The wrong clock: why cash in and out misleads
In the previous guide you met the income statement as a whole — the top line of revenue, the costs stacked below it, and the layers of profit it produces. But we glossed over a question that turns out to decide every number on it: *when* does a sale belong to this period, and *when* does a cost? The tempting answer is to watch the bank account — count a sale when the money lands, count a cost when the money leaves. That is the cash basis, and it is exactly how your own checking account thinks. It is also the wrong clock for measuring whether a business performed well.
Picture a small consultancy. In December it does $50,000 of hard, finished work for clients, but those clients pay on 30-day terms, so the cash does not arrive until January. Meanwhile in December the firm pays last month's office rent and an old supplier bill. On the cash basis, December looks like a disaster — money flowed out, almost none came in — and January looks like a triumph, when in truth December was the month the firm actually earned its keep. The bank balance and the real performance have drifted completely out of step. The cash clock told a lie, not because anyone cheated, but because earning and collecting are simply two different events.
Accounting's answer is to throw away the cash clock for measuring performance and adopt a second one: the [[cash-basis-vs-accrual-basis|accrual basis]]. It records revenue when it is *earned* and expense when it is *incurred*, no matter when the cash happens to move. This is not a minor bookkeeping preference — it is the foundation everything in this rung stands on, and it is why GAAP and IFRS require it for almost every real company. The rest of this guide unpacks its two halves (when is revenue earned, when is expense incurred) and then the rule that binds them together.
Revenue is earned, not collected
Start with the top line. The rule of [[revenue-recognition|revenue recognition]] is short to state and easy to misuse: you record revenue in the period you actually deliver the goods or perform the service — the moment you have done what the customer paid you for. Cash can arrive before, after, or at that moment, and it simply does not matter for *timing* the revenue. The consultancy above earned its $50,000 in December because that is when it finished the work; the January cash is just the settling-up of a debt the client already owed.
The trickier case is when cash arrives *first*. Imagine a gym that sells a $1,200 annual membership and collects the whole sum on January 1st. It is tempting to crow that January earned $1,200 — but the gym has barely unlocked the door. It still owes the member eleven more months of service. So on January 1st it has not earned revenue at all; it has taken on an obligation. The honest entry records $1,200 of *cash* received and $1,200 of a liability called unearned revenue, then converts $100 of that liability into earned revenue each month as the service is actually delivered. Revenue is tied to performance, never to the calendar date the money cleared.
Expense is incurred, not paid
The bottom half of the income statement obeys the mirror-image rule. An [[expense|expense]] is a resource the business used up in order to earn its revenue — flour burned, electricity consumed, staff hours spent — and it belongs to the period in which it was *used*, not the period its bill was paid. A December electricity bill settled in January is still a December expense, because that is when the lights were on. If a shop consumes $5,000 of supplies, owes $3,000 of rent, and runs up $7,000 of wages in a month, it has $15,000 of expense that month regardless of which invoices have actually cleared the bank.
Here lies a trap that catches almost everyone: not every payment is an expense, and not every expense is a payment. When the shop buys a $40,000 delivery truck, no expense happens that day — it has merely swapped one asset (cash) for another asset (a truck). The expense appears slowly, year by year, as the truck wears out: that yearly slice of used-up value is [[depreciation|depreciation]], and only that slice hits the income statement. Conversely, the wages your staff earned in the last week of December are an expense of December even if payday falls in January — an *accrued* cost recorded before any cash leaves. Cash out and expense are two different rhythms.
The matching principle: lining cost up with the sale it created
Now we can state the rule that gives the whole income statement its meaning. The [[matching-principle|matching principle]] says: recognize an expense in the same period as the revenue it helped to generate. The flower shop buys $1,000 of roses on Monday and sells them all on Friday. When did the roses truly cost it something? Not Monday — buying them only converted cash into inventory. The cost belongs on Friday, lined up against the revenue from the sale it made possible. Record the cost on Monday and Monday looks terrible while Friday looks unrealistically wonderful; match them and each day tells the truth.
Matching works in three flavours, depending on how tightly a cost can be tied to a sale. Some costs attach *directly* to specific revenue: the cost of the roses, or more generally the [[cost-of-goods-sold-line|cost of goods sold]] — these wait in inventory until the matching sale happens, then expense at that instant. Some costs benefit *many* periods and are spread across them, like depreciating a machine over its useful life. And some costs cannot be tied to any particular sale and are simply recognized as they are incurred, like the monthly office rent, which keeps the lights on for whatever sales happen to occur. The table below shows the rose example as a tiny journal sketch.
Monday (buy roses, cash -> inventory; NOT yet an expense)
Dr Inventory ............ 1,000
Cr Cash ................... 1,000
Friday (sell roses for 1,800 cash)
Dr Cash ................. 1,800
Cr Revenue ............... 1,800 <- earned now
Dr Cost of goods sold ... 1,000 <- cost matched to THIS sale
Cr Inventory ............. 1,000
Friday's profit on the roses = 1,800 - 1,000 = 800Why accrual beats cash — and where it asks for judgement
Put recognition and matching together and you see why accrual accounting is a better measure of performance than counting cash in and out. The income statement now answers a clean question — *in this period, what did we earn, and what did it cost us to earn it?* — instead of the noisy question *what days did money happen to move?* A great month of work shows up as a great month, even if customers pay late; a truck bought with a year's savings does not crater a single month's profit, because its cost is spread across the years it serves. Profit stops lurching with the accidents of billing and becomes a fair report card on the period itself.
Accrual accounting buys its truthfulness at a price worth naming plainly: it leans on estimates. To match a machine's cost across its life you must guess how many years it will last; to match the cost of products you expect customers to return, you must estimate how many will come back; to recognize revenue on a long project you must judge how far along it is. The income statement is therefore a careful, reasoned judgement, not a measured certainty — which is exactly why the line between honest estimation and manipulation is one of the things auditors are paid to police. Knowing where the judgements live is what separates a reader who trusts the bottom line blindly from one who reads it wisely.