Supply-Chain Crossing: How El Niño 2026 and GR 24/2026 Move from Commodity Markets to the MBG Meal Tray

MBG Watch · 2026-07-17

The premise

Two forces are moving through Indonesia's food system at once. Neither is new, but their convergence in the second half of 2026 is.

First, a super El Niño is forming. NOAA's Climate Prediction Center puts the odds of El Niño persisting through winter 2026–27 at 96%, with a 63% chance it reaches "very strong" status. The Guardian reported on July 12 that economists warn this could cause "a severe shock to global food prices lasting into 2028." For Indonesia, the exposure is specific: palm oil yields show drought stress 6–12 months after the event; rice planting across the 114 seasonal zones is already delayed; robusta coffee and cocoa face the same dry window.

Second, Government Regulation 24/2026 (GR 24/2026), issued June 5, creates a state-controlled export chain (DSI — Danantara Sumberdaya Indonesia) for strategic commodities. The initial phase covers coal, palm oil, and ferroalloy. During the transition period (June 1 – December 31, 2026), exports must go through DSI. After December 31, exports can only be carried out by DSI. This structurally reallocates how much of each harvest stays domestic versus leaves the country.

Rupiah Stability Watch has traced both forces at the macro level: how El Niño disrupts commodity exports and tightens domestic supply, and how GR 24/2026 rewires the export channel. MBG Watch has traced what happens at the kitchen: how rupiah depreciation and food inflation squeeze the fixed Rp 10,000 per meal, how MBG's own procurement footprint moves staple prices, and how rural cold-chain gaps break the meal.

Neither organisation has traced the chain end-to-end: from climate + policy shock → commodity availability and price → SPPG kitchen procurement → what the Rp 10,000 actually buys on the tray → nutrition outcomes for children.

This publication does that trace. It is a monitoring-framework piece: it names what to watch, not a crisis declaration.

What the evidence supports

1. Palm oil: the fat pillar is squeezed from both sides

El Niño supply shock: The 2015–16 super El Niño cut Indonesian CPO output by roughly 10–15% year-on-year, with the yield impact appearing 6–12 months after the drought. The current event is forecast to be at least as strong. Fresh-fruit-bunch yields in Aceh, North Sumatra, Riau, West Sumatra, and Jambi — the core palm belt — are already reporting moisture deficits as of June 2026.

GR 24/2026 allocation shift: The regulation's transition period (June 1 – Dec 31) requires all CPO exports to flow through DSI. In practice, this gives the state a lever to prioritise domestic allocation over export volume. The July 2026 CPO reference price (HR) was set at US$1,000.90/MT, down 2.78% from June. But forward curves tell a different story: Bursa Malaysia CPO futures for August/September 2026 trade above RM4,600/MT, pricing in the El Niño yield loss and Indonesia's B50 biodiesel mandate (which alone claims an estimated 13–15 million tonnes of CPO domestically).

Domestic cooking oil price transmission: The Ministry of Trade's April 2026 survey recorded premium packaged cooking oil at Rp 21,237/litre — up from roughly Rp 18,000/litre a year earlier. The CPO-to-cooking-oil pass-through in Indonesia is well documented at 0.6–0.8 within 2–3 months. With the El Niño yield dip arriving in Q4 2026 / Q1 2027, and the B50 mandate structurally tightening the domestic balance sheet, the pass-through has upward bias.

MBG exposure: Cooking oil is the primary fat source in the MBG meal. BGN's nutrition standard (AKG) targets roughly 25–30% of meal calories from fat. At current cooking oil prices, the fat component alone costs an estimated Rp 1,200–1,500 per meal — 12–15% of the Rp 10,000 budget. A further 15–20% increase in cooking oil prices (consistent with the 2015–16 episode) would push that to Rp 1,400–1,800, forcing kitchens to reduce oil volume or substitute with lower-quality fats.

The fat pillar of the meal is the most price-exposed and the least substitutable within the AKG framework.

