Alliances are drafted for contingencies their signatories hope will remain theoretical. The Makkah Joint Defense Agreement – concluded in August by Saudi Arabia, Turkey and Pakistan – was one such instrument, yet early this week, less than two months after it was signed, the three states announced the activation of its collective deterrence measures. What set it in motion was not an invasion, nor the act of a rival state, but a campaign by a non-state movement based in the northern highlands of Yemen. It may prove the first of many state responses in this particular area.

The risk differs from that which markets have watched at Hormuz, a problem of leverage in which any actor able to threaten a narrow gateway can impose costs far beyond its physical power. Bab al-Mandab is where the workaround runs. Saudi loadings at Yanbu, on the Red Sea, rose more than sixfold in March, from roughly 0.8 million barrels per day to about 5 million bpd, once Hormuz was effectively shut. Around 80% of that volume initially moved south through the Bab al-Mandab Strait, until that route was jeopardized by the Houthi campaign launched in July.

This presents a different vulnerability – asymmetric escalation – where a limited action by an actor of modest means and loose control draws disproportionate responses from many more powerful ones, none of which intends to engage in a wider war. The asymmetry has three dimensions: of scale, since a narrow act produces broad consequences; of actor, since a movement acts while states and alliances must respond; and of control, since no one involved can ultimately determine where the sequence ends with certainty.

The same maritime system has produced this pattern before. In 1956 the crisis at Suez owed little to the quarrel over the canal that started it. Britain, France and Israel each came to the same waterway by a different road, the superpowers were drawn in behind them, and an Egyptian dispute ended as a verdict on the limits of post-war European power. Suez is the best template for what follows, with one difference: the party that began that quarrel was a state that could be negotiated with and pressured. The Houthis are not that.

The sequence
Houthi missile salvos fired at Israel in March widened over the summer into a campaign that broke the militia’s four-year ceasefire with Saudi Arabia. On 20 July the Houthis declared an embargo on Saudi vessels while indicating that other shipping could pass. Most vessels, reportedly, still pass Bab al-Mandab without incident, but weekly transits through the strait fell from 354 – the highest total since December 2023 – to 269 in the first week of the embargo.

In September the campaign changed in kind. Drone strikes on 10 September, which Riyadh attributed to launches from Iraqi territory, damaged pumping stations on the East-West pipeline, the kingdom's principal export artery since Hormuz was constrained. The line restarted on 22 September at reduced rates. In mid-September the Houthis also reportedly took control of the port of Mocha and Perim Island. Yemeni government forces say they retook Mocha on 5 October, but this is refuted by the Houthis. This is no longer only a maritime problem; it is a national security question for the state targeted.

The shoreline is of vital strategic importance. Perim divides the strait's narrowest point into two lanes, the Large Strait of about 17 kilometers and the Small Strait of 2.5 to 5 kilometers. A position on the island would let the Houthis rely on cheap, replenishable weapons, such as sea mines and explosive drone boats, and could allow them to impose transit fees, as Iran has attempted at Hormuz. So far, the pressure has been selective. That is the asymmetry of scale: inexpensive means, expensive consequences.

The actor matters as much as the geography. Iran has helped develop the Houthis' capabilities, but sponsorship is not traditional command and control. Tehran can enable and influence Houthi action without determining every decision. A sovereign government weighs its territory, economy and diplomatic relationships before it uses force; a non-state movement carries a different ledger. Even if the blockade is a bargaining move rather than a bid for wider war, a bargaining move by one party can become the trigger for another. This is the asymmetry of actor: in 1956 Washington's financial pressure on the British pound halted the Anglo-French operation against Egypt within days, but no comparable lever exists over a movement in the Yemeni highlands.

Why responses compound
The Makkah agreement shows the sequence is in motion, at least politically. It went unused for two months while Pakistan and Turkey, which had positioned themselves as mediators in the wider conflict with Iran, weighed their roles. The commitment made on 5 October was an announcement pertaining to collective deterrence measures, not a deployment. No troop numbers or dates have been given, Pakistan's forces were already in Saudi Arabia in supporting roles, and Turkey's parliament has yet to ratify the pact, although statements from Islamabad and Ankara on 8 October may define the commitment and accelerate it.

