RESEARCH DOSSIER
Container Cargo Theft Handling and Damage
Abstract. Containerized trade transforms goods into standardized transport units that move through consecutive maritime, port, inland, warehouse, customs, and documentary systems. This study examines theft, handling damage, stowage failure, and container-integrity events as interacting but analytically distinguishable causes of insured cargo loss. It integrates the Institute Cargo Clauses of 2009, English marine insurance legislation, selected maritime-carriage regimes, container safety and securing instruments, industry claims guidance, contemporary theft intelligence, and peer-reviewed cargo-crime research. The central finding is that the probability of a physical event, its insurance characterization, and the recoverability of an indemnity or recourse claim belong to different evidentiary inquiries. A sealed container can conceal shortage; an apparently damaged container can contain intact merchandise; and documentary custody may diverge from physical custody. The 2025 BSI–TT Club incident intelligence records truck involvement in 70% of analyzed cargo-theft incidents and sea transport in 4%, demonstrating why a maritime cargo policy must be investigated across inland and terminal stages. World Shipping Council estimates of 1,478 containers lost overboard in 2025 address a separate incident population and provide no theft-frequency denominator. The study develops a comparative account of causal reconstruction, proof preservation, contract interpretation, accumulation exposure, and recovery economics, with explicitly hypothetical calculations. It proposes no new proprietary theory or doctrinal system. Its contribution is a source-grounded integration of technical, legal, evidentiary, and insurance mechanisms that can support scientific assessment across jurisdictions.
Keywords: Marine cargo insurance; containerized freight; cargo theft; handling damage; stowage; container seals; CTU Code; Institute Cargo Clauses; multimodal transit; causation; claims evidence; cargo liability; subrogation; cargo crime statistics; risk engineering.
Introduction: The Scientific Problem and Method
The container is simultaneously a standardized physical enclosure, an item of carrier equipment, a location in which individually owned goods may be stowed, and a point of reference for commercial and transport records. These identities intersect but must be analyzed separately. An insured shipment can be stolen while enclosed in an undamaged container; merchandise can be crushed without a visible breach of the container wall; and damage to the container itself may be a cost borne by an equipment owner rather than a loss under an insurance contract covering the goods. Marine cargo insurance is therefore concerned with the protected interest in the subject matter and the insured cause of loss, while claims against carriers, terminals, and inland operators require distinct contractual and statutory reasoning.
This dossier concentrates on three scientific questions. First, how do actual theft techniques and custody transfers affect the identification of time, place, and manner of loss? Second, how do packaging, weight, container fitness, lifting, securing, and the division of responsibilities affect physical damage and contractual consequences? Third, what evidence can establish causation and quantified indemnity while preserving possible recovery against a responsible third party? The inquiry covers containerized general merchandise and selectively discusses high-value goods, temperature-sensitive cargo, and valuable pharmaceuticals where the same security and evidence principles apply. Specialized reefer engineering, dangerous-goods regulation, detailed marine war coverage, and individual national customs procedures receive only treatment necessary to explain the common system.
The method is comparative documentary analysis. Primary sources include Institute Cargo Clauses (A) 2009, the Marine Insurance Act 1906, the 2015 insurance warranty provisions, the Hague–Visby implementing schedule, the joint CTU Code of Practice, and the IMO's container gross-mass requirements. They are analyzed alongside standards, treaty status records, scientific literature, and incident reporting. Contract language is treated as effective only where actually incorporated, while legislation and conventions are applied according to their jurisdiction, date, and legal scope.
Every statistical finding below is identified with its population and methodological restrictions. Professional reports from insurers and security organizations supply operational intelligence, but their case collections are not automatically random samples of worldwide freight movements. Where physical evidence or contract wording is unavailable, a case remains a hypothetical illustration rather than an adjudicated factual event. Monetary examples exclude currency effects and taxes unless specified. The six substantive chapters integrate published research without creating a separate named model or doctrine.
Scientific Concepts and Historical Origins
Standardization and the chain of custody
Containerization makes international movement possible through consistent dimensions, lifting interfaces, securing arrangements, and transferable transport records. The IMO Safe Containers Convention of 1972 establishes an international framework for container structural approval, safety plates, examination, and maintenance. Approval of the container as transport equipment addresses structural safety. It does not certify the quantity, identity, condition, or commercial value of goods concealed behind its doors. The distinction is decisive when the only external evidence at delivery is a container number and a seal that appears intact.
The technical fabric of a modern general cargo container is described in ISO 1496-1:2013, with published amendments in 2016 and 2024, and methods for handling and securing series 1 containers are specified in ISO 3874:2017. These catalog records establish existence and stated scope of the standards; complete proprietary standards were not examined as free full text. Conformity with a design or handling standard is not proof that particular goods were competently blocked, braced, distributed, and protected against their foreseeable transport environment. Physical testing criteria and cargo-packing practice occupy different functional levels.
The joint 2014 IMO/ILO/UNECE CTU Code supplies a cross-mode practical reference for packers, handlers, operators, receiving parties, and trainers. It addresses distribution of mass, internal cargo securing, defects, documentation, and safe unpacking. The code is a nonmandatory international code of practice; its significance in a dispute depends on incorporation, statutory recognition, trade practice, and factual relevance. The official IMO record differentiates the code, circulated as MSC.1/Circ.1497, from supporting informative material under MSC.1/Circ.1498. A proposed later revision under consideration in 2026 must not be silently substituted for the approved 2014 source when reconstructing a historical shipment.
Physical possession, contractual custody, and insurable interest
A typical journey involves a cargo seller, buyer, freight forwarder, container lessor, trucker, depot, terminal, ocean carrier, customs authority, consignee, warehouse, insurers, and sometimes multiple intermediaries. Their functions can overlap. The shipper identified on a bill of lading may be a forwarding agent; the entity stuffing the container may be a subcontracted warehouse; and the party exposed to economic loss may differ from the party physically holding the unit when loss occurs. A chain-of-custody reconstruction must therefore map physical custody, contractual custody, legal title, risk of loss under the sale, and insured interest independently.
The ICC Incoterms 2020 explanation distinguishes allocation of trade obligations from insurance procurement. Under CIF, the default contractual insurance obligation is generally associated with a narrower Institute Cargo Clauses (C) level, subject to agreement for more cover; under CIP, the seller's default obligation uses the broader Clauses (A) level. The chosen rule, named place, policy endorsement, and contract of sale matter. An Incoterms allocation does not itself resolve claims under a bill of lading or establish that a theft or handling accident is insured. The value of the trade term is to identify which commercial party was expected to arrange cover and when sale risk transferred.
