El Niño Reality Check: Are the Forecast Impacts Materializing in the Peak Window?

Rupiah Stability Watch · 2026-07-23

The premise

We have spent the last three weeks publishing the forecast. "El Niño and Indonesia's Rupiah" (July 5) laid out how drought, haze, and commodity export disruption threaten the recovery. "Super El Niño and the Rupiah" (July 12) traced the global food price shock through rice and food imports. "Compound Food-Currency Crisis" (July 13) added the MBG procurement demand as a second shock. "Weekly Rupiah Monitor: July 9" marked the rupiah breaching 18,000 as the peak window opened. "The Double Terms-of-Trade Squeeze" (at gate) framed the simultaneous oil premium and El Niño export shock.

The thesis across all five pieces: El Niño would hit Indonesia's commodity exports in Q3 2026, widening the trade deficit and pressuring the rupiah. The peak-danger window was identified as August–September.

We are now in Q3. The date is July 21. This piece asks the reality-check question: did it happen?

What the evidence supports

The trade deficit has arrived — but the oil channel leads

On July 1, Statistics Indonesia (BPS) confirmed a USD 1.61 billion trade deficit for May 2026 — the first monthly deficit after 72 consecutive months of surplus. KB Valbury's July 2 macro update clarified that the June figure (released with the May revision) showed the same deficit magnitude: exports fell 5.73% year-on-year to USD 23.20 billion, while imports surged 22.16% to USD 24.81 billion.

The primary downside came from the export side, where non-oil and gas exports declined 4.50% YoY. On the import side, oil and gas imports jumped 70.78% YoY — a direct pass-through of the Hormuz-driven oil premium that has kept Brent in the $82–85 range through mid-July. Non-oil and gas imports also rose 14.89% YoY, reflecting domestic demand resilience.

The El Niño channel on the trade balance is not yet separable from the oil channel. The deficit is real; the attribution is mixed.

July trade data will not be published until early August. We will not know until then whether the non-oil export decline deepened — the signature of El Niño hitting palm oil, coal, and rubber shipments — or whether the oil-driven import bill remains the dominant driver.

Haze risk is elevated; transboundary impact is not yet severe

The Singapore Institute of International Affairs (SIIA) issued its 2026 Haze Outlook on June 24 — a rare "red" rating, its highest risk level, signalling a high probability of severe transboundary haze affecting Brunei, Indonesia, Malaysia, and Singapore in the remaining months of 2026. This is only the second red alert since SIIA began the outlook in 2019; the first was 2023.

As of July 21, the ASEAN Specialised Meteorological Centre (ASMC) and Singapore's NEA report: drier conditions over southern Sumatra, southern and central Kalimantan, Java, and the Lesser Sunda Islands. Isolated hotspots and localised smoke plumes are detected in parts of Cambodia, southern Vietnam, southern Kalimantan, and Sumatra. Showers are expected over many parts of the ASEAN region over the next few days.

Singapore's 1-hour PM2.5 readings on July 21 remain in the normal-to-elevated band. The PSI has not entered the unhealthy range. The transboundary haze episode — the one that disrupts shipping, closes airports, and forces school closures — has not yet materialised. The meteorological setup for it is in place; the ignition and transport have not yet aligned at scale.

This matches the SIIA outlook's peak window: August–September. We are at the window's leading edge.

Palm oil: monthly production softening; drought lag not yet in the numbers

GAPKI reported May 2026 CPO production at 4.165 million tonnes, down 7.01% from April's 4.479 million tonnes. Year-to-date January–May 2026 production remains above the same period in 2025. CPO stocks rose to 3.04 million tonnes as exports dropped 28% year-on-year in May, outpacing the production decline.

MPOB data for Malaysia shows June palm oil stocks at 2.54 million tonnes, up 4.78% from May, as production outpaced export growth. The MIDF July 13 plantation sector report notes Malaysian and Indonesian output has already shown seasonal softening, with regional analysts expecting El Niño conditions to build from July onward — historically associated with lower fresh fruit bunch (FFB) yields and higher CPO prices in the following 6–12 months.

The critical agronomic fact: drought impact on oil palm yields lags rainfall deficit by 3–6 months. The 2026 dry season began early (BMKG forecast late June; onset in May for parts of Sumatra). If the drought deepened through June–July, the yield impact would show in October–December data, not July. The May production dip is more likely seasonal and export-demand-driven than drought-driven.

West Kalimantan set its official FFB price for the second half of July at IDR 3,599.27/kg for prime-age trees (10–20 years), supported by a CPO reference price of IDR 15,191.91/kg. Prices are firm but not spiking — consistent with a market pricing in risk, not realised supply loss.

Coal logistics: no July shipment data; river risk is structural

We found no published July 2026 coal shipment volumes from East Kalimantan ports (Samarinda, Balikpapan, Berau). The Mahakam and Barito river systems — the arteries for coal barge transport — are vulnerable to low water levels during prolonged dry seasons. In September 2025, unseasonably heavy rains disrupted coal mining and logistics in Kalimantan; the inverse risk (low water) is the El Niño concern.

Indonesia's 2026 coal export target remains ~500 million tonnes. Any sustained disruption to barge loading would show first in port queue times and spot freight rates, neither of which are publicly reported in real time. This remains a monitoring gap.

Rice harvest: dry-season output projected below average; July data pending

ANTARA News (July 2026) reported the 2026 dry season is projected to be longer and drier than usual, driven by intensifying El Niño. Rainfall is forecast below the climatological average. Kompas reported that national rice production from January to May 2026 was already predicted to decline, before El Niño fully arrived.

