Energy, conflict, climate and trade risks are converging
Physical energy disruption is the immediate constraint; concentrated food and industrial dependencies are the main multipliers.
- Evidence reviewed
- 30 September 2026
- Reading time
- 32 min read
The report separates immediate effects from second- and third-order consequences, then identifies the indicators that would change the assessment.
Executive assessment — physical energy disruption is the immediate constraint; concentrated food and industrial dependencies are the main multipliers
Five developments shaped the week, but they matter because they are drawing on the same limited buffers: energy inventories, transport capacity, food stocks, defence production, public finance, foreign exchange and humanitarian access.
The Strait of Hormuz remains the binding global constraint. Before the conflict, the route carried 20.9 million barrels of oil per day—about 20% of global petroleum-liquids consumption—and more than 20% of global LNG trade. Saudi and UAE bypass pipelines can carry about 4.7 million barrels per day, equivalent to only 22.5% of the pre-war oil flow, before accounting for existing use, security or operational limits. In August, more than 10 million barrels per day of Gulf output remained shut in, Gulf exports were around half their pre-war level and observed global oil inventories had fallen by 507 million barrels since February, according to the IEA’s September market report.
The food outlook is more balanced than the original report implied. El Niño is exceptionally strong, but a global food crisis is not the base case. FAO expects 2026 cereal production to fall 2.0%, while the global stocks-to-use ratio remains a comparatively comfortable 31.6%. The vulnerability lies in traded supply: India, Vietnam and Thailand represent about 64% of projected rice exports, while Indonesia and Malaysia produce about 82% of palm oil. A moderate harvest loss could therefore become a larger price shock if concentrated exporters restrict supply or major importers accelerate stockbuilding.
Ukraine has developed substantial domestic drone capacity, but money and industrial capacity do not automatically become usable defence. Ukraine reports annual defence-production capacity of $50 billion and a 2026 target above seven million drones. Post-cutoff evidence reinforces the bottleneck: on 28 September, Ukraine identified $27 billion of additional defence needs, with roughly one-third required as advance payments for early-2027 drone production. Air-defence missiles remain tied partly to allied stocks and future production. The financing burden is structural: the IMF estimates Ukraine’s 2026 deficit excluding grants at about 21% of GDP and its 2026–2029 financing requirement at $140.3 billion.
In Tigray, the access shock is already observable even though new casualty and displacement totals remain incomplete. The United Nations confirmed the seizure of three airports and subsequently reported a strike on a food-assistance truck, interrupted road movements and communications failures. This is occurring against a pre-event baseline of approximately 749,500 internally displaced people recorded in accessible Tigray locations and an Ethiopian external buffer of only 2.1 months of imports.
The US–China state visit reduced the risk of unmanaged escalation but did not reverse fragmentation. The proposed tariff treatment covers $30 billion of non-sensitive goods in each direction and remains a recommendation rather than enacted relief. The combined $60 billion represents about 14.5% of 2025 bilateral goods trade. Meanwhile, US imports from China fell approximately 19.4% year on year in January–July 2026, while investment and production links continued to shift through Mexico and ASEAN.
The central conclusion is therefore narrower and stronger: the world is not experiencing one undifferentiated crisis. It is managing several shocks through buffers that differ sharply by product, country and time horizon. The risk rises when those buffers are assumed to be interchangeable or unlimited.
Five evidence-backed conclusions
- Hormuz is a physical supply and inventory problem, not only a price shock. Alternative routes can replace only a fraction of normal flows, and refined products are tighter than headline crude inventories suggest.
- Food risk is concentrated rather than global—for now. Aggregate cereal stocks remain a buffer, but rice and palm-oil trade depends on a small number of climate-exposed producers.
- Ukraine’s binding constraint is the conversion of finance into timely capability. Reported factory capacity, funded orders, delivered systems and battlefield availability are different measures.
- Tigray’s access deterioration is observed; its new humanitarian loss is not yet measurable. Pre-existing displacement and economic fragility must not be presented as September losses.
- US–China dialogue is institutionalising managed rivalry, not restoring integration. Direct trade is declining while indirect dependency through connector economies persists.
The risk dashboard separates observed conditions from conditional outcomes
| Risk | Current status | Scale or buffer | Transmission speed | Impact if persistent | Confidence |
|---|---|---|---|---|---|
| Hormuz disruption | Observed physical disruption | 20.9 mb/d pre-war exposure; 4.7 mb/d Saudi-UAE bypass capacity; global oil stocks down 507m barrels since February | Days to weeks | Critical and global | High on direction; medium-high on exact flows |
| Russia–Ukraine war | Observed escalation and long-duration financing pressure | $195bn direct damage; $588bn reconstruction need; 21%-of-GDP 2026 deficit excluding grants | Weeks to years | High and structural | Medium-high |
| El Niño and food | Observed climate signal; food outcome conditional | Cereal stocks-to-use 31.6%; top three rice exporters hold 64% of projected trade | Months | High if production and policy shocks combine | Medium |
| Tigray conflict | Observed access disruption; losses incomplete | About 749,500 pre-event IDPs in assessed Tigray sites; Ethiopian reserves cover 2.1 months of imports | Days to months | Extreme locally; material nationally | High on access; medium on macro transmission |
| US–China fragmentation | Observed dialogue inside continuing controls | Proposed $60bn framework equals about 14.5% of 2025 goods trade | Months to years | High structurally; lower immediately | High on policy structure; medium on implementation |
The categories are not a mechanical composite score. They distinguish impact, speed and evidence strength because a local humanitarian emergency and a global economic shock cannot be ranked credibly on one scale.