Insurance contracts commonly insure merchandise and its declared value rather than the freight container as equipment. Goods insurance, the carrier's liability cover, a terminal operator's liability cover, and equipment physical-damage cover may respond to different interests arising from the same incident. The proper unit of observation for scientific analysis is consequently an identified shipment or consignment with dated custody events and contract layers, not an undifferentiated 'container claim.'
Historical and contemporary scale
The contemporary container risk environment is shaped by dense networks, time-sensitive handoffs, and uneven terminal or roadside security. UN Trade and Development's Review of Maritime Transport 2025 documents rerouting, congestion, freight-rate volatility, and operational strain; these factors can change dwell time and the timing of custody transfers without themselves establishing individual theft or damage causation. A long wait at a terminal increases exposure time in a physical sense, but the observed number of crimes depends additionally on security conditions, reporting practices, and adversarial choices.
Ocean carriage is a critical stage of container transport, while a substantial part of cargo theft occurs away from a ship. In the BSI–TT Club 2025 incident report, trucks account for 70% of classified theft modality, facilities 20%, sea 4%, rail 2%, and other modalities 4%. Those are proportions among incidents documented by a compiled intelligence system, not probabilities that a shipment will be stolen in each mode. This distinction prevents a common analytical error: treating a sectoral incident mix as if it were mode-specific exposure-adjusted risk.
The World Shipping Council separately estimates 1,478 containers lost at sea in 2025, compared with 576 in 2024, against approximately 280 million container movements in the newer year. The estimated 2025 proportion is approximately 0.000528%, rounded by the publisher to 0.0005%. It refers to lost containers overboard, not stolen cargo, minor handling damage, or insured claims. One event accounted for 640 of the 2025 lost containers; the concentration illustrates why event clustering matters for catastrophe and accumulation analysis.
Scholarly interpretation
Ekwall and Lantz's transport-chain cargo theft study examines theft probability and impact as distinct dimensions using TAPA incident records. Their subsequent work on product and location interactions cautions against a uniform security response across commodities and custody stages. Liang and colleagues' Bayesian analysis of UK theft incidents uses 9,316 recorded cases from 2009–2021 and identifies product, year, location, modus operandi, and region as important variables in that dataset. Its modeling results demonstrate methodological possibilities, not transferable global theft probabilities. Together these studies support the conclusion that risk requires identification of both the susceptible goods and the places and procedures through which they pass.
Theft Pathways and Container Integrity
Opportunistic access and organized diversion
Container theft encompasses the removal of the entire loaded unit, extraction of selected goods, fraudulent collection under false transport authority, diversion of a truck or trailer, and insider-enabled substitution or concealment. The physical entry point differs across methods. A genuine pickup can be followed by a robbery; an apparently routine pickup can be unauthorized from its inception; or a container can remain under the same equipment identification while its contents are reduced during an unscheduled stop.
The BSI–TT Club 2025 global report reports that 22% of its analyzed global incidents involved insider participation. Its category charts include hijacking, theft from vehicles and facilities, theft of vehicles, theft by drivers, theft by employees, and unclassified methods. Classification of theft by an employee and insider participation is not necessarily equivalent: an insider may facilitate a fake pickup while another person takes possession. The empirical denominator is a heterogeneous set of observed and collected incidents. Consequently, such findings guide operational inquiry and inspection priorities, but should not be applied as actuarial frequency estimates without exposure-adjusted data.
Strategic or fictitious pickup exploits authority rather than only physical barriers. A person presents transport credentials, a carrier identity, a booking number, or a legitimate-looking instruction and receives custody of freight. More elaborate schemes may use compromised email, identity impersonation, or intermediary chains. The 2026 IUMI–TAPA cargo crime assessment identifies increasing sophistication and digitally enabled fraud. These operational patterns require evidence from both the physical gate and the digital instructions by which an appointment, release, or route change was authorized.
A freight unit's container number may be correct while the party holding it is not authorized. Accordingly, matching the container number to a bill of lading does not prove lawful custody. Investigators should independently match driver identity, tractor and trailer registration, verified carrier authority, dispatch communications, terminal release credentials, gate timestamps, camera footage, geolocation, and receipt signatories. Controls should be proportionate to legal and data-protection requirements. A security measure's value turns on what false claim it can detect, how reliably it can be audited, and whether a colluding insider can alter its evidence.
Mechanical seals and their limits
ISO 17712:2013 classifies and specifies acceptance and withdrawal procedures for mechanical freight-container seals. Its catalog states that the 2013 edition was confirmed in 2023. A properly selected seal offers evidence of interference, with its value depending on manufacture, handling, numbering, inspection, and records. Sealing is an evidentiary practice embedded in a supply chain; a seal is not a forensic guarantee that the cargo was correctly packed, continuously inaccessible, or delivered in the shipped quantity.
The TT Club's seal-integrity guidance emphasizes the investigative importance of seals and recorded inspections. A seal number should be captured at stuffing, release, transshipment where inspection occurs, and delivery, alongside a dated image showing location and installation. Conflicting seal numbers can indicate authorized customs intervention, clerical error, accidental replacement, or deliberate substitution; the difference becomes meaningful only after each possible explanation is examined.
Seals may be defeated without conspicuous damage. Door hardware can be manipulated; identity plates may be imitated; numbered components may be substituted; or goods may be removed through other parts of a unit. An intact-looking seal thus supports a narrow observation about appearance at a particular time. It does not logically prove continuity of internal contents. Conversely, a broken seal is a circumstance requiring investigation, not direct evidence of theft or the amount lost.
A scientific claim file should separate the observations 'seal present,' 'number matched,' 'seal physically examined,' and 'original seal continuity established.' These are different strengths of evidence. The most robust inference depends on joining serial records, contemporaneous images, documented access events, cargo counts, packaging traces, and independent transportation telemetry.
Where the loss may occur
A theft window is bounded by the last reliably established intact-cargo event and the first reliable indication of shortage. If a factory packed and photographed twenty pallets, a depot later noted the seal unchanged, and delivery found only seventeen pallets, the location of loss remains a temporal interval until other evidence narrows it. Electronic gates, GPS positions, route exceptions, weighbridge records, depot inventory, video, and communications are useful because they can turn one long interval into separately testable subintervals.
Location inference is vulnerable to alternative explanations. A weight difference may result from weighing methods, tare inaccuracies, substituted packaging, moisture changes, fuel or equipment differences, or a transcription error. A route deviation may indicate traffic, customs direction, breakdown, or deliberate concealment. Broken packaging may be caused by rough handling without theft. The task is to establish whether the joint observations favor shortage at origin, intervention during carriage, handling loss, or erroneous destination counting.
Research on nonsecure parking theft and on seasonal cargo-crime patterns illustrates why location and time are important dimensions of incident analysis. These studies examine reported patterns, with geographic and reporting limitations, and do not provide universal event probabilities. Security controls should therefore be assessed at a corridor- and commodity-specific level rather than translated into a single worldwide theft-risk score.
