In late 2023 and early 2024, two maritime corridors failed simultaneously. Houthi attacks on commercial shipping in the Red Sea corridor caused the Bab-el-Mandeb strait to lose roughly 70% of its transit volume between December 2023 and April 2024. At the same time, a severe drought in Central America forced the Panama Canal to cut its daily transit allowance in half, with liquefied natural gas carriers seeing a 66% fall in transits.
Neither event was in the Chokepoint Dependency Model's data. The model was built on 2023 UN Comtrade bilateral trade flows and locked at its initial parameters before either disruption began. It had no information about the Houthi campaign or the drought severity. What it did have was a set of structural predictions: which importing countries depended most heavily on each corridor, which commodities dominated that exposure, and what rerouting would cost per voyage if the corridor failed.
Two disruptions then tested those predictions against live data. What follows is an honest account of what the model got right, what it got partly right, and what it was never designed to predict.
What the model predicts, and what it does not. The Chokepoint Dependency Model is not a forecasting tool. It does not predict when a corridor will fail, how severe a disruption will be, or what freight rates will do in response. What it predicts is exposure: which importers carry the highest dependency on a given corridor, in which commodities, at what bilateral trade value, and what the per-voyage rerouting cost will be if the corridor closes. Those are the three dimensions tested below.
The model predicted $1.4 million per voyage via the Cape. Actuals came in at $1.0–1.4 million. This is the cleanest result.
When a loaded tanker can no longer transit Bab-el-Mandeb and the Suez Canal, the alternative is the Cape of Good Hope route around the southern tip of Africa. The detour is long: for a large oil tanker travelling from the Persian Gulf to a European destination, rerouting via the Cape adds approximately 14 days of sea time.
At the vessel operating costs built into the model ($55,000 per day in fuel and $45,000 per day in time charter equivalent), 14 additional days produces a per-voyage premium of $1.4 million. This calculation was made before the Red Sea disruption and published as a model output.
UNCTAD's impact assessment of the Red Sea disruption, published in early 2024, reported that rerouting added approximately $1 million per voyage in fuel costs alone, with total additional costs reaching $1.0 to $1.4 million depending on vessel class. World Bank analysis produced consistent figures. The model's $1.4 million estimate sits at the top of the reported range, which is consistent: the model uses the largest tanker class (VLCCs, which carry around 2 million barrels of crude) as its reference vessel, while the published averages cover a broader mix of vessel sizes.
| Metric | Model prediction | Reported actual (2024) | Source | Verdict |
|---|---|---|---|---|
| Cape detour duration | +14 days | +10–14 days | UNCTAD, World Bank, J.P. Morgan | ✓ Correct |
| Per-voyage cost premium | $1.40M | ~$1.0–1.4M | UNCTAD; World Bank | ✓ Within range |
| Container freight premium | Not modelled | +40–60% sustained; +256% spot peak | Xeneta, Freightos Baltic Index | Out of scope |
The model covers rerouting cost per bulk voyage. It does not model aggregate container market pricing, which reflects fleet capacity dynamics rather than individual voyage economics.
The model ranked Greece as the most exposed European crude importer. The disruption confirmed it.
The model uses real bilateral trade data (actual country-to-country import flows from UN Comtrade) to calculate each importer's dependency on a given corridor. For Bab-el-Mandeb crude, the European ordering was stark: Greece at 54.6% dependency, France at 13.4%, Germany at 6.6%. Greece's exposure was more than four times France's and more than eight times Germany's.
The underlying reason is geography and sourcing. Greek independent refiners source heavily from Persian Gulf producers (Saudi Arabia, Iraq, the UAE) whose cargoes transit Bab-el-Mandeb on the way to Europe. French and German operators have more access to Atlantic Basin alternatives: Azerbaijani crude via the Bosphorus, Norwegian crude from the North Sea, pipeline routes from Kazakhstan. When the Red Sea corridor came under attack, those alternatives provided meaningful insulation. Greece had fewer of them.
This is precisely what the disruption produced. Greek independent refiners reported acute supply chain pressure through the first half of 2024. Greece's shipping industry, the world's largest commercial fleet by deadweight tonnage, bore a disproportionate operational cost burden as tanker operators were forced around the Cape. France and Germany reported materially lower disruption consistent with the model's much lower dependency scores for both countries.
| Importer | Model dependency | Annual exposure | Observed 2024 impact | Verdict |
|---|---|---|---|---|
| Greece | 54.6% | $8.52B | Acute refiner pressure; highest per-capita shipping cost exposure in Europe; well-documented in industry reporting | ✓ Confirmed #1 |
| Indonesia | 4.2% | $6.30B | Modest crude supply disruption via Bab-el-Mandeb; Indonesia's primary corridor exposure is Malacca and Lombok | ✓ Lower-than-expected confirmed |
| France | 13.4% | $4.62B | Moderate impact; Atlantic Basin substitution accessible; Bosphorus-routed Azerbaijani crude unaffected | ✓ Lower impact vs. Greece |
| Germany | 6.6% | $4.03B | Low impact; TAL pipeline unaffected; Kazakhstani crude via Bosphorus insulated | ✓ Low impact confirmed |
Tier 1 bilateral data. UN Comtrade 2023. Observed impact characterisations from World Bank, UNCTAD, EIA, and industry reporting through H1 2024.
