Supply Risk
Tightening
Net score +3.0 · ≥ +2 tightening / ≤ −2 easing
+1.0
Concentration tightening
K exports near-monopolistic (HHI 0.70) — fragile supply base (understated — opaque Russia/Belarus add hidden concentration)
0.0
Availability neutral
swing-exporter shipments +37% vs seasonal normal — within the normal range for this month (z=+0.7, band ±1.0)
+2.0
Cost-push tightening
gas is 78% of the realizable urea price — marginal EU capacity at curtailment economics; margin $+12/t
This is the supply-side lens: not how exposed any one importer is, but whether the swing
exporters that move world supply are flowing or withholding. It blends how concentrated
each nutrient's exports are (few hands = fragile), whether the tap is flowing right now
(shipments vs normal), and whether the marginal producer is being squeezed out on cost.
Weigh the components, not just the headline.
Top concentration
HHI 0.70
K · top Canada
Supply availability
+37%
shipments vs seasonal normal
Gas cost share
78%
of realizable price · curtails >70%
EU marginal margin
$12
vs import parity
World urea
$400
Pink Sheet FOB $/t
Herfindahl-Hirschman index of each nutrient's world exports (1 = single-source monopoly; the dotted line at 0.45 marks near-monopoly). Computed over 2025 — the most recent year for which a full twelve months is on the record for every one of the 10 exporters counted, so all of them are measured over the same window. A part-year would not be comparable: an exporter that has not reported yet would simply drop out of the denominator and inflate everyone else's share. Potash is the most concentrated — and understated here, since Russia and Belarus are absent from the trade data (see note below). Source: UN Comtrade.
Potash is the most concentrated nutrient — K HHI
0.70, led by Canada — and this understates the true
figure, because Russia and Belarus stopped reporting export data after 2022. Their supply
is real but dark; the availability section below reads it through importer records instead. Nitrogen is far
more competitive (many gas-based exporters), so its risk lives in policy and cost, not concentration.
| Nutrient | Export HHI | Top exporters (2025) |
Exports (value) |
|---|
| Nitrogen (urea, 3102) | 0.22 | China 39%, Nigeria 15%, Saudi Arabia 13% | $14,496M |
| Phosphatic (SSP, 3103) | 0.42 | Morocco 61%, China 17%, Israel 11% | $3,804M |
| Potash (MOP, 3104) | 0.70 | Canada 82%, Germany 14%, Egypt 2% | $7,860M |
| DAP / NPK complexes (3105) | 0.27 | China 35%, Morocco 27%, Saudi Arabia 25% | $20,099M |
| Origin | Embedded (tCO₂e/t urea) | CBAM cost ($/t) |
|---|
| Qatar | 1.370 | $42 |
| Egypt — parity basis | 1.390 | $43 |
| Algeria | 1.410 | $45 |
| Nigeria | 1.460 | $49 |
| Russian Federation | 1.470 | $50 |
| Saudi Arabia | 1.470 | $50 |
| Trinidad and Tobago | 1.610 | $62 |
| United States | 2.290 | $120 |
| China | 2.850 | $168 |
Context, not a margin input. These costs are real and legally
operative since 2026-01-01, but they are deliberately EXCLUDED from the EU margin above: CBAM is zero
across every month we can validate against, so carrying it would put an unbacktestable ~$43/t inside a
published number. It is shown here in full instead. Official Commission default values (CN 3102 10 19,
direct + indirect — fertilizers carry indirect emissions), Annex I of IR (EU) 2025/2621 as corrected in
full by IR (EU) 2026/1740. Cost = (embedded − 0.902 benchmark ×
(1 − CBAM factor)) × EUA, so only emissions ABOVE the benchmark are charged — which is why a
2.5% factor in 2026 still bites hard on a coal-route origin. Two things the table shows that
a single number cannot: the gas-based suppliers sit in a narrow band, and China's coal route
is in a different category entirely — CBAM does not tax it more, it prices it out. Note the
declaration basis moves the charge MORE than the origin does: filing verified actuals instead
of accepting the default is worth about
$25/t on the parity origin, against roughly
$20/t across every gas-based
supplier.
Combined value of fertilizer the tracked swing exporters shipped to Brazil and the US, by month — the current 'is the tap flowing' pulse, and the only view that captures Russia and Belarus (importers still report buying from them). Below the seasonal normal = supply throttled. Source: Comex Stat + US Census, pivoted by origin.