Amid such escalation,  each response may be proportionate: a defensive deployment is read as preparation for offense, a naval escort as a widening commitment, a commercial diversion as evidence that disruption is permanent. Suez is again the guide. Each of the 1956 participants acted on what looked like a sound reading of its own interests, and the sum was a canal blocked by sunken ships, a disrupted European oil supply and a British prime minister who resigned within months. No one chose the whole. That is the asymmetry of control: each decision is reasonable on its own terms, and their sum is not.

The same asymmetry also supplies reasons for restraint. A cheap provocation is expensive to answer, and the stronger parties know it. Saudi Arabia has little appetite for another prolonged confrontation with the Houthis, and Turkey and Pakistan have even less reason to be drawn into direct combat, which is consistent with the supporting roles they have so far described. Suez points the same way: the escalation was bounded within days, and Britain's prime minister paid with his office for pushing it.

Where the cost lands
Each workaround has become the next chokepoint. When Hormuz was effectively shut after the war began on 28 February, Middle East crude loadings fell from about 19.4 million barrels per day in February to about 10 million in March, and the Red Sea became the main outlet for Saudi crude. As the Houthi campaign pushed vessels north, away from the southern route, Egypt's Sumed pipeline, with 2.5 million bpd of capacity, absorbed much of the diversion. Crude is offloaded at the Red Sea end, piped across Egypt and reloaded on the Mediterranean coast. The Suez Canal cannot do that job on its own, because only vessels up to Suezmax size can transit it fully laden, so the largest crude carriers loading on the Red Sea either depend on Sumed or must still pass south through Bab al-Mandab. A barrel rerouted away from one chokepoint inherits the risk of another.

The adaptation is not temporary, and Suez set the precedent. When Egypt closed the canal in 1967 it stayed shut until 1975, and shipowners responded by building ever larger tankers that made the long route around the Cape of Good Hope economic. Sumed itself was completed in 1977 as a consequence. The detour became the standard route, and the ships built for it could no longer use the canal fully laden even after it reopened. Vessel tracking shows the same pattern in LNG today. In 2023, about a third of US LNG exports to Asia went through Suez. After mid-January 2024, the trade virtually stopped using the canal, and by 2025 a dominant 88% went via the Cape.

The cost of the cascade shows in prices. Brent traded above $108 in mid-September and hovered between $98 and $101 in the first week of October. July price scenarios spanned from roughly $65 under full resolution to about $148 if fighting escalated and Bab al-Mandab were disrupted; some of this is being borne out now. Freight has changed too, as shipowner earnings on a direct Saudi-to-China VLCC voyage passed $1 million a day in September, and the shuttle arrangements then in use lifted the cost of landing a barrel in China to about $26, from around $2 before the conflict.

Resilience, in other words, is relative: a bypass that funnels through its own chokepoint is less protective than it looks on a map, so the useful question is how many actors a route depends on. That favors diversified supply and shorter, less contested routes to market. Gulf producers remain indispensable. Atlantic Margin basins are a complement, not a replacement, with constraints of their own but less exposure to a non-state actor able to set off a cascade that no one fully controls.

Containment
Three conclusions follow. The actor is of limited means and loosely controlled, so deterrence works imperfectly. Responses are individually proportionate and collectively not, so every defensive step alters the calculations of others. And markets can absorb a longer route but have yet to price a cascade. The central case remains a shock without a reset, as rerouting, spare pipeline capacity and longer voyages have so far absorbed the strain. Duration, however, is the variable. The longer the crisis persists, the greater the chance that a shock becomes a reset, as Suez did. Whether the cascade is widening will show first in the energy flows rather than in the communiqués – in Yanbu loadings recovering toward their pre-attack level, in Sumed throughput, and in southbound transits by VLCCs.

The Makkah agreement was drafted for a contingency the alliance hoped would not materialize. It has now been invoked; how far it is tested is yet to be seen. Suez ended as a verdict on the limits of power, and Bab al-Mandab may yet deliver another verdict on the limits of control. The glass has not broken, and fatigue, cost and the example of Suez all work to keep it intact. But it has cracked in plain view, and the pressure is spreading through every connection the region has. If it does shatter, it may well be on the rocks of the aptly named Gate of Lamentation.

Disclaimer: The opinions expressed in this article are solely those of the author and do not necessarily represent the views of Rystad Energy.