Your delivery van was supposed to last five years. At year four, the mechanic tells you it has at least four more good years in it. In the same month, your accountant suggests switching your inventory costing from LIFO to FIFO. Both feel like "changes to the books" — but under ASC 250, they land on opposite ends of the accounting universe. The van's longer life affects only this year and the years ahead; nobody touches the depreciation you already booked. The LIFO-to-FIFO switch rewrites every prior year you present, as if you had used FIFO all along. Confuse the two and your comparative statements are wrong in a way an auditor, a lender, or a buyer will notice. This guide shows how to tell them apart, what each one requires you to do, and the hybrid cases that trip up even experienced bookkeepers.
The Core Distinction: New Information vs. a New Method
ASC 250, Accounting Changes and Error Corrections, sorts every accounting change into buckets, and the bucket decides the direction of the adjustment:
- Change in accounting estimate — you learned something new about an uncertain amount. Useful lives, salvage values, bad-debt percentages, warranty rates. Estimates are, by definition, revised as new information arrives, so the change applies prospectively: in the period of change, and in future periods if they are affected too. Prior periods are never restated for an estimate change.
- Change in accounting principle — you swapped one acceptable method for another. LIFO to FIFO, completed-contract to percentage-of-completion, one revenue recognition policy to another. Because comparability across periods is the whole point of financial statements, the change applies retrospectively: prior periods presented are restated as if the new principle had always been used.
- Correction of an error — you find that prior statements were wrong under the principles you had already chosen (a math mistake, a misapplied rule, an overlooked fact that existed at the time). Errors are also fixed retrospectively, through restatement — but they are not "changes" at all, and calling one a change understates the problem. More on that trap below.
The logic is straightforward once you see it. An estimate change says "the world turned out different from what I reasonably guessed" — and you cannot fairly rewrite history for a guess that was reasonable when made. A principle change says "I am measuring the same history with a better ruler" — and a better ruler applied to only one year makes year-to-year comparison meaningless.
Changes in Estimate: The Van That Refuses to Die
Return to the delivery van. You bought it for $60,000 with a five-year life and no salvage value, booking straight-line depreciation of $1,000 a month. After 36 months you have accumulated $36,000 of depreciation and a book value of $24,000. Then comes the new information: the van will run four more years, not one.
Under ASC 250, you spread the remaining book value over the revised remaining life. Remaining book value is $24,000; revised remaining life is 48 months (the 12 you had left plus 36 new ones). New monthly depreciation is $500, starting now. The $36,000 already booked stays exactly where it is — no catch-up entry, no restatement, no apology footnote beyond disclosure of a material effect.
The same pattern covers every classic estimate:
- Raising your allowance for doubtful accounts from 2% to 4% of receivables after a rough collections quarter.
- Extending (or shortening) software amortization when a platform's viable life changes.
- Revising warranty accrual rates after actual claims run hotter than history suggested.
One subtlety matters for small businesses: when only the current period is affected, the whole effect lands in the current period. When future periods are affected too, you allocate going forward. Either way, the direction is forward, never back.
Changes in Principle: Why LIFO-to-FIFO Rewrites History
Now the inventory switch. You have costed inventory under LIFO for years, and you want FIFO — perhaps because falling input prices make FIFO income more representative, or because you are courting investors who compare you against FIFO peers. A costing method is an accounting principle, so ASC 250 requires retrospective application:
- Restate each prior period presented as if FIFO had always been used — cost of goods sold, ending inventory, income taxes, and net income all move.
- Book the cumulative effect on periods before those presented as an adjustment to the opening balance of retained earnings (or other appropriate equity) of the earliest period presented. The FASB was explicit that this catch-up never runs through current-period income: none of those effects relate to the current period.
- Disclose the nature of the change, why the new principle is preferable, the effect on each affected line item for the current and prior periods, and the cumulative effect on retained earnings.
That preferability justification deserves emphasis. You may not flip principles at will: a voluntary change must be justified as preferable, and SEC registrants go further, filing a letter from their independent auditor — the preferability letter — stating that the new principle is preferable in the circumstances. Private companies face no letter requirement, but your lender and your auditor will still ask "why now," and "it raises this year's income" is not an answer.
The famous one-way exception: changing TO LIFO
Here is the asymmetry worth memorizing. LIFO-to-FIFO restates the past, because your LIFO reserve lets you reconstruct what FIFO history would have looked like. But a change to LIFO from any other method is treated prospectively — the beginning inventory under the old method simply becomes the opening LIFO base. Reconstructing LIFO layers for years you never tracked them is considered impracticable, so ASC 250 does not ask you to. Direction matters: FIFO-to-LIFO goes forward, LIFO-to-FIFO goes back.
The Hybrids That Trip People Up
Two boundary cases cause most of the real-world confusion, and both resolve in favor of prospective treatment.