1. Hormuz remains the immediate global constraint because secure alternatives cover only a fraction of normal flows
What changed
Iran proposed a conditional seven-day reopening of the strait and a return to nuclear talks; President Trump rejected the proposal on 26 September. Diplomatic movement did not restore insured commercial passage. By 16 September, the International Maritime Organization had confirmed 80 attacks and 22 seafarer deaths. Post-cutoff evidence reviewed for this edition raised the count to 86 attacks and 24 deaths by 29 September.
Exposure summary
| Exposure | Assessment | Evidence and reason |
|---|---|---|
| Asian importers | High | Asia received 89% of Hormuz crude and condensate; China, India, Japan and South Korea received 74% |
| Route-dependent Gulf exporters | High | Export inability can reduce volumes and revenue even when prices rise; pipeline and port alternatives differ sharply by country |
| Fuel- and cereal-importing vulnerable economies | High indirect | UNCTAD identifies 61 economies exposed to both fuel and cereal imports, with currency, debt and safety-net constraints |
| Most exposed sectors | High | Refining and diesel-intensive transport; LNG-dependent power and chemicals; fertiliser, farming and food distribution |
| Strongest buffer | Material but finite | Coordinated stocks, demand adjustment, non-Gulf production and country-specific reserves |
| Largest uncertainty | High | Safe and insured shipping capacity; AIS silence and spoofing make transit counts incomplete |
First order — the direct constraint is lost export and refined-product capacity
Observed — high confidence. Hormuz carried 20.9 million barrels per day in the first half of 2025, equal to roughly 20% of global liquids consumption. In August 2026, more than 10 million barrels per day of Gulf output remained shut in and total Gulf oil exports were around 13 million barrels per day, nearly half their pre-war level. Refined products were tighter: Gulf diesel and gasoil net exports were only just over one-quarter of normal.
This distinction matters. A market can have crude oil in storage and still lack the right diesel, jet fuel, LNG or feedstock in the required location. In early September, US diesel exceeded $200 per barrel, 94% above its pre-war level, according to the IEA.
Second order — emergency stocks and substitution are absorbing the shock by transferring it into inventories, margins and public balance sheets
IEA members made 400 million barrels available. At a constant gross loss of 10 million barrels per day, that equals 40 nominal days of cover. This is an arithmetic sensitivity, not a forecast: restored production, demand reduction, crude quality, product composition and release logistics change the result.
The buffer is being consumed. Observed global oil inventories fell 507 million barrels between February and August, an average of about 2.8 million barrels per day. The US Strategic Petroleum Reserve stood at 284.6 million barrels on 18 September, 29.9% below its year-earlier level; US distillate inventories were 12.7% lower.
Physical redundancy is also limited. Saudi and UAE bypass pipelines have about 4.7 million barrels per day of combined capacity—22.5% of the pre-war Hormuz flow. Capacity is not spare throughput, and the Saudi East-West route itself became a target.
Third order — energy security is shifting from stock volume to usable system design
If disruption persists, governments and firms will place more value on protected pipelines, ports outside chokepoints, refinery flexibility, regional product stocks and the ability to release the correct fuel quickly. Strategic-reserve policy will increasingly be judged by location, product type, draw rate and replenishment, not only headline barrels.
The 2019 Abqaiq-Khurais attack removed about 5.7 million barrels per day, roughly 6% of global supply, but rapid repair and inventories prevented a prolonged shortage. The present shock is more serious because several exporters, infrastructure assets and a maritime route remain affected over a longer period.
The food channel is conditional and product-specific
The original analysis overstated the uniformity of fertiliser pressure. The World Bank’s August benchmarks put urea at $390 per tonne, 46.3% below March, while phosphate and potash prices remained above their 2025 averages. The immediate risk is therefore not a universal fertiliser-price surge. It is uneven availability, gas and feedstock exposure, shipping cost and persistent pressure in specific nutrients and markets.
Assessment: the energy-to-food cascade has not fully activated. It requires persistence through planting, input access, crop yields, trade policy, currencies and household income.
Evidence that would change the assessment
The energy outlook would improve if Gulf oil exports remain above 70% of their pre-war level for four weeks, refined-product exports recover, global inventory draw falls below one million barrels per day and attacks and insurance costs decline together. It would worsen if Gulf exports fall below the August floor, another bypass route or major facility is disabled, or commercial cover becomes unavailable.
