Houthi Presence Near Perim Raises Bab el-Mandeb Risk as Hormuz Traffic Remains Severely Constrained

Bottom Line

The probability of a distinct reciprocal US-Iran kinetic episode by 01:18Z on 12 September is 65 per cent, likely, down from the prior 80 per cent. The prior definition remains in force: a separate reciprocal kinetic episode involving US and Iranian forces, excluding the reported 8-10 September maritime and Al-Azraq incidents and excluding Houthi, Iraqi militia or other partner attacks unless US or Iranian forces directly participate in the same episode. No qualifying episode is independently verified in the supplied record before the cutoff, which falls 57 minutes after the end of the collection window. Confidence is moderate.

The Strait of Hormuz is very likely to remain severely constrained through 21 September, with an 82 per cent probability. Severe constraint means daily Kpler traffic below 50 per cent of the 10-day average, or UKMTO movement near one-quarter of the pre-conflict norm. Kpler recorded seven transits on 10 September against an average of 15, while UKMTO assessed movement at roughly 20 per cent of normal. The measures are not directly comparable, and QatarEnergy-controlled Al Marrouna completed an LNG movement, so a complete closure remains unlikely at 20 per cent.

Houthi forces very likely retain Mocha and Dhubab. Houthi personnel reaching Perim is likely at 70 per cent, sustained presence is roughly even at 55 per cent, and control is unlikely at 35 per cent. Four Yemeni government sources reported the Perim advance, but no independent imagery or evidence of anti-ship systems, radar or established resupply is supplied. The immediate assessed effect is increased exposure for Saudi-linked traffic approaching Bab el-Mandeb, not verified Houthi interdiction capability.

Key Developments

• US-Iran maritime exchange, reported fact: The supplied reporting attributes five Iranian tanker sinkings to the United States, 10 ship attacks near Hormuz to Iran and damage to two US warships to the IRGC. US Central Command and Marco Rubio deny that a US vessel was hit. The outlet for the tanker-sinking claim and independent hull-loss evidence are not identified in the supplied record.

• Hormuz, reported fact: Kpler recorded seven transits on 10 September against a 10-day average of 15. UKMTO assessed overall movement at roughly 20 per cent of the pre-conflict norm. A separate UKMTO series recorded 108 outbound and 102 inbound movements over seven days. These series cover different populations. Al Marrouna completed an LNG delivery from Ras Laffan to Pakistan.

• Perim, reported fact: Four Yemeni government sources reported that Houthi forces reached Perim after taking Mocha and Dhubab, while Saudi-backed Yemeni government forces reportedly withdrew. The supplied reporting does not identify the Yemeni units involved and provides no independent imagery. The identity of the sources affects confidence because all four are connected to a party to the conflict.

• East-West Pipeline, reported fact: Saudi Ministry of Energy reporting during 10-11 September described damage near two pumping stations, injuries and a precautionary shutdown. Satellite imagery showed smoke and fire. The station names, coordinates, exact incident time, mechanism, shutdown duration and restart status are not reported. US officials attributed the attack to drones launched from Iraq; other reporting attributed it to the Houthis.

• Ukraine, reported fact and assessment: Reporting on 11 September described Russian jet-powered drone attacks on Kyiv petrol stations that killed two people, damage to a nine-storey residential building and damage to the central SBU building. Ukrainian reporting also attributes strikes against Novorossiysk, Volgograd and Russian energy infrastructure to Ukraine. Another Russian long-range strike against Ukraine by 26 September is assessed at 80 per cent, very likely, with high confidence.

• Europe, reported fact and assessment: German federal authorities attributed the Leipzig/Halle Airport drone incident to Russia, but public evidence remains limited in the supplied reporting. UK prosecutors allege that Joshua Cammidge conducted GRU-directed surveillance of four British drone-manufacturing sites. EU ambassadors had not agreed on renewing the Russia sanctions regime before 15 September. Continued Russian-linked hybrid activity through December is assessed at 80 per cent, very likely; failure or delay in sanctions renewal is assessed at 45 per cent, roughly even chance.

• Sudan, reported fact and assessment: UN reporting describes increased long-range drone use by both the Sudanese Armed Forces and the Rapid Support Forces. UN, MSF and International Organization for Migration reporting describes attacks on civilian infrastructure, clinic closures and displacement in Blue Nile and around El Obeid. Further drone attacks on civilian sites or infrastructure during 12 September to 12 October are assessed at 85 per cent, very likely, with high confidence.

Analysis

Analytic continuity

Maritime interpretation and mechanism

The maritime record supports severe constraint rather than verified closure. Kpler's seven-vessel count is a daily vessel-transit measure against a 10-day average of 15. UKMTO's roughly 20 per cent figure is an assessment of overall movement against a pre-conflict baseline. The separate 108 outbound and 102 inbound figures are a seven-day aggregate, while approximately 20 US-facilitated southern-route vessels per day and six AIS-derived movements measure different traffic sets. AIS-off passages reduce visibility. These measures therefore support the direction of the judgement without being added together.