UN Comtrade lags a year or more, so the current read comes from what the tracked exporters actually
shipped into two big buyers we track monthly — Brazil and the US. Right now those shipments are
+37%
versus the seasonal normal. This mirror is also how we see Russia and Belarus — they no
longer report exports, but their customers still report buying from them.
China's monthly fertilizer exports. The collapses are export quotas biting — Beijing throttles urea and DAP/MAP to protect domestic supply, then reopens. This is the policy state made measurable, without parsing any announcement. Customs publishes ~5 weeks after month-end, so the recent months are the current state of the tap, not just its history. Source: UN Comtrade through 2024-12, then China General Administration of Customs (stats.customs.gov.cn) from 2025-01. The two agree to within 0.04% on the 24 months where both publish, which is what licenses the join; the dotted line marks the handover.
China is the swing exporter to watch: its quota episodes show up as the export collapses above (the policy
state, made measurable). When Beijing throttles urea or DAP/MAP to protect domestic supply, the world
balance tightens — and the importers in our other three reports feel it first.
Where these numbers come from. China's UN Comtrade record ends at
2024-12. From 2025-01 the series continues on
China's own customs statistics (General Administration of Customs, stats.customs.gov.cn), which publish
roughly five weeks after month-end. The join is not an assumption: across the
24 months both sources cover, customs tonnage as a share of
Comtrade tonnage runs N 99.96% · P 99.98% · K 100.00% · NPK 100.00% — so the two records describe the same goods, and the recent months are the
live state of the export tap rather than a forecast of it. Every other exporter in this report remains on
Comtrade alone, which is why the concentration anchor above stays on a completed year.
Editorial overlay — the supply taps. The supply taps that move every importer. China is the swing N+P exporter: recurring urea and DAP/MAP export quotas (provisional windows from 2021; tight through 2023-25) — its export VOLUME collapsing is the quota biting, the report's clearest availability signal. The potash triangle is Canada (Nutrien/Mosaic, the price-stable giant) + Russia (Uralkali) + Belarus (Belaruskali, sanctioned since 2021 and transit-cut) — the latter two opaque post-2022. Morocco (OCP) holds a near-monopoly on phosphate rock and is the swing DAP/MAP exporter when China withholds. Russia is a top-3 exporter of all three nutrients but reports no Comtrade data post-2022, so its tap is read through partner imports + editorial.
The last tonnes of world nitrogen come from the high-cost European producer, built on TTF gas. Europe is a
net nitrogen importer, so that producer realizes import parity — the landed cost of the
marginal import — not the bare FOB benchmark: $400 FOB +
$40 freight + $43
CBAM = $440/t. Against an EU urea cash cost of
$428/t plus $2/t of carbon the
EU producer now pays itself — the phase-out that charges imports strips free allowances from domestic
plants too, so both sides of the trade carry it — its margin is
$12/t. A healthy margin keeps marginal capacity
running; an underwater one curtails it and tightens world supply — the same European curtailment engine the
Brazil report watches from the buyer's side, here read as a supply-side squeeze.
Who wins the gas?
UREA
spread $-1.17/MMBtu · 2026-09-09
Power can pay
$14.78
/MMBtu · core EU €127/MWh
Urea can pay
$15.95
/MMBtu · urea 2026-08
Crossover
€135
/MWh — power outbids urea above this
Both claims on the molecule, normalized to $/MMBtu of gas — the only basis on which power and urea compare. Where the red line is above the blue, gas-fired power can outbid urea production; the crossover sits near €135/MWh at the current urea price. Power uses the MARGINAL generating unit (45% CCGT), not the fleet average: when power spikes because supply is short, the marginal unit is the least efficient on the stack. Carbon charged at the EUA price. This is a CONFIRMATORY read, not a curtailment forecast — a power win has always coincided with curtailment, but curtailment often runs without one.
Electricity cannot be stored and blackouts are politically intolerable; urea can be stored and its
production is deferrable. So wherever gas is contested, fertilizer is the shock absorber — in administered
systems because a ministry says so, and in market-priced Europe because power outbids it. Pricing both
claims per MMBtu of gas: European power is not currently outbidding urea production for the marginal molecule, at
$14.78 against urea's $15.95 —
a spread of $-1.17/MMBtu. The crossover sits near
€135/MWh: above that the molecule is worth more in a turbine than in an
ammonia reactor, whatever the producer's own margin says.
Read this as confirmation, not forecast. Backtested against observed European ammonia
curtailment, a power win has always coincided with curtailment — no false positives in 139 months
— but curtailment frequently runs without one, because urea prices spike in response to the shortage the
curtailment itself creates. Presence is meaningful; absence is not. The curtailment signal is the gas cost
share above, not this.