The Kra Land Bridge is a $29 billion project built for containers. Malacca's exposure sits in crude and gas.

What Thailand is proposing, and where it stands.

The Kra Isthmus is the narrow neck of the Thai peninsula. Thailand has proposed a canal or land crossing here on and off since the 19th century. The current plan is a land bridge connecting a deepwater port at Chumphon on the Gulf of Thailand to one at Ranong on the Andaman Sea, joined by roughly 90 kilometres of dual-track rail and a motorway. It is costed at about 997.68 billion baht, roughly $28–29 billion, and structured as a 50-year public-private concession under a single private operator.

Promoters argue it could cut transit times by up to four days and shipping costs by around 15 percent for cargo that would otherwise sail around Singapore. As of mid-2026 the project was under a 90-day review announced by Thailand's finance minister, with bidding and construction targeted for 2026 and a first phase for 2030 in earlier official timelines. Independent reporting notes that no major shipping line has committed to use it. Treat the timeline as a promoter's schedule, not a forecast.

The Hormuz closure that began on 28 February 2026 gave the proposal fresh geopolitical momentum. If any single corridor can be closed, the argument goes, every alternative becomes more valuable. That is true. What it leaves open is whether this particular alternative can carry the cargo that makes Malacca a risk problem in the first place.

A land bridge is not a canal. Every tonne has to come off one ship and onto another.

A canal lets a vessel sail through. A land bridge does not. Cargo has to be discharged at one port, moved across the isthmus, and loaded onto a second vessel. For containers that is expensive but routine, because containers are built to be lifted. For bulk cargo it is a different kind of operation.

Dry bulk handling at a modern terminal typically costs around $8–12 per tonne for discharge, storage and reloading. A Capesize carrying about 170,000 tonnes of iron ore would therefore pay roughly $1.7 million in handling alone at $10 per tonne, before rail freight, port dues or the extra port-call days. The alternative, a four-day detour via Lombok, costs about $200,000 at the assumptions below (four days at roughly $32,000 bunkers plus $18,000 charter per day). On those numbers handling would have to fall to about $1.20 per tonne, roughly a tenth of typical terminal rates, before the land crossing matched a short sea detour.

Liquid bulk is harder again. A VLCC carries roughly 270,000 tonnes of crude. Moving that through shore tanks and onto a second tanker needs dedicated crude terminals, with handling at perhaps $15–20 per tonne plus evaporation and contamination management. We found no crude or LNG transshipment capacity in the published project design, which is described around container ports, rail and motorway. That is the main reason we treat crude and LNG as outside what the project is built to carry. It is a statement about the published design, not a proof that no one could ever build such terminals.

Land bridge versus the sea detour, by vessel class.

The Narrows model costs a reroute from three inputs: charter rate, bunker consumption and detour days. For the land bridge it adds double-handling cost per tonne (or per TEU) and an estimated 1.5 days of additional port time at each end. The cards below compare the two for a laden vessel on a representative Asia route.

Containers are the only class where the project's own design points at the cargo. Even there, for ultra-large container ships the double-handling bill runs several times the cost of a short detour, which is why the plausible users are smaller feeder vessels and time-sensitive cargo where days saved carry real value.

Where Malacca exposure actually sits: crude, refined products and gas, not iron ore.

A common framing of the Kra case is that Japanese and Korean steelmakers need a Malacca alternative for iron ore. The data does not support that. In CDM v3, iron ore bound for Japan and South Korea is routed around Malacca already: Lombok carries 98.9 percent of Japan's iron ore and 98.7 percent of Korea's, so their Malacca dependency for that commodity is 0.8 percent each. The Malacca exposure for those economies is in energy: crude petroleum at 95.7 percent for Japan and 75.4 percent for South Korea, refined products for Korea at 81.8 percent, and LNG at 31.1 percent and 48.9 percent respectively. China's crude dependency is lower, at 46.2 percent, but on a much larger base at $151.1 billion a year.

Dependency figures from the Narrows Chokepoint Dependency Model v3 (multi-vintage UN Comtrade blend, 2022 and 2024). "Annual flow via Malacca" is the annual trade value assigned to the corridor, not a loss estimate. China crude's per-exporter attribution is known to be unreliable because of documented transshipment relabelling; the total China crude exposure is retained as directionally reliable.

A geopolitical hedge and a commercial route are different tests.

Dismissing the project as commercially weak misses what governments are actually weighing. A project does not need to be competitive in steady state to be worth building. It needs to have value in the tail scenario it is designed for, a prolonged closure or heavy disruption of Malacca, and the people paying for that option are sovereign planners, not charterers. The reported feasibility study puts the economic return well above the financial return, which is the usual signature of infrastructure justified on strategic grounds. Whether the option is worth $29 billion is a sovereign risk judgement we are not positioned to make.

What the cost model can say is how the option looks in a hard scenario. Take a full Malacca closure with Lombok and Sunda also constrained, and compare a long sea reroute of 18–22 extra days with a land crossing, using the same per-day assumptions as above. For a Capesize at roughly $50,000 a day, 18–22 extra days costs about $0.9–1.1 million, still below the ~$1.7 million handling bill, so bulk does not switch to the land bridge even then. For an ultra-large container ship at roughly $165,000 a day, the same days cost about $3.0–3.6 million, which is close to the ~$3.2 million handling estimate. That is a break-even, not a clear win, and it is consistent with the project's container orientation. Treat these as illustrations of where the crossover sits, not as a forecast of routing behaviour.

The land bridge does not change the dependency numbers. It changes, at the margin, one tail scenario.

For an accumulation underwriter, a crude desk or a lender looking at a tanker or Capesize book, the question is the same as before: how much of a specific counterparty's flow runs through Malacca, and what does disruption cost at a given severity and duration? On the model's evidence the land bridge does not move the crude, refined product or LNG dependency figures, because those cargoes are not what it is designed to carry, and it does not change dry bulk routing because Lombok is already the primary route.

What it could change is the container tail. A closure scenario that includes a functioning land crossing is somewhat less severe for container flows than one without it, subject to capacity actually existing and shipping lines actually using it, neither of which is established today. That is a narrowing of one tail, not a reduction of base-case exposure, and it only matters if the project is built.

So the practical point for anyone writing cargo policy, a P&I accumulation model or a trade finance facility now is that Malacca dependency should be modelled as it stands. A bridge that does not yet have an operator, a bidder or a committed shipping line is not a mitigant for any exposure being written today.