The software company that created FT Mercati Go to ftmercati.com

Guide · Commodities and data

Metal price indexation clauses: a practical guide

How a metal price indexation clause works, what it must contain, a worked example of the calculation and how to automate it in your ERP.

By FT Support Updated on 9-minute read

The short answer

A metal price indexation clause links the price of a supply contract to a reference price, usually the monthly average of the LME Official Cash Settlement or a Milan Index, according to a formula written in the contract. Automating it means having the ERP, the company's management software, calculate with data from an authorised source and fixed rules what is done by hand in Excel today: updated price, adjustment and supporting document.

  • The typical formula: reference price plus premium, converted into euros, multiplied by the metal content, plus the conversion charge.
  • Six elements must be written down: reference, period, premium, exchange rate, metal content with conversion charge, rounding and revision.
  • The calculation stays deterministic: it is done by code, not by an AI model.
  • AI helps to read the clauses in contracts and to flag anomalies.
  • If the reference ceases, as ASSOMET's Milan prices did in 2026, the clause must say what happens.

Note. This guide is for information and practical purposes. The wording of a contract clause should always be agreed with your own legal adviser. The numbers in the example are invented and serve only to show the calculation.

What an indexation clause is for

A company that sells a product in which metal accounts for 60 or 80 per cent of the cost cannot hold its price for a year: one month of rising prices would be enough to leave it working at a loss. The buyer, on the other side, does not want to pay increases that do not depend on the metal.

The indexation clause solves the problem transparently: the part of the price that depends on the metal follows a public, verifiable price, while the part that depends on processing stays fixed or follows other rules. Neither side has to trust the other: the formula is enough.

The six elements a clause must contain

  1. The reference price. With name, source and side of the price. For LME metals it is normally the Official Cash Settlement. For many Italian products, such as copper for semi-finished products, brasses and bronzes, the Milan Indexes are used.
  2. The quotation period. The average of the delivery month, of the month before, of the quarter, or the price on a specific day. It is the element that causes the most disputes when it is written vaguely.
  3. The premium. The surcharge in dollars per tonne that covers transport, duties, financing and the shape of the metal, usually set for a quarter, a half-year or a year.
  4. The exchange rate. Which euro-dollar exchange rate, of which day or period, from which source, to how many decimal places. The two common choices are the fixing recorded at the LME and the ECB reference rate.
  5. The metal content and the conversion charge. How much of the product follows the metal (for an alloy: the percentages of its components) and how much is processing.
  6. Rounding, thresholds and revision. How to round, whether there is a tolerance threshold below which the price does not change, how often it is recalculated and what happens if the reference ceases to be published.

An example calculation

Invented numbers, to show the mechanism. A contract for a brass semi-finished product provides for:

  • metal content: 63% copper and 37% zinc;
  • reference: the monthly average of the LME Official Cash Settlement for the month before delivery;
  • premium: USD 200 per tonne on copper, USD 250 on zinc;
  • exchange rate: the monthly average of the LME euro-dollar fixing for the same month, four decimal places;
  • conversion charge: EUR 1,400 per tonne, fixed for the year.

If in the reference month the copper average is USD 9,500, the zinc average USD 2,800 and the average exchange rate 1.1000:

StepCalculationResult
Copper with premium, in euros(9,500 + 200) / 1.1000EUR 8,818.18/t
Zinc with premium, in euros(2,800 + 250) / 1.1000EUR 2,772.73/t
Value of the metal in the alloy0.63 × 8,818.18 + 0.37 × 2,772.73EUR 6,581.36/t
Supply price6,581.36 + 1,400EUR 7,981.36/t

Each step is trivial. The problem is doing it every month, for dozens of contracts, each with its own formula, without getting the period, the exchange rate or the rounding wrong.

Where the errors come from

  • Wrong period: the delivery month instead of the month before, or vice versa.
  • Average calculated over calendar days instead of trading days.
  • One day’s price converted with another day’s exchange rate.
  • Multiplying by the exchange rate instead of dividing.
  • Last quarter’s premium left in the spreadsheet.
  • Different formulas for different customers, all in one person’s head.
  • No record of the data used: when the dispute arrives, the calculation has to be redone from scratch.

