Coffee is not a single price. It is a bundle of them — green beans traded on exchanges, roasted bags on grocery shelves, Arabica versus the cheaper Robusta, farm-gate prices in Vietnam versus futures in New York. Any honest answer to "what happened to the price of coffee" has to say which price, in which market, and over what period.

The most authoritative single yardstick is the International Coffee Organization's Composite Indicator Price, a blend of Arabica and Robusta reported in US cents per pound. ² Alongside it sit government forecasters (the USDA's Foreign Agricultural Service), data platforms (Trading Economics), national statistics agencies (the US Bureau of Labor Statistics), and a crowd of trade firms and brokerages with commercial stakes in the market.

The backdrop to this investigation's window is a genuine price spike that peaked before it began. Trading Economics records an all-time high of 440.85 in February 2025 — a level driven by earlier weather and crop-cycle concerns. ¹ The question here is what happened afterwards, from August 2025 to August 2026, and why the answer depends heavily on which price you track.

Start with the raw bean, where the numbers converge and the direction is clear. The ICO composite averaged 297.05 cents per pound in August 2025, itself up 14.6% on the month, and green coffee was running 24.3% above a year earlier. ¹ Prices climbed further into late 2025 — the composite reached 330.44 cents in November — and then fell. ¹ By 9 June 2026 the composite had dropped to 231.96 cents, its lowest in nearly two years. ¹ The USDA, drawing on the same ICO index, reported a 25% fall over the seven months to July 2026. ¹ The analyst's arithmetic confirms the November-to-June move works out to about 29.8%, a steeper fall than the USDA's 25% because the USDA measures a seven-month span rather than peak to trough. ²

By early August 2026 the commodity had bounced but not recovered. Trading Economics reported 325.14 cents per pound on 4 August, still down 7.09% over the prior month. ¹ Two other origins independently placed Arabica December futures at 314.65 cents around 1 August. ¹ A roughly 10-cent gap over three days is small relative to the price level, so the market sat around 315–325 cents — well below the November peak but above the June trough. Compared with August 2025's 297-cent composite, the early-August 2026 snapshot looks superficially flat or slightly higher; the crash-and-partial-recovery in between is the real story.

The retail shelf tells almost the opposite tale. US roasted ground coffee reached $9.72 per pound in April 2026 — the highest since the Bureau of Labor Statistics began tracking it in 1980. ¹ The BLS coffee index rose 18.5% year-on-year, with roasted up 17.3% and instant up 22.8%. ¹ The analyst notes these are not comparable to the commodity: retail prices bundle roasting, packaging, distribution and margin, and they track green-bean moves with a lag and a markup. So both facts are true at once. The bean got cheaper through mid-2026; the packaged product on the shelf did not.

The most extreme figures in circulation belong to neither layer cleanly, and they come from the single unverified source in the record. The retail brokerage Just2Trade reports an Arabica record of $4.3795 per pound in October 2025, a 37.8% crash to $2.3885 by 9 June 2026, and a 49.5% surge to $3.57 by 6 July 2026, which it calls the largest weekly gain since 2000. ¹ Two of those three — the October record and the weekly-gain claim — are unverified, resting solely on a single unverified brokerage source with a commercial stake in the market. Their June low, however, does line up: at roughly 238.85 cents, it sits within about 3% (roughly 7 cents) of the ICO's confirmed 231.96-cent composite on the very same date, which corroborates the timing of the trough even if not the exact Arabica figure. ²

Open: What was the Robusta price in August 2026 from a primary authority, rather than a single trade relay of $3,775 per ton [C-011]?; Does an independent exchange record confirm or refute the claimed October 2025 Arabica peak of $4.3795 per pound [C-001]?