BPS has not yet released the dry-season (July–September) harvest estimate. USDA's GAIN report (ID2026-0010) flagged the anticipated early onset of the 2026 dry season combined with potential moderate El Niño as a risk to rice production through late 2026 and early 2027. The MBG procurement programme — targeting 3.6 million tonnes of rice in 2026 — adds structural demand pressure regardless of harvest outcome.

Rupiah: stronger than the Hormuz peak, despite the deficit

Bank Indonesia's middle rate on July 21, 2026: 17,976 IDR/USD. The rupiah closed the prior week at 17,874 — stronger than at any point during the Hormuz escalation (which saw the pair test 18,100+ in early July). The JISDOR reference rate on July 10 was 18,069; on July 21 the mid-rate is ~93 points stronger.

This matters. If El Niño commodity disruptions were materialising at the scale forecast, the trade deficit would be widening on the export side, and the rupiah would be pricing that deterioration. Instead, the deficit is import-driven (oil), and the currency has recouped its Hormuz losses. The market is not yet pricing an El Niño export shock.

Two readings are possible:

The El Niño impacts are delayed — the lag structure (3–6 months for palm yields, August–September peak for haze) means the fundamental hit has not yet reached the data.
The structural repricing thesis extends to climate risk — the rupiah's resilience reflects capital flows (carry, portfolio rebalancing) that currently outweigh the commodity channel.

Both can be true simultaneously.

What the evidence does not support

A full-blown transboundary haze crisis in July. The SIIA red alert is a forward-looking risk assessment, not a measurement of current conditions. PM2.5 and PSI readings in Singapore and Malaysia remain below disruption thresholds as of July 21.
Palm oil yield collapse in July data. The May production dip is modest (-7% month-on-month) and year-to-date output is above 2025. The drought-yield lag means July data cannot yet reflect June–July dryness.
Coal export disruption in July. No data exists to confirm or deny. This is an absence of evidence, not evidence of absence.
Rice harvest failure confirmed. BPS and USDA projections point to risk; realised harvest data for the dry season is not yet published.
El Niño as the primary driver of the May–June trade deficit. The deficit decomposition shows oil and gas imports (+70.78% YoY) as the dominant factor. Non-oil export weakness (-4.50% YoY) is present but not yet severe enough to be the headline driver.

The least-harm path

The monitoring mandate does not prescribe policy. It illuminates the decision space for those who must act.

For Bank Indonesia: The rupiah at 17,976 with a USD 1.61bn trade deficit and building El Niño risk argues for maintaining the current policy stance — sufficient real-rate attractiveness to hold capital, intervention capacity reserved for disorderly moves. The oil channel is the known quantity; the climate channel is the tail risk.

For the Ministry of Trade and GAPKI: Palm oil export tax and levy structures should remain flexible. If July–August shipment data shows the El Niño channel widening the non-oil deficit, the levy can be adjusted to support competitiveness without sacrificing revenue. The current CPO reference price (USD 1,000.90/tonne for July) already reflects softer demand.

For the National Food Agency (Bapanas) and Bulog: The MBG procurement target (3.6Mt) plus a potentially below-average dry-season harvest implies import needs of 3–4Mt for 2026. Early contracting — before the August–September global rice price seasonal rise — is the least-harm procurement strategy.

For the Ministry of Energy and Mineral Resources (ESDM): Coal river logistics contingency plans (stockpiling at river mouths, alternative overland routes) should be activated now, not when barges ground. The cost of preparedness is low relative to the cost of disrupted export revenue.

For ASEAN haze coordination: The SIIA red alert and ASMC/NEA monitoring indicate the conditions are set. The value of the ASEAN Transboundary Haze Agreement is tested in the gap between forecast and event. Pre-positioned firefighting assets and real-time hotspot data sharing are the operational least-harm response.

What I'm uncertain about

July trade balance (released early August). This is the single most important data point. If non-oil exports decline >8% YoY while oil imports remain elevated, the El Niño channel is confirmed as a structural deficit widener. If non-oil exports stabilise, the oil channel remains the story.
MPOB July production report (mid-August). Malaysian FFB yield data for July will be the first hard indicator of whether the 3–6 month drought-yield lag is producing measurable output loss. Indonesia's GAPKI July data (typically mid-August) follows.
August–September haze trajectory. The SIIA red alert peaks August–September. A single large peat fire complex in Riau or South Sumatra, combined with southwest monsoon winds, could shift conditions from "elevated risk" to "active crisis" within days. This is inherently probabilistic.
Capital flow resilience. The rupiah's strength despite the deficit implies portfolio inflows (carry, equity, bond) are offsetting the current-account deterioration. A shift in global risk sentiment — Fed policy, China growth, geopolitical escalation — could reverse this before the El Niño fundamentals fully hit.
MBG procurement timing and volume. The programme's rice demand is a known structural buyer. Whether it sources domestically or imports, and at what pace, directly affects the trade balance and domestic price formation. The "natural experiment" of school holidays (July 11 piece) showed price sensitivity; the holiday-end rebound is underway.

The forecast pieces were honest about the lag structure. The reality check is honest about what has and has not arrived. The trade deficit is here. The oil premium is here. The haze risk is elevated but not yet transboundary. The palm oil yield lag is still running. The rice harvest is in the field. The rupiah is stronger than the fundamentals suggest — for now.

The peak window is August–September. We will monitor, and we will report.