These are report-defined monitoring thresholds, not official forecasts.
2. Ukraine has scaled drone production, but air-defence stocks and advance financing are now the binding constraints
What changed
Ukraine conducted its largest reported drone attack on Moscow within a wider wave of more than 1,000 reported drones during 20–25 September. The count relies partly on Russian military reporting and is evidence of scale, not an independently verified measure of successful strikes.
EU members also agreed how to mobilise €6.6 billion through the European Peace Facility. The total includes reimbursement, training and €1 billion for joint procurement; it is not €6.6 billion of new equipment already delivered.
Post-cutoff evidence strengthens the central conclusion. On 28 September, Ukraine identified $27 billion of additional defence needs, with approximately one-third required as advance payments for early-2027 drone production. The EU reported €28.3 billion of approved procurement plans, including Patriot missiles from future US production, but no new interceptor-stock pledge at the 28 September meeting.
Exposure summary
| Exposure | Assessment | Evidence and reason |
|---|---|---|
| Ukraine | Extreme | Direct destruction, displacement, a 21%-of-GDP deficit excluding grants and reliance on external military and budget finance |
| Russia | High and mixed | Large defence burden and long-war cost, partially buffered by scale, domestic production and energy revenue |
| European NATO members | High strategic | Procurement, stock-replenishment and fiscal trade-offs; highest security exposure in eastern members |
| Most exposed sectors | Extreme to high | Air defence, drones and electronics; energy and logistics; sovereign finance and reconstruction |
| Strongest buffer | Material but dependent | Ukrainian defence production plus external finance; end-June reserves covered 5.2 months of imports |
| Largest uncertainty | High | Usable interceptor stocks, weapon-consumption rates and delivered output are not independently disclosed |
First order — attacks consume scarce defensive assets while expanding the reconstruction burden
The latest joint damage assessment records more than $195 billion in direct damage and almost $588 billion in ten-year recovery and reconstruction needs. Fourteen percent of housing had been damaged or destroyed by the end of 2025, affecting more than three million households. At the end of June 2026, about 3.87 million people remained internally displaced.
The direct operational effect of large drone and missile waves is therefore cumulative: new damage, civilian disruption and the consumption of interceptors occur on top of an already large repair and population burden.
Second order — reported capacity is substantial, but financing timing determines output
Ukraine reports annual defence-production capacity of $50 billion, more than half of military equipment needs met domestically, four million drones produced in 2025 and a 2026 target above seven million. These are official Ukrainian figures and should be treated as attributed capacity and targets, not independently audited delivery.
The request for advance payments exposes the practical bottleneck. Manufacturers need working capital and component orders months before delivery. A fast procurement channel can deliver in-stock drones in days, but it cannot compress the production cycle for Patriot missiles, explosives or specialised electronics.
The closest historical comparison is Europe’s ammunition programme. The EU expected to deliver 524,000 rounds—52% of its one-million-round target—by the original March 2024 deadline, before exceeding one million later that year. Limited stocks, components and investment lead times delayed delivery. The case does not prove a current shortage in every system; it demonstrates why authorised funding and theoretical capacity are not the same as timely capability.
Third order — the war is embedding defence production and external finance into Europe’s economic structure
SIPRI estimates Ukraine’s 2025 military spending at $84.1 billion, or 40% of GDP; Russia spent about $190 billion, or 7.5% of GDP; and European military spending rose 14% in real terms to $864 billion.
Ukraine’s fiscal dependence is equally important. The IMF estimates a 2026 deficit excluding grants of about 21% of GDP and a $140.3 billion financing requirement for 2026–2029. It recorded $52.3 billion of firm non-IMF commitments for the following 12 months. Commitments are not disbursements: end-June reserves reached $51.3 billion, or 5.2 months of imports, only after more than $15 billion of official financing arrived that month.
The long-term system is therefore moving toward permanent joint production, larger defence budgets, stock-replenishment programmes and closer integration of Ukrainian and European defence industry. That transition competes with reconstruction and other public priorities.
Evidence that would change the assessment
The constraint would ease if near-term interceptor stock transfers are formally committed, the €1 billion joint-procurement component produces dated contracts, and the advance-payment requirement for 2027 drones is funded early enough for component orders. It would worsen if funding remains committed but undisbursed, reserves fall materially below the June buffer or attack and interceptor consumption continue to outpace delivered systems.
3. Tigray fighting has already disrupted aid routes; prolonged closure would expose Ethiopia’s thin economic buffer
What changed
On 23 September, the United Nations confirmed the seizure and occupation of the airports at Mekelle, Axum and Shire, intensified fighting and the spread of hostilities into neighbouring areas.
The access impact is no longer hypothetical. By 25 September, an official UN briefing reported a strike on a food-assistance truck, road closures interrupting planned movements and phone and internet failures affecting humanitarian operations and economic life. Responsibility for the opening attack and several battlefield claims remains contested.