The clearest engine behind coffee's price arc is Brazil, the world's largest producer, and the way its output lurches between "off" and "on" years in a well-documented biennial rhythm. In the off-year 2025/26 season, the trade firm Efico put Brazil's total harvest at 62.3 million bags, down 5.3% year-on-year, with Arabica falling 18.4% to 36.5 million bags from 44.7 million the season before. ¹ The analyst's check confirms that 18.4% figure is internally consistent. ¹ These are single-source trade estimates, and other bodies disagree on the exact total — the USDA cited around 63 million bags and Brazil's own Conab as low as 56.54 million, a spread of roughly 11%. The divergence reflects different methodologies and estimate timings rather than a dispute over direction: all three agree the off-year dipped. ¹

Then the cycle turned. The USDA forecasts Brazil's 2026/27 total at 71.9 million bags, a 14% jump, with Arabica alone surging 25% to 47.5 million bags. ¹ The implied off-year base — 71.9 divided by 1.14, about 63 million — lines up closely with Efico's and USDA's 2025/26 numbers, so the on-year rebound story is coherent. ² Globally, the USDA projects a record 2026/27 world crop of 189.7 million bags, 10.8 million more than the prior year, a figure relayed by multiple platforms tracing to that one primary forecast. ¹

Not every origin recovered. Colombia's production fell 33.5% in the first quarter of 2026, to 2.51 million bags from 3.78 million a year earlier, according to the growers' federation Fedecafé. ¹ Vietnam, by contrast, was forecast to edge up 2.5% to 32.5 million bags in 2026/27. ¹ The dominant fundamental, though, was the Brazilian rebound feeding a record global total — a supply picture pointing down for prices even as individual origins struggled.

Open: What did Brazil's final 2025/26 Conab outturn confirm, and how much of the Arabica shortfall was weather-driven versus cyclical [C-018][C-019]?; Did Colombia's Q1 2026 collapse persist through mid-2026, or was it a seasonal artifact [C-025]?

Trade policy is where the demand-and-market side of this story is best documented. On 6 August 2025 the United States imposed a 50% tariff on Brazilian coffee imports — a fact confirmed by four independent origins — Efico, Global Coffee Report, the ICO, and the Brazilian soluble-coffee association ABICS. ¹ The tariff landed exactly as prices spiked: Efico reported Arabica futures leaping from $2.80 to $3.74 per pound into August 2025, and the ICO composite jumping 14.6% that month. ² The timing is suggestive, though the evidence here establishes coincidence rather than proven causation.

Several frictions added upward pressure. StoneX reported Shanghai–Rotterdam container freight surging 23% week-on-week to about $3,850 per twenty-foot unit in April 2026, the highest since October 2025, a real but relatively modest slice of landed cost. ¹ The same firm relayed Cecafé data showing Brazilian exports down more than 20% year-on-year in the first quarter of 2026. ¹ And ICE-certified exchange stocks fell to a two-and-a-half-year low of 264,179 bags by early August 2026, a tightness corroborated in trend by a second origin. ¹ Both the freight and export figures rest on a single trade source and are graded as reported.

On the demand side, consumption held up. The USDA expects global consumption to reach a record 179.7 million bags in 2026/27, with the biggest gains in the EU and the US. ¹ In the US, a National Coffee Association survey found 85% of past-day coffee drinkers consumed it at home in April 2026, the highest share since 2012 — a shift toward the kitchen that tracks the record retail prices squeezing café habits. ¹

The crucial point is that these pressures pushed against a bigger force. The trade press emphasised tariffs, the Colombian shortfall and freight, but the market was looking past current tightness to the record harvest ahead. ¹ That is the tension at the heart of why futures fell while stocks stayed thin.

Open: What role did the Brazilian Real's exchange rate against the dollar play, given multiple sources reference it but no claim quantifies it?; Did speculative fund positioning, as opposed to fundamentals, drive the sharp June–July 2026 reversal — a question no CFTC positioning data in the record can answer?

The headline record prices hide a segment that barely moved. While mainstream US roasted coffee jumped 17.3% year-on-year, the Specialty Coffee Association's survey showed specialty retail up just 0.8% — and the highest-priced specialty coffees actually fell 3.4%. ¹ The analyst offers the likely mechanism: specialty roasters often lock in forward contracts and lean on brand positioning, insulating their shelf prices from short-term green-bean swings in a way commodity-grade blends cannot. This is the single clearest piece of evidence that "the price of coffee" did not uniformly rise.

