Scope and transparency of war exclusions under German Insurance Law

War is a classic accumulation risk. A single outbreak of hostilities may cause losses to a large number of insured businesses at almost the same time. Unlike many ordinary insured perils, the probability, scale and geographical spread of war-related losses are difficult to model with actuarial confidence. Once physical damage occurs, further risks may follow that an insurer can influence only to a very limited extent. 

This combination of accumulation, unpredictability and limited controllability explains why property, marine cargo, aviation and other commercial policies frequently exclude war and related perils. The commercial rationale is to separate extraordinary conflict risks from the “normal” risks priced into peacetime cover. In some markets and classes, however, war risks may be bought back through specialist extensions or separate cover. 

Policy wording matters 

War exclusions are not uniform. German market wordings illustrate the range. Fire policies may exclude loss caused by “war, warlike events, civil war, revolution, rebellion or insurrection”. Aviation hull clauses may extend much further to hostile acts, riots, civil commotion, terrorism, sabotage, hijacking and governmental seizure. Marine cargo wordings may exclude not only war and civil war, but also dangers arising from the hostile use or continuing presence of weapons of war. 

These differences are not merely stylistic. Terms such as “caused by”, “arising out of”, “in connection with” and “resulting from” may produce materially different arguments on causation.  

What counts as “war”? 

Under the German insurance-law approach discussed here, “war” does not depend on a formal declaration of war. The focus is generally on an actual state of armed conflict, especially armed hostilities between states. An exclusion may also expressly cover civil war, rebellion, revolution, insurrection, hostile acts or terrorism, thereby reducing the practical importance of drawing a precise boundary around the term “war”. 

The distinction remains important where the wording is narrow. A single terrorist attack abroad, for example, is not automatically transformed into an act of war merely because the organisation responsible is also involved in armed conflict elsewhere. Classification depends on the wording and the factual context. 

Cyber operations and hybrid conflict 

A traditional exclusion referring only to “war” or “warlike events” may not clearly capture a stand-alone cyber operation conducted without armed force. From the perspective of an ordinary policyholder, the conventional meaning of war is still closely connected with (physical) armed hostilities. A purely digital attack that disables critical infrastructure will therefore fall outside a traditional war exclusion. However, the analysis becomes more difficult where cyber operations form part of a wider armed or hybrid conflict.  

Causation: how close must the connection be? 

An exclusion normally requires a causal link between the excluded event and the loss. A building damaged by shelling presents a direct case. A building that collapses because of an unrelated earthquake during wartime does not become a war loss merely because hostilities were taking place at the same time. 

The difficult cases involve longer chains of events: disrupted supply routes, power failures, governmental measures, loss of access, cyber incidents or damage outside the theatre of operations. Some German commentary accepts that indirect causation may suffice. But an unlimited “but for” analysis could allow almost any remote consequence to be linked back to a conflict. The wording and commercial purpose of the exclusion must therefore play a central role in defining the boundary. 

The German transparency test 

Where German law governs, standard insurance terms are interpreted from the perspective of an average policyholder who is reasonably attentive and seeks to understand the policy, but has no specialist insurance-law knowledge. Exclusions are generally construed narrowly and no further than required by their wording and commercial purpose. 

Section 307 of the German Civil Code adds a transparency requirement. A standard term may be ineffective if it is not clear and comprehensible or if it fails to make the clause’s economic disadvantages sufficiently apparent. In the insurance context, the policyholder must be able to understand the cover that remains after the exclusion is applied. This may create difficulties for clauses that exclude any loss merely “connected with” war without providing a workable limit to that connection. 

This is a specifically German-law control mechanism. International readers should not assume that the same test, remedy or burden of proof applies under another governing law. The broader comparative point is nonetheless important: expansive language does not always produce certainty, and an exclusion may be vulnerable if the policyholder cannot reasonably identify its boundaries. 

The purpose of the exclusion may limit its reach 

German case law has long linked the purpose of a war exclusion to the exceptional danger created by war: a danger whose occurrence or development is unpredictable and cannot be managed through ordinary means. On that reasoning, it is not enough that war appears somewhere in the historical chain of events. The individual loss should reflect the distinctive, materially increased danger created by the conflict. 

This distinction is relevant to delayed consequences. If a risk has stabilised over time and become sufficiently predictable to be priced as an increased ordinary risk, the special rationale for the war exclusion may weaken. The detonation of unexploded Second World War ordnance decades after the end of hostilities illustrates the point. Although the object originated in war, the contemporary risk may be treated as a calculable peacetime hazard rather than an extraordinary wartime accumulation risk. 

Practical takeaways for policyholders 

War exclusions should never be assessed by their label alone. Policyholders facing an exclusion should examine the exact trigger language, the full causal chain, concurrent causes, geographical provisions, definitions, write-backs and any specialist cyber or terrorism clauses. The governing law and dispute-resolution clause may be decisive. 

Evidence should be preserved early. Technical reports, incident timelines, attribution material, governmental orders, supply-chain records and contemporaneous communications may determine whether the loss arose from an excluded conflict peril or from a separate insured cause. When commissioning root-cause investigations, policyholders should also consider privilege and the way findings may later be used in a coverage dispute. 

For multinational programmes, local and master policies should be reviewed together. Differences in wording, admitted cover, sanctions restrictions and local mandatory law can produce different outcomes for the same event. 

Conclusion 

The reach of a war exclusion is always fact- and wording-sensitive. Direct physical war damage will often fall squarely within the exclusion, while cyber operations, hybrid conflict and remote economic consequences raise much harder questions. Under German law, causation, the commercial purpose of the exclusion and the transparency of standard terms may all limit an insurer’s ability to rely on broadly drafted language. 

When a war exclusion is invoked, policyholders should test rather than assume its application. Even in the shadow of a major conflict, the relevant loss may have resulted from a distinct insured peril, an insufficiently connected cause or a clause whose scope is narrower than its title suggests. 

This is an abridged and automatically generated translation of an article by Johannes Laiblin which was first published in Die VersicherungsPraxis 11-2023, p. 26 ff.