Hormuz Traffic Stays Severely Restricted as Houthi Reports Put Mokha Under Pressure

Reporting cutoff: 5 September 2026, 0119Z. Forecasts run from that cutoff.

Bottom Line

US-Iran military exchanges are likely to continue within 72 hours, while commercial passage through the Strait of Hormuz is very likely to remain severely restricted through 15 September. Reports of Ansar Allah ground, missile and drone operations near Mokha raise the risk to Bab al-Mandab shipping, but independent evidence does not establish Houthi control of positions overlooking the strait.

Key Developments

Hormuz

Reporting: Kpler recorded four commodity-vessel passages through the Strait of Hormuz on 3 September and four by 04:55 GMT on 4 September. Rystad Energy estimated wartime oil and product flows at roughly 4 million to 6 million barrels per day. Vice-President JD Vance separately claimed approximately 15 million barrels per day and restored traffic.

Assessment: Commercial passage is very likely to remain severely restricted through 15 September, assessed at 85 per cent with moderate confidence. Severe restriction means that the matched Kpler vessel category remains below half of its pre-war baseline for three consecutive days, with crude-tanker or LNG movements also materially below baseline, or that confirmed detention, convoy or insurance restrictions persist. A complete closure, defined as zero authorised commercial passage for 48 consecutive hours, is unlikely at 20 per cent. Normal operations, defined as at least 80 per cent of the matched pre-war Kpler baseline for three consecutive days without detention, convoy or insurance restrictions, are very unlikely at 10 per cent.

Basis and caveat: The four-vessel count cannot be compared directly with the reported pre-war total of roughly 125 large vessels per day because the categories, observation windows and coverage differ. Kpler is high-proximity commercial tracking data for observed movements, but AIS gaps, delayed passages, escorted movements and incomplete coverage prevent direct conversion of a snapshot into total oil-flow volume. Rystad provides an energy-market estimate. Vance's statement is a first-party political claim. These sources do not measure the same variable.

Red Sea and Bab al-Mandab

Reporting: Local Yemeni accounts and trade reporting dated 3 to 5 September describe Ansar Allah missile, drone and ground operations in Taiz governorate, activity around Maqbanah, Mawzaa and Jabal Habashi, and an effort to isolate Mokha. Some reports claim Houthi control of elevated positions overlooking the Red Sea. Independent imagery confirming permanent control of those positions is not present in the supplied material. Chinese shipping companies and other operators continue to divert some Asia-Europe services around the Cape of Good Hope, but the material does not identify individual carriers.

Assessment: The threat to commercial shipping near Bab al-Mandab is likely to remain elevated over the next 14 days, assessed at 70 per cent with moderate confidence. Continued diversion or suspension of some Asia-Europe services is likely at 65 per cent. A full physical closure of Bab al-Mandab is unlikely at 25 per cent.

Basis and caveat: The forecast relies on reported new kinetic activity, existing Houthi anti-shipping capability and commercial rerouting, not on a confirmed front-line change. The territorial-control component has low confidence because the reports are local, partially repetitive and not independently geolocated. Hodeidah is a separate humanitarian issue. The port handles approximately 80 per cent of food and humanitarian aid destined for Yemen, but the supplied evidence does not establish that operations have stopped.

Ukraine

Reporting: The Ukrainian Air Force and Ukrainian regional authorities reported that Russia launched 121 strike drones and one Kh-31P missile during the night of 3 to 4 September. Ukrainian authorities reported 103 drones destroyed or suppressed, impacts at 15 locations and damage or casualties in Kyiv, Odesa and other areas. The reported strike on the Security Service of Ukraine headquarters in Kyiv wounded 12 people. Odesa reporting recorded one death and at least five injuries.

Assessment: Russia is very likely to conduct at least one further high-volume strike against Ukrainian urban, energy, rail or port infrastructure during the next 14 days, assessed at 85 per cent with high confidence. Ukraine is likely to conduct a retaliatory long-range strike against a Russian military, aviation, energy or logistics target during the same period, assessed at 65 per cent with moderate confidence. A mutual operational pause is unlikely at 25 per cent.

