The cash flow statement

The income statement says what the period earned. The balance sheet says what the cash level is on the last day. Neither says how one became the other, and that is the only question the cash flow statement answers. It is a reconciliation with a fixed start, a fixed end, and a middle that has to add up.

That structure is what makes it checkable. If the four moves do not land on the change in the bank balance, a line is missing, and the amount by which you miss is the size of what you left out.

Each row the engine returns already carries every figure the statement needs. There is no separate cash flow object, because the statement is a way of reading the row rather than a fifth thing to compute.

No credentials of any kind are involved. If an example ever needs one, the string is the word placeholder.

Net income is the figure the income statement finished with, after tax and after interest. It is already net of every cost the period incurred, which is exactly what makes it the right place to begin: everything still to come is a correction for timing, not a new cost.

Marchmont's twelve month figure is £445,631, and every month of it is positive. A statement that begins here and ends below zero has not found a loss. It has found three timing differences.

Depreciation is the accounting recognition that an instrument bought three years ago is being used up. It reduced profit this month. No money left the bank this month because of it, so it has to come back out of the bridge.

Over twelve months that is £102,025. It is the only add back in the Marchmont model, which is unusual: real statements often carry several, and the test for each one is identical.

Session 3 built the three balances. Here they become one number. Receivables rising means revenue recorded and not collected. Inventory rising means money spent on instruments not yet sold. Payables rising works the other way, because an unpaid supplier is a cost recorded without money leaving.

Over the year the three together consumed £664,607, which is the single largest line on the bridge and larger than the profit above it. Receivables alone grew by £539,454.

Capex is money paid for something that will be used over years, so the income statement refuses to treat it as this period's cost and spreads it through depreciation instead. The bank does not spread anything. The full £180,310 came out during the year.

Notice where the subtotal sits. Operating cash flow of minus £116,951 is struck before capex, so it describes trading alone. Free cash flow of minus £297,261 is struck after it. The same twelve months, two legitimate summary figures, nearly £180,000 apart, and which one gets quoted is a choice somebody makes.

Anybody reading a set of accounts has the income statement and the balance sheet in front of them. The indirect method begins with a number they can already see and moves it using changes they can already compute from two balance sheets. That makes the statement auditable from outside, without access to the ledger.

The direct method's first line, total cash received from customers, cannot be reconstructed by a reader at all. It has to be taken on trust, which is the opposite of what a reconciliation is for.

Both methods would tell you Marchmont's bank balance fell. Only one tells you that £664,607 of it is recoverable money sitting in invoices and stock rather than money that has gone. Those two readings lead to completely different decisions, and the second one is a financing conversation rather than a performance one.

This is also the honest case against the indirect method. Starting from an accrual figure and adjusting it means the statement is built from estimates rather than from bank records, and an error in the accruals flows straight through. The direct method has no such exposure. That is a real trade and it is the reason the argument has never fully closed.

A bridge read from the bottom is a diagnostic. Free cash flow of minus £297,261 with capex of £180,310 says a little over half the drain was investment, which is a choice that can be paused. The other half is operating, which cannot.

Within the operating half, £664,607 went into working capital while profit contributed £445,631. A business whose operating cash flow is negative purely because of a working capital build is in a different position from one whose profit is negative, and the bridge is what distinguishes them.

Profit grows faster than the working capital build does, so the gap between them narrows every month. That is the improvement. But the gap is still a gap: minus £3,165 in month 11 and minus £427 in month 12 are smaller drains, not deposits.

A level only rises when the flow into it is positive. Reading an improving flow as a recovering level is the most common misreading of a cash flow statement, and it is comfortable because the direction of travel looks right.

Opening receivables, inventory and payables are set consistently with month 1 activity, so the working capital delta in month 1 is exactly zero and operating cash flow that month is £22,062. That figure is real output from the model and it is not evidence about Marchmont.

Any model with a constructed opening position has a month 1 like this. Quoting it as a finding is a mistake worth making once, in private, rather than in a board pack.

The third expression is the check worth keeping. A subtotal that cannot be rebuilt from the lines above it is a subtotal that has picked something up on the way.

One side says interest is a cost of running the business as it is currently funded, so it belongs with trading and operating cash flow should be struck after it. On that view, moving interest out of operating flatters a leveraged company and makes it look like a lightly geared one.

The other side says interest is a payment to a capital provider, exactly like a dividend, and that operating cash flow should describe what the assets produce regardless of who financed them. On that view, leaving interest in operating makes two identical businesses look different purely because one borrowed.

Both are coherent and the frameworks have not converged. In the Marchmont engine interest is deducted before net income, so it is inside the operating figure, and the minus £116,951 is an after interest number. Saying which convention a figure used is more useful than defending either one.

There is no single agreed definition of free cash flow. Subtract all capex and you get minus £297,261. Subtract only the capex needed to keep the existing instruments running and you get something less negative, because part of Marchmont's £180,310 is buying capacity for growth rather than replacing what wore out. Both figures are honestly labelled free cash flow.

The useful move is not to pick a winner. It is to ask what was subtracted to get there, which is a fair question to put to your own team as well as to somebody else's filing. A free cash flow figure with no stated subtraction list is not wrong, it is unreadable.

Marchmont's debt never moves in this model, no shares are issued and no dividend is paid, so the financing section is empty and free cash flow happens to equal the change in the bank balance exactly. That equality is a property of these assumptions, not a general rule, and it will stop holding the moment session 8 adds borrowing.

It is worth noticing in a real statement too. A company whose financing section is empty is living on what it generates. One whose operating section is negative and whose financing section is large is living on somebody else's money, which is a position rather than a verdict.