A change in depreciation method is accounted for as a change in estimate. Switching a machine from double-declining-balance to straight-line looks like a principle change — it is, technically, a different method. But ASC 250 treats it as a change in estimate effected by a change in principle: the new method reflects revised expectations about how the asset's benefits will be consumed, which is new information about an uncertain pattern. So you apply it prospectively to the current book value, exactly like the van example. The general rule behind it: when a principle change is inseparable from an estimate change, the whole thing is treated as an estimate.
Impracticability converts retrospective to "earliest practicable." When you cannot determine the period-specific effects on one or more prior periods — records are gone, assumptions cannot be reconstructed without hindsight — you apply the new principle to asset and liability balances as of the beginning of the earliest period for which it is practicable, with the offsetting adjustment to opening retained earnings of that period. This is a genuine evidentiary standard, not a convenience opt-out: "recreating the workpapers would take weeks" is not impracticable, but "the receiving logs from 2019 no longer exist" may be.
The Tax Side Is a Separate Filing
Everything above is financial reporting. Your tax return follows different rules, and a book change never, by itself, changes your tax method.
Changing how you report an item to the IRS — including switching inventory valuation methods — generally requires the Commissioner's consent via Form 3115, Application for Change in Accounting Method. Many routine changes ride an automatic-consent procedure: you attach the form to your timely filed return for the year of change (plus the required copy to the IRS office named in the instructions) instead of waiting for a private ruling. The cumulative tax catch-up is the Section 481(a) adjustment, which generally spreads the right way by statute rather than by your preference.
Two warnings specific to this topic. First, LIFO carries the conformity rule: if you use LIFO on your tax return, you generally must use it in your financial statements too — so a LIFO-to-FIFO book change often forces a coordinated tax change, each with its own paperwork. Second, the IRS does not care that your estimate change was prospective for GAAP; tax timing follows tax law (for example, depreciation follows Section 168 regardless of the useful life you book). Expect book-tax differences, track them deliberately, and never assume one filing covers both systems.
Booking It in Plain-Text Books
If you keep your books in Beancount, estimate changes are pleasantly mechanical: you just start posting the new amount. Nothing already posted moves. For the van, with a comment recording the reasoning for your future self and your auditor:
; Van placed in service 2024-01-15 at 60,000.00 USD, 5-year life.
; Original depreciation: 1,000.00 USD/month through 2026-12 (36,000.00 USD accumulated).
; 2027-01 revision: 4 more years of service expected. Remaining book value
; 24,000.00 USD over 48 months = 500.00 USD/month, applied prospectively.
2027-01-31 * "Van depreciation - revised estimate"
Expenses:Depreciation:Equipment 500.00 USD
Assets:Equipment:Van:Accumulated-Depreciation -500.00 USDA principle change is where plain-text accounting shows its character. Beancount has no "restate" button, and that is a feature: your ledger is an append-only history, so the years as originally reported stay visible forever. Retrospective application happens in your comparative reports — you regenerate the prior-year statements under the new principle and present them alongside the current year, with the opening-equity catch-up disclosed exactly as ASC 250 requires. Keep the restated comparative reports next to the as-filed originals, and document the principle change, the preferability reasoning, and the retained-earnings adjustment in the same commit or memo so the audit trail is one place, not three. If you are new to this style of bookkeeping, the documentation walks through ledger structure and reporting from first principles.
Mistakes That Draw Auditor Comments
Restating for an estimate change. The single most common error: recomputing three years of depreciation at the new life and booking a catch-up. It feels thorough; it is wrong. Estimates go forward.
Running a principle change's catch-up through current income. Booking the cumulative effect as this year's gain or loss inflates (or craters) the current period with amounts that belong to history. It goes to opening retained earnings of the earliest period presented.
Labeling an error correction as a change. Discovering you expensed a capital asset for two years is not a "change in capitalization policy" — it is an error, and errors get restatement plus prior-period-adjustment disclosure, not preferability language. Auditors treat mislabeling as a control signal, and not a good one.
Flipping principles without a preferability story. Voluntary principle changes need a written "why this is better" rationale contemporaneous with the change. Write the memo when you decide, while the business reasons are fresh — reconstructing preferability during audit fieldwork is how defensible changes start looking opportunistic.
Forgetting the tax return. The book entries post, the comparative statements get restated, and nobody files Form 3115. Months later the IRS notice arrives treating the return as filed on the old method. Put the tax filing on the same checklist as the book entries.
Your Decision Checklist for the Next Change
The next time something in your books needs to move, run it through this order: Is the prior treatment wrong under rules you had already chosen? That is an error — restate and disclose it as one. Is it new information about an uncertain amount? That is an estimate — adjust prospectively from the remaining balance. Is it a different acceptable method for the same history? That is a principle — restate every period you present, catch up through opening retained earnings, justify preferability in writing, and file the tax paperwork as its own workstream. Three questions, asked in that order, get the direction right every time.