How to automate it

  1. Write the formula in structured form. For each contract the ERP stores the parameters: reference, period, premium, exchange rate, percentages, fixed charge, rounding. No longer free text, but fields.
  2. Connect the prices from an authorised source, with the history and the averages already calculated. This is the subject of the guide on how to connect LME prices to your ERP.
  3. At the end of the period the system calculates the updated price for each contract and, where needed, the adjustment on deliveries already invoiced.
  4. The evidence is saved. Next to each result the price, exchange rate, dates and source are kept: the supporting document for the customer is generated from there.
  5. A person approves. The system proposes; issuing price lists, credit notes or debit notes goes through a confirmation.
  6. Discrepancies are flagged. A supplier invoice that does not match the formula, an expired premium, a contract with no valid reference.

Where artificial intelligence (AI) comes in

In this process AI has a precise role, and it is not doing the sums.

  • Reading the contracts. The clauses are written in PDFs, in Italian and in English, each in its own way. An AI agent reads them and proposes the structured parameters, which a person checks. It is the step that makes automation practicable even with hundreds of contracts already signed.
  • Explaining. “Why did the October price go up by 4%?” The answer comes in plain words, with the numbers used.
  • Flagging. Ambiguous clauses, references no longer published, invoices that do not follow the formula.

The calculation remains entrusted to deterministic code: same formula, same data, same result, every time. It is the rule we apply to all our AI agents.

The case of the Milan prices

For decades many Italian contracts cited the “Milan prices” calculated and published by ASSOMET, Italy’s non-ferrous metals industry association. Since 1 August 2026 ASSOMET no longer publishes them. FT Mercati continues the calculation independently with the Milan Indexes: same criteria, full history preserved, series that run through 1 August without a break. There are thirteen daily indexes, from copper for semi-finished products to brasses, from bronze to silver.

What to do about existing contracts:

  • indexed contracts are not left without a reference: monthly averages, comparisons and revision clauses continue to work;
  • it is worth updating the source citation to “FT Mercati Milan Index” (“Indice Milano FT Mercati” in Italian) and agreeing the change with the counterparty;
  • it is an opportunity to add to the clauses what is often missing: what happens if a reference ceases.

All the details are in the guide to the Milan Indexes.

Checklist for writing or reviewing a clause

  • Is the reference price stated with name, source and side of the price?
  • Is the quotation period defined without ambiguity, public holidays included?
  • Does the premium have an expiry date and a renewal rule?
  • Does the exchange rate have a source, a day or period and a number of decimal places?
  • Are the metal percentages and the conversion charge written down?
  • Are there rules for rounding, tolerance thresholds and frequency of revision?
  • Does it say what happens if the reference price ceases to be published or changes its name?
  • Who calculates, by when, and with which document is the result communicated?

If the answer to all of them is yes, the clause can be automated. If you want to do it, let’s talk.

Sources

  1. FT Mercati, Guide to the LME price: monthly averages and the reference for physical contracts
  2. FT Mercati, Guide to LME premiums: the LME-plus-premium formula, quotation period
  3. FT Mercati, Guide to the euro-dollar exchange rate in metal prices
  4. FT Mercati, Guide to the Milan Indexes: continuity after 1 August 2026

Checked on 17 September 2026. Rules and deadlines change: always check the official source before you decide.

Frequently asked questions

The short answers

What is a metal price indexation clause?

It is the part of a supply contract that sets out how the price changes as the price of a metal changes. It states the reference price, the period over which it is calculated, the premium, the exchange rate to use, the share that does not depend on the metal, and the rounding and revision rules.

Which price is usually used as the reference?

For metals traded on the London Metal Exchange the most common reference is the monthly average of the Official Cash Settlement for the delivery month or the month before, to which the agreed premium is added. In Italy, for copper, brasses, bronzes and other products, many contracts use the Milan Indexes, which continue the long-established Milan prices.

What happens to contracts that cited ASSOMET's Milan prices?

ASSOMET, Italy's non-ferrous metals industry association, stopped publishing the Milan prices on 1 August 2026. FT Mercati continues the calculation with the Milan Indexes, with the same criteria and the full history, so the series have no breaks. The advice is to update the source citation in the contracts, agreeing the change with the counterparty.

Can the calculation of an indexation clause be automated?

Yes. You describe the formula in structured form for each contract, you connect the prices from an authorised source, and at the end of the period the ERP calculates the updated price and the adjustment, saving price, exchange rate, date and source. A person approves before the documents are issued.

What is artificial intelligence for in this process?

Not for doing the sums, which remain entrusted to deterministic code. It is for reading the clauses in PDF contracts and proposing their parameters, for explaining to a colleague or a customer how a price was arrived at, and for flagging anomalies, for example an invoice that does not match the formula.

Do you have a process you would like to see working better?

Tell us about it in a first conversation, with no obligation: together we work out whether and where AI brings a measurable benefit, and which step to start with.