Moore v IAG New Zealand Limited

IN THE COURT OF APPEAL OF NEW ZEALAND

I TE KŌTI PĪRA O AOTEAROA

CA356/2019
[2020] NZCA 319

BETWEEN	GRAEME JOHN MOORE Appellant
AND	IAG NEW ZEALAND LIMITED Respondent

Hearing:	7 May 2020
Court:	Cooper, Gilbert and Courtney JJ
Counsel:	N R Campbell QC and P A Cowey for Appellant C M Meechan QC, C M Laband and S K Swinerd for Respondent
Judgment:	31 July 2020 at 9.30 am

JUDGMENT OF THE COURT

AThe appeal is allowed.
BThe answer to the preliminary question set out in [8] is “no”.
CThe respondent must pay costs to the appellant for a standard appeal on a band A basis and usual disbursements. We certify for second counsel.
____________________________________________________________________

REASONS OF THE COURT

(Given by Gilbert J)

Introduction

As is well-known, the Christchurch region was hit by three major earthquakes almost a decade ago — on 4 September 2010, 22 February 2011 and 13 June 2011. The appellant, Mr Moore, owns a home in Scarborough, a suburb of Christchurch. His house was not damaged at all in the September 2010 earthquake, but it sustained serious damage in the February 2011 earthquake and again in the June 2011 earthquake. The estimated reinstatement cost is said to be approximately $2.08 million for the damage caused by the February earthquake and $2.77 million for the damage caused by the June earthquake. It is common ground that these losses are separate in that none of the losses caused by the February earthquake are subsumed in the losses assessed as having been caused by the June earthquake.
Mr Moore insured his home with IAG New Zealand Ltd (IAG) for the 12‑month period 13 November 2010 to 13 November 2011 (the policy period). The sum insured under the policy was $2.5 million. The question on this appeal is whether Mr Moore has cover up to $2.5 million for each loss or whether that is the limit of cover for both losses, even though these occurred four months apart. The answer turns on the interpretation and operation of the aggregation clause in IAG’s policy. Aggregation clauses are common in insurance policies. Their purpose is to enable separate losses to be treated as a single loss in certain circumstances for the purposes of setting the limit of cover. Such clauses also determine whether one or more deductibles are to be borne by the insured for losses arising in a given set of circumstances.

The policy

The primary insuring clause in part C of the policy reads:
We cover sudden accidental loss to the house during the period of insurance.
The words in bold are defined terms. “Accidental” is defined to mean “unexpected and unintended by you”. “Loss” means “physical loss or physical damage”. “You” means the “Insured”.
Part C excludes loss which is caused by, among other things, earthquakes, unless cover is provided in part D of the policy. The relevant clause in part D reads:
We pay for sudden accidental loss to the house which is caused by any of these:
1. earthquake …
IAG’s limit of liability for any loss is the sum insured. This limit generally applies to each successive insured loss occurring during the policy period. However, this general rule is subject to an aggregation provision (clause C2), which aggregates losses for the purposes of calculating the limit of cover in certain circumstances. The interpretation of this clause lies at the heart of this appeal. Clause C2 relevantly reads:
The most that we pay for any loss (or any series of losses caused by one event) is the sum insured shown in the schedule.
“One event” is defined in the policy as “a single event or a series of events which have the same cause”.

Preliminary question

While the issue of quantum has not been resolved, the primary issue dividing the parties concerns the proper interpretation and operation of the aggregation clause. The parties therefore sought determination by the High Court of the following preliminary question:
On a proper interpretation of the aggregation clause in this policy, can it be said that the most IAG is required to pay for the loss on 22 February 2011 and the loss on 13 June 2011 is the sum insured?

High Court judgment

Dunningham J found that the losses to the house caused by the February and June 2011 earthquakes were a “series of losses” caused by “a series of events” which had the same cause, being the September 2010 earthquake.[1] This meant that even though the September 2010 earthquake caused no damage to the house and occurred outside the policy period, IAG’s liability under the policy was capped at $2.5 million for all losses caused by both the February and June earthquakes. The answer to the preliminary question was found to be “yes”.[2]

The appeal

[1]Moore v IAG New Zealand Ltd [2019] NZHC 1549 [High Court judgment].

[2]At [59].

Mr Moore appeals. He contends:
(a)Two losses four months apart do not make a “series of losses”.
(b)Two earthquakes four months apart do not make a “series of events”.
(c)Even if there were such a series, the aggregation clause applies only if the series of events (acting together) caused the series of losses, which was not the case.
(d)IAG did not prove that the September earthquake was the proximate cause of both the February and June earthquakes.

Approach to interpretation

It is common ground that the aggregation clause is to be interpreted in accordance with normal principles of contractual interpretation. An objective approach is required to ascertain the meaning that would be conveyed to a reasonable person with knowledge of any relevant background and in the context of the contract as a whole. The text is, of course, of central importance. If the language used has an ordinary and natural meaning, that provides a strong indication of what the parties must be taken to have meant.[3]
[3]Firm PI 1 Ltd v Zurich Australian Insurance Ltd [2014] NZSC 147, [2015] 1 NZLR 432 at [60]–[63].
Because of the way the appeal was argued, we will start with the words used in the clause, recognising however that the correct interpretation is not the product of a close examination of the literal meaning of words but, rather, the meaning the clause would convey to a reasonable person reading it as a whole and in context.

