The text below is an edited version of an Insight originally published on CRU Online, dated 10 March 2026. For the full version, contact us here.
Aluminium flat-rolled products are not immune to the latest market turmoil. We updated our Monte Carlo simulation to assess how the current risk environment could push European conversion fees higher. Using the same ensemble decision-tree regression framework we used previously for monthly 1050 common-alloy conversion fees, this update shifts the focus to upside risks from soaring metal premia, elevated metal prices, and persistent energy pressures.
What we updated and why it matters
We refreshed the model with the latest for six-monthly inputs – European stainless‑steel sheet price, the LME 3‑month aluminium price, Rotterdam ingot duty‑paid premium, the USD/EUR exchange rate, the Economic Sentiment Index (ESI) and inflation (HICP).
Feature‑importance in the updated model has moved away from the LME 3‑month aluminium price alone and now places greater weight on stainless steel sheet prices and the Rotterdam ingot duty‑paid premium. USD/EUR retains only a marginal influence, which is why we allowed the exchange rate to vary within the stress framework rather than imposing a separate foreign‑exchange shock.
This shift matters because it shows that downstream cost and premium conditions are now more determinant of European conversion fees than the LME in isolation. In the present mix (high metal prices alongside sharply wider premia) fee outcomes are more sensitive to premia and downstream price conditions.
Key behavioural relationships
Our earlier exploratory work remains valid – the 1050 conversion fee tracks the common sheet alloy 5754 closely on an ex‑premium basis, so 1050 Monte Carlo simulations provide a useful directional guide to 5754 when adjusted by the average long-term spread. Plate conversion fees behave differently – owing to higher value-added content, plate fees are a separate category and are generally better placed to withstand short‑term turbulence.
We ran three scenarios
We ran three Monte Carlo scenarios using the trained regression model:
- Base scenario: Inputs vary according to their historical distributions and dependence structure.
- Single‑factor stressed scenario: Ingot premium fixed at an extreme historical level while other inputs vary.
- Multi‑factor extreme scenario: High LME and ingot premium combined with elevated inflation and moderately high stainless‑steel prices, with USD/EUR and ESI varying conditional on historical dependence.
What the Monte Carlo simulations show
At current market fee levels, the 1050 conversion fee sits slightly above the mean that is predicted by both the base and the single-factor ingot stress runs. In other words, a high ingot premium on its own does not mechanically translate into a much higher ex‑premium conversion fee. By contrast, the multi-factor extreme scenario produces a materially higher mean predicted fee and a distribution that shifts well above current rates, signalling strong upside risk if high metal prices, sharply wider premia and persistent inflation occur together.
Current conditions only partially align with that extreme set‑up: LME prices and ingot premia are elevated, but stainless steel prices and inflation have not reached the stress levels used in the extreme case. This scenario should be treated as a forward-looking stress test rather than a reflection of today’s market.
We see clear upside risk to European conversion fees if multiple cost pressures persist in tandem. We can help clients quantify these risks through scenario analysis, benchmark modelling and procurement stress testing, enabling better planning and pricing decisions in an uncertain market. If you’d like a tailored run of these scenarios for your exposure, contact us here.
