How the Three Statements Link

One model, three views

Every transaction is recorded twice, and the three statements are the two sides grouped by what they are measuring rather than by account. The income statement measures performance over a period — revenue, cost, profit. The balance sheet measures position at an instant — assets, liabilities, equity. The cash flow statement measures movement over the same period as the income statement, restricted to the one asset everyone watches. Articulation is what links them. Net income flows into retained earnings, so the income statement explains part of the change in equity. Cash from all three sections flows to the closing cash balance, so the cash flow statement explains the change in one line of the balance sheet — and it has to reconcile exactly, because it starts from net income and adjusts for everything else that moved. The reason this matters beyond bookkeeping is forecasting. If you assume revenue grows 10% and leave receivables days unchanged, receivables grow 10% too, and the cash flow statement has to absorb that increase. Model the income statement alone and you will predict profits the business cannot fund; model the cash flow statement alone and you will not know why. A three-statement model is just the discipline of making the three agree. The trace, in order — Transaction: a $20m sale on 30-day terms, at a 40% gross margin: Revenue and gross profit rise · Income statement: Operating profit up $12m · Balance sheet: Receivables up $20m, retained earnings up $12m, inventory down $8m · Cash flow statement: Net income up $12m, working capital subtracts $12m, cash unchanged ← · Thirty days later: Receivables become cash in operating activities A transaction can be profit without being cash, or cash without being profit. Depreciation, working capital and capex are the three places the two diverge in practice, and F4 works each of them.

The four traces worth being able to do in your head

Depreciation. Add $10m of depreciation and operating profit falls $10m, retained earnings fall $6.9m after tax, property falls by the accumulated amount, and the cash flow statement adds the $10m back — which is why a non-cash charge can leave cash untouched while lowering the tax bill. The tax saving, $2.1m at a 21% rate, is the only cash effect, and it arrives through a lower tax payment. A sale on credit. Revenue and profit rise, receivables rise by the full invoice, and the cash flow statement subtracts the increase in receivables. Nothing has arrived. Thirty days later, when the customer pays, receivables fall and operating cash rises by the same amount, with no further effect on profit — which is why a year of high growth reports profit well above cash flow and a year of slowing growth can report the reverse. A machine bought on finance. Property and debt rise by the same amount, equity is untouched, and the investing section shows nothing while the note discloses it. Depreciation arrives over the machine’s life, and the repayments arrive in financing, now split between interest in operating and principal in financing. A dividend. Cash falls, retained earnings fall by the same amount, and nothing touches the income statement at all. This trace is the one that explains why a company can be profitable every year and still be worth less than it was: dividends are not an expense, they are a distribution of the equity that the profit built. Cash does not appear twice. A common error in a hand-built model is to subtract capex in investing and also to leave it out of the depreciation schedule; the cash leaves once, and the asset it bought depreciates separately, over years.

The fourth statement nobody reads

Three statements explain almost everything, and the fourth catches what they miss. The **statement of changes in equity** reconciles the equity on last year’s balance sheet to this year’s, line by line: net income added, dividends and buybacks taken out, shares issued for employee compensation put in. It is the only statement where the change in share count is stated in financial terms rather than buried in a note, which makes it the fastest place to see whether reported profit is being paid for with dilution. It also carries the items that bypass the income statement entirely. **Other comprehensive income** collects unrealised movements — the translation of foreign subsidiaries when the dollar moves, unrealised gains on securities held for sale, changes in the value of hedges. None of it passes through profit, all of it changes equity, and it can be large: a company with substantial overseas earnings can post a solid year of profit and a flat or falling book value because a strong dollar shrank the translated value of everything abroad. The practical use is as an audit. If the equity line moved by an amount that the income statement, the dividends and the buybacks cannot explain, this statement names which of the other things did it — and the fourth one is almost always the one the learner has never heard of. • Net income raises equity; dividends and buybacks lower it • Shares issued as compensation raise it in dollars and dilute the count • Other comprehensive income moves equity without ever passing through profit • A buyback is not value creation on this statement — it is a transfer from equity to shareholders This is where a multinational or emerging-market investor watches currency: a translation loss can shrink a group’s book equity for a year without a single operating decision being made, and it reverses when the currency does (MR18).

The three errors articulation catches

Articulation is not a formality performed at the end of a model. It is the test that finds the three mistakes that make a forecast look convincing and be wrong, and each of them has a specific signature. **The plug.** If the balance sheet does not balance, the lazy fix is a line called something like “other assets” or “funding adjustment” that absorbs the difference. Every real model has a moment where this temptation appears. The honest fix is to find the transaction that was recorded once: a dividend paid without cash falling, capex without the asset rising, a share issue that never reached equity. A plug of a few percent of assets can move a valuation by more than the entire growth assumption. **Cash that does not move when it should.** The cash flow statement starts from net income and adjusts for everything else. If you increase revenue and leave receivables days unchanged, receivables must rise and cash from operations must fall by the same amount — that is not a modelling choice, it is the identity. A model where revenue grows and free cash flow does not move is not optimistic, it is broken. **And the tax line.** Tax appears in the income statement, inside deferred tax on the balance sheet and in the cash flow statement, and a change in the rate has to land in all three. When it does not, the effective rate and the cash paid drift apart with every forecast year. The signature of each error — A growing “other assets” line: a plug absorbing an unrecorded transaction ← · Revenue up, receivables days flat, cash flat: working capital never modelled ← · Net income and cash from operations diverging for no stated reason: an accrual with no counterpart · The effective tax rate drifting from the cash tax rate: the rate change landed in only one statement · Equity not reconciling to the prior year plus profit less distributions: something bypassed the statement of changes in equity (F5) A three-statement model is worth building precisely because it fails loudly. A revenue-and-margin spreadsheet agrees with you no matter what you type into it.

