IAS 2: Why IFRS Banned LIFO (And US GAAP Didn't)

Why the oldest surviving standard tells us the LIFO debate was never really about accounting.

The Oldest Layer

When IFRS 18 takes effect for annual periods beginning on or after 1 January 2027, IAS 1 will be withdrawn. Its going concern requirements will move to IAS 8, and the number "IAS 1" will disappear from the active literature. From that point on, the lowest-numbered standard still in force will be IAS 2 — Inventories.

IAS 2 was issued in 1975, in the earliest wave of standards produced by the International Accounting Standards Committee. It has been revised, most substantially in 2003, but it remains the oldest surviving layer of the IFRS architecture.

Standards this old are worth reading carefully, because their shape is rarely the product of a single design decision. It is the product of history — and in the case of IAS 2, the decisive chapters of that history were written not by accountants, but by tax law.

This first installment starts from the mechanics — what the cost formulas actually are, and what they do to the numbers — and then traces how IAS 2 arrived at its current form. The destination is the most famous divergence between IFRS and US GAAP: the treatment of LIFO, which turns out to be a story about tax systems, not measurement theory.

Three Formulas, Defined

A cost formula answers one question: when identical units were purchased at different prices, which cost goes to the income statement when a unit is sold, and which cost stays on the balance sheet?

FIFO — first-in, first-out. The units purchased earliest are deemed sold first. Cost of goods sold carries the oldest costs; closing inventory carries the newest costs. Note the word deemed: FIFO is a cost-flow assumption, not a claim about which physical box left the warehouse.

Weighted average. Each sale is costed at the average cost of all units available, weighted by quantity. In a periodic system the average is computed once for the period; in a perpetual system it is recomputed at each purchase (the moving average familiar from every ERP). Both income statement and balance sheet carry a blend of old and new costs.

LIFO — last-in, first-out. The units purchased most recently are deemed sold first. Cost of goods sold carries the newest costs; closing inventory carries the oldest costs — and if inventory levels are maintained, those old costs can sit on the balance sheet for decades, in layers dating back to the year LIFO was first adopted.

IAS 2 today permits the first two (plus specific identification for items that are not interchangeable). It prohibits the third. Why it does so is the subject of this article — but the argument only makes sense once you see the numbers move.

The Numbers, Side by Side

Consider a company in a rising-price environment:

• Opening inventory: 100 units at 10

• Purchases during the year: 100 units at 14

• Sales during the year: 100 units at 20

Total cost available is 2,400 for 200 units. Each formula splits that 2,400 differently between the income statement and the balance sheet:

Same physical business, same cash flows. The figure in the middle of the diagram never changes: 2,400 of cost, waiting to be divided. The formula decides only where it goes — and therefore what two very different audiences get to see.

The first audience is investors, who read gross profit. The second is the tax authority, which reads the same number and sends a bill. The table's last two rows are the same choice viewed by the two audiences: report 1,000 of profit and pay 250, or report 600 and pay 150.

A cost formula, in other words, is not a computation technique. It is a decision about how you want your profit to appear — to the people who value your company, and to the people who tax it. The catch, of course, is that in some jurisdictions both audiences must be shown the same number.

Hold that thought, because the entire history of IAS 2 — and the survival of LIFO in the United States — is the story of that catch..

A Note on Scale

Our example has one product, two purchase lots, and one sale. A real inventory has tens of thousands of SKUs, each with thousands of receipts and issues a year, spread across plants, warehouses, and valuation areas. At that scale, nobody applies a cost formula; a system does.

This changes what the choice actually is. Selecting FIFO or weighted average is not a policy statement to be applied by an accountant at year-end — it is a configuration decision in the ERP's inventory valuation logic, made once, at implementation, and thereafter executed automatically on every goods movement. The moving average recalculated at each receipt, the FIFO layers consumed in sequence, the standard costs with their variance accounts: these are the forms in which paragraph 25 of IAS 2 exists in the real world.

It is worth stating plainly, because it inverts the usual mental picture: the cost formula is not something applied to the transaction data after the fact. It is the rule by which the transaction data is created. By the time the financial statements are prepared, the choice has already been executed millions of times.

Three Dates: 1975, 1993, 2003

The original IAS 2 of 1975 permitted FIFO, weighted average, and LIFO alike. This was not a compromise but a method: the IASC of the 1970s worked by cataloguing the treatments in use across major jurisdictions and writing standards broad enough to accommodate them. A standard built this way answers "what do companies do?" rather than "what should the number mean?"

The 1993 revision, part of the Comparability project, demoted LIFO to an allowed alternative — a treatment the standard tolerated but no longer endorsed, and a status everyone understood as transitional.

