Aug Concentration: Avg Refine Top-1 72% / Top-3 88% — Cu Mine 23% vs Smelt 50% as Cathode +36%

Our Q3 concentration lens answered the distribution question after IEA Global Critical Minerals Outlook 2026 restated midstream shares: average top refining-country share (ex-REE) at 72%, Top-3 near 88%, median Top-1 across the tracked table in the low-70s, and 10 stages clearing ≥70% Top-1 — with gallium refined still near 99% and rare-earth separation the rare durable ease at 85%. The August theme update then qualified the flow side of the same system: World Bank Pink Sheet (Aug 4, 2026; July monthly) printed tin +55.5% YoY, copper +36.2%, metals & minerals index +25.2%, and copper spot treatment charges near −$90/t against a $0/t 2026 settle, with a China smelter cut plan around 10%. This August concentration lens asks the desk question those two vintages imply together: how concentrated is the system at the top of the distribution once Pink Sheet stress sits beside the held Top-1 / Top-3 / HHI ladder — and where is supply thin enough that the economy’s “always available” assumption breaks?

The interactive dashboard above is built as that overlay. Toggle Binding tip, Stress × share, Mine → plant, and Price & HHI. Filter by stage, stress band, and minimum Top-1. Metric control flips Top-1, Top-3, and an analytical Binding score (Top-1 × Aug stress weight). The punchline is deliberately two-sided. On shares, the Q3 ladder is held — we are not pretending July prices restated country capacity. On stress, copper, tin, aluminum, zinc, and fertilizer gates show that thin spots can tighten without the Top-1 percentage moving a point. Analysts who quote only mine shares miss smelter tips; analysts who quote only price YoY miss that gallium and graphite anode stay extreme even without a Pink Sheet print.

The Aug concentration scoreboard

Read the table as a family of market shares plus stress meters, not one index. Top-1 answers who leads a stage. Top-3 answers how thick the tip is. HHI answers how monopolistic the country ladder feels. Pink Sheet YoY answers whether the price gate is screaming on the same physical input. Binding score is an analytical product of the first and the last — useful for ranking, not for citing as a disclosed IEA number.

Binding tip: Top-1 still owns the desk, stress reorders the middle

Open Binding tip and leave the metric on Top-1. The ranked bar still puts gallium, graphite anode, battery recovery, REE separation, and the cobalt / nickel midstream cluster at the top of the distribution. Flip to Top-3 and the tip barely softens — many of those stages clear ≥90% in the top three. Flip to Binding and copper smelting / refining and tin climb because Aug stress weights them harder even though their raw Top-1 is “only” the 48–50% band.

That is the first hinge for August. Concentration architecture did not need a new IEA census to stay binding. What changed is which middle-of-the-table stages become operationally hot when cathode, solder, and fertilizer prices print double-digit YoY on top of already-elevated shares. Filter Min Top-1 ≥70% and you still see ten stages. Filter Stress = extreme and copper joins the gallium / graphite / REE club for a different reason: not because Chile’s mine share jumped, but because China smelter capacity at 50% meets record-adjacent LME copper and broken treatment charges.

Producer seats remain China-heavy on the midstream tip, with Indonesia’s nickel growth-capture and DRC cobalt mine as the clearest non-China Top-1 seats. Sector cards show batteries and semiconductors carrying the highest median Top-1 and the most “hot price” hits where Pink Sheet maps exist.

Stress × share: price YoY is not a share restatement

Switch to Stress × share. The scatter puts Pink Sheet YoY on the vertical axis and held Top-1 on the horizontal. Tin sits high on price (+55.5%) with a mid-40s Top-1 estimate — stress outrunning the disclosed share tip. Copper sits at +36% with smelt / refine Top-1 50%. Aluminum (+20%) and zinc (+25.5%) occupy the structural-metal band. Nickel is the revision story: +9.8% YoY but −11.5% from the May peak, while Indonesia’s refined-supply growth capture still prints ~75% Top-1 on the Q3 share ladder.

The concentration curve beside it is the share geometry the Q3 lens already scored: cumulative Top-1 mass rises far above the equal line. August does not flatten that curve. It paints which points on the curve are also pricing stress. Toggle the equal line off if you want the raw tip heaviness without the 45-degree reference.

Caveat in one sentence, then expand later: price Δ ≠ share Δ. A copper mine Top-1 of 23% can coexist with a +36% cathode print because the binding scarcity is midstream fees and capacity geography, not Chilean pit share.

Mine → plant: the stage flips Pink Sheet makes expensive

Open Mine → plant. Lithium still shows the largest gap in the battery stack — Australia mine Top-1 37% versus China chemical refining 70% (+33 pp). Copper is the August poster child: Chile-led mine Top-1 23% versus China smelting capacity 50% (+27 pp), now paired with LME copper $13,543/mt in July and spot TC near −$90/t. Graphite’s gap is smaller because the mine is already concentrated (77.8% → 90% anode). Cobalt barely flips at all (74.2% → 75%) — both ends concentrated. Rare earths separate hotter than they mine (69% → 85%).

