Futures Correct, but Physical Supply Risks Are Rebuilding (11 September 2026)
The coffee market is entering the second half of September with an increasingly pronounced disconnect between futures pricing and conditions in the physical market. New York arabica has remained under pressure as speculative selling, expectations for improved Brazilian supply and the prospect of fresh coffee entering ICE warehouses have encouraged a more defensive positioning. December arabica settled at 285.70 cents/lb on September 11, losing 3.35% over the shortened trading week. London robusta behaved differently, with November closing at $3,525 per tonne and remaining comparatively resilient as nearby Vietnamese supply stayed tight.
This divergence is important because arabica and robusta are currently responding to different parts of the supply story. New York is increasingly trading the possibility of future supply relief, while London continues to reflect restricted immediate availability. Beneath both markets, however, weather risk is becoming more relevant again. The result is not a straightforward bullish or bearish setup. Coffee is moving into a phase in which the timing, quality and commercial availability of supply may matter just as much as headline production estimates.
Brazil: Rain Is Becoming a Quality and Availability Issue
Persistent rainfall across parts of Brazil's coffee belt has complicated the final stages of the 2026 harvest. Industry observations indicate that rainfall exceeded 100 mm in several areas of São Paulo and Minas Gerais during only the first eleven days of September, with approximately 120 mm reported around Garça, 105 mm around Caconde and similarly high accumulations across parts of southern Minas Gerais and the Cerrado.
The production impact remains difficult to quantify precisely, but the nature of the risk is becoming clearer. Rain during crop development can be beneficial, yet repeated precipitation during the closing phase of harvest has a very different effect. It can delay picking, complicate mechanical harvesting, increase cherry drop and reduce the proportion of coffee that can be dried under optimal conditions. Coffee that remains on the tree for too long, or falls onto wet ground, is also more exposed to quality deterioration. For that reason, the market increasingly needs to distinguish between total Brazilian production and the volume of commercially usable coffee that ultimately meets exporter, roaster and exchange specifications.
The September rainfall also comes after wetter conditions during June, July and August in several producing regions, making cumulative exposure more significant than any single weather event. At the same time, early rains have triggered flowering in parts of the coffee belt, encouraging discussion of a potentially large 2027 crop. That interpretation should be treated cautiously. Flowering is only the first stage of the production cycle, and a strong bloom does not automatically translate into a large harvest. Fruit setting, follow-up rainfall, temperature conditions and subsequent moisture availability will determine whether current flowering develops into meaningful production potential.
The market is therefore attempting to price the possibility of a large crop more than a year in advance while the current harvest is still facing weather-related collection and quality risks. That creates a considerable forecasting gap between visible supply today and expected supply further forward.
Record Exports Do Not Necessarily Mean Comfortable Supply
Brazilian exports provide another example of why headline data require context. Cecafé reported that Brazil shipped more than 4.15 million 60-kg bags of coffee in August, up sharply from the previous year and representing an exceptionally strong monthly export performance. Green coffee shipments reached approximately 3.82 million bags, including around 2.87 million bags of arabica and close to 954,000 bags of canephora coffee.
At first glance, these figures support the bearish argument. Brazil has harvested a large crop and is clearly capable of supplying the international market. However, strong exports also represent strong physical draw. When shipments accelerate, coffee leaves origin inventories more quickly, and this matters particularly when carry-in stocks are not abundant. Record exports can therefore signal both improved current availability and faster depletion of that same availability.
September data reinforce this tension. By September 11, shipments had reached roughly 710,000 bags, while certificate requests for future export were materially higher. The issue is therefore not whether Brazil can export coffee; it can. The more relevant question is how much readily available inventory remains once domestic consumption, export commitments, producer retention and quality segregation are taken into account.