Exposure summary
| Exposure | Assessment | Evidence and reason |
|---|---|---|
| Tigray | Extreme | Fighting, three seized airports, large displaced and returnee populations and impaired humanitarian movement |
| Amhara–Afar border areas | High | Hostilities and aid corridors can cross regional boundaries already affected by insecurity |
| Ethiopia nationally | High | Security and relief costs can weaken debt restructuring, fiscal reform, foreign exchange and investment confidence |
| Most exposed sectors | Extreme to high | Food and humanitarian assistance; aviation, roads and telecommunications; sovereign finance and public services |
| Strongest buffer | Conditional | Pretoria peace framework and established humanitarian logistics capacity |
| Largest uncertainty | High | Whether fighting becomes sustained federal-regional war and how much new displacement has occurred |
First order — access has deteriorated, but the new human loss remains unmeasured
Airports matter because northern Ethiopia has long distances and damaged services. They are only one part of the system: food, fuel, medical evacuation, staff rotation and commerce also require secure roads, working communications, protected vehicles and operating authority.
Current casualty and post-23-September displacement totals remain incomplete. The defensible conclusion is that access has deteriorated and civilian risk has increased—not that a precise new humanitarian loss has already been measured.
Second order — a large pre-existing vulnerable population raises the cost of each access failure
The latest detailed IOM assessment identified approximately 749,500 internally displaced people and 1.12 million returning IDPs in accessible Tigray sites and villages before the offensive. A European Commission baseline separately recorded more than 500,000 aid-dependent IDPs. The numbers use different definitions and coverage; both are exposure baselines, not September losses.
Ethiopia has significant humanitarian capacity. WFP’s national network moves more than 600,000 tonnes of food annually to over 6,000 distribution points. Capacity, however, does not equal access. During the previous war, 115 loaded aid trucks were at one point waiting for clearance to proceed to Mekelle while millions of people were behind battle lines. The historical case tests the route-and-clearance mechanism; it does not estimate current need.
Third order — prolonged conflict could interrupt a recovery that remains externally fragile
Ethiopia enters this shock with improving but thin buffers. The IMF estimates reserves of $5.9 billion, equivalent to 2.1 months of imports, expected tax revenue of 9.5% of GDP and 13.4% headline inflation in May. The country remains in external and overall debt distress.
Reforms since 2024 lifted reserves from around two weeks to more than two months of import cover, narrowed foreign-exchange shortages and reduced inflation. Prolonged conflict could weaken those gains through security spending, delayed reform, investor caution, food-price pressure and renewed currency demand. The national effect will depend on duration and geographic spread; it should not be inferred from battlefield control alone.
Evidence that would change the assessment
Risk would ease if all three airports resume scheduled civilian or humanitarian operations, major road corridors sustain two-way movement for at least 14 days, communications recover and AU-mediated talks produce a monitored timetable. It would worsen if road or airport disruption persists across two reporting cycles, post-event displacement rises, aid tonnage falls or external involvement is independently confirmed.
4. El Niño is an extreme climate signal, but concentrated traded supply—not global stock scarcity—is the main food risk
What changed
NOAA recorded a +1.8°C Niño-3.4 anomaly in August, a greater than 90% chance of a very strong event and a 75% chance that October–December strength exceeds the post-1950 record threshold. These probabilities describe Pacific warming, not the probability of crop failure.
Exposure summary
| Exposure | Assessment | Evidence and reason |
|---|---|---|
| South and Southeast Asian rice exporters | High | India, Vietnam and Thailand account for about 64% of projected world exports |
| Indonesia and Malaysia | High | The two countries produce about 82% of palm oil |
| Rain-fed and import-dependent low-income economies | High household impact | Climate sensitivity, currency weakness and limited fiscal space can turn moderate global moves into acute local stress |
| Most exposed sectors | High | Rice, palm oil and tropical crops; fertiliser-intensive farming and food processing; hydropower and irrigation |
| Strongest buffer | Material | Global cereal stocks-to-use of 31.6%, large producer reserves and partial substitution across crops and oils |
| Largest uncertainty | High | Whether severe weather affects several major exporters in the same crop window |
First order — the climate signal is strong, but physical outcomes remain local
El Niño changes the probability of drought, heat, fire and heavy rainfall. It does not create the same outcome everywhere. Crop exposure depends on rainfall timing, irrigation, soil moisture, planting decisions, the Indian Ocean Dipole and local preparedness.
The current global balance provides a buffer. FAO forecasts cereal production of 2.980 billion tonnes in 2026, 2.0% below 2025, with stocks of 947.2 million tonnes and a 31.6% stocks-to-use ratio. Rice production is forecast at 553.1 million tonnes against use of 559.0 million tonnes, leaving a 5.9-million-tonne gap—about 1.1% of annual use—to be met from stocks.
Second order — concentrated trade and policy can amplify a manageable harvest loss
USDA projections put India, Vietnam and Thailand at approximately 64.2% of world rice exports; the top five exporters account for about 78.6%. Indonesia and Malaysia produce 82.1% of palm oil. These are UFOQ.AI calculations from institutional data, and their importance is concentration rather than false precision.