Security controls and displaced risk
Theft prevention requires credible carrier onboarding, independent verification of route instructions, separation of duties at high-risk release points, secure holding areas, tracked seal inventories, exception reporting, and careful allocation of responsibility for after-hours stops. Procedures that merely relocate unprotected dwell time may transfer exposure from a guarded terminal to a roadside parking area. Ekwall's cargo-theft displacement analysis discusses how offenders may adjust to protective measures. The paper supports investigating displacement effects; it cannot prove any particular intervention's net benefit without a comparison that includes changed routes and incident reporting.
Controls are also evidence-generating investments. A gate camera that records only a container's side may help identify equipment but fail to establish the person who authorized release. A GPS unit may prove where a trailer traveled but not who removed goods inside a locked trailer. A verified seal-number photograph and signed pallet count before and after custody transfer can resolve a different uncertainty. Choosing controls by the propositions they can prove produces stronger claims evidence than counting installed technologies alone.
Handling Events, Stowage, and Governing Risk Allocation
Physical mechanisms of handling damage
Handling damage is a mechanism-dependent category. A container may be dropped, struck by mobile equipment, improperly lifted, overloaded, stacked under unsuitable conditions, moved with damaged corner fittings, or exposed to water through a defective door gasket or roof panel. Its contents may suffer compression, lateral movement, puncture, toppling, abrasion, shock, or water damage. The exterior may remain superficially sound because loads transmit through the structure into internal dunnage, pallets, and merchandise. Conversely, a dented exterior does not establish measurable damage to the particular goods insured.
The container's construction, inspection, and method of handling are relevant physical constraints. The CSC safety-approval framework concerns testing, safety plating, and examination; the ISO 3874 handling standard concerns methods for lifting and securing loaded and empty units. They create useful technical reference points for investigation, but compliance assessments must be based on the actual edition and full applicable text. A catalog abstract cannot establish whether a damaged spreader operation complied with a clause that has not been examined.
Internal cargo damage may arise from a combination of acceleration, mass distribution, inadequate restraints, package geometry, and mechanical fragility. The 2014 CTU packing guidance emphasizes competent loading and securing across different modes. Its causal significance is strongest when an investigator documents how the observed restraint layout could or could not resist the accelerations encountered. Displaced pallets alone cannot distinguish an extreme handling event from underdesigned blocking. Stowage photographs, package construction specifications, damage orientation, and movement traces are needed to separate the mechanisms.
A properly packed container may still be damaged by an extraordinary external impact. A poorly secured load may be damaged during ordinary movement even when the carrier performs its duties. Between these cases lie concurrent causes. Insurance analysis should resist treating 'poor packing' and 'rough handling' as mutually exclusive descriptions before the evidence identifies whether one, the other, or their combination produced the insured damage. The answer also turns on whether the insurance wording treats particular packing defects as exclusions and on who performed the packing.
Verified gross mass and its limits
The IMO explains that SOLAS verified gross mass requirements took effect on July 1, 2016. The shipper is responsible for obtaining and documenting VGM before a loaded container is accepted for loading. The two permitted methods include weighing the packed unit or summing the masses of packages, contents, securing materials, and the container tare mass under an approved certified method. VGM supports safe ship stowage and stack stability.
The verified gross mass identifies a total mass; it is not a certificate of content identity, commercial value, correct center of gravity, packing adequacy, or structural integrity on delivery. For a missing-goods claim, a departure VGM provides a mass observation tied to one method and one point in time. If a later weighbridge measurement is to be compared with it, the investigator must reconcile scale tolerances, tare, equipment configuration, and any legitimately removed cargo. The inference of theft from a gross-mass difference is only as reliable as those reconciliations.
Institute Cargo Clauses and scope of coverage
The 2009 Institute Cargo Clauses (A) insure against risks of loss of or damage to the insured subject matter, subject to specified exclusions and operative policy conditions. 'All risks' is a conventional coverage description; it still requires a fortuitous loss within the insured transit and compliance with the actual policy terms. Theft of the insured merchandise is conceptually within the broad risk grant unless policy modifications or specific operative terms alter the result. Handling damage may likewise fall within the general grant when a fortuitous insured cause is established. Evidence of physical deterioration is necessary to distinguish a covered accident from inherent vice, routine loss, and nonfortuitous commercial quality deterioration.
Clause 4.3 of the published 2009 Clauses (A) is particularly important. The packing exclusion refers to insufficiency or unsuitability of packing or preparation to withstand ordinary incidents of the insured transit when that packing or preparation was carried out by the insured or its employees, or before the insurance attached. For that provision, packing includes stowage in a container, and employees are distinguished from independent contractors. This wording is materially narrower and more fact-dependent than an assertion that all loss involving inadequate packaging is excluded. The disputed causal path, the entity that packed, and the timing relative to attachment must all be recorded.
Clause 5 separately deals with unfitness of a container or conveyance for safe carriage subject to stated knowledge, timing, and person-related conditions. A claimant's knowledge cannot be inferred merely because the container was hired for the journey. An investigator must establish who selected or inspected the unit, whether the insured or its employees were privy to unfitness when loading, and whether loading occurred before attachment. These particulars can control an otherwise apparently straightforward water-ingress or structural-failure claim.
The 2009 Clauses (A) also exclude ordinary leakage and wear, inherent vice, and loss caused by delay, even where the delay arises from a covered risk, subject to the stated treatment of general-average-related expenses. The Marine Insurance Act 1906, section 55 provides an English statutory proximate-cause framework, subject to contrary policy terms. Proximate cause in this context is an interpretive legal question of effective causation, not merely the latest chronological event. Where stolen goods are recovered with measurable handling damage, two losses may require distinct coverage and valuation findings.
Clauses (B) and (C) are structured around enumerated insured perils rather than the broad Clauses (A) grant. An insurer's published ICC clause collection documents the availability of the 2009 A, B, and C forms. The actual wording, endorsements, and certificate decide coverage: a policy procured at a Clauses (C) default under CIF should not be described as providing general theft insurance by virtue of the trade term alone. Extensions such as theft, pilferage, non-delivery, loading and unloading, concealed damage, or storage can materially change the analysis where included; their presence must be established from the executed insurance documents rather than assumed from a product label.
Temporal boundaries and policy attachment
A container can be packed before insurance attaches, wait in a depot after commercial risk has passed, move under a carrier's custody, and remain in an inland warehouse after ocean discharge. The insurance voyage and transport contracts do not necessarily begin and end at the same physical events. Marine cargo clauses have transit and termination provisions, while carriers have limited periods of responsibility under particular regimes. Thus an investigator should create a chronology containing: initial assembly and quality inspection; packing; tender to the first custodian; insurer's attachment trigger; warehouse and terminal gate events; vessel loading and discharge; inland handoff; arrival; and delivery or final storage.