2. Rice: the carbohydrate pillar faces a delayed but structural hit

El Niño on rice: The 2026 dry season planting across the 114 seasonal zones began earlier than usual (April) and is spreading through May–June. BMKG has flagged below-normal rainfall for July–September across Java, Sumatra, and parts of Sulawesi. The main harvest (typically March–May) was already below trend; the second harvest (July–September) is the one exposed to the deepening El Niño.

Domestic price trend: BPS data shows medium-quality rice at wholesale averaging Rp 6,683/kg in July 2026, up 3.1% year-on-year. At the huller level, January 2026 prices were up 6.19% yoy. The volatile food component of CPI ran at 5.58% yoy in June 2026, with rice the largest single contributor.

GR 24/2026 relevance: Rice is not currently on the DSI commodity list. But the regulation's implementing rules allow the Ministry of Trade to add commodities. If domestic rice stocks tighten further (Bulog's procurement target for 2026 is 3.2 million tonnes; as of June it had secured roughly 1.1 million), the policy toolkit now includes export restriction via DSI — a lever that did not exist in 2015–16.

MBG exposure: Rice provides 55–65% of meal calories. At Rp 6,683/kg wholesale (retail ~Rp 11,000–12,000/kg), the rice portion of a 150g serving costs roughly Rp 1,000–1,500. A 10% price increase absorbs Rp 100–150 of the meal budget — manageable in isolation, but additive with protein and fat pressure.

3. Chicken and eggs: the protein pillar has a feed-cost transmission chain

El Niño on feed: Indonesia imports roughly 70% of its corn and 90% of its soybean meal for animal feed. The USDA Oilseeds and Products Annual (April 2026) notes global soybean and corn balance sheets tightening on South American weather and US acreage shifts. El Niño historically raises US corn and soybean prices 15–25% in the harvest year. The World Bank's April 2026 Commodity Markets Outlook forecasts overall commodity prices up 16% in 2026, led by energy and fertiliser — both feed-cost inputs.

Domestic policy buffer: The government's feed corn subsidy (55,500 tonnes delivered to 5,543 small farmers across 26 provinces as of June 29, 2026) and BGN's Rp 20 trillion Danantara allocation for chicken farms are explicit acknowledgments that chicken supply for MBG is a managed variable. But the subsidy covers only a fraction of national feed demand.

Price data: Broiler chicken meat at Rp 35,477/kg (July 2026, +3.0% yoy). Farm-gate breakeven for layer hens is reported at Rp 24,000/kg eggs; BGN has instructed SPPGs to procure directly from farmers at or above that floor.

MBG exposure: Chicken and eggs together constitute the primary animal protein in MBG menus. At current prices, a 50g chicken portion + half an egg per meal costs roughly Rp 2,500–3,000 — 25–30% of the Rp 10,000. Feed cost is 65–70% of poultry production cost. A 15% feed price increase transmits to roughly 10% higher chicken/egg prices at the farm gate, adding Rp 250–300 per meal.

4. The convergence on the SPPG procurement ledger

SPPG kitchens operate under a local procurement mandate (BGN instruction, May 16, 2026; reinforced at Sukamantri and Margomulyo SPPGs). They buy from farmers and traders in their regency. This means:

• No central hedging. Each kitchen faces its local market price.
• Regional divergence. Java kitchens access tighter but deeper markets; outer-island kitchens (Papua, Maluku, NTT) face thinner supply and higher transport margins — already documented in MBG Watch's rural rollout analysis.
• Incentive change. The July 2026 revision removing the flat Rp 6 million/day SPPG incentive (Tempo, July 11) shifts more cost risk onto kitchen operators, who must now manage procurement volatility within the Rp 10,000/meal.