Importers are already responding. The Philippines announced plans to import up to five million tonnes of rice, primarily from Vietnam and Thailand. Planned purchases are not evidence of global shortage, but simultaneous precautionary buying can tighten an export market before aggregate production fails.
Input conditions are mixed. Urea prices fell sharply from March to August, while phosphate and potash remained elevated and fuel, freight and local access costs continued to differ by market. Historical IMF modelling estimates that a 10% oil-supply price shock or a 10% fertiliser-price increase can raise international cereal prices by about 2% after three quarters. This is an illustrative historical relationship, not a 2026 forecast.
Third order — policy response may matter more than the initial crop loss
During the 2015–16 El Niño, Southern African cereal output fell 24%, from 41 million to 33 million tonnes, creating severe regional need while abundant global supply contained the broader price effect. In 2007–08, by contrast, export restrictions by a small number of suppliers materially amplified the rice-price shock despite broadly adequate global production.
The distinction is central. A severe local humanitarian crisis can occur without a global shortage, while a moderate aggregate loss can become a global price shock if concentrated exporters restrict supply.
Over time, repeated volatility will support irrigation, storage, climate-resilient seed, monitoring and insurance. It may also encourage strategic stockbuilding and export controls, reducing the reliability of open food markets for import-dependent economies.
Evidence that would change the assessment
The global outlook would worsen if FAO or USDA cuts rice output by another five million tonnes or more, at least two major exporters suffer material crop deterioration, or a top-three rice supplier restricts exports. It would improve if exporter stocks remain available, rainfall normalises during critical crop windows and substitute suppliers offset Asian losses.
These are analytical alert levels. The base case remains local production and food-access stress with manageable aggregate global supply.
5. US–China guardrails reduce accident risk, but direct decoupling is becoming indirect dependency
What changed
The 24–25 September state visit produced an advanced-AI incident channel, trade and investment boards and recommendations for more favourable tariff treatment for $30 billion of non-sensitive goods in each direction. USTR states that the US-side recommendation covers about 30% of US exports to China.
The measures remain recommendations. The Chinese Foreign Ministry confirmed the broader discussions but did not publish the same detailed commercial package. Existing tariffs, licensing rules and technology controls remain in place.
Exposure summary
| Exposure | Assessment | Evidence and reason |
|---|---|---|
| United States and China | High direct | More than $400bn in 2025 goods trade, with concentrated tariff, exporter and consumer effects |
| Mexico and ASEAN | High mixed | Gain orders and investment while remaining exposed to Chinese inputs, origin enforcement and alignment pressure |
| Advanced-technology nodes | High strategic | Taiwan, South Korea, Japan and the Netherlands sit between US controls, Chinese demand and semiconductor chokepoints |
| Most exposed sectors | High | Semiconductors and AI compute; electronics and telecoms; EVs, batteries and critical minerals |
| Strongest buffer | Material | Supplier diversification, selective licensing and new communication channels |
| Largest uncertainty | High | Chinese value added embedded in connector-country exports and whether the new channel works during a crisis |
First order — targeted dialogue improves, but the commercial package remains limited
US goods exports to China were $106.0 billion and imports $308.7 billion in 2025. The proposed combined $60 billion framework therefore equals about 14.5% of 2025 bilateral goods trade. It represents roughly 28% of US exports to China but about 10% of US imports from China. These are coverage calculations, not estimates of tariff savings or additional trade.
The immediate value is institutional: a channel for AI incidents and a process for non-sensitive trade can reduce misunderstanding and give firms limited policy visibility. Its effectiveness will be demonstrated only by use, implementation and compliance.
Second order — bilateral trade continues to contract while costs and production shift unevenly
US Census data show January–July 2026 imports from China down approximately 19.4% year on year, exports broadly flat and total goods trade down about 14.5%. Demand, front-loading and other policies also contributed, so this is not a tariff-only estimate.
Historical evidence shows the mechanism. The US International Trade Commission estimated that Section 301 tariffs reduced affected imports from China by 13%, raised affected US production by 0.4% and raised affected US product prices by 0.2% during 2018–2021. The result was material sourcing change, modest average production gain and some price cost—not costless reshoring.
Third order — fragmentation is redirecting value chains rather than cleanly separating them
China’s share of US goods imports fell from 22% in 2017 to 14% in 2023. Yet IMF research found that, across connector economies, a 1% gain in US import share was associated with a 1.6% increase in China’s export share and a 0.7% increase in China’s FDI share. These correlations do not prove transshipment; they show that direct decoupling can coexist with deeper upstream Chinese linkage.
ASEAN’s share of US electrical-goods imports exceeded 20% by 2023, while its share of global FDI reached almost 15% in 2024. Some of this reflects real production expansion. Economy-wide reshoring remains limited: OECD experimental estimates show export-weighted domestic value added across 41 economies increasing only from about 77.0% in 2022 to 77.6% in 2024.
