Extraordinary Events: The One-Off Items That Distort the Benchmark
Extraordinary events defined: the one-off items — litigation, restructuring, spectrum charges, natural events — that distort a comparable’s PLI, and the adjust-versus-exclude discipline.
Definition
Extraordinary events (the extraordinary items) are the one-off, non-recurring entries in a comparable’s financials that distort its PLI away from the recurring run-rate the benchmark is trying to measure — the comparability difference of the income-statement kind, in contrast to the working capital adjustment’s balance-sheet kind. The forms: the litigation outcome (the one-year settlement gain or the provision charge), the restructuring cost (the plant closure, the redundancy package, the one-off impairment), the regulatory one-off (the spectrum charge, the licence fee, the fine), the natural event (the flood, the disaster loss, the one-year demand shock), and the transactional one-off (the asset-sale gain, the acquisition cost, the one-off listing expense). Each is extraordinary in the benchmarking sense — not part of the recurring economics the tested party’s comparables are being measured for — and each, left in the comparable set’s values, pulls the arm’s length range in the direction the one-off went. The discipline is the standing fork: adjust (the item is isolable and measurable — the PLI is computed on the adjusted figures, the item and the adjustment stated) or exclude (the item is not isolable, or the distortion is structural — the comparable drops out of the set, the [accept-reject matrix] (/docs/glossary/accept-reject-matrix) carrying the reason).
An extraordinary event, in one handling:
1. The item (the one-off — the litigation charge, the spectrum fee, the restructuring cost)
2. The test (one-off and non-recurring? isolable from the recurring operations?)
3. The handling (adjust — the adjusted PLI, the item and the adjustment stated / exclude — the recorded reject)
4. The record (the matrix row: the item, the amount, the treatment, the source)
| The element | The content |
|---|---|
| The item | The one-off entry (the litigation, the restructuring, the regulatory charge, the natural event, the transactional gain) — identified in the qualitative screening |
| The test | One-off and non-recurring (not a normal cost of the business — the recurring litigation provision of an active-litigation company is not extraordinary, the distinction stated) and isolable (the amount separable from the recurring operations) |
| The handling | Adjust (the PLI on the adjusted figures — the item removed, the adjustment stated) or exclude (the comparable out of the set — the reason recorded) |
| The record | The matrix row: the item, the amount, the treatment (adjust/exclude), the source (the filing note, the announcement) |
The working read (the comparability adjustments guide): the extraordinary event is the adjustment family’s most frequent member — the qualitative screening reads the filing notes for them company-by-company, and the handling decision (adjust vs exclude) is the per-company judgment the defending the matrix guide rehearses. Two disciplines stand out. The recurrence test: the recurring version of an “extraordinary” item (the active-litigation company’s annual settlement provisions, the cyclical industry’s regular inventory write-downs) is part of that company’s economics — adjusting it out would understate its true recurring position, and the correct handling is often the exclude (the company’s risk profile is not the tested party’s) with the reason stated, not the adjust. The direction honesty: the adjustment is applied the same way regardless of the direction it moves the range — the one-off gain that inflates the comparable’s margin is adjusted out as readily as the one-off charge that deflates it, and the benchmarking mistakes checklist names the direction-selective adjustment (the one-offs adjusted only where they flatter) as a line item.
Example
A TNMM study (OP/OC) on an Indian services company. Three comparables, FY25, each with a one-off:
| Comparable | The item (from the filing notes) | The test | The handling |
|---|---|---|---|
| C1 | ₹42 cr litigation settlement charge (one disputed contract — the notes state the one-off) | One-off, isolable (the note isolates the amount) | Adjust — the FY25 PLI on the charge-removed figures; the item and the adjustment in the matrix |
| C2 | ₹18 cr annual litigation provision (the company is in active, recurring disputes — the provision is a standing line) | Recurring (the company’s economics include the dispute cost) | Exclude — the risk profile is not the tested party’s; the reason recorded (not adjusted — the provision is the company’s cost of doing business) |
| C3 | ₹90 cr spectrum charge (the regulatory one-off, the notes state the non-recurring character) | One-off, isolable | Adjust — the charge removed, the PLI on the adjusted figures |
The contrast is C1/C3 (adjust — the one-off, isolable, removed with the amount stated) against C2 (exclude — the recurring version, the company’s economics, the risk mismatch recorded). The [IQR] (/docs/glossary/interquartile-range) is built on C1’s and C3’s adjusted values and the remaining clean members; C2 is out with its reason. The file’s benchmarking annex shows the three rows — the item, the amount, the treatment, the source — and the TPO’s follow-up (“C2: why excluded?”) is answered by the recurrence test, on the record.
See also
- Comparability Adjustments Beyond Working Capital
- Qualitative Screening
- Benchmark Refresh
- Defending Your Accept-Reject Matrix
FAQ
Is “extraordinary” a statutory term in this context? No — the accounting extraordinary-item concept (the old IAS 1 “items of income and expense arising from unusual or infrequent activities”) was removed from the standards, and the Indian scheduled framework does not carry a TP-specific “extraordinary item” list. The benchmarking sense is a comparability concept: the entry is extraordinary for the purpose of the benchmark — one-off and non-recurring relative to the recurring run-rate the PLI measures. The discipline is the stated test (the one-off character, the isolation), not a schedule — which is why the [handling decision] (/docs/benchmarking/comparability-adjustments) (adjust vs exclude) and its record carry the defensibility, not a label.
What if the extraordinary item is in the tested party’s own results? The tested party’s PLI is the unadjusted one (the tested party is the anchor — see the [working capital adjustment] (/docs/glossary/wc-adjustment) FAQ on the direction). Where the tested party’s own year carries a one-off, the handling is the multi-year averaging (the 3-year pool dilutes the one year’s item — the averaging as the tested-party-side treatment) with the item disclosed in the file (the one-off stated, the averaging’s effect on it noted) — not an adjustment of the tested party’s own PLI, which would be the method error (the anchor moved to fit the range). The disclosure is the defence: the examiner sees the item, the averaging, and the stated interaction.
Adjust or exclude — the decision rule in one line? Isolable and one-off → adjust (the amount removed, the adjusted PLI, the item stated); recurring or not isolable → exclude (the company’s economics differ, the reason recorded) — and the decision is applied without regard to the direction the handling moves the range. The [defending the matrix guide] (/docs/benchmarking/defending-accept-reject) treats the adjust/exclude column as the one the examination reads most closely, because it is where the benchmarking judgment (not the arithmetic) is on the record.
Run the screens as a study, not a spreadsheet
Quartyl applies the method, PLI and screening steps above as a pipeline — and keeps a documented reason for every exclusion.
Related docs
Handling Extraordinary Events in Comparability
How to handle extraordinary and non-recurring events in Indian comparability analysis — one-off gains, COVID-style years, restructuring, and the treatment that survives TPO review.
Read docBenchmark Refresh: When and How to Re-run the Comparables Study
The benchmark refresh defined: the re-run of the comparables search and screens on current data — the triggers (data age, the range position), and the record that makes a refresh defensible.
Read doc