← Quantery Blog

How to read goodwill and impairment charges

October 12, 2026 · 8 min readfundamentalsgoodwillacquisitions

Goodwill is the part of an acquisition price that can't be assigned to identifiable assets after liabilities are counted. It records the premium paid for a business's expected customer relationships, assembled workforce, market position, and future earning power. An impairment charge says some of that recorded premium is no longer supported by the company's estimate of fair value.

The charge cuts goodwill, operating income, net income, and equity. It doesn't send cash out when recorded. The cash left when the acquisition closed. That timing is why impairment can look harmless in a cash-flow screen and still carry useful information about management's acquisition record.

This starts from the balance-sheet distinction between book value and tangible book value. Tangible book strips out purchased intangibles. Now we'll examine what put the largest one there, and what it means when part of it comes off.

How does an acquisition create goodwill?

Suppose a buyer pays more for a company than the fair value of its identifiable net assets. The purchase accounting has to place the price somewhere. Cash, receivables, inventory, property, assumed debt, and identifiable intangible assets are measured first. The residual becomes goodwill:

goodwill = purchase price - fair value of identifiable net assets

Identifiable intangibles are assets that can be separated or tied to contractual rights. Purchased technology, customer relationships, and trademarks can land here. Goodwill is the remainder. It can't be sold apart from the acquired business.

Don't read that remainder as a bag of cash or a separately appraised brand. It is a record of what the buyer paid beyond the net assets that accountants could identify. Some premium may prove sensible. Some may reflect expected synergies that arrive. Some may be the price of winning an auction.

Internally built reputation works differently. A company generally expenses the cost of building its own brand and customer base as it goes, so no matching goodwill asset appears. Acquire a rival with similar economics and a large goodwill balance can arrive on day one. The accounting therefore remembers the transaction. It doesn't capture every valuable quality of the business.

Goodwill has an unusual life after that day. It isn't run through a fixed amortization schedule under the accounting used by US public companies. Instead, management tests it for impairment at least annually and again when events suggest the carrying value may have fallen, as Microsoft's accounting policy describes. A carrying value is the amount recorded on the balance sheet. Fair value is management's estimate of what the reporting unit is worth under the required valuation framework. When carrying value exceeds supported fair value, the excess is written down within the accounting limits.

That estimate leans on forecasts. Expected revenue, margins, long-term growth, and the discount rate all matter. A small revision can matter a lot when the gap between carrying value and estimated fair value is thin. Read the valuation assumptions and sensitivity language around a large goodwill balance. The balance-sheet line alone can't tell you how much room is left.

What does an impairment charge do to the statements?

An impairment is an expense on the income statement. It reduces the goodwill asset and shareholders' equity on the balance sheet. Under the indirect cash-flow method, it is added back while reconciling net income to operating cash flow because no current cash payment accompanied the charge. Microsoft's filing shows both sides: $6.193 billion as operating expense and the same amount added back in operating cash flow.

That creates a sharp split between earnings and cash for the period. A company can report a large loss while operating cash flow changes little because of the write-down. A thesis that requires positive trailing net income may drop the company. A thesis driven by operating cash flow may barely move.

Neither reading settles the question. The net-income rule can treat an old acquisition error as if it were a fresh operating cash cost. The cash-flow rule can erase evidence that management paid for earning power that didn't arrive. Decide which claim you're testing before you choose the adjustment.

An impairment also lowers the denominator in ratios built from assets or equity. Future return on assets can rise after assets are written down, even if the operation earns exactly what it did before. A goodwill-excluding return-on-capital ratio won't take the same mechanical jump because goodwill was already absent from its capital base. The return-on-capital guide explains that trade-off: exclude goodwill to study tangible operating efficiency, then leave it in when studying acquisition discipline.

The charge is usually backward-looking. It recognizes that earlier expectations no longer support the recorded asset. It doesn't reveal the date the economics first deteriorated, and it may arrive well after the market has revised its view. Treat the filing date as the moment the accounting became public, not the moment the acquisition failed.

What did Microsoft's aQuantive charge reveal?

Microsoft's fiscal 2012 filing is unusually clean because it connects a large write-down to an acquisition by name. The company said the charge related mainly to goodwill from its 2007 acquisition of aQuantive in the Online Services Division.

The thesis was online advertising. By the annual test, Microsoft had lowered its forecasts for the division. The filing points to slower projected growth in search queries, search advertising revenue per query, and display revenue.