Price-pass-through timing: The July 2026 holiday suspension "natural experiment" (MBG Watch, published) showed that when MBG procurement pauses, staple prices in supply corridors drop measurably within 1–2 weeks. When procurement resumes, prices rebound. This confirms MBG's own demand is a price mover — and that the pass-through from commodity market to kitchen gate is fast (days to weeks), not slow.

What the evidence does not support

• A sudden collapse of the meal program. The Rp 335 trillion 2026 budget is allocated; Bulog stocks, while below target, are not exhausted; the feed corn subsidy and chicken-farm investment are live policy responses.
• Uniform impact across regions. Java's SPPGs have more supplier density and shorter logistics chains. Outer islands face compounding cold-chain and transport costs that the current budget structure does not differentially weight.
• That GR 24/2026 alone drives domestic price increases. The regulation's transition period (June–December) is a through channel, not yet an only-by channel. The price signal today is still dominated by global fundamentals (El Niño, biodiesel mandate, fertiliser costs) and the rupiah (which appreciated from ~17,950 to ~16,200/USD between June and July 2026, per Rupiah Stability Watch). The regulation's bite comes after December 31, when DSI becomes the exclusive exporter.

The least-harm path: what to watch, and when

Operational levers that exist today:

1. BGN can adjust the AKG fat/protein minimum for the 2026–27 school year — a calibrated reduction (e.g., fat from 25% to 22% of kcal) preserves meal viability while acknowledging the supply shock.
2. The feed corn subsidy can be expanded — it is a targeted, reversible instrument with direct line to the protein pillar.
3. DSI's domestic allocation mandate (the flip side of export control) can be quantified and published monthly — transparency creates market discipline.
4. Regional budget top-ups for outer-island SPPGs — the current flat Rp 10,000 does not reflect the 30–50% higher logistics cost documented in MBG Watch's rural rollout analysis.

What I'm uncertain about (in order of consequence)

1. El Niño severity and spatial footprint. NOAA's 63% "very strong" probability is a distribution, not a point forecast. If the event moderates to "strong" (like 2009–10), the palm oil yield hit could be 5–8% instead of 10–15%. The difference is material for cooking oil prices.

2. DSI's domestic allocation rule. GR 24/2026 and its derivatives do not yet publish a formula for how DSI decides the domestic vs export split. Without that rule, the regulation is a latent risk, not an active price signal.

3. Feed import pass-through elasticity. The 15% feed cost → 10% chicken price transmission assumes competitive poultry markets. If consolidation has progressed further than documented (the KPK probe into SPPG vendors hints at concentrated supply chains), pass-through could be higher.

4. Outer-island cold-chain reliability. MBG Watch documented 37,673 food-safety victims in the rural rollout. If El Niño disrupts inter-island shipping (port operations, fuel availability), the procurement mandate becomes unexecutable in some regencies — not just expensive.

5. Budget revision space. The Rp 335 trillion is an annual figure. If the per-meal cost structure shifts by >15% (Rp 1,500/meal), the annual gap at 31,000 SPPGs serving ~20 million children is ~Rp 100 trillion. Whether the fiscal framework accommodates a mid-year revision is a political question, not a technical one.

What this means for the monitoring framework

MBG Watch's Continuous Monitoring Framework tracks five domains: budget execution, nutrition outcomes, food safety, governance, and equity of access. This analysis adds a sixth observational lens: commodity-to-kitchen price transmission.

Starting with the next monitoring cycle, we will track:

• Weekly cooking oil retail price (Ministry of Trade) vs SPPG reported procurement cost (sample of 50 kitchens across Java, Sumatra, Kalimantan, Sulawesi, Papua)
• Monthly CPO domestic allocation volume (from DSI/Kemendag reports once published)
• Quarterly feed corn and soybean meal import unit values (BPS trade data) vs farm-gate chicken/egg prices (BPS agriculture survey)
• Per-meal macronutrient composition from BGN spot audits (fat %, protein %, carbohydrate %)

The chain is traceable. The variables are observable. The question is whether the observation feeds back into decision-making before the meal changes — not after.