grow with it, and that growth consumes cash. A model where receivables are grown at an arbitrary rate misses the entire relationship between growth and cash, which is usually the most important insight the model produces.

Step 6: Build the cash flow statement

Now the question I was answering earlier. Build it last, because it is derived from the other two statements rather than forecast independently.

The structure has three sections.

Operating activities

Start with net income, then reverse out anything non-cash and adjust for working capital movements.

Net income
+ Depreciation and amortisation
+ Other non-cash charges
- Increase in accounts receivable
- Increase in inventory
+ Increase in accounts payable
= Cash from operating activities

The sign logic, which is where beginners lose hours. An increase in an asset is a use of cash. If receivables rise, you have made sales but not collected the money, so cash is lower than profit suggests. An increase in a liability is a source of cash, because you have received goods or services and not yet paid.

Practical rule: assets up means cash down, liabilities up means cash up.

Take these movements directly from your balance sheet by subtracting the prior period from the current period. Do not calculate them separately, because that is how the two statements drift apart.

Investing activities

 - Capital expenditure
+/- Acquisitions or disposals
= Cash from investing activities

Capital expenditure comes from the fixed asset schedule. For a straightforward model this section is often just capex.

Financing activities

+ New borrowing
- Debt repayments
+ Equity issued
- Dividends paid
= Cash from financing activities

These come from the debt schedule and your dividend assumption.

Reaching ending cash

 Opening cash
+ Cash from operating activities
+ Cash from investing activities
+ Cash from financing activities
= Ending cash

Ending cash now links back to the cash line on the balance sheet. That single link is what closes the model.

Step 7: Close the loops

Two connections finish the build.

Interest. Now that you have cash and debt balances, go back to the income statement and calculate interest expense from the debt schedule and interest income from the cash balance. Complete the income statement down to net income.

Retained earnings. Confirm that net income is feeding retained earnings on the balance sheet correctly.

At this point the model is circular in concept though not in Excel, provided you used opening balances for interest. Change any assumption and the effect flows through all three statements.

Step 8: The balance check

Add a row at the bottom:

 Check = Total assets - (Total liabilities + Total equity)

It must equal zero in every period. Format it so a non-zero value is immediately visible, using conditional formatting in red.

This row is not decoration. It is the model's only reliable proof of correctness, and no professional model ships without one.

When it does not balance

It will not balance the first time. That is normal and the diagnosis is systematic rather than mysterious.

First, find the pattern. Look at the size and behaviour of the error.

If the error equals net income, retained earnings is not linking to the income statement, or it is linking to the wrong period.

If the error equals depreciation, depreciation is missing from either the fixed asset schedule or the add-back in operating cash flow.

If the error appears in the first forecast period only, an opening balance is wrong or a formula is pulling from the wrong column.

If the error grows each period by the same amount, something is being double counted or omitted consistently. Check whether a balance sheet line is being forecast independently and also appearing in the cash flow statement.

If the error is exactly double a figure you recognise, you have a sign error. You added where you should have subtracted, which produces an error of twice the amount.

If the error equals a working capital movement, your cash flow statement is not pulling the movement from the balance sheet, or the sign is inverted.

The systematic approach. Check in this order:

Does opening cash equal the prior period's ending cash?
Does ending cash on the cash flow statement equal cash on the balance sheet?
Does retained earnings equal opening plus net income minus dividends?
Does closing debt on the balance sheet match the debt schedule?
Does closing fixed assets match the fixed asset schedule?
Are all working capital movements taken from balance sheet differences?

Nine times out of ten the error is in that list. Work through it in order rather than staring at the whole model.

Common beginner mistakes

Building the cash flow statement first. It cannot be forecast independently. It is derived.
Hardcoding numbers inside formulas. The model stops being a model the moment an assumption cannot be changed in one place.
Calculating interest on average balances in a first model. The circular reference will confuse you before you understand it. Use opening balances.
Forecasting working capital as a growth rate. Use days. The relationship to revenue is the entire point.
Inconsistent formulas across a row. Build column D, then copy right. If you edit a single cell mid-row, note it prominently.
No balance check. Without it you have no idea whether the model is right, and neither does anyone reviewing it.
Forgetting that increases in assets consume cash. The most common sign error in the whole exercise.

What to do once it works

Add scenarios. Build a switch using CHOOSE driven by one input cell, so you can move the whole model between base, upside, and downside cases:

 =CHOOSE($C$4, base_value, upside_value, downside_value)

Add a sensitivity table. A two-way data table showing how a key output moves across a range of two assumptions. In finance, presenting a single number where a range belongs is a substantive error.

Then build a valuation on top. Once you have projected free cash flow, a discounted cash flow model is the natural next step, and it reuses everything you have built.

Then do it again from scratch for a different company in a different industry. The second build is when the logic actually becomes yours, because you are no longer following steps, you are making choices.

Why this is worth the effort

Beyond the obvious use in valuation and planning, the three-statement model is the most efficient way to genuinely understand how a business works. You cannot build one without confronting the fact that growth consumes cash, that profit and cash are different things, and that a capital-intensive business behaves nothing like a service business.

It is also directly testable in hiring. Modelling questions and Excel exercises are standard in finance interviews, and being able to say you have built a full three-statement model from real filings, and to walk through the linkages under questioning, is stronger evidence than any course certificate. Two or three completed models constitute a portfolio, which matters a great deal if you are early in your career or moving into finance from another field.

Presenting that work correctly on paper is its own skill, and our guide on how to write a finance CV that passes ATS screening covers how to turn projects like this into quantified achievement bullets rather than a vague skills claim.

The bottom line

Build the income statement down to EBIT. Build the fixed asset and debt schedules. Build the balance sheet except cash. Derive the cash flow statement. Link ending cash back to the balance sheet. Close the interest loop. Add the check row.

That sequence is the model. Everything else is refinement.

Your first attempt will not balance, and that is not evidence you cannot do this. It is the ordinary experience of everyone who has built one. Work the diagnostic list, find the broken link, and fix it. The moment the check row turns to zero, you will understand the three statements in a way that no amount of reading achieves.

Financial Modelling for Beginners: Building Your First Three-Statement Model