Basis and caveat: The persistence forecast rests on Russia's demonstrated launch capacity, repeated recent attacks and stated military objectives, not solely on the latest strike. Ukrainian Air Force reporting is close to the event. Regional authorities and major media provide additional corroboration, although casualty and damage figures remain less certain than the reported launch activity.

Sudan

Reporting: The UN Independent International Fact-Finding Mission reported on 3 September that foreign personnel, weapons, training, military technology and logistics are sustaining RSF operations, including networks involving Colombia, the UAE, Chad, southeastern Libya and Bosaso. The UAE and Chad deny supporting RSF operations. The mission and other reporting describe foreign-supplied drones used by both the RSF and SAF, but the evidence is stronger for specific RSF-linked networks than for equivalent external support to both sides.

Assessment: RSF and SAF drone attacks across Darfur, Kordofan and Blue Nile are very likely to remain a principal source of civilian harm during the next 30 days, assessed at 85 per cent with moderate confidence. The RSF is likely to retain access to reported external personnel and logistics during the next 1 to 3 months, assessed at 75 per cent with moderate confidence. Effective interruption of those networks is unlikely at 20 per cent.

Basis and caveat: The UN mission is a high-reliability institutional source for documented patterns, but its findings do not establish direct UAE government responsibility. Local casualty accounts provide proximity but vary in reliability and independence. This judgement does not imply that foreign support to the RSF and SAF is equally established.

Europe

Reporting: German authorities reported more than 165 probable sabotage cases, 747 suspicious drone detections involving 1,068 flights over critical infrastructure, military bases and defence-industrial sites, and 112 espionage-related offences during 2026. Germany attributes the 4 August Leipzig-Halle operation to Russian state actors, while other accounts dispute aspects of that case and attribute separate incidents to climate activists, criminal actors or lone perpetrators. Recent reported cases include Jänschwalde, Dormagen, Weisweiler and a Rhode & Schwarz facility in Munich.

Assessment: Sabotage, espionage and suspicious-drone activity affecting European defence, energy, transport and military-support sites is likely to continue through 4 December, assessed at 70 per cent with moderate confidence. A further incident receiving credible, independently corroborated Russian state attribution during the next 30 days is assessed at a roughly even chance with low confidence. An overt Russian conventional attack on NATO territory remains very unlikely at 10 per cent.

Basis and caveat: German official statistics establish the volume of reported incidents but not a common Russian command structure. Attribution to a specific Russian intelligence service is not established in the supplied material. The broad activity forecast is therefore stronger than the Russian-directed campaign assessment.

Analysis

The principal 24 to 72 hour risk is a further US-Iran exchange. The capability evidence is direct: Iran has launched missiles and drones at US-linked facilities, and the United States has struck IRGC-associated air-defence, radar, naval and communications sites. Intent is supported by Iranian retaliation statements and US conditions for talks. Preparation is indicated by continuing force deployments and regional air-defence activity. The main interruption pathway is a verified diplomatic pause accompanied by a halt in launches and strikes from both sides for at least 48 hours.

Hormuz traffic requires separate treatment from oil-flow estimates. A vessel-transit count records tracked movements, not barrels per day. Oil-flow estimates include different vessel classes, loading dates, pipeline movements, storage releases and tracking assumptions. The current evidence supports severe restriction, but it does not resolve the difference between Vance's reported 15 million barrels per day and Rystad's 4 million to 6 million estimate. The most reliable near-term indicator is a matched Kpler series, supplemented by port, operator and insurance reporting.

The Houthi assessment is narrower than a finding that Ansar Allah has seized the approaches to Bab al-Mandab. The reported ground offensive indicates intent and local combat activity. Existing missile and drone capability indicates means. Commercial rerouting provides an independent consequence indicator. The missing element is independently verified, durable control of firing or surveillance positions near Mokha. The central forecast therefore concerns continued threat and routing disruption, not confirmed territorial consolidation.

The Ukrainian strike forecast has higher confidence than the Houthi territorial assessment. Ukrainian military reporting, regional authorities and major media broadly corroborate the 121-drone attack, although they do not independently establish every damage claim. A further Russian strike is supported by the campaign's recent persistence, available launch capacity and the absence of a reciprocal pause. Ukraine's retaliatory forecast is weaker because the evidence is based mainly on stated intent and does not establish preparations.

