GSX Coffee Market Update

Any market or trading views are in no way are to be considered investment advice.

December Coffee Market Review

ICE NY Arabica (Mar) saw a 42.4c/lb high-to-low decline as we moved into Christmas, falling from 381.45c on Dec 1 to 339.05c on Dec 22. The move appeared driven by net spec liquidation and CAM selling in typically light holiday volumes.

Contract flow and spreads:

The Dec 2025 contract, despite minimal open interest, held its Dec/Mar premium.
Three major stoppers, each initially near a ~500-lot max position limit, ultimately stopped only about half of that starting size, even though visible NY certified stocks were sufficient to meet delivery.
Working assumption: the longs likely had a commercial need for coffee, while the shorts were a mix of underpriced roasters and trade house arb shorts that overstayed. Roasters needing to price during the notice period continues to underscore a hand-to-mouth approach.
During this three-week flat price decline, the Mar/May inversion softened by roughly 6% of flat price, from about 17c to just under 15c.

Physical balance sheet updates:

ECOM maintained its 2026/27 Brazil crop estimate at 75.8m bags (48.0 arabica, 27.8 conilon).
ECOM marginally reduced global S&D surpluses:
Marex’s latest 2026/27 Brazil estimate: 73.8m bags (48.5 arabica, 25.3 conilon).

Additional observations:

Brazil trade reports point to decent to good farm conditions. However, Sep to Dec 2025 SdM and ZdM precipitation totals are underperforming versus 5 to 10-year history, running roughly 60% to 80% of normal. Is that sufficient? Early yield feedback from upcoming 2H January crop tours will be important.
ICE NY certified stocks rose +46.2k bags in December. Grading activity improved, with 98k bags passing, about 80% from Brazil and Peru. Tariff elimination should lessen demand for certs.
ICE rule changes for tenderable origin levels versus Mar 2026 delivery:
January index rebalance and reweighting looks minimal: net purchases around +1% of open interest.

Outlook At the end of November, we expected a shift from Brazil dominating supply-side flow to a broader mix of origin selling, alongside a negative shift in arabica paper balance sheet flow. December’s price action reflected some of that, but we still view the market tactically rather than with a near-term trend call.

NY increasingly behaves like two separate markets:

2025/26: broadly balanced, within +/-1%
2026/27: shifting back to an approximate 5% surplus after 4, possibly 5, years of deficits

The remaining 2025/26 paper market, particularly Mar 2026 to Jul 2026, may stay choppy.

Variables we already know:

Destination stocks remain historically low.
Roasters’ priced position improved from November, but pricing remains hand to mouth.
Brazil farmers are well capitalized and, with the 2025/26 crop well sold, should stay disciplined scale-up sellers.
Fine cup naturals that could blend into certification are offered begrudgingly, and we do not expect that to change.
Large trade houses remain heavily short differentials in major origins. Credit availability has improved versus Dec 2024 and Jan 2025, which should support a higher differential floor as shorts cover.
Mar/May may have stoppers positioned. Unless there is another physical sale of certs off Mar 2026, the Mar/May upside potential may be pre-FND.
HG Honduran differentials near -10/46 versus March look close to tenderable parity if logistics align in time, but levels remain premium versus May.

Longer-term perspective The trade continues to expect the balance sheet transition to surplus to translate into significantly lower futures prices. A functional futures market should anticipate that transition. Tariff elimination has also enabled US roasters to plan forward again, yet the first two nearby months still represent roughly 70% of open interest.

The timing of the shift from balance to surplus, and the bearish impact on paper, is likely to be driven by specs more than origin selling. The Mar 2026 to Mar 2027 spread ended Dec 2025 at 10.77%, down from the 15% area, but not low enough (yet) to pull non-commercials into meaningful net short positioning. We are watching the forward curve closely for the spec selling inflection point.