Full explanation
Overview
A thousand identical units in one warehouse. Two completely different profits — from nothing but a policy choice.
Inventory looks like the simplest thing on a balance sheet. Count the boxes, add up what you paid. But IAS 2 is where two identical companies can report profits two thousand four hundred euros apart, and both be completely correct.
Assets you hold in order to sell
Inventories are assets held for sale in the ordinary course of business, assets in the process of production for that sale, or materials and supplies to be consumed in production. Finished goods, work in progress, raw materials. And notice what defines them: not what they are, but that you intend to SELL them. A building is inventory to a property developer and property, plant and equipment to everyone else.
The boundaries first
Before the mechanics, the boundaries — because IAS 2 hands several things to other standards. Work in progress on construction contracts goes to IFRS 15. Financial instruments go to IFRS 9. Biological assets and agricultural produce at the point of harvest go to IAS 41 — and then, at that fair value, become IAS 2 inventory. Spare parts are inventory unless they meet the definition of property, plant and equipment under IAS 16. And two groups measure at fair value less costs to sell instead of the normal rule: producers of agricultural and mineral products, and commodity broker-traders.
Two of those cost rules do real work against manipulation. Here's the first.
What you may capitalise
What goes into cost? Three buckets. Purchase cost: the price, plus import duties and non-recoverable taxes, plus transport and handling — less trade discounts and rebates. Conversion cost: direct labour, plus production overheads. And other costs, but only those incurred in bringing the inventory to its present location and condition. Then the exclusions, which matter just as much. Abnormal waste. Storage, unless it's a necessary part of production. Administrative overheads. Selling costs. None of those may sit in inventory — they go straight to expense. Borrowing costs can only be capitalised where IAS 23 lets them.
Absorb on normal capacity , not actual
And one rule inside conversion cost does real work against manipulation. Fixed production overhead is absorbed into units based on NORMAL capacity — not on what you actually produced. So if a factory that normally makes ten thousand units only makes five thousand this year, it cannot push the whole year's fixed overhead into those five thousand units and park it on the balance sheet. The unabsorbed half is expensed immediately. Without that rule, a company could hide the cost of idle capacity inside inventory. IAS 2 also permits standard cost and the retail method — but only where the result approximates actual cost.
So that's cost. But cost is only half the rule — and the other half is what stops inventory being overstated.
The lower of cost and NRV
Inventory is measured at the LOWER of cost and net realisable value. Net realisable value is the estimated selling price in the ordinary course of business, less the estimated costs of completion and the estimated costs necessary to make the sale. Note that it's an entity-specific figure — it is not fair value, and the standard says so explicitly. So the test is simple: what did it cost you, what will you actually net from selling it, and you carry the lower of the two.
Compare item by item
Two things about that write-down. First, you compare item by item, not across the whole warehouse — you cannot offset a profitable line against a loss-making one. Second, the write-down is not permanent. If the circumstances that caused it change and net realisable value recovers, IAS 2 requires you to REVERSE the write-down — capped at the original cost, so you can never write it back above what you paid. That reversal is a requirement, not a choice. And note that IAS 36 has nothing to do with any of this — impairment of inventory lives entirely inside IAS 2, through the net realisable value test.
That's the rule. Now where it actually bites, industry by industry.
Prudence becomes a number
So where does this actually bite? Net realisable value is where prudence stops being a principle and becomes a number, and it lands differently depending on what you sell. Technology and electronics: obsolescence is the killer. The moment a new chip generation launches, last year's model cannot be sold at list price, and net realisable value collapses while the goods are still physically perfect. Fashion and apparel: it is the calendar, not the condition. Winter coats sitting in a warehouse in March will be sold at markdown, so net realisable value is the expected markdown price less the cost of selling it — and the judgement is how deep that discount has to go. Food and pharmaceuticals: a shelf life is a clock. Stock approaching expiry may have no net realisable value at all, and if you have to pay to destroy it, the deduction for costs necessary to make the sale can wipe out the entire carrying amount. And slow-moving spare parts: nothing is wrong with them, they simply are not selling, so companies provide by age bucket rather than item by item.