Why articulation is a fraud detector

The three statements are not three reports of the same year; they are three views of one set of transactions that have to stay consistent, and that constraint is what makes them hard to fake. A company can choose an estimate, move a cost, or recognise revenue early, but whatever it does has to land in all three places at once, and each landing point leaves a mark. Putting a fake sale in the income statement raises revenue and profit; putting the cash in requires a matching asset, which means an invented receivable; running the cash through means an invented bank balance, a cash flow statement that does not tie to the balance sheet, or a real counterparty who later denies the transaction. The larger the invention, the more of the three statements it has to touch, and the more surface area there is to check. This is why the practical forensic checks in this subject are all articulation checks rather than ratio checks. Compare the growth of revenue with the growth of receivables; if receivables outrun revenue for several quarters, either the business has changed its terms or some of the revenue is not yet real. Compare cumulative net income with cumulative operating cash flow over three to five years; a persistent gap is the accrual build that eventually has to reverse. Compare the balance sheet to the statement of cash flows at the level of the cash balance itself — it is the one line that both statements contain, and a company whose cash does not tie is telling you something. None of these requires an estimate or a forecast, which is exactly why they are powerful: they are internally referenced. The skill worth building is reading the statements **as a single machine** rather than as three documents to be read in sequence. When something moves, the question is not “what did this do to profit” but “where did the other side of this go” — and the answer is always somewhere, in inventories, receivables, payables, deferred revenue, an intangible, a lease liability, or the cash balance itself. An analyst who can trace the other side has a check on every line at once, and a check is worth more than any single ratio, because a ratio tells you what happened and articulation tells you whether it happened at all. • Every transaction must land in all three statements, and each landing point leaves evidence. • Receivables outrunning revenue, and cumulative profit outrunning cumulative cash, are the two classic tells. • The cash balance appears in two statements — if they disagree, stop and read the notes. • Ask “where did the other side go” about every change; the answer is always somewhere.

The plumbing: circularity, the revolver and the cash sweep

A three-statement model is not merely three linked statements. As soon as you model interest on the debt balance, it becomes **circular**: interest expense depends on how much debt is outstanding, the debt balance depends on how much cash the business generated after interest, and the cash generated depends on the interest expense. Spreadsheets announce this as a circular-reference warning, and the two wrong ways to handle it are common. The first is to switch on iterative calculation and never check whether it converged, which produces numbers that depend on how many times the engine happened to loop. The second, and worse, is to break the loop by pasting hard-coded values over the interest formula — a model that produces a balanced balance sheet and a false income statement, and that quietly stops updating when an assumption changes. The principled fix is to model the **revolver and the cash sweep**, which is what the majority of real models do because it also describes how a company actually funds itself. The mechanism has three parts. A minimum cash balance, because a business cannot operate on zero. A short-term borrowing facility — the revolver — that is drawn when the cash flow after mandatory debt repayments would push cash below the minimum. And a sweep: any positive cash flow above the minimum and after mandatory repayments is used to repay the revolver first. With those three rules the loop is bounded rather than circular, because the ending revolver balance is determined by the cash flow before interest, and interest can then be computed on the average of the opening and closing balances without changing them. The check that the loop has converged is mechanical: cash never falls below the minimum, the revolver never goes negative, and interest equals the rate multiplied by average debt in every period. If all three hold, the model is consistent; if the revolver is negative, the model is repaying debt it never drew. The cash sweep also reveals why so many models that balance are still wrong. A model with no revolver has nothing to absorb a funding deficit, so the deficit shows up as negative cash on the balance sheet — which still balances, because the identity does not care about the sign — or as an unexplained plug. A model with a sweep but no minimum cash lets the company operate on a zero balance, which flatters the interest line and can make a distressed business look solvent. And a model that computes the sweep on end-of-period balances instead of average balances will pay the wrong interest in a declining debt schedule, which matters most in exactly the leveraged situations where the model is being used. One layer sits underneath all of it and belongs to the same conversation: the **classification judgements** that decide what lands in operating cash flow. Interest paid is reported in operating activities under U.S. GAAP and can be presented in financing under IFRS. Capitalised interest moves a financing cost into the investing section and out of the income statement. Leases, since the accounting changes of the late 2010s, sit on the balance sheet as a liability with a matching asset, and the interest portion can land in either operating or financing depending on the election. Stock compensation is added back as a non-cash charge inside operating cash flow while the dilution arrives in the equity section. None of these moves the business, and every one of them moves the cash-flow statement that a screen reads — which is the practical reason the articulation checks of this lesson have to be built into the model rather than applied to the output. • Interest on debt makes the statements circular: the fix is a bounded loop, not an override. • Revolver plus minimum cash plus a sweep determines the debt balance from cash flow before interest. • Convergence check: cash at or above the minimum, the revolver never negative, interest on average debt. • A model with no revolver balances with negative cash — the identity is satisfied and the story is not. • Classification choices — interest, capitalised interest, leases, stock compensation — move cash flow without moving the business. The link to the fraud section of this lesson is direct: a model that needs a hard-coded plug to balance is the spreadsheet version of exactly the accounting a company would use to hide a problem. If your own model cannot converge honestly, that is information about the assumptions rather than about the software.

Sources

• The articulation of the three statements and the articulation of cashStandard financial modelling practice; CFA Institute FSA curriculum
• Retained earnings as the bridge between profit and equityFASB ASC 220 / ASC 505
• Why a change in a forecast must be traced through all three statementsStandard three-statement modelling practice (e.g. Rosenbaum & Pearl, "Investment Banking")