12. Incoterms, risk & landed cost

Picture a procurement team negotiating a unit price with a supplier in Vietnam. The supplier comes back with two quotes:

• EXW Hai Phong: $22.50 per unit. Buyer arranges everything from the factory door onward.
• DDP Dallas: $26.80 per unit. Supplier handles ocean freight, US customs, drayage, last-mile, the whole thing.

The procurement team picks DDP because it’s “cleaner”: one number, one supplier, no logistics complexity to manage. Six months later, the logistics team realizes they have no visibility into the lane, no carrier of record, and no way to enforce SLAs. The unit price difference ($4.30) was never the point. The Incoterm choice gave away control of 10–20% of total landed costillustrative hidden inside the supplier’s freight markup.

Mode-vs-mode comparisons collapse the moment you try to compare them on freight cost alone. The comparison only works at total landed cost: the sum of unit cost, freight, duty, compliance, carrying cost, and the cost of variability. And landed cost only holds up if the Incoterm is explicit about where risk, cost, and control transfer. This chapter is the contractual and financial spine of mode selection.

Starting cold? Incoterms are the standardized rules that say who pays for what and where risk transfers between seller and buyer. Landed cost is the all-in number per unit that lets you compare modes apples-to-apples.

Incoterms 2020 — the 11 rules

Incoterms are ICC ‘s Incoterms® 2020 rules [ICC Incoterms 2020]. They specify who arranges, who pays, and at what point risk transfers from seller to buyer. They do not transfer title (that’s a separate contractual matter), and they do not impose a taxation framework (customs and VAT rules are independent).

Four rules work for any mode; seven are sea/inland-waterway only.

Any-mode rules (use these for multimodal / containerized / air)

• EXW — Ex Works. Seller makes goods available at its premises. Buyer does literally everything else, including loading at seller’s dock. Maximum buyer obligation. Rarely the right choice for cross-border inbound; the buyer ends up responsible for export clearance in the seller’s country, which is operationally and legally ugly.
• FCA — Free Carrier. Seller delivers goods, export-cleared, to a carrier nominated by the buyer, at an agreed named place. Risk transfers when delivered to that carrier. This is the right any-mode rule for most containerized inbound. Handles the EXW export-clearance problem cleanly.
• CPT — Carriage Paid To. Seller contracts and pays for main carriage to a named destination; risk transfers to buyer when goods are handed to the first carrier. Cost and risk transfer at different points, an important nuance.
• CIP — Carriage and Insurance Paid To. Same as CPT, plus seller must procure insurance at Institute Cargo Clauses (A) level (all-risks) as of 2020. Under Incoterms 2010 it was Clauses (C); the 2020 revision raised the bar.
• DAP — Delivered at Place. Seller delivers to the named place, ready for unloading. Seller arranges everything except destination import clearance and unloading. Risk transfers at the named place.
• DPU — Delivered at Place Unloaded. Same as DAP, but seller also unloads. Replaced “DAT” in the 2020 revision.
• DDP — Delivered Duty Paid. Seller handles everything, including import clearance and duty payment at destination. Maximum seller obligation.

Sea / inland-waterway only rules

Use these for non-containerized ocean cargo (bulk, breakbulk, ro-ro). Don’t use them for containerized cargo. The risk-transfer points (ship’s rail, vessel side) don’t correspond to operational reality when a container is handed over at a CY/CFS well before it reaches the vessel.

• FAS — Free Alongside Ship. Seller delivers goods alongside the vessel at the named port.
• FOB — Free on Board. Seller delivers onto the vessel. Risk transfers when the goods are on board (the 2010 revision clarified this; it used to be “across the ship’s rail”).
• CFR — Cost and Freight. Seller pays ocean freight to named destination port; risk transfers at origin on-board.
• CIF — Cost, Insurance & Freight. Same as CFR plus insurance (at Clauses C, same as CIP 2010; the 2020 revision did not upgrade CIF’s insurance level).