The model identified crude oil as 78% of Bab-el-Mandeb exposure. The disruption confirmed crude dominance, with an honest caveat on containers.
Of the model's total Bab-el-Mandeb bilateral exposure across covered importers ($74.7 billion), crude petroleum accounts for $58.3 billion: 78%. Vehicles, palm oil, and copper ores make up most of the remainder. The model's prediction: if Bab fails, crude supply chains bear the largest financial impact.
The disruption confirmed this in dollar terms. The Red Sea corridor's share of global seaborne crude volumes fell from approximately 12% in April 2023 to 6% in August 2024, a halving of crude transit consistent with the scale of exposure the model identifies. The EIA, IEA, and Kpler all confirmed significant declines in seaborne crude via the corridor.
Where the picture is more complex: container tonnage at Suez fell 82% at its February 2024 trough, a steeper proportional decline than crude in volume terms. The model covers container goods at Bab in lower-confidence estimates (vehicles, electronics, apparel), but its non-crude coverage is narrower than the full containerised goods universe transiting the corridor. The model's $74.7 billion total should be read as a floor on trade at risk, not a ceiling.
The Panama drought hit LNG hardest: transits fell 66%. The model ranked LNG as the largest Panama exposure by dollar value. This is the second clean validation.
The Panama Canal drought cut daily transits from 36–38 in July 2023 to 18 per day by February 2024. Among all commodity classes transiting the canal, liquefied natural gas fell hardest: the Canal Authority reported a 66% decline in LNG transits, the single largest proportional drop of any cargo type.
LNG carriers (vessels that move gas cooled to minus 162 degrees Celsius in insulated tanks) are largely captive to the Panama route for US Gulf to Asian Pacific deliveries. Unlike bulk vessels, they cannot easily divert to alternative routes without transforming the economics of the voyage entirely. Japan and South Korea, which import significant volumes of US LNG through the canal, reported elevated LNG spot prices and accelerated purchases from alternative suppliers through the first half of 2024.
The model's Panama rankings, calibrated against Panama Canal Authority tonnage data, place LNG and LPG as the largest dollar-value exposure: China at $8.3 billion, Japan at $5.1 billion, South Korea at $4.8 billion. The 2024 drought confirmed LNG as the single most impacted commodity class, precisely matching the model's ranking.
| Commodity | Model adj. exposure | Observed 2024 impact | Verdict |
|---|---|---|---|
| LNG / LPG | $23.6B | LNG transits minus 66%. Largest proportional decline of any commodity class. Japan and South Korea worst affected on US LNG route. | ✓ Confirmed #1 |
| Soybeans | $8.7B | Dry bulk transits severely curtailed. Brazil–Asia soybean route impacted; partial redirection via Suez documented. | ✓ Major impact confirmed |
| Coal | $2.4B | Dry bulk restrictions hit coal transit. US coal exports to Asia via Panama affected. | ✓ Impact confirmed |
| Corn | $2.3B | US corn to Asia disrupted; some diversion to Suez for Latin American origins. | ✓ Impact confirmed |
| Wheat | $0.7B | Smaller Panama wheat flow; Black Sea / Suez dominant; relatively lower impact at this chokepoint. | ✓ Lowest grain impact |
Panama LNG / LPG adj. exposure includes China ($8.3B), Japan ($5.1B), South Korea ($4.8B), Mexico ($1.9B), Indonesia ($1.5B) and others. ACP FY2024 annual data confirms LNG as top impacted commodity class.
Three outcomes were outside the model's scope. Being clear about this is part of how the model earns trust.
A validation exercise is only credible if it honestly identifies what was not predicted alongside what was. Three categories of 2024 outcomes were outside scope.
Rerouting cost and Cape detour duration: UNCTAD OSGINF2024d2 "Impact to global trade of disruption of shipping routes in the Red Sea region" (2024); World Bank report (February 2024). Container freight rates: Xeneta; Freightos Baltic Index; Kpler "Update on Red Sea trade flow impacts" (October 2024). Bab-el-Mandeb transit decline (minus 70%): European Parliament EPRS Briefing (2024). Container tonnage at Suez (minus 82%): UNCTAD. Egypt canal revenue loss: IMF blog "Red Sea Attacks Disrupt Global Trade" (March 2024). Greek refiner impact: Hellenic Petroleum and Motor Oil operational filings H1 2024. Panama Canal: ACP FY2024 annual data. LNG transit decline (minus 66%): ACP FY2024; EIA Today in Energy (December 2024). Japan / South Korea LNG spot impacts: IEA Gas Market Report H1 2024. Bilateral trade data: UN Comtrade 2023. Chokepoint parameters locked before November 2023 disruption onset. All model assumptions are adjustable; contact fysh@narrows.io to run scenarios against specific inputs.