At its May 2012 annual test, Microsoft estimated the division's fair value using discounted cash flow. The company recorded a $6.193 billion goodwill impairment, reducing that division's goodwill from about $6.4 billion to $223 million. The filing calls the charge non-cash and non-tax-deductible.

The figures expose the sequence. Cash went out with the acquisition. Goodwill preserved much of the premium on the balance sheet. Later, revised expectations forced most of that recorded balance through operating expense. The charge didn't cause the economic disappointment. It measured management's revised view of it.

That makes the case useful for two separate research questions. If you're studying current operating cash generation, adding back the charge can be defensible because it wasn't a current cash outflow. If you're studying capital allocation, deleting it would throw away the point. The original purchase price and the later write-down belong in the record.

How can a stock thesis read goodwill?

Start with the balance sheet, then compare it with a prior filing. Quantery carries goodwill and other intangibles as statement fields. A large balance tells you that acquisitions shaped the accounting asset base. A large year-over-year decline gives you a filing to open.

The shape below is illustrative, and the bundled templates remain the reference for exact fields. The Quantery DSL documentation covers the expression vocabulary.

# "Acquisition accounting check" - illustrative
params:
  goodwill_heavy: 0.30       # proposed share of assets
  writedown_large: 0.20      # proposed year-over-year decline

features:
  assets_now: newest(total_assets)
  goodwill_now: newest(goodwill)
  goodwill_prior: lag(goodwill, 4)
  goodwill_share: if(assets_now > 0, goodwill_now / assets_now, null)
  goodwill_change: if(goodwill_prior > 0,
    (goodwill_now - goodwill_prior) / goodwill_prior, null)

criteria:
  acquisition_exposure:
    rules:
      - { when: "is_null(goodwill_share)", score: 0, flag: no_goodwill_read }
      - { when: "goodwill_share >= $goodwill_heavy", score: 2, flag: goodwill_heavy }
      - { when: "goodwill_share > 0", score: 1 }
      - { else: 0 }
  possible_writedown:
    rules:
      - { when: "is_null(goodwill_change)", score: 0, flag: no_comparison }
      - { when: "goodwill_change <= -$writedown_large", score: 2, flag: open_filing }
      - { else: 0 }

gate:
  mode: score
  min_score: 2

The thresholds are starting parameters. They aren't accounting rules. Move them and see whether the research set changes smoothly.

More important, a decline in goodwill is only a prompt. Acquisitions, divestitures, foreign-exchange translation, and reporting-unit reorganizations can move the balance too. The filing note has to establish that an impairment occurred and explain which business was affected. Structured fundamentals can find the footprint. They can't supply the cause.

Keep lag(goodwill, 4) tied to the same fiscal quarter in the prior year. Comparing a year-end balance with the previous quarter can confuse a completed deal with a write-down. Use newest() so a missing current value stays missing; a function that scans backward could pull an old goodwill figure into a new balance sheet.

Which backtests answer useful questions?

One design asks what happened after a material goodwill write-down became public. Use the filing date as the signal date, require that the decline be confirmed in the filing note, and study later returns. Any result is hypothetical and excludes costs. It can show whether these disclosed events were followed by a pattern in your tested sample. It can't tell you whether impairment caused the returns.

A second design tests acquisition exposure before any charge. Rank goodwill as a share of total assets, then pair it with operating performance and leverage. That asks whether balance sheets carrying large acquisition premiums behave differently, without waiting for management's valuation model to concede a loss.

A third compares two definitions of quality. Run return on capital once with goodwill removed and once with goodwill included. If a serial acquirer looks strong only after acquisition premiums disappear from the denominator, you've found the exact assumption doing the work. Testing one rule at a time keeps that comparison readable.

Don't backfill an impairment signal to the quarter when you think trouble began. The later filing benefits from forecasts and judgments unavailable then. Point-in-time backtesting has to wait until the charge and its explanation were public.

Goodwill records a bet management already made

Goodwill doesn't tell you that an acquisition was bad. It tells you how much of the purchase price remains after identifiable net assets were measured. Impairment doesn't recover the cash, and it doesn't create the operating weakness. It marks a downward revision to the value accounting can still support.

So keep two lenses. Strip goodwill out when the question is how efficiently tangible operations earn. Put it back when the question is what management earned on all the capital it chose to spend. Quantery lets both versions sit in plain text, one editable line apart. Run both. The disagreement between them is often the research result.

Want to try this on your own rules? Quantery is free for 14 days: the full app, no card required.

← All articles