The emerging system is managed globalisation: selective trade relief in non-sensitive products alongside persistent controls, subsidies, origin requirements and duplicated compliance in strategic sectors.
Evidence that would change the assessment
A broader thaw would require enacted tariff schedules, additional product coverage and actual removal—not suspension—of selected measures. Managed rivalry remains the base case if non-sensitive relief proceeds while semiconductor controls, Section 301 tariffs and strategic subsidies remain.
The key 90-day tests are implementation of the $30-for-$30 recommendations, renewal or lapse of Section 301 exclusions due in November, use of the AI incident channel and changes in BIS licensing or Chinese critical-mineral controls.
The combined risk is measurable: persistent energy pressure can amplify climate and financial stress, but the chain has not fully activated
The most important interaction remains:
Restricted energy and product flows → higher or more volatile fuel, gas, freight and selected fertiliser costs → weaker farm margins and higher import bills → concentrated crop losses or precautionary buying → export controls and tighter traded supply → local-currency food inflation → subsidy, reserve and debt pressure → weaker consumption, investment and political stability.
Each arrow must be tested separately.
| Link | Evidence | Current judgement |
|---|---|---|
| Energy → fertiliser | Natural gas can represent 80–90% of ammonia production cost; Middle East producers are important urea and ammonia exporters | Active exposure, but not uniform price escalation: urea eased sharply by August |
| Inputs → cereal prices | Historical IMF estimates suggest a 10% oil or fertiliser shock can raise international cereal prices by about 2% after three quarters | Credible lagged mechanism, not a 2026 forecast |
| Global → domestic prices | Historical IMF work finds faster fuel pass-through than staple-food pass-through and stronger transmission where currencies weaken | High risk in import-dependent economies; low uniformity across countries |
| Prices → fiscal pressure | The IMF estimated that the 2022 food and fertiliser shock added $9bn to the import bills of 48 highly exposed countries and that targeted protection required $5–7bn | Credible mechanism; 2026 cost requires country-specific modelling |
| Crop loss → policy amplification | Historical rice and wheat episodes show export controls can increase the traded-market effect beyond the physical loss | Conditional; no broad 2026 restriction wave is yet established |
The OECD’s September downside exercise shows the potential scale of interaction. Persistently higher energy and food prices combined with tighter financial conditions reduce 2027 global growth by 0.7 percentage point and raise inflation by 1.1 percentage points relative to baseline. This is an illustrative model scenario, not the OECD forecast and not an isolated estimate of any one event.
Three 90-day pathways define what should be monitored
| Pathway | Likelihood | Impact | Activation conditions | Evidence that weakens it |
|---|---|---|---|---|
| Persistent but contained strain | Medium-high | High but uneven | Hormuz exports remain impaired; inventories decline more slowly; El Niño losses remain dispersed; no broad food-control wave; security conflicts persist without major geographic expansion | Sustained energy normalisation, stable crops and reliable aid access |
| Operational relief | Low-medium | High positive | Gulf exports and refined products recover for four weeks; attacks and insurance costs fall; crop and stock data remain stable; trade and crisis channels are implemented | Renewed infrastructure attacks, export restrictions or delayed implementation |
| Cross-event escalation | Low-medium | High | Fertiliser or fuel pressure reaccelerates; at least two major exporters cut crop outlooks; a leading exporter restricts supply; Tigray or Ukraine demands rise while finance tightens | Stable exporter stocks, lower inputs, no restrictions and timely external finance |
Likelihood, impact and confidence are kept separate. These pathways are conditional judgements, not numerical forecasts.
The monitoring dashboard focuses on operating evidence rather than announcements
| Indicator | Baseline | Deterioration signal | Source route |
|---|---|---|---|
| Gulf oil and refined-product exports | August exports about half pre-war; diesel just over one-quarter | Fall below August level or fail to recover for another month | IEA Oil Market Report |
| Global oil inventories | Down 507m barrels since February | Draw remains above 2 mb/d | IEA; EIA |
| Shipping security and insurance | 86 confirmed attacks by 29 September | Continuing attacks or unavailable commercial cover | IMO; corroborated industry data |
| Fertiliser mix | Urea $390/t in August; other nutrients elevated | Fertiliser index above 170 or urea above $600/t | World Bank Pink Sheet |
| Global cereal balance | 31.6% stocks-to-use; 5.9m-tonne rice production-use gap | Additional rice downgrade of 5m tonnes or more | FAO; USDA |
| Food trade policy | No broad restriction wave | A top-three rice exporter restricts supply or several large importers accelerate purchases | FAO policy monitoring |
| Ukraine funding-to-delivery | $27bn additional need; advance payments required | Financing or contracts delayed into late Q4 | Ukraine MoD; EEAS; IMF |
| Ukraine air defence | Future procurement without new 28 Sep stock pledge | No dated near-term stock transfer while attack pressure persists | EEAS; national disclosures |
| Tigray access | Three airports seized; road and communications disruption observed | Two consecutive reporting cycles of interrupted movement or rising post-event displacement | UN; WFP; IOM |
| Ethiopia external buffer | $5.9bn; 2.1 months of imports | Reserve loss, FX spread deterioration or delayed programme review | IMF; National Bank; World Bank |
| US–China implementation | $30bn each direction recommended | No customs action, new controls or channel failure | USTR; BIS; Chinese authorities |
| Connector-economy dependence | Direct bilateral trade falling | Faster third-country imports without reduced Chinese upstream linkage | Census; IMF; OECD TiVA |
The numeric alert levels are UFOQ.AI monitoring rules anchored to current data. They are not official forecasts and should be revised when the underlying series changes.