The competing interpretation is that the tanker and ship incidents represent interdiction, warning fire or coercive boarding rather than a distinct reciprocal US-Iran combat episode. Vessel-owner reports, hull-loss evidence, geolocated satellite imagery, launch signatures and independent casualty records would discriminate between the interpretations. A confirmed US or Iranian platform loss linked to a reciprocal exchange would raise the direct-conflict probability. Intact vessels, boarding records, warning broadcasts and absence of a reciprocal impact would lower it.

Perim is adjacent to the Bab el-Mandeb approach, but territorial presence alone does not establish effective interdiction. No supplied source reports Houthi anti-ship missiles, radar, persistent observation equipment or a resupply route on the island. The most immediate mechanism is increased observation and threat potential against Saudi-linked traffic, especially vessels approaching or departing the Red Sea. A sustained military presence supported by weapons and logistics would create a more direct risk to Saudi-linked vessels. A temporary landing without resupply would have a smaller effect.

East-West Pipeline and energy implications

The Saudi Ministry of Energy reported a precautionary shutdown after damage near two pumping stations during 10-11 September. Satellite imagery showed smoke and fire, including extensive fire damage at one location. The supplied record does not identify the pumping stations or provide a verified throughput loss. US officials attributed the incident to drones launched from Iraq, while other reporting attributed it to a Houthi attack. The reported physical damage is more credible than either attribution because the imagery and official reporting partially converge, but the duration and supply effect remain unresolved.

The operational consequence for Saudi Aramco depends on restart time and measured flow. A restart within 72 hours with stable throughput would reduce the supply-risk assessment. A shutdown beyond 72 hours, lower Jazan refinery utilisation or a confirmed fall in East-West Pipeline flow would increase it. Gulf crude exports above 10 million barrels per day indicate partial supply continuity, not normality. Brent prices above US$100 per barrel and elevated VLCC rates reflect both Hormuz and Red Sea risk, so the contribution of the pipeline incident cannot be isolated.

Oman arrangement and failure modes

The reported 14 September Oman meeting involving Iran, Iraq and Gulf Arab states is a maritime deconfliction channel rather than evidence of a broad ceasefire. A workable arrangement could establish vessel notification, designated routes, inspection rules or limited escort procedures. It could exclude US-linked or Saudi-linked shipping, lack enforcement, or permit one party to defect while preserving the appearance of an agreement. A joint communiqué with participants, route rules and enforcement provisions would support a substantive arrangement. Continued attacks during or immediately after the meeting would indicate that the mechanism is either narrow or ineffective.

Stakeholder consequences

For the US Fifth Fleet, a verified US vessel loss, confirmed Iranian attack on US personnel or 48 hours without an independently verified commercial Hormuz transit would increase the force-protection and maritime-surveillance burden. For the Jordanian Armed Forces, another confirmed missile launch against Muwaffaq Salti Air Base or a fall in interception effectiveness would increase base-defence pressure. These are consequences, not recommendations.

For Saudi Aramco, the key decision-relevant facts are East-West Pipeline restart time, pumping-station throughput and Jazan refinery utilisation. For QatarEnergy, a second verified LNG movement through Hormuz by 21 September would demonstrate continued access for selected cargoes, while a missed scheduled movement would increase delivery uncertainty. For Japanese and South Korean LNG buyers, two consecutive missed cargoes or a delay exceeding 48 hours would indicate scheduling stress, but the current record does not establish a physical shortage.

For Maersk, MSC, Hapag-Lloyd, BP and Frontline, three consecutive Kpler counts below eight transits, or a confirmed Houthi attack on a commercial vessel near Bab el-Mandeb, would support continued rerouting and higher chartering costs. The reported 26 commodity vessels through Bab el-Mandeb on 10 September show that the passage remains active, so a full Red Sea shutdown is not the present assessment.

Alternatives

• Compartmented conflict continues through 26 September, with restricted Hormuz traffic and Houthi pressure on Saudi-linked shipping: 45 per cent, roughly even chance. This fits the current combination of severe but incomplete maritime disruption, Houthi gains and no broad ceasefire mechanism.

• The Oman meeting produces a narrow navigation arrangement without ending US-Iran hostilities: 25 per cent, unlikely. The arrangement could cover notification or designated routes and exclude US-linked or Saudi-linked vessels. The lack of an enforcement mechanism and incompatible positions reduce the probability.

• A confirmed direct US-Iran episode or mass-casualty attack on a US or major commercial vessel produces wider escalation: 15 per cent, unlikely. The pathway exists, but independent confirmation of US ship damage is absent and Washington has reportedly avoided direct Houthi strikes.

• The Perim report proves temporary or inaccurate, the Yemeni government recovers the island and traffic recovers: 15 per cent, unlikely. This requires no sustained Houthi presence and Kpler traffic reaching at least 50 per cent of its 10-day baseline for two consecutive days.

Indicators & Warnings

• Qualifying US-Iran episode: Seek matching US Department of Defense or US Central Command and Iranian or IRGC accounts, supported by vessel-owner reporting, geolocated imagery, satellite evidence or independent casualty reporting. A Houthi or Iraqi militia attack alone does not qualify. Horizon: by 01:18Z on 12 September, with confirmation through 13 September.