A delayed discovery of shortage is not itself proof that the theft occurred after the policy ceased. Conversely, the existence of an insurance certificate does not prove the loss was suffered within its insured period. When a container arrives with an intact seal but contains fewer units than declared, the central dispute may concern when the missing units first ceased to be present. Packing records and contemporaneous verification at origin become more probative than a general statement that the container 'arrived short.'
Carriage of goods: maritime and inland responsibility
Carrier liability and first-party cargo insurance must be examined on separate legal tracks. The UK's Carriage of Goods by Sea Act 1971 schedule implements the Hague Rules as amended by the 1968 Brussels Protocol (Hague–Visby Rules). The reproduced rules address the carrier's duties in loading, handling, stowage, carriage, care, and discharge, together with defenses, notification, package and weight limitation, and time-bar mechanisms. Their scope depends on the governing contract, connecting factors, and applicable domestic enactment. Under the conventional tackle-to-tackle scope in Article I, cargo custody before loading and after discharge can be addressed differently from the sea leg.
For carriage subject to United States law, the Carriage of Goods by Sea Act of 1936 is reproduced in the historical statutory note to Title 46 and generally provides a $500 per-package or customary freight-unit limitation unless the value was properly declared, subject to legal classification and the shipment's terms. Its interpretation in container cases can be consequential because identification of a 'package' depends on the bill of lading, nature of the cargo, and controlling case law. It would be scientifically unsound to substitute a container count for a package count without reviewing the shipping document and the relevant court decisions. The same incident can therefore have materially different liability ceilings under different governing carriage regimes.
The Hamburg Rules entered into force on November 1, 1992, and apply where their treaty and domestic conditions are satisfied. They establish a different international sea-carriage liability regime. By contrast, the Rotterdam Rules treaty status lists five parties and twenty actions required for entry into force as checked in November 2026. The Rotterdam Rules must be analyzed as a prospective harmonization instrument or incorporated contractual reference rather than treated as a generally operative international carriage convention. Road legs may instead engage national law or a road-carriage convention; the correct instrument must be identified rather than inferred from the presence of a shipping container.
Physical damage could occur at a terminal between regimes, and a multimodal bill of lading may have its own network-liability language. For each claim, identify the contracting carrier, performing carrier, terminal, and inland operator; the effective legal regime; the responsible period; exemptions and limits; and the contractual evidence needed to bring suit. The mere presence of a carrier's liability does not mean the cargo insurer can recover the full insured indemnity. Carrier liability may be capped, disputed, procedurally barred, or economically unrecoverable.
Insurance warranties and operational conditions
Security protocols sometimes enter the insurance policy as warranties, conditions, exclusions, or representations. For contracts governed by relevant UK law, Insurance Act 2015, section 10 changed the former rule of automatic permanent discharge for breach of warranty into suspension of liability during an unremedied breach, subject to statutory details. Section 11 on relevant terms restricts reliance on breach of certain risk-reducing terms where the insured shows that noncompliance could not have increased the risk of the loss actually occurring. Its applicability depends on the term and any relevant contracting-out regime. The two sections should be read in their operative form and context; they do not make every failure to follow security advice harmless.
A requirement to park overnight only at approved facilities, use a specified seal type, or verify a collecting driver's identity might have a different effect depending on its drafting and the legal regime. Researchers should identify whether the wording defines the risk, imposes a warranty, specifies a condition of recovery, or forms part of an exclusion. The practical effectiveness of a control and its legal status are different questions; both are necessary to evaluate a disputed claim.
Claim Evidence, Adjustment, and Recovery
Establishing the insured loss
A cargo claim begins by identifying the precise economic subject matter and its pre-loss condition. The policy schedule, certificate, commercial invoice, packing list, purchase contract, goods description, serial and lot numbers, insured valuation basis, bill of lading, terminal records, warehouse documents, and first inspection at destination establish the documentary framework. Physical evidence then tests that framework: photographs, independent tally, broken seal, package deformation, water traces, contamination findings, laboratory testing, damage survey, security records, and transport telemetry.
The crucial comparison is between demonstrably present or sound goods at a verified prior event and goods found missing or damaged at a later event. A packing list created before loading is a statement of intended contents unless supported by actual count, scan, weight, or audit. A clean bill of lading generally records apparent external order and condition and may not establish hidden internal contents of a sealed container. A delivery receipt annotated 'seal intact' records a seal observation, not a count of palletized goods. Careful investigation assigns each document a specific evidentiary proposition.
The industry Marine Cargo Claims Handbook (2025) distinguishes common documents from additional proof needed for sea, road, and storage damage. It discusses surveys, movement records, packing and stowage photos, and temperature information. This is a professional guide rather than a binding legal code; a contractual notice or suit limitation may require additional action. Its value lies in translating physical-cause hypotheses into a practical request for evidence.
For theft, the initial record should capture immediate discovery, persons present, physical condition, seal IDs, contemporaneous tally, transport instructions, route and custody exceptions, and reports made to carriers and relevant authorities. For handling damage, records should specify packaging orientation, impact direction, crushed or punctured components, moisture measurements, cargo securing arrangement, and evidence of repeated versus discrete impacts. A survey that labels the cause 'handling' without identifying an impact or mechanism is a tentative classification, not a scientific reconstruction.
Chronology, evidentiary continuity, and alternative explanations
Evidence must survive both technical and legal scrutiny. A photograph should be linked to its device, time, location, observer, and original unedited file; metadata can be wrong or modified, so independent custody records remain important. An electronic log should retain event times with time zones and system-clock offsets. A camera gap should be recorded as an absence of observation, not as proof of wrongdoing. A seal removed for customs inspection should be logged with the authority, old seal, new seal, and corresponding images. Each handoff record should record who observed which condition and whether that observer could view the cargo itself.
A practical test uses competing explanations. Suppose an insured claims that five of fifty cartons were stolen in inland transport. The observed facts are a matched destination seal number, a shortage found during unloading, and a driver stop outside the intended route. At least four explanations require examination: the cartons were never loaded; they were loaded but removed while sealed or after seal substitution; they were moved into another consignment at an intermediate facility; or the destination count was erroneous. Independent loading scans, vehicle and seal records, warehouse reconciliation, and destination count procedures can distinguish these possibilities. The route stop alone cannot select one.
