Ferroalloy supply and ferroalloy demand are not in the same places, and they are not there for the same reasons. Production follows energy. Consumption follows steelmaking. The distance between those two maps is the entire reason a trading desk exists.
Supply follows power, not steel
A submerged arc furnace is, economically, a device for converting electricity into alloy. Silicon-based alloys consume on the order of 8,000 to 9,000 kWh per tonne. At that intensity, a few cents per kWh is the difference between a profitable furnace and an idle one.
The consequences are structural:
- Production clusters around cheap, firm power — hydro basins, subsidised industrial tariffs, stranded generation looking for a load
- Supply is seasonal where power is hydrological. Wet and dry seasons show up in output, independent of demand
- Capacity is slow and lumpy. A furnace is a multi-year, capital-heavy commitment. Supply does not respond to a good quarter
- Policy is a supply variable. Power allocation, export duties and licensing move available tonnage faster than any physical constraint
Demand follows steel, and steel is moving
Consumption sits where crude steel and castings are made, which is a different and shifting map. Three shifts matter to anyone sourcing forward.
Route. The growth of electric arc furnace steelmaking raises ladle-side alloy work per tonne and raises sensitivity to alloy quality.
Grade. Mills moving up the value ladder shift the basket toward microalloys — niobium, vanadium, titanium — where a small tonnage carries high value and a long qualification.
Geography. New steel capacity is not being added where the old capacity is. Each new regional cluster creates a sourcing problem before it creates a supply one.
The structural fact of this market is that the lowest-cost producer of an alloy and the mill that needs it are usually in different countries, often on different continents, and frequently separated by a duty.
What keeps the two sides apart
Four frictions sit between a producer with tonnage and a buyer who needs it.
| FRICTION | WHAT IT ACTUALLY MEANS |
|---|---|
| Qualification | A mill cannot simply switch producer. Trial cargoes, furnace trials and internal sign-off take months |
| Specification drift | A producer's "FeSi 75" and a buyer's "FeSi 75" are the same name and frequently not the same material, once sizing and traces are compared |
| Trade measures | Duties, quotas and licensing make some origin-destination pairs uneconomic regardless of production cost |
| Payment and risk | Cross-border terms, documentary requirements and counterparty risk rule out counterparties who could otherwise trade |
None of these are solved by finding a cheaper number. They are solved by knowing which producers have already cleared them for a given destination.
Balance is a local condition
"Global balance" is a reporting convenience. For a buyer, the only balance that matters is the one at their gate: which producers can deliver their specification, to their port, under their terms, at a price that works. A globally long market can be locally tight if the producers who are long cannot clear the duty, the spec or the qualification.
That is why two buyers of the same alloy in the same month can have genuinely different experiences of the market, and why both are right.
Practical implications
- Qualify early, not on need. The time to approve a second producer is while the first one is performing
- Watch the origin map, not just the price. A duty change can remove your cheapest source overnight without any producer's cost changing
- Treat power policy as supply news. Tariff changes and hydrology in producing regions are leading indicators
- Separate term from spot deliberately. Term covers the baseline that must not stop; spot takes the opportunity