Between harvests, Arabica basis takes control

01

ICE Benchmark: use Coffee C, but price the contract month

ICE Coffee C is the global Arabica benchmark and the starting point for this report. The conventional contract is quoted in cents per pound, with March, May, July, September, and December delivery months. Q3 buyers therefore need to identify which futures month truly matches their shipment and pricing period. A headline “Coffee C price” without the month can conceal spread and roll risk.

The latest complete ICO month available at publication is June: New York Arabica averaged 256.75 US cents per pound, down 4.3% from May. That is a historical anchor, not a live Q3 quotation. The purpose of the number is to establish the market entering the quarter; executable decisions should use current contract data and a physical offer with the same timestamp.

Q3 basis can firm even if Coffee C rallies, soften, or trade sideways. If an exporter sells approved inventory but cannot replace the same cup until the next northern Sumatra harvest, the replacement differential rises. If sellers carry ample commercial inventory and need turnover, basis can soften even when the screen is volatile. The screen and the differential answer different questions.

Certified Arabica stocks near 0.41 million bags at the end of June also argue against complacency about global buffer availability. The inventory is not Indonesian supply and does not dictate an origin premium, yet thin certified stocks can make Coffee C more reactive. That raises the value of separating hedge decisions from physical allocation decisions.

02

Indonesia Fundamentals: six origins, three supply conditions

USDA estimates Indonesian Arabica production near 1.45 million 60-kilogram bags for 2025/26. Within that total, Q3 can be grouped into three practical conditions. First, Gayo and Mandheling are moving away from their first harvest and toward the second. Second, Java and Kintamani programs are working through their own seasonal availability. Third, Toraja and Flores remain attractive but often smaller, more fragmented, and logistics-sensitive.

For northern Sumatra, the core question is how much approved coffee remains after Q2 commitments. Visible warehouse stock is not identical to saleable stock. Coffee may already be allocated, awaiting final preparation, outside the buyer’s cup tolerance, or held for a different shipment. Replacement basis should be calculated against a like-for-like lot, not any green coffee in the building.

Java and Kintamani can diversify both cup and calendar. Washed preparation, estate or group identity, and cleaner sensory profiles can serve buyers who do not need the traditional wet-hulled expression. Toraja and Flores offer distinct character and story but require realistic assumptions about aggregation and movement from origin. Their commercial value increases when buyers plan lot size and shipment early rather than asking them to behave like a large northern Sumatra pool.

Specialty Arabica remains a qualification business. Commercial Arabica can be assembled around objective grade and cup tolerances, while specialty programs add identity, process, sensory target, and often certification or documentation. Q3 scarcity will appear first in the narrowest specifications.

03

Macro → Micro: currency and finance magnify replacement decisions

Bank Indonesia indicators in July placed the policy rate at 5.75%, June consumer inflation at 3.34% year on year, and JISDOR at roughly Rp17,909 per US dollar on 22 July. These readings are snapshots, not fixed Q3 assumptions. Their relevance is the path from a dollar-denominated Arabica market to rupiah working capital and farmer purchasing power.

A weaker rupiah can make a stable dollar offer more valuable locally, supporting farmer and collector bids. It can also raise dollar-linked logistics and imported input costs. Meanwhile, elevated coffee values increase the nominal amount financed per container. Even an unchanged interest rate becomes more expensive in absolute terms when inventory value rises.

At origin, microeconomics decides whether the global signal reaches the producer. Collectors compete on cash, location, quality tolerance, and speed. Farmers decide whether to sell cherry, parchment, or dried bean and how much labor to invest in selection. Specialty qualification can add revenue, but only if the premium exceeds additional picking, processing, separation, and rejection costs.

04

Origin Yield: read productivity with qualification rate

BPS 2024 dried-bean productivity for smallholders provides a public province-level baseline. The numbers include provincial coffee and are not a species-pure Arabica census. They are planning proxies rather than forecasts or farm guarantees. Good commercial analysis combines them with tree age, weather, productive area, process, and the share of output that meets the contract.

A useful farmer dashboard tracks productive hectares × saleable yield × realized price, then subtracts labor, inputs, processing, logistics, finance, and quality loss.

Productivity and premium can compensate for one another but are not substitutes. A high-yield plot that produces inconsistent cup may fail a specialty program. A low-yield plot with an exceptional profile may still struggle if the premium does not cover labor and agronomic renewal. Buyers and suppliers create a healthier chain when premiums reward verifiable work and agronomy improves saleable yield.

05

The Q3 Arabica path, month by month

July: inventory quality becomes visible. The market enters the quarter with Q2 crop in warehouses and commitments. Buyers should ask how much coffee is unallocated, what proportion has passed cup approval, and whether the same profile can be replaced. A broad stock number can overstate physical flexibility.

August: the bridge can tighten. Northern Sumatra is further from its first peak and not yet fully into its second. Commercial basis may remain balanced where specifications are broad and inventory is adequate. Narrow specialty profiles, certified programs, and prompt shipments face greater risk of firmness. Java, Kintamani, Toraja, and Flores can provide alternatives, but not as silent substitutions; the cup and supply model are different.

September: second-harvest expectations enter pricing. Early indications from flowering, cherry development, and farmer behavior will influence forward offers. The market may price confidence before significant export-ready volume exists. Buyers should distinguish a forecast of future cherry from an executable shipment of prepared Arabica.

06

Commercial versus specialty: different Q3 bottlenecks

Commercial Arabica bottlenecks are usually quantity, grade, moisture, defects, and shipment. Buyers can improve execution by defining acceptable ranges and giving suppliers room to assemble efficiently. Specialty bottlenecks are narrower: consistent cup, traceability, process, certification, and lot identity. The more attributes a program fixes, the smaller its replacement pool becomes.

The correct comparison is landed utility, not the smallest differential. A slightly higher basis can be cheaper if it reduces rejection, blending correction, delay, or quality variance. Conversely, a specialty label adds little value unless the lot delivers the promised sensory and documentary difference.

07

Q3 scenarios and indicators

Base case: Coffee C stays volatile, commercial Indonesian Arabica basis is broadly balanced, and scarce specialty basis firms through the inter-harvest bridge. Tighter case: approved stock is lower than expected, currency supports rupiah bids, logistics slow, or second-harvest expectations weaken. Softer case: Q2 carry is ample, sellers prioritize turnover, quality conversion is good, and early second-harvest confidence improves replacement offers.

• Monitor matched Coffee C month, spreads, and roll dates—not only the nearby headline.
• Measure approved and unallocated inventory by origin, grade, and cup.
• Track second-harvest indications in Gayo and Mandheling without treating cherry as exportable coffee.
• Use Java, Kintamani, Toraja, and Flores as intentional origin strategies, not anonymous substitutes.