For damage, the causal alternatives may be a pre-shipment defect, inadequate palletization, ordinary vibration acting on fragile merchandise, extraordinary impact, deficient stowage, water ingress from container defects, or prolonged delay. Each has different coverage implications under Clauses (A), different possible recourse defendants, and different evidentiary tests. A chemical corrosion analysis might establish saltwater versus freshwater ingress; an impact recorder might narrow the time of shock but not prove which operator caused it; a supplier's product testing may establish pre-shipment weakness. A well-documented negative finding is as useful as a positive one when it eliminates a plausible route of causation.
Notice, cooperation, and time bars
Cargo insurance requires prompt, accurate notice under the actual policy, and contracts of carriage can impose shorter notice or litigation periods. The U.S. COGSA statutory text provides a framework for notice of apparent and concealed damage and a one-year suit period in its specified maritime context. The applicable Hague–Visby implementation likewise contains notice and liability-discharge provisions. These are legal requirements with jurisdictional qualifications; they should be mapped onto the event chronology when a claim is opened. Notice to a cargo insurer and a letter reserving rights against a carrier perform different functions. Neither substitutes automatically for timely suit.
Policy clauses may require reasonable action to preserve rights against carriers and other bailees. The standard Institute Cargo Clauses (A) 2009 contain duties of minimizing covered loss and preserving claims against third parties. Under Marine Insurance Act section 78, a sue-and-labor clause may provide supplementary reimbursement of reasonable protective expenditure, subject to its terms, and the assured has a statutory duty to take reasonable measures to avert or minimize loss. Recovery of a salvage or protective cost therefore depends on the applicable legal and policy terms; it is not simply all money spent after discovery.
A preservation package normally includes a dated timeline, source manifest, original files and cryptographic digests, responsible custodians, photographic inventory, survey plan, relevant policy terms, applicable limitation deadlines, claim correspondence, and evidence of reasonable salvage or security decisions. For recurring claims, comparing the quality of these packages may reveal systemic reporting gaps that are otherwise misclassified as low recovery rates.
Quantity, value, damage, and salvage
Loss of goods requires an auditable quantity and value basis. For a total theft of identified goods, the gross commercial value may differ from the insured valuation; a policy can use invoice cost plus agreed uplift, valued policy provisions, or other specific terms. For partial theft, the starting point is documented units missing multiplied by the appropriate insured unit value, with adjustments required by the particular policy. For damaged goods, an insurer may consider repair, replacement, diminution in value, residual salvage, and reasonably incurred expenses. Any deductible, coinsurance, limit, or aggregation term belongs to the insurance contract, not to a generic notion of fairness.
A claim of shortage should be reconciled against returns, over-shipments, legitimate substitutions, production variances, and packing errors. A claim of damage must distinguish preexisting manufacturing quality issues from transit-related deterioration. In a partial-loss adjustment, evidence of what can still be sold matters as much as evidence of what cannot. Salvage disposition should preserve applicable export, environmental, food-safety, security, and contractual controls; an apparent salvage value is economically meaningful only if realization is lawful and feasible.
Consider a wholly hypothetical consignment of 400 machines insured at $1,200 each. A verified theft removes 30 units; 20 additional machines are damaged by a documented impact; repairs cost $250 per damaged unit and restore full function. Before policy terms, the direct shortage is 30 × $1,200 = $36,000 and the repair component is 20 × $250 = $5,000. If the policy applies a single $2,500 deductible to the combined occurrence and contains no other relevant limitation, an illustrative insured payment is $38,500. If the loss is instead treated as two separate occurrences with separate deductibles, or if the damaged units require a different valuation, the result changes. This example is algebraic and establishes no typical cargo price, premium, frequency, or claims practice.
Subrogation, reimbursement, and net cost
After indemnity, the insurer may acquire rights of recovery against a responsible carrier, terminal, warehouse, or other party to the extent allowed by applicable law and contract. Carrier recovery requires a viable cause of action, identifiable counterparty, proof that loss happened during its responsibility, compliance with notice and suit deadlines, and assessment of any package or weight limitations. The 2009 Clauses (A) expressly state that the insurance is not intended to benefit carriers or other bailees. This aligns the preservation of recourse with the first-party insurance structure, while the exact law of subrogation and distribution of proceeds remains jurisdiction-specific.
A simplified recovery illustration shows why policy indemnity and final insurer burden differ. Suppose $100,000 has been validly indemnified, and a carrier action has an estimated gross recovery of $40,000, contingent legal expenses of $10,000, and an independently assessed 60% probability of successful collection. An expected net recovery calculation of 0.60 × ($40,000 − $10,000) = $18,000 yields expected insurer net claim cost of $82,000. This assumes zero expense on failure and full ownership of proceeds, assumptions that must be revised where legal expenses are sunk, subrogated recovery is shared, or insured uninsured losses have priority. It is an explanatory calculation, not a recommendation to litigate or a valuation of a real case.
Evidence controls therefore have economic value beyond proving the insured's initial claim. A signed delivery tally can reduce a dispute about quantity; independent custody logs can establish a possible defendant; and a securely preserved original recording can support settlement. Conversely, a carrier limitation or limitation-period expiry can make recovery uneconomic despite high confidence in physical causation. Research must distinguish evidentiary proof, legal enforceability, and collectability.
Research Evidence Quality, Quantification, and Sensitivity
The denominator problem
Cargo-theft intelligence frequently draws on reported insurance claims, police records, customs intelligence, trade-association incident systems, and commercial media. Each captures a different part of the risk process. A theft report may describe one container, one road vehicle, one consignment, or one criminal episode affecting several consignments. A cargo insurance database may count claims by policy or occurrence rather than theft events. A shipping operator may count twenty-foot equivalent units (TEUs) rather than physical boxes; a household-goods insurer may have a unit of exposure defined by annual turnover. These units cannot be combined arithmetically without a documented mapping.
The 2025 BSI–TT Club report is especially useful for directional identification of products, places, and methods. It is not a random draw from all global freight movements, and the publisher describes compilation from law enforcement, businesses, associations, open media, and professional specialists. The 70% truck share and 4% sea share describe observed incident composition. A mode-specific rate would require reliable numbers of exposed consignments or unit-time by mode, comparable reporting coverage, and consistent case definitions. Without these, any statement that an individual truck journey is a specified multiple more dangerous than a sea journey is unsupported.
The World Shipping Council's 2026 container-loss update supplies a different numerator and an approximate global movement denominator for lost containers at sea. Its 2025 estimate is based on survey participation representing around 90% of global container-vessel capacity, with extrapolation for remaining capacity. The percentage quantifies one narrow category of physical events. It does not translate into cargo-theft probability, damage-claim probability, or insurer loss ratio. It also does not assign a common monetary value to lost boxes. The source identifies one large event as responsible for 640 of 1,478 estimated lost containers; this concentration indicates that average annual counts obscure correlated losses in marine catastrophes.
