Coffee Prices Erase Early Gains as the Dollar Rallies

March arabica futures fell $3.10 (-0.83%) and March robusta lost $11 (-0.28%) after a dollar-index rally to a four-week high triggered long liquidation. Prices have been supported by below-average rainfall in Brazil (Minas Gerais 47.9 mm, 67% of normal) and tightening ICE arabica stocks (as low as 398,645 bags), while bearish pressure stems from surging Vietnamese exports (+17.5% y/y to 1.58 MMT) and larger global supply forecasts (USDA FAS projects world coffee up 2.0% to 178.848m bags with robusta +10.9%). Mixed signals from inventories, weather and trade/tariff dynamics leave coffee markets volatile and sensitive to near-term flow and FX moves.

Analysis

Market structure: Weather-driven arabica tightness (Minas Gerais receiving ~67% of normal rain last week) plus sub-2-year ICE arabica inventories create asymmetric upside for arabica vs robusta, while Vietnam export growth (+17.5% y/y in 2025) and FAS +10.9% robusta supply forecast compress robusta pricing power. Winners: Brazilian growers and premium-grade arabica longs; losers: robusta-heavy processors and spot market short-speculators. Exchanges (ICE/NDAQ) benefit from volume/volatility but face directional flow swings tied to DXY moves.

Risk assessment: Near-term (days) price moves will be dominated by USD strength (DXY breakouts) and weekly inventory prints; expect 10–20% intramonth swings if Brazil sees further dry spells or Vietnam logistics hiccups. Tail risks: a sharp Brazil frost or renewed US tariff disruption could double price moves; conversely sustained favorable weather + record global production (USDA projects +2% to 178.8M bags) could push prices 15–25% lower over quarters. Hidden dependence: coffee prices are tightly FX-sensitive (BRL/VND) and exposed to shipping/container availability.

Trade implications: Tactical long bias to arabica vs short robusta is justified for weeks–months, using futures or options to control risk; volatility spikes make call spreads (bull) and put spreads (bear on robusta) efficient. Cross-asset: expect modest negative correlation with USD and slight risk-on link to EM debt spreads if large exporter FX flows reverse. Key catalysts: weekly Somar/Conab updates, monthly Vietnam export figures, ICE inventory releases, next USDA supply report.

Contrarian view: The market may be underestimating arabica tightness because ICE inventories recovered only from a 1.75-year low to a 2.5-month high—still lean relative to multi-year averages; robusta oversupply narrative ignores potential export/logistics bottlenecks from Vietnam that can transiently tighten front-months. Reaction to a single DXY spike is likely overdone; use spreads to capture mean reversion. Historical parallels: 2013–2014 weather-driven arabica squeezes show rapid 20–40% rallies on short covering, so size accordingly.

Key Decisions for Investors

• Establish a 2% notional long position in March arabica futures (KCH26) or equivalent exposure via JO ETN, sized to portfolio notional; hedge with a 60-day bull call spread (pay one near-the-money call, sell one 7–10% OTM) to cap premium; set tactical stop-loss at -12% and take-profit tranche at +25% or if Conab/USDA materially increases arabica output.
• Implement a relative-value pair: long KCH26 (2% notional) vs short RMH26 (1.5% notional) for 3–6 months to capture structural arabica tightness vs robusta oversupply; rebalance monthly and unwind if the spread narrows below historical mean by >75% (or if FAS revisions reduce arabica decline forecast).
• Buy 90-day puts on RMH26 sized to 1% portfolio (or buy a put spread to limit cost) to profit from continued Vietnamese export pressure; sell 30–45 day call spreads against this position if premium is rich to achieve net-zero/credit entry when IV spikes above 30% implied.
• Reduce non-core exposure to broad agricultural commodity ETFs by 1–3% and rotate into selective coffee exposure (as above) for 1–3 quarter horizon; monitor DXY—if it sustains above 105 for >5 trading days, pause gross long additions and favor option-defined risk structures.