Commodity Spread Monitor — March 2026

View

Domestic crush is the constructive lever in March, but quietly. USDA raised 2025/26 crush by 5 mb to 2.575 bb on a +5 mb import addition routed entirely into processing. Exports stayed at 1.575 bb and ending stocks at 350 mb, holding stocks-to-use flat at 8.2%. The cleanest read is on the meal side: oil share of crush value at 51.9% (week ending Mar 6) — the highest since August — coexists with a USDA cut to soyoil-for-biofuel of −800 mn lb on RVO/45Z policy uncertainty, partly offset by +750 mn lb routed into food/feed/other.

Signal

Watch the next two prints to size the conviction. NASS January Fats & Oils printed a record 227.86 mb crush (+~7% YoY); two more above-trend monthly prints would force USDA to keep raising 2025/26 crush, with the rotation eating into the export forecast at April 9. Brazil's record 178 mmt harvest near completion makes that rotation more likely, not less.

What changed

• USDA raised 2025/26 U.S. crush from 2.570 bb (Feb) to 2.575 bb (Mar), routing a +5 mb import addition straight into processing. Exports unchanged at 1.575 bb; ending stocks unchanged at 350 mb.
• January 2026 NASS crush printed a record 227.86 mb (vs ~213 mb Jan 2025, +~7% YoY) — well above pre-report estimates.
• USDA cut 2025/26 soyoil-for-biofuel by −800 mn lb on RVO/45Z policy uncertainty, partly offset by +750 mn lb routed to food/feed/other (net −50 mn lb usage).
• CONAB (Mar 13) put Brazil's 2025/26 harvest at a record 177.85 mmt; USDA holds 180 mmt unchanged. Harvest is near complete in Mato Grosso, Paraná, MS and GO.

View

Late March's energy setup is two-layered. The single-week WPSR (week ending Mar 20) showed crude flat versus the 5Y average and distillate building +3.0 mb. But EIA's March STEO explicitly forecasts distillate inventories below the 5Y average across the entire forecast horizon, and refinery utilization climbed from 90.8% (week ending Mar 6) to 92.9% (week ending Mar 20) as a comparatively light spring turnaround season completed. The trade is on the structural framing, not the week.

Signal

The Iran–Hormuz risk premium is in flat price, not in the crack — that's the cleanest read. Watch whether the next two WPSR prints rebuild distillate beyond +3 mb (which would invalidate the structural setup) or stay near current levels (which validates STEO). OPEC+ confirmed +206 kb/d unwind for April, easing crude tightness exactly when products need to do the work.

What changed

• WPSR (week ending Mar 20) showed crude stocks at 456.2 mb (~0.1% above 5Y average — flat) after a +13 mb build across the month from 443.1 mb (week ending Mar 6).
• Refinery utilization climbed to 92.9% (single-week, 4w avg 91.1%) as a comparatively light spring turnaround season completed.
• Four-week product supplied diverged: distillate 3.932 mb/d (+1.3% YoY) but gasoline 8.796 mb/d (−0.9% YoY) — diesel demand running ahead of gasoline.
• EIA's March STEO held its forecast that distillate inventories will remain below the 5Y average through the forecast horizon, even as the latest week drew distillate to about 0.4% below the 5Y average.
• OPEC+ JMMC (Mar 1) confirmed continuation of the 1.65 mb/d voluntary cut unwind, with a +206 kb/d adjustment for April.

View

March is when concentrate market tightness becomes operational. The 2026 annual benchmark TC/RC settled at $0/t (Antofagasta–Jiangxi, Dec 19 2025) and SMM's imported concentrate index broke −$60/dmt for the first time on Mar 13. China's CSPT formally agreed to cut 2026 refined output by more than 10% via longer and more frequent maintenance windows — the first coordinated smelter discipline of this cycle. At the same time, LME copper stocks at 359k t (Mar 24) and SHFE at 306k t (Mar 19) alongside near-record COMEX inventories sit above 1 Mt total, and Yangshan premium collapsed to ~$34/t in mid-March, signaling no Chinese appetite for refined imports. The combination is the cleanest mid-stream margin transfer in the note.

Signal

The signal is that economics has flipped from smelter to miner without flat price doing the work. Watch whether spot TC/RC keeps deepening into CESCO Week (early April) and whether the announced CSPT cuts show up in monthly Chinese refined output. Grasberg's late-March phased restart adds a swing variable — Freeport's Q1 results in late April will quantify the 2026 supply path.

What changed

• 2026 annual copper concentrate benchmark settled at $0/t and 0.0 c/lb (Antofagasta–Jiangxi, Dec 19 2025), down from $21.25/t and 2.125 c/lb in 2025 — first ever zero settlement.
• SMM imported copper concentrate index broke −$60/dmt for the first time on Mar 13 2026 (close: −$60.39/dmt), continuing a slide from −$45 in early January.
• CSPT (China Smelter Purchase Team) formally agreed to cut 2026 refined output by more than 10% via longer/more frequent maintenance windows.
• LME copper stocks rose to 359k t (Mar 24) and SHFE stocks at 306k t (Mar 19) alongside COMEX ~415k st — total visible inventories above 1 Mt, a multi-year high.
• Yangshan premium collapsed to ~$34/t in mid-March, signaling no Chinese appetite for refined imports.
• Freeport-McMoRan announced a Grasberg phased restart from late March 2026 after the September 2025 force majeure; 2026 guidance cut to ~318 kt (from 1.7 → 1.1 → 0.7 Bn lb).