The Sudan judgement has been narrowed since the previous draft. The evidence supports reported foreign networks associated with RSF operations and reported foreign-supplied capabilities on both sides. It does not establish equivalent external support, direct UAE government responsibility or an effective supply system for every reported route. The UN mission's institutional reporting raises confidence in the existence of external networks, but denials by the UAE and Chad and the absence of claim-level sourcing reduce confidence in attribution and future disruption forecasts.

European hybrid activity is also separated from attribution. The German figures establish a broad security workload. They do not establish that all 165 sabotage cases, 747 drone detections or 112 espionage offences are Russian-directed. A Russian state role in individual cases remains an important hypothesis, but continued incidents without forensic attribution remain the more defensible alternative.

Stakeholder implications

• CENTCOM and host governments: Jordan, Kuwait, Bahrain, Iraq and the UAE face continued risk of missile and drone alerts within 72 hours. A confirmed new impact would affect base operations, air-defence inventories and force-protection measures. The supplied material does not establish current US interceptor stocks.
• QatarEnergy and European LNG importers: Severe Hormuz restriction through 15 September increases the risk of delayed Gulf LNG movements. The prior reporting that QatarEnergy cancelled 29 cargoes and extended force majeure supports a delivery-risk assessment through early November, but the current packet does not provide a new cargo schedule.
• Asia-Europe shipping operators: Operators that continue Suez routing face exposure to Houthi missile and drone attacks near Bab al-Mandab. Operators diverting around the Cape of Good Hope face longer voyages, higher fuel use and insurance costs. The supplied reports do not identify individual carriers, so no company-specific routing assessment is made.
• Ukrainian Railways, port authorities and energy operators: Another Russian strike involving energy, rail or port infrastructure within 14 days would increase reliance on alternative routes and create additional service interruptions. The latest attack already affected Kyiv, Odesa and other locations, but the supplied material does not quantify aggregate network downtime.
• WFP, OCHA and humanitarian agencies: El Fasher and Hodeidah require separate monitoring. Closure of the remaining El Fasher-Zamzam route would restrict assistance to displaced people in North Darfur. Damage or cargo suspension at Hodeidah would affect Yemen's main aid-entry system. Neither full stoppage is established in the current reporting.
• German and Baltic security authorities: The volume of suspicious drone, sabotage and espionage reporting indicates continuing investigative and force-protection pressure. A forensic link to a Russian service would change attribution confidence; additional incidents without such evidence would not establish a coordinated Russian campaign.

Analytic continuity

Source quality and limitations

The packet contains 86 records in the highest reliability category, 72 in the next category, 36 in a middle category, 27 low-reliability records, six very low-reliability records and one unusable record. The supplied material does not map those aggregate grades to individual claims and does not provide article-level links or publication identifiers. The assessments therefore rely on the named source type, proximity, independence and corroboration described in the material.

Kpler and Rystad provide high-proximity commercial data and analysis for vessel movements and energy flows, but they measure different variables. The Ukrainian Air Force, CENTCOM, the Iranian military, Gulf governments and German authorities provide direct access to their own operations or security reporting, but each has an incentive to shape public interpretation. The UN Fact-Finding Mission provides the strongest institutional basis for the Sudan external-support assessment. Local Yemeni and Sudanese reporting improves proximity but has weaker independent verification. Major-media reporting adds corroboration when it does not simply repeat an official or local account.

