Lesson 2 of 8
Costs meet their revenue
A cost is recognised in the period of the sale it helped earn. What that rule requires you to estimate, and what happens to costs that match no sale at all.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- A cost is charged to the period of the revenue it earned, which requires knowing which revenue that was.
- Costs that cannot be traced to a specific sale are charged to the period they relate to instead, which is most operating costs.
- The rule forces two estimates: how much of a cost belongs to the future, and when the revenue was earned at all.
Buying $30,000 of paper is not a cost. Using $26,000 of it to bind books that were sold is. The remaining $4,000 sits on the balance sheet until the books it becomes are sold too.
That is the matching idea in one example, and almost every accrual in existence is a version of it.
One purchase, split by when it was used
- Materials bought$30,000
- Used in this year's sales−$26,000
- Still on the shelf at the year end= $4,000
Three ways a cost finds its period
A sales commission is the cleanest case: it exists because a specific sale happened, so it belongs in that sale's period. A receptionist's salary cannot be attributed to any particular sale, so it is charged to the month it was worked.
What revenue recognition has to settle first
Matching cannot start until the revenue is placed. Under ASC 606 a company recognises revenue when it transfers control of a good or service to the customer, in the amount it expects to be entitled to.
That sentence contains two judgements. When control transferred, which for a shipped product is usually obvious and for a multi-year service contract is not. And how much the company expects to collect, which depends on discounts, returns and refunds that have not happened yet.
Both judgements then propagate into every cost matched against that revenue. A company that recognises revenue earlier necessarily recognises the associated costs earlier too.
The costs that match nothing
Some spending creates future value that cannot be attributed to any particular future sale. Research is the standard case: money spent on it may produce a product in four years or nothing at all.
US rules resolve this by requiring most research and development to be expensed as incurred, which is a deliberate choice of reliability over matching. The cost lands in the year it is spent, and the revenue it eventually earns lands in whatever year that turns out to be.
The result is a known distortion. A company that spends heavily on research looks less profitable than an otherwise identical one that bought the same capability through an acquisition, because the acquired version becomes an asset on the balance sheet and is amortised, while the built version was expensed. That asymmetry appears again in the accounting standards course.
Why the rule matters to a reader
Two consequences worth carrying.
Profit is an opinion about timing, and the opinion is disclosed. Every significant matching judgement a company makes appears in its accounting policies note, which is why that note is the first thing to read in an unfamiliar filing rather than the last.
And a cost that has been deferred to the balance sheet is a cost that has not gone away. Prepaid rent, capitalised software and inventory are all future expenses sitting as assets, and a growing balance in any of them is a growing charge queued up for later periods.
Check yourself
A company buys $30,000 of materials and uses $26,000. What is the cost for the year?
- $30,000, because that is what was bought
- $26,000, because that is what the year's sales consumed
- $24,000, because that is what was paid
Show the answer
Answer: B.
Matching charges the portion consumed by the revenue recognised. The unused $4,000 stays an asset until the sales it serves happen.Why are most research costs expensed as incurred rather than matched to future revenue?
- Research produces no value
- The future revenue it will earn cannot be reliably identified, so the rules choose reliability over matching
- Research is not a real cost
Show the answer
Answer: B.
Matching requires knowing which revenue a cost earned. Where that cannot be established, US rules charge the cost immediately, accepting a known distortion in exchange for a figure that is not guesswork.What does a rising prepaid expenses balance represent?
- Cash the company has recovered
- Costs already paid that will be charged against profit in later periods
- An error in the matching process
Show the answer
Answer: B.
A deferred cost is still a cost. Holding it on the balance sheet moves when it hits profit, never whether it does.Sources
- ASC 606, Revenue from Contracts with Customers, on recognising revenue when control of a good or service transfers to the customer, at the amount the entity expects to be entitled to.
- ASC 730, Research and Development, on expensing research and development costs as incurred.
What this leaves out
- The materials example uses the invented figures from the bookkeeping course, chosen so the split can be checked by hand.
- ASC 606 is summarised in a sentence here. It is a five-step model with extensive application guidance, and industries with long contracts apply it very differently from a retailer.
- Software development costs, film costs and certain contract costs follow their own capitalisation rules that differ from the research treatment described.
- Nothing here is advice about any security.