GSX Coffee Market Update - January 2026

During early January, lingering concerns about Brazil’s elevated temperatures and suboptimal rainfall helped support the NY coffee market.

The January 7–8 rally, during which March 26 NY traded up to 382/383, collapsed differentials in Honduras and Nicaragua. The paper market encountered substantial volumes of physical coffee. Estimates suggest Honduras may have sold 0.5 million bags or more at differentials of -18/-20 cts/lb FOB against March. There was also gossip that 0.25 million bags could be shipped to Houston versus the March board. While considerable volume traded, the extent of the trade’s pre-existing short positions is difficult to assess. Previously, shorts have exaggerated estimates of coffee moving to the board. Inversions will do that.

As January progressed, Brazilian rain forecasts turned more favorable, indicating improved bean-filling conditions. Consensus estimates for Brazil’s 2026/27 crop remain 70–75 million bags, nearer 75. Recent field observations and photos indicate excellent crop health, with NDVI readings in Minas and São Paulo showing significant improvement.

Speculators were caught offside: buy high, sell low.

Brazilian farmers remained largely inactive through December, reluctant to incur additional tax liabilities. In January, differentials for fine cup MTGB naturals, often used in Brazil’s certified offerings, strengthened significantly into positive territory. Fine cup differentials firmed from -10/-15 last October to +10/+20 by late January. Farmers are reluctantly selling the remainder of their 2025/26 very good cup (VGC) and fine cup (FC) coffees. It may be premature, but absent a significant rally, we would expect fine cup differentials to remain at premiums until the new crop becomes available. This would then limit new Brazils to the board, unless an entity chose to certify at a negative margin.

Natural demand for NY-certs, used in decaf and blending, is estimated at 30,000–60,000 bags per month. In the absence of meaningful new Brazilian certs until possibly September 2026, questions arise as to whether Central American and African gradings will be sufficient to stabilize NY certified stocks in the 400,000–500,000 bg range over the next six months.

The May NY–London (LDN) arbitrage, for the fifth consecutive month, struggled to sustain moves toward $1.75–$1.80. The LDN Nov25 stopper appears to be accumulating additional robusta cert stocks in January. The arb narrowed nearly 50 cts from early January highs, with the NY–LDN ratio slipping to 1.72, a neutral range. On an NY basis, the ratio would appear undervalued between 1.2 and 1.4.

Demand remains a point of contention within the trade. Overall, North American and Brazilian demand appears flat to lower. ABIC estimates Brazilian consumption down 2.3% in 2024/25 (Nov/Oct). In contrast, Asian demand remains steady, with Luckin Coffee expanding aggressively.

Destination stocks remain quite low, unsurprising given that Brazil’s exports in January–December 2025 were 10 million bags below the prior calendar year. The market now anticipates a shift from multiple deficit years to a notable surplus. Brazil’s cost of production estimates naturally vary significantly by farm and region, but overall arabicas are around R$750/bg and conilons R$400–500/bg. Even if NY and LDN dropped significantly further, there should still be a positive return.

Technically, NY appears weak. Barring climatic issues, the highs seem established. A significant KC2 level would be the 270 lows of July/August 2025. But this transition from deficit to surplus may well occur in fits and starts. The KC2–KC7 spread holds a 7% inversion, while the new crop Sep 26–Sep 27 spread exceeds 6%. London’s one-year inversion is close to 10%. In our view, a break in the inversions to at least a flat board may be necessary to trigger major systems short selling.

The paper market encountered significant physical flows near $3.80 against March. At current levels, the paper is the cheapest form of the commodity. No significant origin is near tenderable parity.

Stepping back from the granularity of arbs, ratios, and inversion percentages: at what price levels will major roasters extend paper coverage in a significant fashion for the next 6–12 months? Not at current levels, but do they scale down if the price begins with a 2? Or do they wait for something truly discounted, $2.00–$2.50?