Raw materials are not written down automatically
But here is the trap that catches people in every one of those industries. If you hold raw materials and their market price falls, you do NOT automatically write them down. Paragraph thirty-two is explicit: materials held for use in production are not written down below cost if the finished products they go into are still expected to sell at or above cost. So a manufacturer whose copper drops in value keeps the copper at cost, provided the finished cable still sells profitably. You only write the materials down when the fall in their price signals that the finished product itself will now sell below cost. Take Meridian: if its raw material cost eight thousand euros and the market for that material halves, but the finished unit still sells at a profit, the eight thousand stays. Prudence is not the same as pessimism — IAS 2 asks for a judgement about the finished product, not a reflex about the input.
Now the part that actually moves reported profit — and it's just a choice of formula.
The numbers, before the formulas
Let's actually calculate it, because this is where the formulas stop being names and start being numbers. Meridian Limited buys a thousand units in two lots. Four hundred units at fifty euros — twenty thousand. Then six hundred units at seventy euros — forty-two thousand. So a thousand units on hand, costing sixty-two thousand euros in total. It then sells seven hundred of them at eighty euros each, so revenue is fifty-six thousand. The question is which seven hundred units you just sold — and that is the only thing the formulas disagree about.
Same 700 units. Two methods.
First-in first-out assumes you sold the oldest stock first, so you work down the layers. The first four hundred units came in at fifty euros — that is twenty thousand. You still need three hundred more, and the next layer is at seventy — three hundred times seventy is twenty-one thousand. Add them: cost of sales is forty-one thousand. What is left on the shelf is the three hundred newest units, all at seventy, so closing inventory is twenty-one thousand. Notice the two add back to sixty-two thousand — every euro you paid is either in cost of sales or still on the balance sheet, never lost and never invented. Weighted average does something completely different. It refuses to track layers at all. You take the total cost, sixty-two thousand, and divide it by the total units, a thousand, giving one blended rate of sixty-two euros a unit. Then every unit is that price. Seven hundred sold times sixty-two is forty-three thousand four hundred of cost of sales, and the three hundred left are eighteen thousand six hundred. Those also add back to sixty-two thousand.
So those are the numbers. Here's what they do to the accounts.
A €2,400 swing from one assumption
So put the two side by side on the same seven hundred units. First-in first-out gives cost of sales of forty-one thousand, so gross profit is fifteen thousand. Weighted average gives forty-three thousand four hundred, so gross profit is twelve thousand six hundred. A difference of two thousand four hundred euros in reported profit, from nothing but the assumption about which units left the warehouse. And look at what it does to the balance sheet: first-in first-out leaves twenty-one thousand of inventory, weighted average leaves eighteen thousand six hundred. In a rising market first-in first-out always reports the higher profit and the higher inventory, because it charges the old cheap costs to profit and leaves the new expensive ones on the balance sheet. Weighted average smooths it. Neither is wrong. Both are permitted. They simply tell you a different story about the same warehouse.
Which formulas IAS 2 allows
So which formula may you use? IAS 2 permits first-in first-out and weighted average cost. It also permits specific identification, but only for items that are not ordinarily interchangeable. And it prohibits last-in first-out outright — it has since the 2003 revision. The reasoning is straightforward: last-in first-out rarely reflects the actual physical flow of goods, and in a rising market it leaves the oldest, cheapest costs sitting on the balance sheet, understating what the inventory is really worth. Whichever formula you choose, you must apply it consistently to all inventories of a similar nature and use — you cannot switch between them line by line to flatter a result.
And that write-down we talked about? Which formula you chose changes how big it is.
