Japan is both a developed insurance market and an unfortunate hotspot for natural disasters. As we approach the fourth anniversary of the devastating Tohoku earthquake and tsunami in north-eastern Japan, Mr Ian Roberts and Mr Nicholas Sykes of Clyde & Co Clasis Singapore, highlight Japan’s unsettled law of causation, particularly in respect of concurrent independent causes of business interruption loss.
Stating the obvious, where an insured sustains a loss, it is to be indemnified if that loss is caused by a peril covered by the policy (or in the case of all risks insurance, that the loss was accidental and not caused by an excluded peril). This may not prove too challenging to determine in a single loss event such as a fire destroying a factory.
The question of causation
However, in respect of wide area damage often resulting from natural disasters, several perils are likely and the identification of which caused the loss may prove more challenging when assessing an insured’s business interruption losses.
Those losses may result from a number of concurrent independent perils, for example: (i) damage to the insured’s premises; (ii) damage to its suppliers’ premises; (iii) prevention of access to its premises; (iv) a lack of public utilities; and/or (v) damage to its customers’ premises, one or more of which may be uninsured. In those circumstances, how are insurers to determine the correct indemnity under Japanese law?
The law of causation under Japanese law is unsettled. Legislation is limited and case law is undeveloped. Japanese law does recognise the principle of causation and, in particular, for the purpose of business interruption indemnities at least implicitly, the distinctions between independent and interdependent, and concurrent and consequential causes of loss.
Some of our readers will be more familiar with English law rather than Japanese law. We first reconsider the English law position in order to highlight the distinction in approach under Japanese law.
The English law position
Under English law, causation must be established in both fact and law. Enquiries must be made as to whether the cause is both “proximate” to the loss and that that loss would not have occurred “but for” the incidence of the peril.
English law defines proximate as the “dominant”, “effective” or “real” cause of the loss. Again, using the example of a total loss caused by a factory fire and assuming an absence of any other peril, the fire was the dominant cause (as no other peril existed) in destroying the factory and such destruction would not have occurred but for that fire. If fire is an insured peril, the insured will be indemnified for its loss.
In respect of business interruption losses, the proximate cause can be more challenging to determine. Assume a factory is damaged by an earthquake which interrupts the insured’s business (an insured peril), the insured’s suppliers’ premises are also damaged (an uninsured peril) and the insured is prevented from accessing its premises to operate its business (an insured peril). These three perils independently result in financial loss to the insured’s business and all three are possible proximate causes which but for their occurrence, the insured would not have suffered a loss.
“But for” test
When determining concurrent independent causes of loss under English law, the English High Court decision of Orient Express Hotels Ltd. v Assicurazioni Generali S.p.A.  Lloyds Rep I.R. 531 (OEH) provides helpful guidance as to how insurers should proceed.
OEH clarifies that the “but for” test remains applicable in respect of business interruption losses. Generally, if there are multiple concurrent independent causes of loss, one of which is uninsured and which would have caused the loss in the absence of the other insured cause(s), the loss is irrecoverable.
In that instant, the insured’s claim failed the “but for” test as absent the insured peril, the insured would have still sustained the loss as a result of the uninsured peril. However, the court in OEH did recognise that in exceptional cases, it is at least possible that fairness and reasonableness may require a different test, but the court did not elaborate when a case would be exceptional, or what test would be appropriate in those circumstances (which will no doubt be the subject of future case law).
Notably, in the US, commercial parties have developed anti-concurrent causation clauses (which are not enforceable in all States), which, if written into the policy, prevent an insured from recovering its losses from insurers in circumstances where the loss was caused by two or more perils and at least one of which is uninsured.
Japanese law: the distinction
Generally, causation is an unsettled area of Japanese insurance law. Japanese law (which is governed by a Civil Code and is not common law), like English law, looks for the proximate cause of loss.
In respect of concurrent causes, limited Japanese case law arising out of the 1995 Great Hanshin earthquake is of some assistance in respect of interdependent causes of loss, in which lower district courts “apportioned” losses between the insured and uninsured causes on the bases of either the general principle of good faith or sufficient legal cause under Japanese law (there is no clear legal principle of apportionment under Japanese law). However, we are aware of no reported cases on concurrent independent causes and whether the same principle of “apportionment” would also be applied. Equally, the Japanese courts are not bound by the decision in OEH.
How then can an indemnity be determined in respect of losses caused by concurrent independent perils?
Subject to policy wording, one approach open to insurers is to proceed on the basis that Japanese law would hold that if one of at least two concurrent independent perils could have solely caused the loss and that one peril is insured and the other is uninsured, the uninsured concurrent peril(s) is ignored and the insured is indemnified in whole (subject to any necessary adjustments such as trends or the application of average). However, in the absence of any case law or legislative guidance, it is not clear whether a Japanese court would agree with this approach.
Another approach we have seen adopted is to “apportion” losses between insured and uninsured causes of loss by applying the general principle of good faith and/or sufficient legal cause (as the Japanese courts have done in respect of interdependent losses). Taking the earthquake damage example above, there are three causes of loss, two of which are insured and, hypothetically, the loss is a result of all causes equally. The losses in that example could be apportioned 2/3rd insured and 1/3rd uninsured and the insured could be indemnified for 2/3rds of its loss.
A further approach could be a more detailed forensic exercise where it is possible to assess the exact and distinct financial losses caused by each peril and then to deduct losses caused by the uninsured peril.
Given the uncertainty in Japanese law, insurers must proceed with caution when assessing concurrent causes of business interruption losses governed by Japanese law and not assume that a similar approach as provided under English law would apply.
The principles of good faith and sufficient legal cause under Japanese law should be considered. A number of suggested approaches are outlined above but readers should not consider that an exhaustive list. Discussions with experts on a case-by-case basis are necessary to determine a suitable approach to adopt under Japanese law. Ultimately, at present, there is no one certain correct approach.
Mr Ian Roberts is a Partner and Mr Nicholas Sykes is a Senior Associate, both at Clyde & Co Clasis Singapore.
We are grateful to Ms Haruka Murata of Messrs. Hibiya-Nakata for her collaboration on section Japanese law: the distinction.