The 2003 revision, part of the Improvements Project undertaken ahead of the EU's 2005 mandate, abolished LIFO outright. The Basis for Conclusions reasons that LIFO rarely reflects the actual flow of goods, and in periods of changing prices leaves the balance sheet carrying amounts unrelated to recent cost levels — precisely the stranded 10 in our example.

Catalogue, hierarchy, rule: three dates, one direction. Viewed from inside accounting theory, the final step was almost unremarkable. Which raises the real question: if the theoretical case was this clear, why does LIFO survive in US GAAP to this day?

The Answer Is in the Tax Code

The persistence of LIFO in the United States has one load-bearing pillar: the LIFO conformity rule in the US Internal Revenue Code. A company that elects LIFO for tax purposes must also use LIFO in its financial reports. The deferred 100 in our example is available only to a company willing to show the 600 gross profit to its investors.

The conformity rule welds the tax election to the financial reporting choice. A US company cannot keep the tax deferral while presenting FIFO-based earnings. So when the IASB banned LIFO, it was not asking US companies to change an accounting policy; it was asking them to surrender a tax position. That is not a debate accounting standard-setters can win, and the FASB — whatever its members' views on measurement theory — was never in a position to converge. The divergence between IAS 2 and ASC 330 is often presented as a disagreement about inventory accounting. It is more accurately described as a place where a tax statute blocked the road.

Three Ways Tax Law Shapes Accounting Standards

Once you see the mechanism, you start to see it everywhere — and it is worth being precise about the fact that the mechanism operates differently in different jurisdictions. LIFO itself provides a clean natural experiment, because LIFO exists in tax law well beyond the United States.

The rigid-coupling model. In the United States, tax and financial reporting are welded together for this specific item by the conformity rule. The coupling runs from tax to accounting: to keep the tax treatment, the company must mirror it in the financial statements. Accounting standards lose the freedom to move.

The dependency model. In jurisdictions where taxable income is computed from the statutory financial statements — book-tax conformity in the broad sense — the coupling runs the other way. Accounting standards can change, but every change flows through to tax outcomes, so preparers, and often the standard-setters themselves, move cautiously. The result is not prohibition but friction: reforms arrive late, exceptions persist, and transition periods stretch.

The two-layer model. Germany and Italy show a third configuration. Both permit LIFO in domestic statutory accounts, and both accommodate it in tax computation. Yet neither has any quarrel with the IFRS ban — because consolidated IFRS reporting and individual statutory-plus-tax reporting are separate layers. A German group can carry LIFO in its HGB individual accounts and its tax filings while its consolidated IFRS statements use weighted average. The tax position and the IFRS position never have to be the same number, so they never collide.

The comparison yields a compact conclusion. The presence of LIFO in a country's tax law tells you almost nothing. What matters is the coupling structure — whether, and in which direction, tax and financial reporting are bolted together. The United States could not follow the IASB because its coupling is rigid and item-specific. Two-layer jurisdictions followed without difficulty because their coupling is severed at the consolidation boundary.

This is also, we would suggest, a generally useful reading habit: when two accounting frameworks diverge and the divergence refuses to close despite decades of convergence effort, look for the tax statute standing behind one of them.

The Verdict

1975 produced a catalogue. 1993 produced a hierarchy. 2003 produced a rule. The standard did not change because inventory changed; it changed because the institution writing it changed — from a committee documenting consensus into a board asserting a measurement philosophy, with a continent's capital markets about to depend on the result.

LIFO is still taught, and still memorized, as the famous inventory difference between IFRS and US GAAP. After tracing the history, we can name what feels wrong about that framing: it is not an accounting difference at all. As a representation of how goods flow, LIFO lost the argument decades ago — the history above is simply the record of that verdict being tested and confirmed, revision by revision. What survives in US GAAP is not a competing view of measurement but a tax election, preserved by a statute that requires an accounting method as its price. File it not under "inventory measurement" but under "what tax law leaves behind in accounting" — and a memorized difference becomes a story you can explain in one sentence.

In Vol. 2, we turn from history to structure: how IAS 2 organizes itself, why measurement occupies nearly the entire standard while recognition takes a single paragraph, and how the net realisable value rules that look so tidy on paper are actually operated on the shop floor and in the ERP.

Premium Slide Deck Available

The key concepts from this article have been turned into a premium, fully editable white-label PowerPoint template, ready for corporate training, CPD, consulting, and client presentations.

https://ifrslabo.gumroad.com/l/IAS2_Inventories_WTD

Practice Questions

Practice questions based on this article are available on IFRS-OneQ. Reinforce what you've read — one question at a time.

Disclaimer

This article is for general educational purposes only and does not constitute accounting, tax, or legal advice. Accounting standards and tax laws are subject to amendment, and references to specific jurisdictions are simplified for illustration. The numerical examples are deliberately simplified. For decisions affecting financial reporting or tax positions, please consult a qualified professional.