Why the C market is whipsawing, and what governs the next move.
Every price in the trip report is market plus differential. The differential is what gets negotiated; the market is what moves underneath it, and in 2026 it has moved 188 cents. This page is the background to that number.
Download the full briefing (PDF)Arabica has not trended in 2026. It has oscillated across a range of 188 US cents per pound, roughly 77% of its own low, and has done so in both directions within a matter of weeks.
The volatility has two causes that are usually conflated. One is fundamental: a record Brazilian crop arriving late, into thin inventories. The other is mechanical: ICE raised margin requirements, liquidity contracted, funds closed positions, and the resulting illiquidity amplifies every move into a one-way run. Reading a 16% session as pure weather news misreads the market.
The sawtooth is the point: arabica made four distinct reversals of 30 cents or more in seven months. Points are individually sourced settlements, not a continuous series.
Hover or focus a settlement to see what moved the market that day.
| Date | KC ¢/lb | What happened |
|---|---|---|
| Jan 20, 2026 | 345 | Lowest since 19 Dec 2025. Brazil estimates unchanged at 70–80 m bags, weather favourable, positive biennial cycle. |
| Feb 04, 2026 | 314.5 | Lowest since Aug 2025. Down 12.31% over four weeks, 21.21% over twelve months. |
| Feb 15, 2026 | 278 | A 15-month low. Record Brazilian harvest forecasts plus a stronger US dollar. |
| Mar 05, 2026 | 281 | n/a |
| Mar 20, 2026 | 320 | Highest since early February. Cooxupé guided its own exports down 10% to ≈ 4.4 m bags; growers withheld sales. |
| Apr 23, 2026 | 302.4 | Up 4.64% on spot scarcity and rising freight, Cape of Good Hope rerouting adding weeks of transit. |
| Jun 01, 2026 | 260.6 | Lowest since Nov 2024. Rain eased, harvest expected to resume in Minas. Localised hail damage confirmed. |
| Jun 08, 2026 | 245.9 | Sixth consecutive losing session; a 20.5-month low. The floor of the current range. |
| Jul 06, 2026 | 349.95 | Up 48.75 cents: a 16.2% single-session gain and a 5.75-month high, on harvest delay and weather risk. |
| Jul 24, 2026 | 313.8 | Up 4.40 (1.42%) after a sharp Thursday sell-off, recovered on pre-weekend short covering. |
Values to April reference the then-nearest contract; June and July reference the July and September 2026 contracts. Roll effects account for a small part of the differences between adjacent points.
Three layers operating at once, and they compound. Volatility of this order is not explained by fundamentals alone.
ICE raised margin requirements for coffee futures. Higher margin raises the cost of carrying a position, so commodity funds cut exposure and liquidity dried up. In a thin book a given order size moves price much further than it would in a liquid one. This is the most underappreciated driver right now, and it means price moves currently carry less information about supply and demand than they normally would.
Brazilian arabica is in a positive biennial year after five seasons of adverse climate, rebounding off a low base: 2025/26 arabica had fallen 13.6% to ≈ 38.1 m bags. But harvest progress is the constraint: Brazil was 64% complete as of 15 July against 77% a year earlier and a 70% five-year average. Heavy June and early-July rain slowed picking and raised questions about bean quality.
El Niño is no longer a forecast: the Japan Meteorological Agency confirmed formation on 10 June, and NOAA puts a Super El Niño at 67% probability. The arabica-specific concern is timing: El Niño may delay Brazilian rains in September and October, exactly when arabica flowers. That affects the following crop, not the one being harvested. Minas Gerais has meanwhile flipped from too wet to too dry, at 20% of average rainfall in the week to 19 July.
The spread between CONAB's 66.2 m and private estimates near 76 m is wide by historical standards. Treat any single figure as one scenario.
| Forecaster | Total | Detail |
|---|---|---|
| CONAB | 66.2m bags | Of which 44.1 m arabica. The conservative anchor. |
| StoneX | 75.3m bags | Raised from 70.7 m in November: a 20.8% annual rise. |
| Sucafina | 75.4m bags | Record projection. |
| Eisa | 75.8m bags | Brokerage estimate. |
| Marex | 75.9m bags | A 15.5% year-on-year increase. |
Both sides are supported by current data. Illiquidity means whichever side receives fresh news will overshoot.
| Level or window | Value | Significance for arabica |
|---|---|---|
| Range floor | 244–246 | The 8 June low. Held once. A break below would signal the market accepting the surplus thesis in full. |
| Current | ≈ 314 | Roughly 29% above the floor and 27% below the 52-week high: near the mid-point of the range. |
| First resistance | 324–350 | The July whipsaw zone, capped by the 6 July high of 349.95. |
| Range ceiling | 432 | The 52-week high. Only a genuine supply shock, most plausibly frost, reopens this level. |
| Now to end Aug | Frost season | The single largest upside risk. Frost in Minas or São Paulo damages trees, not just one crop, and would reprice the curve within days. |
| Aug to Sep | Harvest completion | Confirms or refutes the record-crop thesis. Volume and cup quality both matter after the rain damage. |
| Sep to Feb 2027 | Flowering and rainfall | Sets the 2027/28 crop, which some project above 80 m bags. This is where El Niño would actually bite. |
| Ongoing | ICE margin | If margin requirements ease and liquidity returns, volatility should compress even with unchanged fundamentals. |
A 188-cent range means an unhedged fixed-price commitment can move from profitable to loss-making with no change in the underlying business. Your book was fixed at 246–265 and the market sits at 314 today: the same contracts, 20–21% more expensive, on market movement alone.
Even at the 8 June low of 245.90, arabica sat above its five-year average. Any plan built on a return to pre-2024 levels is working from an obsolete assumption: including any budget that still carries an old cost base.
Thin books widen spreads and worsen fills, so the screen price is a less reliable execution reference than usual. When we fix a differential matters more than it did twelve months ago.
Brazilian arabica producer prices in June were down 24–34% year-on-year: 1,520 reais per bag in Patrocínio, 1,470 in Guaxupé, 1,460 in Franca, the same Franca on this itinerary. Farmgate prices rise more slowly than retail and fall harder on the turn, so producers absorb volatility that traders and roasters are better placed to manage. The real appreciating 7.2% to ≈ 5.10 compresses local receipts further and discourages selling.
Relationship and differential-based contracts insulate specialty buyers from part of the swing, but the elevated floor lifts every cost base. Demand for traceable arabica keeps outgrowing supply, so the premium over commercial grades is widening rather than narrowing, which is the argument for the specialty layer, not against it.
Arabica is trading a wide range between roughly 245 and 350, with the 52-week extremes at 244 and 432. Fundamentals argue for gradual easing as the record Brazilian crop completes and arrives. Illiquidity and an open weather window argue that the path there will be violent rather than smooth. The frost season over the coming weeks is the most consequential near-term variable; El Niño's effect on September and October flowering is the most consequential medium-term one. Absent both, the balance points lower. With either, the upper end of the range returns quickly.
Prepared from public market sources. Informational only; not investment, hedging or trading advice.