The Disciplines of M&A
The acquisition method, purchase price allocation, the goodwill debate, and the places where IFRS 3 and ASC 805 produce genuinely different numbers.
Every business combination reaching a US GAAP or IFRS balance sheet today is accounted for the same way: the acquirer identifies itself, measures consideration transferred at fair value, remeasures the target's identifiable assets and liabilities to fair value, and books the residual as goodwill. This is the acquisition method (US GAAP: ASC 805; IFRS: IFRS 3), and its universality is itself the product of a deliberate rule change, not a historical constant.
Until the early 2000s, a second method existed alongside purchase accounting: pooling of interests. Under pooling, two companies combined at book value. No fair-value step-up, no goodwill, no new basis of accounting at all — the combined entity simply added the two sets of book values together and restated prior-period financials retroactively as if the companies had always been one. Pooling was available only for qualifying stock-for-stock combinations that met a long list of criteria (no material cash consideration, no buybacks or unusual arrangements around the deal, continuity of ownership interests, and more).
That optionality was the problem. Two economically identical acquisitions — one structured to qualify for pooling, one not — produced materially different financial statements. A pooled deal generated no goodwill and no amortization drag, so an acquirer could pay a large premium in stock and report as if nothing had changed. That created a structuring incentive: the stock-financed roll-up waves of the late 1980s and 1990s leaned on pooling specifically because it let acquirers avoid ever showing the cost of what they had paid. The FASB eliminated pooling in 2001 with SFAS 141 (FASB, 2001); the IASB followed with IFRS 3 in 2004 (IASB, 2004).
The abolition changed deal behavior: it made the accounting consequence of overpaying visible regardless of how a deal was financed. Once every deal generates fair-value step-ups and goodwill, the choice between cash and stock reverts to being a financing, control-dilution and tax question rather than an accounting-optics one. It also had a second-order effect — with goodwill under the new purchase method no longer amortized (see below), the dilution math boards and bankers ran for stock deals shifted from a pooling-vs-purchase framing to the accretion/dilution analysis practiced today, built around intangible amortization, interest cost, and share-count dilution rather than pooling eligibility.
Once acquisition accounting applies, the mechanical exercise is purchase price allocation (PPA): total consideration transferred — cash, stock, assumed debt, and the acquisition-date fair value of any contingent consideration — is allocated across the target's identifiable assets and liabilities at fair value, with anything left over booked as goodwill.
Where the purchase price goes: allocation and the amortisation drag
The commercially important part is what "identifiable assets" captures that never sat on the seller's balance sheet. A target that built its franchise organically typically expensed the cost of building it — sales and marketing spend that created customer relationships, R&D that created technology, brand spend that created a trade name — so those assets carry no cost basis at all pre-deal. PPA recognizes them for the first time, at fair value, as separately identifiable intangible assets: customer relationships, developed technology, trade names and trademarks, and in-process research and development (IPR&D — an indefinite-lived intangible until the underlying project is completed or abandoned, at which point it either starts amortizing or is written off). Only the residual — value the acquirer paid for that cannot be attributed to any identifiable asset, typically synergies, assembled workforce, and the premium for control — is goodwill.
This matters commercially because most of what PPA newly recognizes has a finite useful life and amortizes through the P&L for years after close, while goodwill does not (it is tested for impairment instead — see below). Illustrative example: an acquirer pays $500 million for a target with $50 million of net tangible assets. Diligence identifies $150 million of intangibles — $90 million of customer relationships amortized over 10 years, $60 million of developed technology amortized over 5 years — leaving $300 million of goodwill. That allocation alone creates roughly $21 million of annual non-cash amortization expense in the first five years, none of which reflects a change in the business's cash economics. This is precisely why acquirers report adjusted EBITDA and non-GAAP EPS that add back acquisition-related intangible amortization: statutory earnings are mechanically depressed relative to the cash the business generates, for as long as the schedule runs. A dealmaker should always ask what fraction of the intangible base is amortizing over what schedule — short-lived developed technology front-loads the drag; long-lived customer relationships or indefinite-lived trade names spread it thinner or remove it altogether.
Both regimes have used impairment-only accounting for goodwill since the purchase-method reforms of 2001 (US) and 2004 (IFRS): goodwill is not amortized, it is tested at least annually (and whenever a triggering event occurs) and written down if its carrying value exceeds its recoverable or fair value. The mechanics of that test differ between the two regimes, and both boards have been actively reconsidering the model — the state as of 2026 is not settled and dealmakers should track it.
Under US GAAP, the test originally had two steps: compare the fair value of a reporting unit (an operating segment or one level below) to its carrying value, and if it was lower, perform a hypothetical purchase price allocation to measure the implied impairment. The FASB eliminated step two in 2017 (ASU 2017-04), leaving a single-step test: impairment equals the excess of a reporting unit's carrying value over its fair value, capped at the goodwill balance. Under IFRS, the test has always been single-step but structured differently: goodwill is tested at the cash-generating unit (CGU) level — the smallest group of assets generating largely independent cash inflows, which is frequently a finer unit of measurement than a US reporting unit — and impairment is triggered when the CGU's carrying amount (including allocated goodwill) exceeds its recoverable amount, defined as the higher of fair value less costs of disposal and value in use.
