Bab el-Mandeb has printed the ten lowest weeks in seven years of records, all of them since 5 August — and Suez, the same route's other end, has not moved at all.
Traffic through Bab el-Mandeb has fallen by a quarter since the Houthis declared a blockade of Saudi-linked shipping in July, and for the first time none of it has reappeared anywhere else. That is the part that breaks. This site has already published the opposite result: when the Red Sea emptied out in 2023 and 2024, the Cape of Good Hope picked up 98.1 percent of what Bab el-Mandeb lost, almost ship for ship. Run the same arithmetic on this summer and the Cape does not pick up any of it. It goes down.
The fall itself is the deepest on record. In the 2,820 days of AIS-observed transit counts this site reads, from 1 January 2019 to 20 September 2026, exactly ten seven-day readings for Bab el-Mandeb sit below 25.3 transits a day — and all ten of them fall between 5 August and 20 September 2026. The previous low, 25.3, was set on 29 July 2024, at the height of the first Houthi campaign. The worst week of the war that emptied the Red Sea the first time is now the eleventh-worst week in the file.
The strange number is at the other end. Suez and Bab el-Mandeb are the two ends of one route, and for five years they counted the same ships: 60.6 and 61.5 transits a day across 2019 to November 2023, within one percent of each other. Since the Houthi strikes on the Saudi tankers Encelia and Layla on 22 July, Suez has run at 41.2 a day against Bab el-Mandeb's 27.4 — a ratio of 1.50, and 1.58 over the last month. Ships are still entering the Red Sea from the north at an unchanged rate. A quarter fewer are coming out of the bottom. The gap between the two counts, 15.2 ships a day, is nearly four times everything still transiting Hormuz.
Nothing in the file absorbed them. Against the same 20 July break, the Cape of Good Hope fell 4.8 percent, Malacca fell 1.5 percent, and Suez was the only one of the five measured chokepoints to gain anything at all — plus 1.0 percent, at the mouth of the one that lost the most. The honest caveat is that the Cape was already carrying 93.3 a day, the elevated level it reached absorbing the last Red Sea crisis, so it was not starting quiet; but in 2024 it rose 79 percent from a standing start, and this time it fell. The other caveat governs every count here: observed transits are a lower bound, and a ship with its transponder off is not a ship that stopped. After the six deaths on the Tihamah on 11 August, dark sailing is not a small hypothesis.
Two things change. The Cape of Good Hope count has been the standard proxy for Red Sea disruption since 2024, and on this episode it reads clean while the Red Sea posts its worst weeks ever — so what it measures is the availability of a detour, not the presence of danger. And for Hormuz specifically: Saudi Arabia's Red Sea terminals are one of only two pipeline routes that move Gulf crude without passing Iran's guns, and the southern exit from that one is now at the lowest count anybody has recorded. The second strait to die this summer was Kerch. This is the third, and it is dying in the one place that was supposed to be the way around the first.
Data series · 2026-09-20 IMF PortWatch AIS-observed daily transit calls, Bab el-Mandeb Strait, 2019-01-01 to 2026-09-20 (chokepoints.json)
Data series · 2026-09-20 The same PortWatch series for the Suez Canal, used for the two-ends-of-one-route comparison
Data series · 2026-09-20 The same PortWatch series for the Cape of Good Hope and the Malacca Strait, used to test where the missing traffic went
Event · 2026-07-22 Houthis strike the Saudi tankers Encelia and Layla, the first attacks after their 20 July blockade declaration source
Event · 2026-08-11 Houthi missiles kill six on the cargo ship Tihamah at Bab al-Mandeb, the first shipping deaths of the war source
Computed, not asserted — every number here is derived from public data: official EIA Brent closes via FRED, IMF PortWatch transit counts, Polymarket and Kalshi odds, and source-linked curated events. Method and caveats · Atom feed · llms.txt
https://chokepointdata.com/analysis/2026-09-25-red-sea-dead-end/