Conclusion — resilience depends on the usability of buffers, not their headline size
The strongest conclusion from the week is not that every risk is worsening at the same rate. It is that the quality of the buffer determines whether exposure becomes impact.
Hormuz has revealed the difference between pipeline capacity and secure spare throughput. Ukraine shows the difference between announced funding, reported factory capacity and delivered capability. Tigray shows that a large humanitarian network cannot operate without roads, communications and protection. El Niño shows that comfortable global food stocks can coexist with concentrated export risk and severe local hunger. US–China trade shows that a change in customs origin does not necessarily remove upstream dependency.
This is the practical meaning of the three-order analysis. The first order identifies the physical or policy shock. The second tests how actors respond and whether buffers work. The third asks which responses persist long enough to change institutions, investment and system design.
The next 90 days should therefore be judged against operating evidence: restored flows, replenished inventories, financed production, delivered equipment, accessible aid routes, revised crop balances, enacted trade rules and functioning crisis channels. Diplomatic announcements matter, but they do not substitute for implementation.
Method and evidence discipline
- Observed: Direct measurement, official record or verified operating condition.
- Estimated: Modelled result, attributed figure or calculation with a stated denominator.
- Inferred: Causal interpretation supported by evidence but not directly measured.
- Scenario: Conditional pathway activated only if specified triggers occur.
Material calculations state their inputs and limitations. Exposure is kept separate from realised impact. Historical events test mechanisms rather than predict identical outcomes. Media are used mainly for chronology; central claims rely on official, multilateral or institutional sources. Post-cutoff evidence is used to test the report and is identified where material.
Sources and limitations
Source approach
The report prioritises operational records, statistical agencies, regulators, multilateral institutions and official decisions. Research papers are used for historical causal estimates. Government statements from parties to a conflict are attributed and are not treated as independently verified. Journalism is used mainly for chronology where an official record is unavailable.
Sources were reviewed through 30 September 2026. A source published after the 27 September event cutoff is identified in the body when it changes or tests the assessment.
Principal sources
Energy, Hormuz and cross-system transmission
- US EIA — World Oil Transit Chokepoints: pre-war Hormuz oil and LNG flows, destinations and bypass-pipeline capacity.
- IEA — Oil Market Report, September 2026: Gulf production shutdowns, exports, refined products, prices and global inventory draw.
- IEA — coordinated emergency-stock action: scale and composition of the 400-million-barrel action.
- International Maritime Organization — verified shipping incidents: attacks and seafarer casualties.
- US EIA — weekly petroleum balance sheet: US strategic, commercial crude and distillate inventories.
- World Bank — September 2026 Commodity Price Data: oil, gas and fertiliser benchmarks.
- UN Trade and Development — Hormuz exposure: fertiliser transit and vulnerable importer exposure.
- OECD — Interim Economic Outlook, September 2026: baseline assumptions and the combined energy, food and financial downside exercise.
- IMF — international-to-domestic commodity-price transmission: historical fuel and food pass-through estimates.
- IMF — Tackling the Global Food Crisis: historical import-bill and household-support estimates.
Russia–Ukraine war
- World Bank, EU, UN and Ukraine — fifth damage and needs assessment: direct damage, housing impact and ten-year reconstruction needs.
- Ukraine Ministry of Defence — production capacity: attributed Ukrainian defence-industry capacity.
- Ukraine Ministry of Defence — additional defence needs, 28 September: post-cutoff funding and advance-payment requirement.
- European External Action Service — air defence and European Peace Facility: procurement, reimbursement and delivery timing.
- IMF — Ukraine 2026 Article IV and Extended Fund Facility review: deficit, financing requirement, commitments and reserves.
- SIPRI — 2025 military expenditure: Ukrainian, Russian and wider European military spending.
- UNHCR — Ukraine displacement data: internally displaced population estimate sourced to IOM.
- European Commission — EU military support to Ukraine: aggregate support, training and ammunition-production capacity.
Tigray and Ethiopia
- United Nations — Secretary-General statement, 23 September: confirmed escalation and airport seizures.
- United Nations — daily briefing, 25 September: reported food-truck strike, road interruption and communications disruption.
- IOM — National Displacement Report portal: detailed pre-event Tigray displacement baseline.
- IOM — Ethiopia Crisis Response Plan 2026: national displacement and returnee context.