The amortization question has resurfaced repeatedly and remains unresolved on both sides as of 2026. The FASB studied returning to an amortization-with-impairment model for several years and voted in 2022 to abandon that project, retaining impairment-only (FASB, 2022). It did not stay closed: in February 2026 the FASB revisited goodwill accounting, directing staff to research a range of options — amortization for all entities, expanded voluntary amortization, voluntary write-offs, and targeted changes to impairment testing — without adding a formal project (FASB, 2026). That research led to a vote at the end of July 2026 to pursue "targeted improvements" to the impairment test itself rather than amortization: moving the level of testing from the reporting unit up to the operating segment (a coarser unit that nets stronger- and weaker-performing businesses together, reducing the frequency of triggers) and eliminating the requirement to test annually absent a triggering event (FASB, 2026). Neither change has been finalized into an ASU; cost-benefit analysis is ongoing. The IASB voted in November 2022 to retain the impairment-only model (IASB, 2022), then issued an exposure draft in March 2024, Business Combinations — Disclosures, Goodwill and Impairment, proposing enhanced disclosure of an acquisition's strategic rationale and the targets management uses to monitor it, plus a simplification removing the requirement to use pre-tax cash flows and a pre-tax discount rate in the value-in-use calculation (IASB, 2024). That project remained in redeliberation through late 2025, with disclosure scope and auditability concerns unresolved (IASB, 2025); no amended standard had been issued as of 2026. The takeaway for a dealmaker: impairment-only holds on both sides for now, but the testing mechanics — especially on the US side — are actively moving, and a change to the unit of testing would materially affect how often and how severely write-downs occur.
Both IFRS 3 and ASC 805 treat contingent consideration — earn-outs, milestone payments, clawbacks — as part of the consideration transferred, measured at fair value as of the acquisition date and folded into the PPA (and therefore into the goodwill calculation) at signing. What happens next depends entirely on how the obligation is classified.
If the contingent payment is classified as a liability — which covers the overwhelming majority of earn-outs, since most are cash-settled or have a cash-settlement feature — it is remeasured to fair value at every subsequent reporting date, with the change running through the P&L, under both regimes. If it is instead classified as equity (a fixed number of shares with no cash alternative and no other feature disqualifying fixed-for-fixed treatment under IAS 32 or ASC 815-40), it is not remeasured after initial recognition.
The liability treatment produces a result that surprises people encountering it for the first time: when the acquired business performs well against its earn-out targets, the expected payout rises, the fair value of the liability increases, and the acquirer records a charge running through earnings — precisely when the deal is succeeding. An underperforming acquisition, by contrast, can produce a P&L gain as the earn-out liability is written down. Any acquirer disclosing "earn-out remeasurement" as a line item in its non-GAAP or adjusted-earnings reconciliation is, in effect, telling the market it wants credit for the deal working without taking the accounting hit that comes with it working.
The two standards are closely converged in structure — both use the acquisition method, both allocate to fair value, both cap the measurement period at one year from the acquisition date for finalizing provisional estimates — but several differences flow directly into reported numbers.
Non-controlling interests. Where an acquirer takes less than 100% control, IFRS 3 permits a choice, made separately for each acquisition, between the full goodwill method (NCI measured at fair value, so goodwill is grossed up for the NCI's implied share) and the partial goodwill method (NCI measured at its proportionate share of identifiable net assets, with goodwill recognized only for the acquirer's share). ASC 805 mandates the full goodwill method only — no election. Two otherwise identical partial acquisitions can therefore report different goodwill and different NCI balances depending on the reporting regime and, under IFRS, the policy choice made.
Contingent liabilities. IFRS 3 recognizes a contingent liability assumed in a business combination if its fair value can be measured reliably, regardless of probability — a lower bar than IAS 37 applies outside a business combination. ASC 805 splits contractual and non-contractual contingencies: contractual ones are generally recognized at fair value at the acquisition date, while non-contractual ones are recognized only if more likely than not to meet the definition of an asset or liability. The practical effect: IFRS acquirers tend to book contingent liabilities more readily than US GAAP counterparts with an economically similar exposure.
Definition of a business. Both boards narrowed this definition in recent years, for the same reason: too many economically asset purchases were being swept into business-combination accounting, triggering goodwill recognition, deferred tax on step-ups, and mandatory expensing of transaction costs where an asset acquisition would have simply capitalized them. The FASB's ASU 2017-01 introduced a "screen" test — if substantially all of the fair value of the gross assets acquired is concentrated in a single asset or group of similar assets, the transaction is not a business (FASB, 2017). The IASB followed with a similar concentration test amendment to IFRS 3, applicable to business combinations with an acquisition date on or after 1 January 2020 (IASB, 2018). A deal failing the business definition on either side is an asset acquisition — no goodwill, consideration allocated on a relative-fair-value basis, transaction costs capitalized rather than expensed.