• Hormuz severe constraint: Use Kpler's daily series against the current 10-day average of 15. Three consecutive days below eight transits would support continued severe constraint. Two consecutive days at or above eight, plus a verified crude or LNG movement, would weaken it. Treat AIS-off passages as unobserved. Horizon: 12-15 September and through 21 September.

• Hormuz closure: No Kpler-recorded commercial entry or exit, no UKMTO-confirmed commercial movement and no verified AIS-off crude or LNG passage for 48 hours would support closure. One verified passage resets the observation clock. Horizon: 12-21 September.

• Perim consolidation: Commercial satellite imagery showing personnel, vehicles, fortifications, radar, missile equipment or repeated resupply would support sustained presence. A verified Yemeni government return or no Houthi activity across three imagery collections would weaken control. Horizon: 12-26 September.

• Oman talks: A joint communiqué naming participants and specifying routes, notification, inspection and treatment of US-linked and Saudi-linked vessels would support deconfliction. Failure to convene, continued attacks or a communiqué without operating terms would not. Horizon: 14-17 September.

• Saudi energy infrastructure: Saudi Ministry of Energy, Saudi Aramco or satellite-derived throughput confirmation of a restart and stable flows for 72 hours would reduce risk. A shutdown beyond 72 hours, new fire or smoke, lower pipeline flow or reduced Jazan refinery utilisation would increase it. Horizon: 12-26 September.

• Russia-Ukraine strikes: Russian launch warnings, Ukrainian General Staff air-defence reporting and independently geolocated impacts on Kyiv, Odesa, energy infrastructure or other Ukrainian targets would confirm another strike. A verified seven-day reciprocal pause would weaken the forecast. Horizon: 12-26 September.

• EU sanctions: A unanimous Council renewal before 15 September would break the non-renewal concern. A lapse, formal delay or recorded veto would confirm it. Horizon: by 15 September.

• Yemen and Sudan humanitarian conditions: UN Office for the Coordination of Humanitarian Affairs, International Organization for Migration and MSF reporting of new displacement, detention of aid workers, clinic closures or further drone strikes on civilian infrastructure would confirm deterioration. Horizon: 12 September to 12 October.

Alternatives

The four alternatives above are mutually exclusive by dominant outcome through 26 September and sum to 100 per cent. The Perim probabilities in the Outlook are conditional judgements about the island's status and are not additional scenario probabilities.

Outlook

The probability of a distinct reciprocal US-Iran kinetic episode by 01:18Z on 12 September is 65 per cent, likely, with moderate confidence. Severe Hormuz constraint through 21 September is 82 per cent, very likely. Complete closure is 20 per cent, unlikely. Houthi retention of Mocha and Dhubab through 26 September is 80 per cent, very likely. Houthi personnel reaching Perim is 70 per cent, likely; sustained presence is 55 per cent, roughly even chance; and control is 35 per cent, unlikely.

Houthi attacks on Saudi territory or energy infrastructure through 26 September remain 70 per cent, likely. A durable US-Iran ceasefire by 23 September remains 20 per cent, unlikely. The 14 September Oman meeting producing a limited maritime arrangement is 35 per cent, unlikely. Another Russian long-range strike against Ukraine by 26 September is 80 per cent, very likely. Russian-linked European hybrid activity through December is 80 per cent, very likely. Further SAF or RSF drone attacks on civilian sites or infrastructure during 12 September to 12 October are 85 per cent, very likely.

The overall confidence is moderate because the central US-Iran and Perim judgements depend on disputed or party-derived reporting. Confidence is high for observed Hormuz disruption, continued Russian strike activity and the Sudanese drone-war trend, but those peripheral judgements do not raise confidence in the central attribution and escalation questions.

CrisisBrief Index

The CBX is 75, labelled SEVERE. It uses a fixed daily weighting for comparability: conflict 40 per cent, maritime trade 25 per cent, energy 20 per cent and political risk 15 per cent. Each component is scored from 0 to 100 against the recent baseline in the supplied record, where 0 represents no material crisis activity and 100 represents systemic, multi-theatre disruption. The calculation is 0.40 times 82, plus 0.25 times 78, plus 0.20 times 75, plus 0.15 times 52, which rounds to 75. The descriptive uncertainty is approximately plus or minus five points because traffic populations, casualty reporting and attribution are incomplete.

Conflict is 82 because US-Iran exchanges, Houthi territorial gains, Russia-Ukraine strikes, Israel-Hezbollah activity and Sudanese drone warfare are active. Maritime trade is 78 because Hormuz movement is near one-fifth of the UKMTO pre-conflict norm while Bab el-Mandeb remains exposed to selective Houthi pressure. Energy is 75 because oil is above US$100 per barrel, VLCC rates are elevated and the East-West Pipeline's status is unresolved, although Gulf crude and LNG continue to move. Political risk is 52 because the EU Russia sanctions deadline is unresolved, Oman diplomacy lacks a broad ceasefire mechanism and Sudanese diplomacy has not resolved the nationwide arms-embargo question.