The Ekwall–Lantz research on loss probability and impact illustrates how prevention choices depend on whether an intervention chiefly reduces event frequency or conditional severity. That general research insight is relevant, but its estimates arise from particular reported datasets and classifications. Applying a regression coefficient to another country, mode, insurer, or set of goods would require new validation. The Liang et al. Bayesian network study demonstrates interdependencies among observed factors in UK incident data. Prediction on biased incident-only records should not be confused with calibrated population-event likelihood when the nonincident exposure base is missing or selected differently.
Illustrative frequency, severity, and accumulation
For a defined exposure unit, expected annual insured theft loss can be expressed by an ordinary actuarial identity: E[L] = E[N] × E[X], where N denotes the number of insured theft occurrences during the exposure period and X denotes average covered loss per occurrence, provided definitions and exposure are aligned. This equation is conventional insurance arithmetic, not a newly established theory. It is most useful when analysts report separately the empirical or assumed provenance of the occurrence frequency and severity inputs. In an insured portfolio, contractual deductibles and coverage conditions must be applied at the claim level before a net covered-loss average is estimated.
For a transparent hypothetical example, assume 10,000 annual container-consignment movements, an incident frequency of 0.0015 per movement, and covered cost of $48,000 per occurrence after deductibles. The expected count is 15 events and the expected aggregate is $720,000, implying $72 per movement. These are invented educational inputs, not observed industry estimates. If frequency is 0.0008 instead, expected cost is $384,000. If frequency remains 0.0015 but conditional covered loss rises to $80,000, expected cost rises to $1,200,000. Both the observation period and cargo-value mix must be fixed before these scenarios can be compared.
A further complication is accumulation. One fire or crane-collapse event can damage several containers at once, whereas an organized fraudulent pickup scheme may compromise multiple shipments through a single stolen credential. If annual expected frequency is the same, correlated event severity can produce very different upper-tail capital needs. A simple scenario makes this concrete: an occurrence damaging ten $50,000 insured consignments produces a $500,000 gross event, rather than ten independently timed $50,000 incidents. Deductible aggregation, event definitions, reinsurance attachment, and claims-handling resources can differ even if total nominal goods value is unchanged. The WSC concentration evidence for 2025 supplies a real-world example of event clustering in a related physical loss category, but does not quantify a theft accumulation correlation.
Evidence-quality sensitivity
Claims decisions are sensitive to the availability of credible origin records. Consider a shortage discovered after delivery: a signed manufacturer tally and serial scan at origin plus an independently recorded depot inspection substantially narrow the missing-goods interval. If the only origin record is an automatically generated commercial packing list, then alleged shortage during transport may also reflect inaccurate original declaration. The resulting uncertainty affects liability recovery and claims investigation, even where the insurer ultimately decides to pay on a separate evidentiary standard.
There is also selection bias in claims files. Large losses receive more investigation; some small thefts go unreported or are absorbed as shrinkage; successful fraud detection may change the definition of a recorded event; and commercial claims inventories differ from police records. This selection changes the apparent proportions of causes. When evaluating control effectiveness, a lower count after introducing seal checks could represent less theft, less detection, fewer exposures, reclassification, or changed reporting. Sound comparisons require consistent exposure measures, pre/post monitoring quality, product mix, route allocation, and ideally a credible comparison group.
Physical damage evidence has parallel limitations. An impact indicator may be triggered but not show the damage mechanism; a shock logger can have threshold and calibration problems; moisture sensors can be affected by placement, condensation, or improper activation. Independent testing, original record retention, and a plausible mechanical explanation improve causal assessment. The ISO 1496-1 container specification and ISO 3874 handling classification provide an engineering context, but a published catalog description cannot replace inspection of the full purchased technical standard or expert examination of actual damaged equipment.
A comparative evidence hierarchy in practice
A strong claim is constructed from convergent records. At the base are retrospective assertions without corroboration. More informative are contemporaneous business records with known creation and access controls. Stronger still are independently generated custody records, original photos, itemized scans, reproducible physical surveys, and verified external measurements. The comparative importance of a record depends on the proposition it tests and on possible contamination or collusion; no absolute numerical weight is assigned here. This is an evidentiary ordering for case analysis, not a proposed universal scoring framework.
Evidence may also be negative. A terminal scan showing no exceptional handling, a valid seal custody transfer, or a stable temperature log can exclude certain explanations if the instruments operated throughout the relevant interval. Negative observations are weaker where there are blind spots, interrupted monitoring, or incomplete device custody. Claims reports should explicitly describe both available observations and the missing portions of the record. An honest uncertainty statement is a component of rigorous source attribution, not an indication that causal investigation was unnecessary.
Jurisdictional and contract sensitivity
Legal conclusions about insurance and carriage are unusually sensitive to governing law. The English proximate-cause statute and Insurance Act 2015 section 11 are meaningful within the contracts and legal rules to which they apply. The UK Hague–Visby schedule and U.S. COGSA text operate in distinct legal settings. A claim involving a seller in one state, a carrier in another, a third-country terminal, and a buyer in a fourth requires identification of contract clauses, mandatory law, procedural forum, and any chosen arbitration regime. The use of the phrase 'international maritime law' without identifying the effective instrument would conceal material differences.
Treaty status is also dynamic. The Hamburg Rules status record identifies participating jurisdictions and entry-into-force dates. The Rotterdam Rules status record shows that this later convention has not achieved the required twenty actions for entry into force as of the verification date. The CTU Code's nonmandatory status is an additional distinction between a technical recommendation and a rule with direct binding force. A release-quality claim should attach each normative conclusion to its edition, jurisdiction, incorporation, and relevant transaction date.
Verification limits of this investigation
This study verified publicly available treaty-status records, statutes, regulatory explanations, institutional summaries, ICC(A) wording reproduced by a specialist legal publisher, published ISO catalog abstracts, scholarly publisher or university records, and professional incident reports. It did not purchase restricted ISO full texts or proprietary policy databases; did not obtain original insurer claims files, police case records, raw TT Club incident data, or carrier cargo-level movement histories; and did not independently determine the enforceability of individual contract endorsements in a named national court. Claims about ISO standards are accordingly restricted to the verified catalog scope. The BSI–TT Club report is analyzed at the publicly available full-report level, while other academic contributions with paywalled full text are represented by their verified bibliographic records and publicly available abstracts.
Quantitative examples are deliberately constructed to expose the arithmetic of valuation, frequency, and recovery; none is presented as an empirical estimate. The principal scholarly contribution is original cross-domain comparison and causal explanation, with explicit attention to how a single event enters insurance, carriage, engineering, and evidentiary systems. No independent external peer review has been conducted or claimed.