A write-down three times larger
Now watch what happens when net realisable value bites. Meridian has three hundred units left, and net realisable value falls to fifty-eight euros — seventeen thousand four hundred. Under first-in first-out those units are carried at seventy each, twenty-one thousand, so it writes them down and takes a three thousand six hundred euro loss. Under weighted average they're carried at sixty-two — eighteen thousand six hundred — so the write-down is only one thousand two hundred. Identical physical stock, identical market conditions, and a write-down three times larger under one policy than the other. But here is the part almost nobody points out. Take the write-down off each gross profit and both companies land on exactly eleven thousand four hundred. They converge, and they must — because closing inventory now sits at net realisable value under either formula, so the total cost charged is the same sixty-two thousand less seventeen thousand four hundred whichever route you took. The cost formula moves the SHAPE of your income statement, not the total, in the year net realisable value binds. It moved the total in the good years, before the write-down. That is the honest version of the story.
Where a reader can check all of it
When inventory is sold, its carrying amount becomes an expense in the period the revenue is recognised — that's the matching that cost of sales exists to achieve. And the disclosures are where a reader can actually check all of this: the accounting policy including the cost formula used, the carrying amount by classification, the amount recognised as an expense in the period, and — importantly — the amount of any write-down AND any reversal, with the reasons for the reversal. That reversal disclosure exists precisely because it's a judgement that flatters profit.
One more change is coming, and it affects how all of this appears on the page.
Which selling costs come off NRV?
Now the most recent thing you need to know, because it changed how net realisable value is calculated in practice. In June twenty twenty-one the IFRS Interpretations Committee published an agenda decision on the costs necessary to sell inventories. The question was simple: when you work out net realisable value, do you deduct only the extra costs caused by the sale, or every cost needed to make it? Many companies had been deducting only the incremental ones. The Committee concluded that you deduct ALL costs necessary to make the sale — and that this is not limited to incremental costs. So fixed selling overheads, storage that is needed to get the goods sold, an allocated share of a distribution facility: if the cost is necessary to make the sale, it comes off. That lowers net realisable value, which means more inventory fails the lower-of test and more write-downs get recognised. Note also what IAS 2 itself has NOT done: apart from consequential changes, it has not been substantively amended since the two thousand and three revision. The standard is stable — the movement is in how it is applied.
Buy it · value it · test it · sell it
And IAS 2 completes the asset family. Something physical you use in the business — IAS 16. No physical substance — IAS 38. Held to earn rent or for capital growth — IAS 40. Held to sell in the ordinary course — IAS 2, here. And when values fall, IAS 36 tests all of them — except these. Inventory is the one asset IAS 36 explicitly does not touch, because IAS 2's net realisable value rule already does that job. Buy it, value it, test it, sell it.
IFRS 18 changes where it appears
From January 2027, IFRS 18 replaces IAS 1 and changes where all of this is presented. Cost of sales lands in the operating category, and so does any net realisable value write-down or reversal. IFRS 18 also requires operating expenses to be presented by nature or by function — which directly changes how the cost of inventories appears on the face of the statement. Not one number in IAS 2 changes. The statement it lands on does, and because IFRS 18 applies retrospectively, 2026 comparatives are already being restated.
That's the mechanics complete. Now what it means if you're reading the accounts.
Read the policy note first
What should an investor take from this? Two things. First, read the cost formula in the accounting policy note before you compare gross margins between two companies. Two businesses holding identical stock can report different cost of sales purely because one uses first-in first-out and the other weighted average — and in a rising market that gap widens. Second, watch write-down reversals. A reversal increases profit without a single extra sale, and IAS 2 requires the reasons to be disclosed precisely so you can judge whether the recovery is real or convenient.
Every figure traceable to a paragraph
Meridian on the actual statements, under first-in first-out. Revenue, fifty-six thousand. Cost of sales, forty-one thousand. Gross profit, fifteen thousand. Then the net realisable value write-down of three thousand six hundred, giving a gross profit of eleven thousand four hundred. And on the balance sheet, inventories of seventeen thousand four hundred — three hundred units at the lower of cost and net realisable value. Every figure traceable to a paragraph.