- World Food Programme — Ethiopia operations: planned reach and national logistics capacity.
- European Commission — Ethiopia humanitarian situation: aid-dependent displacement and humanitarian-need baseline.
- IMF — fifth review for Ethiopia: reserves, import cover, inflation and fiscal indicators.
- IMF and World Bank — Ethiopia debt sustainability analysis: debt-distress assessment.
- World Bank — Ethiopia reform and macroeconomic background: reform progress, foreign-exchange conditions and historical conflict context.
El Niño and food supply
- NOAA Climate Prediction Center — ENSO diagnostic: observed ocean anomaly and strength probabilities.
- FAO — Cereal Supply and Demand Brief: global production, use, stocks and stocks-to-use ratio.
- FAO — Food Price Index: observed international food-price movements.
- USDA — Grain: World Markets and Trade: rice production, consumption and trade forecasts.
- USDA — palm-oil production data: producer concentration.
- FAO — agrifood policy highlights: anticipatory import, stock and trade-policy responses.
- World Bank — Commodity Markets Outlook, April 2026: food, energy and fertiliser baseline and risk channels.
- WFP — 2015–16 Southern Africa El Niño response: historical regional production and humanitarian effects.
US–China trade and technology
- White House — Trump–Xi state-visit fact sheet: US account of institutional and commercial outcomes.
- USTR — Board of Trade recommendations: proposed product coverage and share of US exports.
- Chinese Foreign Ministry — state-visit account: Chinese account of the agenda and bilateral discussions.
- US Census Bureau — trade in goods with China: current bilateral trade values.
- US International Trade Commission — effects of Section 301 tariffs: retrospective import, production and price estimates.
- US Bureau of Industry and Security — semiconductor export-licensing policy: current control architecture.
- IMF — Changing Global Linkages: bloc trade, investment and connector-economy relationships.
- IMF — ASEAN and supply-chain reallocation: ASEAN production, investment and US import-share evidence.
- OECD — Global Value Chain Repositioning: experimental domestic-value-added evidence.
- WTO–IMF — Trade Policy Activity Index: wider trade-policy intervention.
Report-specific limitations
| Limitation | Why it matters | Treatment in this report |
|---|---|---|
| Event and evidence cutoffs differ | Evidence published on 28–30 September could be mistaken for part of the 20–27 September event window | Post-cutoff material is identified and used to test the earlier assessment, not rewrite the event chronology |
| Conflict data are incomplete | AIS manipulation, inaccessible territory, operational secrecy and disputed battlefield claims limit precise measurement | Shipping, attack, production and territorial figures are attributed, dated and expressed with confidence caveats |
| Conflict parties report some key figures | Ukrainian production targets and official military claims may reflect advocacy as well as operational data | Capacity, targets and needs are attributed; they are not presented as independently audited delivery |
| Exposure is not realised impact | Trade share, displaced populations, planned funding and infrastructure capacity describe vulnerability rather than loss | The report separates exposure baselines from observed disruption and realised outcomes |
| Capacity is not usable spare capacity | Pipeline, factory, logistics and humanitarian-network capacity may already be occupied or unavailable | The analysis tests security, finance, insurance, inputs, permissions and delivery time before treating capacity as a buffer |
| Global averages hide distribution | Comfortable cereal stocks or aggregate reserves can coexist with severe local scarcity | Country, sector and household exposure is considered separately from global totals |
| Food stocks are not automatically tradable | Stocks may be held by governments, located away from demand or restricted by policy | Concentration, trade rules and accessibility are analysed alongside the global stocks-to-use ratio |
| Data vintages and market years differ | FAO, USDA, IMF and national series may use different calendars, revisions and definitions | Calculations use values from the same source and vintage; incompatible totals are not combined |
| Calculated ratios are analytical estimates | Shares such as bypass coverage, rice concentration and package coverage may appear more precise than the inputs allow | Inputs and formulas are stated, results are rounded and operational limitations are explained |
| Historical coefficients are not forecasts | Pass-through and policy effects vary with currencies, taxes, subsidies, stocks and market structure | Historical research tests the mechanism; it is not mechanically applied to 2026 |
| Scenarios are conditional | Alert thresholds could be misread as probability forecasts or management targets | Likelihood, impact and confidence are separated; thresholds are explicitly identified as UFOQ.AI monitoring rules |
| Causal overlap creates double-counting risk | Energy, war, trade, climate and financial conditions can influence the same prices or growth outcomes | The report tests each causal arrow and avoids assigning the full outcome to more than one event |
| Public data cannot resolve every dependency | Current war-risk insurance, interceptor inventories, Chinese value added through third countries and government food-stock accessibility remain incomplete | These gaps are stated rather than replaced with unsupported estimates |
Interpretation boundary
The report is a risk-foresight assessment, not a prediction, investment recommendation or estimate of guaranteed loss. Its conclusions describe the strongest supported causal pathways as of the evidence cutoff. They should be updated when operating data, policy implementation or independently verified conflict information materially changes.