A sole-source government contract is benchmarked by reconstructing the price a competitive process would have produced, using evidence that exists independently of the counterparty. Four reference methods carry most of the weight: published index pricing, netback or freight parity, comparable transactions, and cost build-up. Each is built separately, then reconciled into a defensible range.
The decisive discipline is sequencing. The benchmark method must be fixed and recorded before the price is negotiated. A benchmark produced afterwards documents the outcome; it does not test it.
What you will learn
- Why benchmarking replaces the price discovery an open tender would have delivered.
- How to build four independent references and reconcile them into a defensible range.
- What auditors test first, and the four failure patterns that discredit an otherwise sound file.
Written by the G2G Deal Design practicepowerabode DMCC's neutral buy-side practice for government-to-government procurement. We hold no position in the transactions we advise on — no cargo, no supplier, no equity. People, standards and certifications.
Why benchmarking replaces the tender
Open competition does two jobs at once. It selects a counterparty, and it discovers a price. When a government contracts directly — because supply security demands it, because a previous tender was compromised, or because only one realistic source exists — the selection job disappears but the price discovery job does not. Someone still has to establish what the deal should cost.
Benchmarking is that substitute. It is not a negotiating tactic and not a post-hoc justification. It is the mechanism by which a buying government forms an independent view of value before it commits, and the record by which it later demonstrates that the view was reasonable.
Four independent reference methods
No single reference settles a sovereign contract. Each method has a blind spot, and the blind spots differ, which is precisely why several are built in parallel.
1. Published index pricing
Most traded commodities have assessed reference prices published by independent price reporting agencies. Linking contract price to a named index, with the assessment window, the publication and the fallback all specified, moves the argument away from opinion and onto a public number. The work is in the detail around the index rather than the index itself: which assessment, averaged over what period, quoted at which delivery basis, and what happens if the publisher changes methodology or ceases publication.
2. Netback and freight parity
An index price refers to a specific location. A cargo delivered somewhere else is worth that price adjusted for the cost of moving it — freight, insurance, losses, port and terminal charges, demurrage exposure and any quality adjustment. Building the netback yourself shows whether the offered differential reflects real logistics or simply captures margin. Where several supply routes are physically possible, parity across those routes is often the single most revealing calculation available to a buyer.
3. Comparable transactions
Contracts of similar volume, tenor, credit profile and delivery structure — whether disclosed by other governments, visible in regulatory filings, or reported by trade press — establish the range within which serious counterparties transact. Comparability has to be argued explicitly. A twelve-month spot-linked arrangement and a fifteen-year take-or-pay commitment are not evidence about each other, and treating them as such is one of the faster ways to lose an auditor's confidence.
4. Cost build-up
Where indices are thin or absent — bespoke infrastructure, specialised equipment, unusual specifications — the remaining route is to rebuild the supplier's economics: input costs, conversion, logistics, financing, and a return commensurate with the risk actually transferred. Cost build-up is the weakest method used alone, because it depends on assumptions the buyer cannot verify. It is valuable as a cross-check on the other three, and as the only available method when they are unavailable.
The benchmark file
Benchmarking produces a document, not a conversation. A file that survives scrutiny contains, at minimum:
- The question the benchmark answers, stated as a price or price mechanism, not as a general assessment of reasonableness.
- The methods selected, and the reasoning for excluding the ones that were not used.
- The sources, dated, with the version or assessment date of each input recorded.
- The calculations, reproducible by a third party from the sources alone.
- The range produced, with the reasons for divergence between methods explained rather than averaged away.
- The date the method was fixed, relative to the negotiation timetable.
That last line does more work than the rest combined. It is the difference between a benchmark and a rationalisation.
What makes a benchmark survive audit
Auditors and public accounts committees rarely dispute arithmetic. They test whether the analysis could have reached a different conclusion, and whether anyone would have noticed if it had.
| Question asked | What answers it |
|---|---|
| Was the method chosen before the price? | Dated method note, approved ahead of substantive negotiation. |
| Could an independent party reproduce it? | Named sources, retained inputs, transparent calculation. |
| Was contrary evidence considered? | Divergent references recorded and explained, not discarded. |
| Who could overrule the analysis? | Documented separation between the negotiating team and the assurance function. |
| Does the contract hold the benchmark? | Index, window, fallback and review mechanics written into the agreement. |
Where benchmarks fail
Four failure patterns recur across jurisdictions and sectors.
- Benchmarking after agreement. The number is settled, then a reference is found that accommodates it. Reviewers detect this quickly, and it discredits analysis that may otherwise have been sound.
- Accepting the counterparty's reference set. A supplier proposing both the price and the benchmark against which it is measured has not been benchmarked.
- Averaging away disagreement. When netback parity and comparable transactions diverge materially, the gap is information. Collapsing it into a midpoint discards the most useful finding in the file.
- Benchmarking price and ignoring mechanics. Indexation, quality adjustment bands, delivery tolerance, penalty regimes and change provisions can move realised value further than the headline number. A benchmark confined to price per unit measures the smaller part of the deal.
Practical sequence. Fix the method and approve it. Build each reference independently. Reconcile into a range and explain the divergence. Negotiate against the range. Write the mechanism — not just the number — into the contract. Retain the file for the audit that arrives years later.
Key takeaways
- Benchmarking substitutes for the price discovery a tender would have delivered — it is not a justification exercise.
- Build four references independently: published index, netback parity, comparable transactions and cost build-up.
- Fix and date the method before the price is negotiated; sequencing is what auditors test first.
- Explain divergence between methods rather than averaging it away — the gap is the finding.
- Benchmark the mechanism as well as the number: indexation, quality bands and penalties often move more value than price.