Integrated Scientific Synthesis and Research Agenda
One physical event, several legal and financial outcomes
The comparative evidence establishes that a container claim unfolds through at least three independently testable questions: what physically happened to the goods, what insured or contractual obligation attaches to the event, and what recoverable value can be established. The separation follows from the underlying documents themselves. The ICC(A) risk and exclusion wording defines one contractual lens, the Hague–Visby duties and defenses define another where operative, and the CTU packing guidance provides engineering context. A coherent explanation must connect these without treating them as interchangeable tests.
For instance, consider a container packed by an independent contractor with inadequately secured appliances, later subjected to an extraordinary impact during terminal handling. A physical expert might conclude that the impact and poor securing both contributed. An insurer examining a policy incorporating 2009 Clauses (A) must analyze clause 4.3's person- and time-specific wording and the effective proximate cause, not simply attach an exclusion label to any movement of goods. A terminal liability claim would require evidence of terminal custody, impact, breach of a relevant duty, and recoverability under the effective handling contract. All three inquiries use the same photographs and movement records, but they answer distinct questions.
The analogous theft case is a container delivered with a visually correct seal and a confirmed shortage. The apparent seal condition does not disprove theft; it also does not disprove incomplete packing. Serial inventories, seal examination, origin counts, and custody records determine the range of plausible events. If a fraudulent carrier collected the unit using altered digital authority, security-control failure may be central even though the container itself was never physically broken. In that setting, the legal characterization of dishonest appropriation and the identity of the responsible transport intermediary may be as consequential as the goods' invoice value.
Prevention and proof as linked operational functions
Packing inspections, seal control, VGM documentation, terminal access verification, and secure carrier onboarding are commonly described as preventive measures. Their evidentiary functions deserve equal attention. Proper records can distinguish loss at origin from in-transit shortage, accident from inherent vice, and authorized customs opening from criminal interference. The ISO seal classification, SOLAS VGM explanations, and TT Club seal guidance support these complementary roles at different levels of authority. None independently certifies the entire chain.
This analysis supports differentiated controls. High-value portable electronics may warrant detailed serial scanning and verified pickup identities; heavy machinery may require lift-plan, tie-down, and impact evidence; food cargo may require separate temperature, contamination, and traceability records. The empirical work on cargo product and location interactions supports such specificity. The objective is to preserve the information necessary for a later reconstruction while reducing realistic exposures, rather than to require a uniform technology bundle for every cargo class.
Boundaries of observed risk
Theft, minor handling loss, lost-overboard containers, and maritime catastrophe each require their own numerator, denominator, time basis, and financial definition. The 2025 theft intelligence can reveal concentrations of methods and geography; WSC's sea-loss counts can describe the rare loss of complete containers in that setting; neither identifies the other phenomenon's probability. For underwriters and researchers, combining them as a single 'container risk rate' would lose the distinction that explains both cause and mitigation.
Exposure-adjusted analysis also requires consistent values and protection terms. A minor pilferage claim in a container of valuable pharmaceuticals can be more costly than complete loss of low-value goods. A carrier's liability limit can make insured loss and recovery materially divergent. Insurance deductibles can suppress observed paid-claims frequency even when physical damage is common. These features make portfolio averages dependent on contract design, product composition, the strength of evidence, and the economics of recovery.
Testable future inquiries
Future empirical research could combine anonymized, shipment-level data on commodity class, declared and insured value, handoffs, place and time of packing, seal and security records, journey distance and dwell time, transport regime, physical observations, paid indemnity, salvage, and third-party recovery. Linkage would require privacy safeguards and common event definitions. The research design should identify a population of exposed shipments, not only claims already made. This would permit meaningful estimates of frequency, conditional severity, reporting bias, and the incremental benefit of individual prevention measures.
A second research direction concerns documentary integrity. Investigators could compare independently observed pallet counts and seal records against electronically recorded pickup authorizations, retaining the full chain of revisions and reason codes. A third direction concerns handling: calibrated movement telemetry and detailed cargo survey records could distinguish extreme impacts from ordinary transport stresses and inadequate securing. A fourth concerns legal recovery: harmonized claim data on jurisdiction, package limits, notice deadlines, and final recoveries could quantify how much of a technically well-proved event survives contractual and procedural barriers. Each proposal is a research agenda, not a newly named theoretical model or a claim of demonstrated causal improvement.
Final interpretation
Container cargo insurance is best understood as financial protection applied to a moving, partly observable physical asset within a changing network of legal custody and operational controls. Theft frequently exploits authorization and access; handling damage follows mechanics and stowage; claims turn on the ability to reconstruct both. Insurance wording determines when a proven physical loss becomes an insured loss, while carriage rules and evidence determine whether an insurer or insured can transfer part of the financial cost to another party. The resulting scientific picture is integrated but preserves each level's distinct evidentiary criteria.
Conclusion
The investigation establishes six evidence-supported conclusions. First, container equipment integrity, internal cargo integrity, and verified commercial contents are different observables. Second, theft assessment requires simultaneous examination of physical access, authority to take custody, seal records, and documentary or digital diversion. Third, handling and stowage losses require mechanical reconstruction supported by actual packing, container-condition, and movement evidence. Fourth, 2009 Institute Cargo Clauses (A), particularly the packing and container-unfitness terms, demand precise identification of who packed, when insurance attached, and what caused the damage. Fifth, first-party cargo cover and third-party carrier liability are governed by separate rules and may produce different recoverable amounts for the same incident. Sixth, reported cargo-crime shares and container-loss-at-sea counts measure distinct populations; their numbers are informative only when the denominator and collection process are respected.
The original synthesis lies in tracing physical causation through a common evidentiary chronology into insurance characterization and then into recoverable economic loss. This approach clarifies why the strongest prevention controls are often also the controls that preserve reliable contemporaneous proof. It also explains why accurate causal attribution cannot be replaced by a seal check, a general 'all risks' label, or an unqualified industry loss percentage. The analysis makes no unsupported claim to have discovered a new risk theory. Its conclusions are grounded in identified documentary sources, comparative interpretation, and transparent hypothetical arithmetic, with legal and empirical boundaries stated throughout.