Export prices in the world's second-largest producer also softened. Daily Coffee News, relaying USDA data, reported Vietnam's average export price at $5,127 per metric ton in the first half of 2025/26, down 9% year-on-year, and in March 2026 at $4,553 per ton, down 22% from March 2025. ¹ Both figures are single relays graded as reported, but they point to falling, not rising, prices at the Vietnamese export gate. These declines are Robusta-heavy export-gate prices measured against a year-earlier peak base and cover different periods and products than the ICO composite's August 2025 year-on-year gain, so the apparent contradiction reflects different market layers rather than a data conflict. ¹

Within Brazil, the picture fractured by product. ABICS, citing official IPCA inflation data, reported that ground coffee (café moído) rose a cumulative 75.25% across 2024–2025 — but soluble coffee rose only 34.32% and the cheap street cafezinho just 24.22%. ¹ The same commodity shock, filtered through different products and contract structures, produced wildly different consumer outcomes. The evidence for this counter-trend is thin — mostly single sources, and none showing an entire national retail market falling in absolute terms — but it is enough to reject any claim that coffee prices rose everywhere.

Open: Were there any European or Asian retail markets where coffee prices fell in absolute terms over the window, given no such data appears in the record?; What contract and hedging practices explain specialty coffee's near-flat pricing, beyond the general mechanism inferred [C-032]?

Why did the commodity price behave as it did? The strongest explanation is the biennial supply story. One reading holds that prices ran high through mid-2025 because Brazil was in its off-year, with Arabica output down 18.4%, reinforced by August 2025's price jump. ² The analyst grades this as supported. Its natural sequel explains the fall: from November 2025's peak the composite dropped roughly 30% to the June low as the market priced in a record 2026/27 harvest led by Brazil's 14% rebound. ³ These two phases — off-year tightness giving way to on-year glut expectations — account for the arc's shape without needing anything exotic. What would sharpen the case is Brazilian meteorological data and CFTC positioning reports, neither of which is in the record.

A second explanation puts the August 2025 spike on the US tariff, which coincided precisely with the price surge. ¹ The analyst rates this supported as well, but the two are not rivals — the tariff plausibly amplified an already-tight off-year market rather than substituting for it. Distinguishing them would require US import-volume data showing how far the tariff redirected trade flows.

A third reading credits speculation. The wild June-to-July 2026 swings — a claimed 37.8% drop then a 49.5% surge within weeks — look faster than crop or inventory data can justify, hinting at fund-driven volatility. ¹ But this rests on figures from a single unverified brokerage source, and one is contradicted by ICE stocks that stayed at multi-year lows rather than swinging. ¹ The analyst grades speculation only plausible, and the missing CFTC positioning data means it cannot be confirmed or dismissed.

Finally, the retail layer needs its own explanation. Green coffee rose 24.3% year-on-year into August 2025 and retail rose 17.3%, yet specialty barely moved at 0.8%. ² The analyst's supported hypothesis is segmented pass-through: forward contracts and brand positioning let specialty roasters absorb what commodity blends could not. This resolves the apparent paradox of falling futures and record shelf prices — they are different markets moving on different clocks.

The evidence forces one firm conclusion at the commodity level: coffee's green-bean price did not rise across this window — it fell substantially. This rests on CONFIRMED primary data, not interpretation. The ICO composite dropped to a near two-year low on 9 June 2026, the USDA independently measured a 25% seven-month fall, and the November-peak-to-June-trough move works out arithmetically to about 29.8%. ² Anyone answering the central question about the raw commodity must say: down, then partly back.

Equally forced is the retail counterpoint. US roasted coffee hit a verified record $9.72 per pound in April 2026, up 17.3% year-on-year. ¹ This is not contradicted by the commodity fall; it is the expected lag and stickiness of packaged goods. The honest verdict is that the two layers diverged — and both statements are simultaneously true. That is why the basis for the central answer is MIXED rather than a clean rise or fall.

On causation, the evidence points most strongly to Brazil's biennial cycle and the anticipated record 2026/27 harvest as the dominant driver of the commodity decline. ² The tariff and the Colombian collapse were real upward pressures that the coming glut ultimately overwhelmed. ² This is a supported conclusion, stated with the confidence primary data allows.

Two cautions are labelled speculative. First, the possibility — reasoning shown — that speculative fund flows drove the extreme June–July swings rather than fundamentals; this is plausible but unprovable on the record, and its supporting figures are unverified. ¹ Second, that the tariff independently caused the August 2025 spike; the timing is striking, but coincidence with an off-year peak is not proof, and no import-flow data isolates the effect. Where the evidence stalemates — the exact commodity peak, the true magnitude of any weekly gain — the report leaves it unresolved rather than adjudicating from a single unverified source.