Producers Are Not Selling Aggressively Into the Decline
Conditions in Brazil's physical market add another layer to the analysis. Despite the arrival of the new crop, industry reports suggest that producer selling has remained relatively restrained following the decline in New York. Buyer interest remains present across several quality categories, but many growers appear unwilling to accelerate sales at lower terminal prices and are instead marketing only the volumes needed to meet immediate financial commitments associated with the end of harvest and crop processing.
This matters because a large theoretical crop does not automatically translate into immediate pressure on the physical market. If producers have sufficient financial capacity to retain coffee, supply can remain relatively tight even when headline production is high. When futures fall more quickly than producers adjust their price expectations, physical liquidity can contract and differentials can remain firm despite weakness in the terminal market.
For the bearish case to strengthen materially, the market therefore needs more than a large crop. It also needs producers to become more willing sellers. At present, those two conditions do not appear to be fully aligned.
ICE Stocks Remain the Critical Bearish Counterweight
The clearest potential source of relief for the arabica market is the effort to rebuild ICE certified inventories. Stocks fell to roughly 218,000 bags by September 11, leaving the exchange with one of its smallest certified buffers in decades. Such low inventory provides little protection against disruptions in physical supply and has been one of the strongest structural supports for New York arabica.
That situation could now begin to change. Large trading companies are moving Brazilian arabica toward exchange warehouses for certification, and market reports indicate that significant volumes may be prepared for potential delivery against the December contract. If a meaningful proportion of those arrivals successfully passes grading and becomes certified, exchange stocks could recover sharply from current extreme lows.
This is potentially the most important bearish development in the near-term arabica balance because rebuilding certified stocks would reduce the scarcity premium associated with immediate deliverability. It could also alter speculative behaviour by weakening the argument that the exchange is operating with an exceptionally thin buffer. However, there is an important distinction between coffee arriving at warehouses and coffee becoming certified stock. Until grading is completed, the full increase remains potential rather than confirmed supply.
The pace of certification over the coming weeks may therefore matter more for New York than the headline size of the Brazilian crop itself.
Vietnam: Tight Stocks Keep Robusta Supported
The physical situation in Vietnam remains different. Domestic coffee prices were trading around VND 95,000–95,500/kg on September 12, supported by limited old-crop availability ahead of the next harvest. Farmer and commercial stocks appear relatively low, restricting the volume available to exporters and helping explain why London robusta has remained firmer than arabica despite broader pressure across the coffee complex.
Weather is now adding another variable. A tropical depression developed near the Hoang Sa area and moved toward central Vietnam, bringing substantial rainfall and rough maritime conditions. The immediate implications for coffee should be interpreted carefully because the core Central Highlands producing areas are not necessarily at the centre of the system. It would therefore be premature to treat the tropical depression itself as a major coffee production event.
The broader issue is the persistence of wet conditions. If excessive rainfall extends into coffee-growing areas as the crop approaches harvest, it can increase cherry drop, complicate field access and slow drying. On the other hand, adequate moisture ahead of harvest can still support bean development. The market risk therefore depends less on one individual storm and more on whether wet conditions persist into a more sensitive phase of the harvest cycle.
With old-crop inventories already restricted, even moderate disruption to harvesting, logistics or quality could have a disproportionate effect on nearby robusta pricing.
El Niño Raises the Tail Risk
The broader climatic backdrop is also becoming increasingly relevant. Forecasts point to a strengthening El Niño through the final quarter of 2026, raising the probability of more pronounced weather anomalies across major producing regions. This should not be interpreted as a deterministic production forecast. El Niño changes the probability distribution of rainfall and temperature outcomes; it does not guarantee drought, excessive rain or crop losses in a specific coffee region.
Its importance lies instead in the interaction between weather variability and already limited inventories. In a market with abundant carryover stocks, temporary weather disruption can often be absorbed without major price consequences. When certified arabica inventories are near multi-decade lows and Vietnamese old-crop availability is already restricted, the same weather event has greater potential to affect price.
The market therefore enters the coming months with reduced tolerance for production or logistical shocks.