The smelter stress strip is the fee-side companion to that copper flip: China capacity share 50%, utilisation ~85% vs ex-China ~70%, 2026 TC/RC settle $0/t, spot path about −$90/t, and a China cut plan near 10% (~961 kt class, with halted planned capacity in the multi-megatonne narrative from the August update). Custom smelters paying to process concentrate while cathode prints record-adjacent is the market’s way of saying the midstream tip is binding even when the mine map looks plural.

Pair this panel with the copper mine vs refinery geography companion when you need the longer spatial story. This post’s job is the concentration meter with an August stress overlay, not a full GIS rewrite.

Price path & HHI: basket stress beside band mix

Switch to Price & HHI. The composed path shows copper climbing from 2025 annual $9,947 through Q1 $12,831 into July $13,543, while the metals & minerals index peaks in May (148.8) and cools to July 140.5 — still +25.2% versus 2025. Nickel’s May spike and July revision sit as a dashed reminder that YoY signs can hide intra-year mean reversion.

The HHI donut answers a different question: how many stages land in extreme (≥5,000), high, moderate, or plural bands. Midstream gallium / graphite / recovery / REE / cobalt / nickel growth and the average-refine headline sit in the extreme-to-high club. Copper mine (~900) is the plural reminder that some pits are diversified even when plants are not. Downstream disruption cards keep IEA’s graphite (~$300B/yr) and REE full-chain (~$6.5T framing) risk dollars next to their Top-1 shares — order-of-magnitude security meters, not precise loss forecasts.

Top-k clearance counts close the panel: stages clearing Top-1 ≥50%, ≥70%, ≥85%, and Top-3 ≥90%. August did not move those counts with a new share census; it asks whether desks still treat “plural mines” as proof the system is safe when the clearance ladder says otherwise.

Who is exposed — and what the Aug overlay buys

Exposed to quoting the wrong stage: readers who cite copper mine Top-1 23% as evidence the metal is “diversified” while China holds ~50% of smelting capacity and spot TC is deeply negative; readers who cite lithium mine plurality while chemical refining sits near 70%.

Exposed to price-as-share confusion: desks that treat tin’s +55.5% YoY as proof of a new monopoly share print, or nickel’s May→July pullback as proof Indonesia’s growth-capture tip eased. Shares are held from Q3; prices are stress.

Exposed to thin future relief: battery and magnet buyers who hoped IEA’s investment pullback (critical minerals −9%, battery metals −20%, lithium −40%) would automatically flatten Top-1 — capital diversion without delivered midstream capacity leaves the tip intact.

Exposed on food-system gates: fertilizer importers who see Morocco’s export Top-1 at “only” 28% and miss DAP +14% and phosphate rock +11.5% as thin-gate stress on a top-3 near 67%. See also the phosphate fertilizer export companion.

What the overlay buys: a single board where Top-1 / Top-3 / HHI stay comparable to the Q3 concentration print, while Pink Sheet and smelter meters flag which thin spots are currently expensive. That is the August contribution relative to both the Q3 concentration companion and the price-stress update.

Caveats and methodology

• Share ladder is held from concentration-2026q3 (IEA GCMO 2026 midstream + MCS 2026 mine anchors). July prices do not restate country capacity shares.
• Price Δ ≠ share Δ. Pink Sheet YoY and spot TC/RC are stress meters on the same chokepoints, not Top-1 revisions.
• Binding score is analytical (Top-1 × stress-band weight). Do not cite it as an IEA or USGS disclosed index.
• Top-3 and HHI on many rows are derived/estimated from disclosed Top-1 plus prior theme ladders; country shares may not sum to 100%.
• Tin, aluminum, and zinc Top-1 figures in this table are approximate midstream / primary gates for stress mappingconfidence tagged estimated.
• Copper Q1 / May path points on the price chart are illustrative bridges between disclosed annual / July Pink Sheet prints and Q3 narrative; July and 2025 annual are the hard anchors.
• Spot TC ~−$90/t follows the IEA / industry Mar-2026 path cited in the August updatefees move faster than annual settles.
• Disruption dollar figures (graphite ~$300B, REE ~$6.5T) are IEA framing orders of magnitude, not loss forecasts.
• US net-import reliance percentages are MCS-linked context, not the concentration numerator.
• Cross-agency comparisons (IEA vs USGS vs Pink Sheet) are directional; scopes and years differ.

The shareable takeaway

August did not redraw who owns the chokepoint ladder — it showed which tips are expensive while the Top-1 / Top-3 architecture stayed intact. Avg refining (ex-REE) still prints 72% Top-1 / ~88% Top-3; 10 of 19 stages clear ≥70% Top-1; gallium remains the extreme at ~99%. Copper is the binding Aug story in one line: mine Top-1 23% → smelt 50%, cathode +36% YoY, spot TC near −$90/t. Tin’s +55.5% reminds you stress can outrun share. Watch shares and stress on the same board — not either meter alone — to see which physical inputs the economy still assumes will always be available.