References
BSI Consulting and TT Club. (2026). 2025 Cargo Theft Report (April 2026; observational intelligence concerning 2025). TT Club. https://www.ttclub.com/fileadmin/uploads/tt-club/Documents/Corporate_Literature/TT_Club_BSI_2025_Cargo_Theft_Report_FINAL.pdf
Clyde & Co. (2025, August). Marine Cargo Claims Handbook. https://online.flippingbook.com/view/950054648/
Ekwall, D. (2009). The displacement effect in cargo theft. International Journal of Physical Distribution & Logistics Management, 39(1), 47–62. https://doi.org/10.1108/09600030910929183
Ekwall, D., & Lantz, B. (2013). Seasonality of cargo theft at transport chain locations. International Journal of Physical Distribution & Logistics Management, 43(9), 728–746. https://doi.org/10.1108/IJPDLM-06-2012-0175
Ekwall, D., & Lantz, B. (2015). Cargo theft at non-secure parking locations. International Journal of Retail & Distribution Management, 43(3), 204–220. https://research.hanken.fi/en/publications/cargo-theft-at-non-secure-parking-locations-2/
Ekwall, D., & Lantz, B. (2016). Supply chain risk analysis and assessment: Cargo theft. Transportation Journal, 55(4), 400–419. https://research.chalmers.se/en/publication/?id=232571
Ekwall, D., & Lantz, B. (2019). The moderating role of transport chain location in cargo theft risk. The TQM Journal. https://research.hanken.fi/en/publications/the-moderating-role-of-transport-chain-location-in-cargo-theft-ri/
Ekwall, D., & Lantz, B. (2020; published online 2021). The relationship between impact and probability in supply chain risk management: A cargo theft example. International Journal of Decision Sciences, Risk and Management, 9(4), 241–260. https://research.hanken.fi/en/publications/the-relationship-between-impact-and-probability-in-supply-chain-r/
International Chamber of Commerce (ICC). (2019/2020). Incoterms 2020: Official ICC overview and insurance provisions. https://iccwbo.org/business-solutions/incoterms-rules/incoterms-2020/
International Maritime Organization (IMO). (1972, as amended). International Convention for Safe Containers (CSC): Official convention overview. https://www.imo.org/en/about/conventions/pages/international-convention-for-safe-containers-%28csc%29.aspx
International Maritime Organization (IMO). (2016). Verification of the gross mass of a packed container: SOLAS VI/2 implementation and methods. https://www.imo.org/en/ourwork/safety/pages/verification-of-the-gross-mass.aspx
International Maritime Organization (IMO). (2014). CTU Code and MSC circulars: Official explanatory record. https://www.imo.org/en/ourwork/safety/pages/ctu-code.aspx
International Organization for Standardization (ISO). (2013). ISO 1496-1:2013, Series 1 freight containers—Specification and testing—Part 1: General cargo containers for general purposes (catalog entry; amendments noted). https://www.iso.org/standard/59672.html
International Organization for Standardization (ISO). (2013). ISO 17712:2013, Freight containers—Mechanical seals (catalog entry). https://www.iso.org/standard/62464.html
International Organization for Standardization (ISO). (2017). ISO 3874:2017, Series 1 freight containers—Handling and securing (catalog entry). https://www.iso.org/standard/63826.html
International Underwriting Association of London and Lloyd's Market Association. (2009). Institute Cargo Clauses (A), 1 January 2009 (reproduced in International Cargo Insurance, legal publisher record). https://www.i-law.com/ilaw/doc/view.htm?id=315920
If P&C Insurance. (n.d.). Cargo conditions and clauses: 1982 and 2009 Institute Cargo Clauses A, B, and C; insurer document collection. https://www.if-insurance.com/large-enterprises/insurance-solutions/marine-insurances/cargo-conditions-and-clauses
Insurance Act 2015 (UK), c. 4, § 10 (breach of warranty). https://www.legislation.gov.uk/ukpga/2015/4/section/10
Insurance Act 2015 (UK), c. 4, § 11 (terms not relevant to actual loss; original enacted text). https://www.legislation.gov.uk/ukpga/2015/4/section/11/enacted?view=extent
Liang, X., Fan, S., Lucy, J., & Yang, Z. (2022). Risk analysis of cargo theft from freight supply chains using a data-driven Bayesian network. Reliability Engineering & System Safety, 226, Article 108702. https://www.sciencedirect.com/science/article/pii/S0951832022003271
Marine Insurance Act 1906 (UK), c. 41, § 55 (included and excluded losses). https://www.legislation.gov.uk/ukpga/Edw7/6/41/section/55
Marine Insurance Act 1906 (UK), c. 41, § 78 (sue and labor). https://www.legislation.gov.uk/ukpga/Edw7/6/41/section/78
Merchant Shipping/Carriage of Goods by Sea Act 1971 (UK), c. 19, Schedule (Hague–Visby Rules). https://www.legislation.gov.uk/ukpga/1971/19/schedules?timeline=false&wrap=true
TT Club. (2024, August 13). TT warns container seals are being ignored to the detriment of cargo security. https://www.ttclub.com/news-and-resources/news/article/tt-warns-container-seals-are-being-ignored-to-the-detriment-of-cargo-security/
TAPA EMEA and International Union of Marine Insurance. (2026, February 3). IUMI and TAPA EMEA warn of escalating cargo theft and freight fraud. https://tapaemea.org/news/iumi-and-tapa-emea-warn-of-escalating-cargo-theft-and-freight-fraud/
UN Trade and Development (UNCTAD). (2025). Review of Maritime Transport 2025: Staying the course in turbulent waters. https://unctad.org/publication/review-maritime-transport-2025
United Nations Commission on International Trade Law (UNCITRAL). (1978, entry into force 1992). United Nations Convention on the Carriage of Goods by Sea (Hamburg Rules). https://uncitral.un.org/en/texts/transportgoods/conventions/hamburg_rules
United Nations Commission on International Trade Law (UNCITRAL). (2026 status checked). Hamburg Rules: Status of treaty participation. https://uncitral.un.org/en/texts/transportgoods/conventions/hamburg_rules/status
United Nations Commission on International Trade Law (UNCITRAL). (2026 status checked). Rotterdam Rules: Status of treaty participation and entry-into-force threshold. https://uncitral.un.org/en/texts/transportgoods/conventions/rotterdam_rules/status
United Nations Economic Commission for Europe (UNECE), International Maritime Organization (IMO), and International Labour Organization (ILO). (2014). IMO/ILO/UNECE Code of Practice for Packing of Cargo Transport Units (CTU Code). https://unece.org/transport/intermodal-transport/imoilounece-code-practice-packing-cargo-transport-units-ctu-code
U.S. Congress. (1936, as amended). Carriage of Goods by Sea Act, reproduced at 46 U.S.C. § 30701 historical and statutory notes. Legal Information Institute, Cornell Law School. https://www.law.cornell.edu/uscode/text/46/30701
World Shipping Council. (2026, June 25). World Shipping Council releases Containers Lost at Sea Report: 2026 Update (2025 events). https://www.worldshipping.org/news/world-shipping-council-releases-containers-lost-at-sea-report-2026-update
World Shipping Council. (2026). Containers Lost at Sea: Historical data and 2026 update. https://www.worldshipping.org/containers-lost-at-sea