The FOB China trap

A large share of Asian-origin containerized inbound is quoted “FOB China” by suppliers who have always done it that way. This is wrong. FOB is a sea-only rule for non-containerized cargo, and using it for containerized freight creates unclear risk transfer if the container is damaged at the CY before loading. The correct rule is FCA [named port]. Many shippers let it slide in contracts because carriers and brokers understand the intent, but for high-value cargo with an insurance claim, “FOB for a container” is the first thing underwriters attack.

Shaded cells show which party pays at each stage. Risk generally transfers at the same step, with named exceptions (CPT/CIP/CFR/CIF: risk transfers at first carrier / on-board even though cost continues).

Rule	Load at origin	Origin inland	Export clearance	Main carriage	Import clearance	Destination inland	Unload at destination	Duty + taxes	Insurance
EXW Ex Works	B	B	B	B	B	B	B	B	buyer arranges
FCA Free Carrier	S	S	S	B	B	B	B	B	buyer arranges
CPT Carriage Paid To	S	S	S	S	S	S	B	B	buyer arranges
CIP Carriage & Insurance Paid To	S	S	S	S	S	S	B	B	ICC (A)
DAP Delivered at Place	S	S	S	S	S	S	B	B	buyer arranges
DPU Delivered at Place Unloaded	S	S	S	S	S	S	S	B	buyer arranges
DDP Delivered Duty Paid	S	S	S	S	S	S	S	S	buyer arranges
FAS Free Alongside Ship sea only	S	S	B	B	B	B	B	B	buyer arranges
FOB Free on Board sea only	S	S	S	B	B	B	B	B	buyer arranges
CFR Cost and Freight sea only	S	S	S	S	B	B	B	B	buyer arranges
CIF Cost, Insurance & Freight sea only	S	S	S	S	B	B	B	B	ICC (C)

Seller pays Buyer pays ICC Publication 723E — Incoterms® 2020

The cost transfer map

Each Incoterm defines ten cost items and whether the seller (S) or buyer (B) is responsible:

Cost item	EXW	FCA	CPT	CIP	DAP	DPU	DDP	FOB	CFR	CIF
Loading at origin	B	S	S	S	S	S	S	S	S	S
Export packing	S	S	S	S	S	S	S	S	S	S
Origin inland freight	B	S*	S	S	S	S	S	S	S	S
Export clearance	B	S	S	S	S	S	S	S	S	S
Main carriage	B	B	S	S	S	S	S	B	S	S
Insurance	B	B	B	S (A)	B	B	B	B	B	S (C)
Destination port / term. handling	B	B	B	B	S	S	S	B	B	B
Destination inland freight	B	B	B	B	S	S	S	B	B	B
Import clearance	B	B	B	B	B	B	S	B	B	B
Duty / taxes	B	B	B	B	B	B	S	B	B	B
Unloading at destination	B	B	B	B	B	S	S	B	B	B

* Under FCA, origin inland freight is seller’s responsibility only up to the named delivery point.

Risk transfer ≠ cost transfer

Imagine a container shipped CIF Rotterdam. The seller paid ocean freight all the way to Rotterdam. But a storm damaged the cargo mid-Atlantic. Whose insurance claim is it?

The buyer’s. Under CPT, CIP, CFR, and CIF, risk transfers at origin (when goods are handed to the first carrier or loaded on the vessel) but cost is borne to destination. If the goods are damaged in transit, the buyer owns the claim, even though the seller paid the freight. The CIF and CIP rules include seller-provided insurance to address this; CPT and CFR do not.

Practically: always insure goods under buyer control even if the Incoterm is CFR or CPT. A cheap all-risks warehouse-to-warehouse marine policy is table stakes for serious inbound operations. The marine insurance market is dominated by Lloyd’s syndicates and global brokers (Marsh, Aon, Willis); shippers with meaningful annual cargo value typically negotiate an open-cover policy that auto-applies to declared shipments rather than a per-shipment policy.

How to pick an Incoterm

A practical framework:

1. Default to FCA (origin named place) for containerized inbound from any non-domestic origin. You control the main carriage, which means you own the carrier relationship, the lane data, the mode flexibility, and the visibility.

2. Use DAP or DDP when the supplier has meaningfully better logistics capability than you do on that lane, or when the volume is too small to justify building your own routing.