Indicators & Warnings

• US-Iran: Within 72 hours, two independent reports of a new Iranian launch against a US-linked facility or a new US strike on an IRGC-associated site would confirm persistence. A verified 48-hour halt by both sides would support the diplomatic-pause alternative.
• Hormuz: Compare identical Kpler vessel categories and time windows. Three consecutive days at or above half of the matched pre-war baseline would indicate partial recovery. Zero authorised commercial passage for 48 hours, confirmed by an independent operator, port or government source, would materially raise the closure probability.
• Bab al-Mandab: Geolocated evidence of permanent Houthi control of coastal artillery, radar or elevated positions near Mokha, or a confirmed attack on a named commercial vessel, would confirm increased maritime exposure within 3 to 14 days.
• Ukraine: Another attack involving at least 100 one-way drones, or combined missile and drone strikes on Ukrainian energy, rail or port infrastructure, would confirm persistence within 14 days. Two successive nights below 30 one-way drones with no missile launch would support a temporary reduction.
• Sudan external support: New independently documented RSF-linked foreign personnel, drones or logistics would support continued network access. A documented interdiction followed by a sustained fall in deliveries would support the interruption alternative within 1 to 3 months.
• El Fasher: Closure of the remaining Zamzam-El Fasher route, suspension of Norwegian Refugee Council cash assistance or a verified halt in aid convoys would indicate worsening access within 7 to 30 days.
• Hodeidah: Port damage, suspension of cargo clearance or a sustained reduction in humanitarian arrivals would indicate worsening Yemen access within 7 to 30 days. These indicators are separate from El Fasher.
• Europe: Two independently corroborated sabotage, espionage or suspicious-drone incidents affecting German, Polish or Baltic defence, energy or transport infrastructure within seven days, followed by forensic evidence linking an identified Russian service, would support the Russian state-directed hypothesis within 7 to 30 days.

Alternatives

• US-Iran diplomatic pause: Unlikely, 25 per cent. A halt in launches and strikes by both sides for 48 hours, followed by matched Kpler traffic increasing towards its baseline, would distinguish this branch. The immediate consequence would be lower base-protection pressure and gradual improvement in tanker and LNG movements.
• Controlled Hormuz recovery without a ceasefire: Unlikely, 30 per cent. Three consecutive days of matched traffic above half of the pre-war baseline, renewed LNG passages and no new detention notices would indicate selective reopening. This would reduce closure risk but would not restore normal operations.
• Limited Houthi gains: Roughly even chance, 45 per cent. Continued attacks and rerouting without independently verified control of firing or radar positions would indicate that the reported ground advance was temporary or overstated. Shipping risk would remain elevated without a new coastal threat position.
• Lower Russian strike tempo: Unlikely, 25 per cent. Two successive nights below 30 one-way drones, no missile launch and no major Ukrainian retaliation would indicate a temporary operational pause, not a substantive ceasefire.
• RSF supply interruption: Very unlikely, 15 per cent. A documented interdiction or sanctions action followed by a sustained fall in new personnel, drones or weapons would support this branch. It would increase pressure on RSF forces but would not end Sudan's war.
• Unresolved European attribution: Likely, 55 per cent. Continued incidents without forensic links to a Russian service would support the criminal, activist or unattributed hypothesis. Security pressure would remain, but a coordinated Russian campaign would not be established.
• Dual-chokepoint maritime shock: Very unlikely, 10 per cent. A confirmed named-vessel attack, multiple operator suspensions and zero authorised passage at Hormuz and Bab al-Mandab for 48 hours would indicate this wildcard. The consequence would be rapid pressure on Gulf oil and LNG deliveries, Cape-route capacity and maritime insurance costs.

Outlook

A further defined US-Iran exchange within 72 hours is likely at 70 per cent with moderate confidence. Severe Hormuz restriction through 15 September is very likely at 85 per cent with moderate confidence. A complete closure is unlikely at 20 per cent, and normal operations are very unlikely at 10 per cent.

The Bab al-Mandab shipping threat is likely to remain elevated over the next 14 days at 70 per cent with moderate confidence. Russia is very likely to conduct another high-volume attack against Ukraine during the same period at 85 per cent with high confidence. Ukraine is likely to conduct a retaliatory long-range strike at 65 per cent with moderate confidence.

RSF and SAF drone attacks in Sudan are very likely to continue during the next 30 days at 85 per cent with moderate confidence. The RSF is likely to retain access to reported external networks during the next 1 to 3 months at 75 per cent with moderate confidence. European hybrid incidents are likely to continue through 4 December at 70 per cent with moderate confidence, but a further incident with independently corroborated Russian state attribution during the next 30 days is only a roughly even chance with low confidence.

CrisisBrief Index

The CrisisBrief Index is 77, labelled SEVERE. It is a comparative daily workload measure, not a probability of global crisis. The index uses a fixed weighted average of conflict at 35 per cent, maritime trade at 25 per cent, energy at 20 per cent and political instability at 20 per cent. The components are scored against the same 0 to 100 scale each day and updated only when the reporting establishes a material change in intensity, geographic reach, disruption or governance risk.