Common-control transactions. IFRS 3 explicitly excludes combinations between entities under common control from its scope, and — unlike almost every other corner of business combination accounting — no IFRS standard fills that gap. The IASB studied the question for years and in 2023 decided not to develop new requirements, concluding that the diversity of practice was tolerable to investors and that a standard's costs would likely exceed its benefits (IASB, 2024). IFRS preparers restructuring within a group therefore choose between acquisition accounting (fresh fair values) and predecessor or carryover accounting (book values, often applied retrospectively), based on local convention and which answer the transaction is meant to produce. US GAAP has explicit guidance instead: ASC 805-50 requires common-control transfers to be accounted for at the transferor's historical cost, with financial statements retrospectively combined as if the transfer had occurred at the start of the earliest period presented — closer in spirit to old pooling accounting, but confined to the common-control fact pattern.
Transaction costs. Both regimes expense acquisition-related transaction costs — advisory, legal, diligence fees — as incurred, rather than capitalizing them into the purchase price or goodwill. This itself was a change from the pre-2001/2004 purchase method, which had capitalized these costs. The one carve-out on both sides: costs to issue debt or equity securities used to finance or fund the deal are not expensed. Debt issuance costs are capitalized against the related liability and amortized over its life; equity issuance costs (underwriting discounts and similar fees on shares issued as consideration or to raise acquisition financing) reduce additional paid-in capital rather than hitting the P&L.
Acquired deferred revenue. Historically, both regimes required an acquirer to remeasure assumed deferred revenue to fair value at the acquisition date — not face value, but the cost to fulfill the remaining performance obligations plus a reasonable margin on that remaining effort. Because the seller had already earned most of its margin acquiring and onboarding the customer, this exercise routinely produced a haircut well below the deferred revenue on the target's own books, mechanically understating the acquirer's reported revenue from that customer base in the first year or two post-close — a recurring source of investor confusion around subscription-business deals. The FASB changed this for US GAAP: ASU 2021-08, issued in 2021, requires an acquirer to recognize and measure acquired contract assets and liabilities as if it had originated the contracts itself under ASC 606, rather than at fair value — a carryover basis that eliminates the haircut (FASB, 2021). IFRS has made no equivalent change; IFRS 3 still requires fair-value measurement of assumed contract liabilities. A live, easy-to-miss cross-standard question follows: an IFRS acquirer's post-close revenue on an acquired subscription book will typically still show the haircut that a US GAAP acquirer, having adopted ASU 2021-08, no longer reports.
Deferred tax on step-ups. Where a deal does not carry over the target's tax basis (common in stock deals and non-taxable structures), the fair-value step-up to intangibles and other assets creates a book-tax basis difference with no corresponding cash tax benefit, requiring a deferred tax liability at the acquisition date under both IAS 12 and ASC 740. That DTL itself increases the amount allocated away from net identifiable assets and pushes more of the residual into goodwill — an acquirer effectively grosses up goodwill by roughly the DTL recognized on the step-ups. Both regimes carve out an "initial recognition exception" preventing this mechanism from applying circularly to goodwill itself.
Replacement share-based payment awards. When an acquirer exchanges a target's outstanding options or restricted stock for its own awards, both IFRS 2 and the combined ASC 718/805 framework require splitting the fair value of the replacement awards between consideration transferred (service already rendered before the deal) and post-combination compensation expense (remaining future service), following the ratio of pre-combination to total requisite service. The result is ongoing stock compensation expense after close that was economically part of what the acquirer paid but never shows up in the purchase price or goodwill.
A quality-of-earnings (QoE) study is a diligence exercise, typically run by an accounting advisory firm engaged directly by the buyer (or, in a sell-side QoE, prepared in advance by the seller's advisors to pre-empt buyer questions). It normalizes historical EBITDA or earnings for non-recurring, non-operating, and one-time items; tests the quality and durability of revenue recognition, customer concentration and churn; and analyzes net working capital trends to establish the peg used in the purchase agreement.
It is not an audit, and the distinction matters commercially, not just technically. An audit expresses an opinion on whether a full set of financial statements is fairly stated under GAAP or IFRS, based on procedures designed to provide reasonable assurance across the statements as a whole, and produces an opinion letter third parties can rely on. A QoE engagement is scoped by the buyer to answer specific deal questions, does not test every transaction, does not opine on GAAP or IFRS compliance, and produces a report for the engaging party rather than a public assurance opinion. It routinely surfaces a "normalized" or "run-rate" EBITDA that differs — sometimes materially — from management's deck, because it strips out add-backs the buyer's accountants judge illegitimate. That normalized number, not the audited historical figure, is usually what the purchase price multiple gets applied to — which is why a QoE finding can move price in the final weeks of a deal even when the audited financials never change.