   Where DDP is the right call: low-volume lanes from suppliers with genuinely better local logistics than yours, complex destinations where your forwarder has no presence, or programs where the procurement cost of building inbound oversight exceeds the freight cost saved. For high-volume strategic lanes, default away from DDP because it hides freight cost in unit price and surrenders lane visibility.

3. Use CIP (not CIF or CFR) when the supplier insists on arranging freight and insurance, because Clauses (A) coverage matters.

4. Avoid EXW unless you have a well-established origin forwarder who can handle export clearance on the seller’s behalf.

   Where EXW can work: short cross-border moves where the buyer’s broker handles export clearance routinely (US-Mexico maquiladora flows, intra-EU pre-Brexit). Operationally fragile on intercontinental container trade.

5. Avoid FOB / CFR / CIF for containers. The risk-transfer points (ship’s rail, vessel side) don’t correspond to operational reality when a container is handed over at a CY/CFS well before it reaches the vessel.

   Why some shippers tolerate it anyway: the practical risk-allocation gap rarely manifests, carriers and brokers understand the intent, and changing the Incoterm on long-running supplier contracts is friction. Tolerable on low-value, high-volume lanes where a claim is unlikely. Not tolerable on high-value cargo where an underwriter will attack “FOB for a container” the first time you file.

The single biggest strategic Incoterm error most large shippers make is drifting toward DDP with consolidated suppliers because it “simplifies” procurement. In reality it hides 10–20% of freight costillustrative inside unit price, surrenders lane visibility, and destroys the ability to run an annual RFP, because you’re not the shipper of record.

Landed cost: the formula that actually matters

Imagine procurement quoted you $22.50/unit EXW Vietnam. Marketing wants to know what to charge the customer. Finance wants to know the gross margin. Neither answer is the unit price. The number you actually need is total landed cost per unit (TLC/unit):

TLC/unit = Unit price
         + Freight share per unit (main carriage + pre/on-carriage)
         + Duty + MPF/HMF + applicable tariffs (e.g., Section 301, ADD/CVD)
         + Brokerage + documentation fees
         + Destination handling (THC, drayage, chassis, lumper)
         + Insurance premium share
         + Inventory carrying cost during transit (daily rate × transit days)
         + Expected D&D and accessorial spend (historical average)
         + Allowance for shrinkage / damage in transit
         + Compliance / partner-government filing cost share
         + Packaging cost delta vs. alternative modes
         − Volume rebates attributable to this lane

Most ERPs carry landed cost accounting features (SAP, Oracle, NetSuite) but the default configuration almost always understates by 8–15%illustrative because accessorials, D&D, and carrying cost are posted to different GL accounts than freight. A periodic forensic landed-cost study (pulling actuals from TMS, WMS, broker system, and AP) often reveals that the cheapest-on-paper mode is not the cheapest in practice.

Carrying cost is not optional

Many shippers set carrying cost to zero in landed-cost comparisons and justify it with “we’re not capital constrained.” This is a financial error. Carrying cost is weighted-average cost of capital × inventory value × (transit days / 365). Even if cash is cheap, it’s not zero, and the sensitivity shows up sharply when you’re comparing a 2-day air lane to a 35-day ocean lane.

Trade policy as an inbound input

The tariff environment has been in continuous flux since 2018, and the rate of policy change has accelerated meaningfully in 2024–26. For any major inbound lane, landed cost should be parameterized on tariff rate so scenario swings can be modeled in under a day when policy changes. Treat the tariff input as an environmental variable, not a constant.

The active policy regimes:

• US Section 301 tariffs on China-origin goods (25% on many HTS codes; list expansions, exclusions, and reinstatements through 2025–26).
• US Section 232 on steel and aluminum, global, with country-level negotiated carveouts.
• EU Carbon Border Adjustment Mechanism ( CBAM ) [EU CBAM] — in transitional phase since 2023, definitive phase from 2026, imposes a carbon price on imported cement, iron & steel, aluminum, fertilizers, electricity, hydrogen, and expanding.
• US IEEPA tariffs. Emergency tariff authority deployed on China, Canada, and Mexico origins in 2025, with rates and scope shifting on policy announcements.
• Antidumping ( ADD ) and countervailing ( CVD ) duties. Commodity-specific, petition-driven, often 20–200%illustrative on specific HTS codes from specific countries. Apply to small parcel shipments as well as containers.
• De minimis crackdowns. US Section 321 ($800 threshold) under active executive-action review in 2024–26, particularly targeting China-origin low-value parcels.

What this means operationally:

• HTS classification audit is the highest-ROI post-entry action. A small classification error compounded across volume reroutes meaningful duty spend.
• Scenario modeling on at least three tariff cases (current, +10% on origin country, +25% on commodity) for any strategic lane.
• Tariff engineering (component sourcing geography, last-substantial-transformation analysis, country-of-origin documentation) is increasingly worth specialist legal/compliance investment.
• Trade policy monitoring as a real function. Weekly scan of USTR, CBP, and FERC announcements rather than waiting for the broker to notice and pass a fee through.

MABD, OTIF, and the penalty economics of landed cost

Imagine a CPG supplier shipping pallets of breakfast cereal to a Walmart DC. The pallets arrive 18 hours late. Walmart’s OTIF program assesses a chargeback equal to 3% of cost [Walmart Supplier OTIF Policy], which on a $200,000 shipment is $6,000. The freight savings the supplier got from picking the cheaper carrier just got wiped out 10 times over.

Retailers (and increasingly non-retail big-box customers) enforce arrival-window compliance via chargebacks:

• MABD — Must Arrive By Date. The last acceptable day of receipt. Early receipt also often penalized.
• OTIF — On-Time, In-Full. Walmart’s program assesses 3% of cost for non-compliant POs (with 95%-in-full / 90-98%-on-time targets depending on supplier type) [Walmart Supplier OTIF Policy]; Target and other big-box retailers run similar programs in the 1–3% of invoice valueillustrative range, enforced carrier-facing or shipper-facing depending on the contract.

For suppliers shipping outbound to retailers, MABD/OTIF penalties belong in the cost-to-serve. For any shipper whose inbound feeds an outbound flow with MABD/OTIF exposure, inbound reliability is not a freight KPI; it’s a revenue KPI. The correct way to value inbound reliability is the expected penalty cost of a day of late arrival, not the freight cost savings of a slower mode.

Putting it together

Landed cost discipline is the single biggest structural gap between mature and immature inbound organizations. The symptoms of the gap:

• Procurement and transportation run parallel RFPs that don’t reconcile.
• Incoterms are chosen by inertia, not by strategy.
• “Freight cost per kg” is a reported KPI; “landed cost per unit” is not.
• Tariff scenarios are recomputed annually, not continuously.

Fix these four and everything downstream (mode selection, carrier negotiation, network design) becomes a tractable problem rather than an argument.

How to think about landed cost on your own lanes

Five decisions worth revisiting:

When you’re picking an Incoterm: ask where you can actually see and act, not just where the unit price is lowest. FCA at a named origin port is the most common containerized choice; understand that it collapses your visibility on the factory-to-named-point leg.

When procurement is drifting toward DDP for “simplicity”: push back. DDP is right for low-volume lanes from suppliers with genuinely better local logistics than yours. For high-volume strategic lanes, it hides freight cost in unit price and surrenders lane visibility.

When you compute landed cost using ERP defaults: rebuild it from actuals once per year. ERPs typically understate by 8–15%illustrative because accessorials, D&D, and carrying cost are posted to different GL accounts than freight.

When tariff policy moves: treat your top-5 lanes as parameterized on tariff rate. Scenario-model +10% origin country, +25% commodity in under a day. Don’t wait for the broker to surface the change.

When inbound feeds an outbound flow with MABD/OTIF exposure: value inbound reliability as a revenue KPI, not a freight KPI. The expected penalty cost of a day of late arrival is the right comparison number, not the freight cost savings of a slower mode.

Pick a single high-volume inbound lane and compute TLC/unit with every component above using six months of actuals. Compare to the number procurement used in the unit cost negotiation. The gap is rarely under 10%illustrative, often over 25%illustrative. That gap is the magnitude of the hidden cost, and the size of the prize for fixing how you compare modes.