Series of losses

The Judge accepted IAG’s contention that, on the plain meaning of the words, a “series of losses” for the purposes of the aggregation clause simply meant “repeated experiences of loss”. Thus, if the insured “suffered more than one loss within a cover period”, that would automatically constitute a series of losses. No other connecting factor was required.[4]
[4]High Court judgment, above n 1, at [32].
Mr Campbell QC, for Mr Moore, submits that a series of losses means three or more losses that are similar in nature and temporally proximate. He relies on the primary dictionary meaning of the word “series” as being “a number of things of which each is similar to the preceding or in which each successive pair are similarly related”.[5] He argues that there are two aspects to the definition. First, each item in a series will be similar in some way. Secondly, a series must have three or more items because it is not otherwise possible to have “successive pairs”.
[5]Tony Deverson and Graeme Kennedy (eds) The New Zealand Oxford Dictionary (Oxford University Press, Oxford, 2005) at 1029.
We reject the numerical argument. While in many contexts, a series will mean three or more, that cannot be what the parties can have intended here. It would make no sense in the present context for the first, third, fourth and subsequent sudden accidental losses, all caused by the same insured event, to be counted for the purposes of the aggregation clause, but not the second (unless there were more than two).
However, we consider Mr Campbell is on firmer ground when he submits that the word “series” in the clause indicates losses that are linked in some way and temporally related, not simply any two or more losses occurring at any time during the 12-month policy period. There are three reasons why we agree with Mr Campbell’s submission. Each of these reasons is developed further below. The first point is that, on a careful analysis, the Judge’s interpretation gives no effect at all to the words “series of”. We consider it unlikely that these words would have been mere surplusage in this critically important and concisely worded provision which defines the limit of cover. Secondly, we consider the words, according to their ordinary and natural meaning, convey to the reasonable reader some form of linkage and temporal proximity. Thirdly, this interpretation is consistent with the basic principle expressed in the clause that the limit applies to any loss. The words in parentheses are subordinate to that primary rule. Accordingly, they should be read consistently with the primary rule and not in an expansive way that would tend to defeat or contradict it.
The aggregation clause applies to all losses insured under the policy, including those covered by parts C and D. The primary insuring clause in part C, and the extended earthquake cover provided in part D, both provide cover for “sudden accidental loss to the house”. As noted at [4] above, “loss” means physical loss or physical damage. We are therefore concerned with sudden accidental physical damage to the house under both insuring clauses. We can demonstrate that the Judge’s interpretation gives no effect to the words “series of” by looking at an example of cover under part C. On the Judge’s interpretation, sudden accidental physical damage to one part of the house caused by a flood on day one of the policy period and separate sudden accidental physical damage to another part of the house caused by a fire on the last day of the policy period would qualify as a “series of losses” because the insured would have “suffered more than one loss within a cover period”.[6] There are two losses in the policy period in this example, but there is no other connection between them. On the Judge’s analysis, all the connecting work in the aggregation clause is achieved by the word “caused” and the word “series” is entirely superfluous — the meaning of the aggregation clause is the same with or without the words “series of”:
The most we pay for any loss (or any ~~series of~~ losses caused by one event [a single event or ~~a series of~~ events which have the same cause]) is the sum insured shown in the schedule.
[6]High Court judgment, above n 1, at [32].
A reasonable person reading the policy would expect that the words “series of” (which are incorporated in the clause in two places taking account of the extended definition of “one event”) were included in the clause for a reason. The clause forms part of IAG’s “Supersurance House” standard policy wording, which is drafted using plain English and with commendable brevity. The expressed aim is to achieve clarity for its policy holders — “so you know what it does cover and what it does not”. The clause itself is self-evidently of critical importance, not just to IAG but to its many home owning policy holders. The clause is headed “HOW MUCH WE PAY”. It is inconceivable that IAG would not have paid very close attention to the drafting of this provision in its standard form policy which is widely used throughout New Zealand. All of this indicates that the words “series of” were intended to mean something.
What would the reasonable reader take the words to mean? We consider the word “series” would be understood in the present context as to convey its natural and ordinary meaning of a linked sequence of sudden, accidental losses, occurring proximately in time.
While care must be taken in comparing the wording used in different contexts, the authorities which have examined the word “series” in aggregation clauses in insurance policies tend to support the conclusion that the word suggests a temporal sequence of occurrences having some form of linkage. For example, in Distillers Co Bio‑Chemicals (Australia) Pty Ltd v Ajax Insurance Co Ltd, the High Court of Australia was required to determine whether a number of negligence claims by or on behalf of infants born with deformities due to their mothers taking the drug thalidomide during pregnancy were claims “in respect of or arising out of all occurrences of a series consequent on or attributable to one source or original cause”.[7] Stephen J considered the word “series” in this clause meant “a number of events of a sufficiently similar kind following one another in temporal succession”.[8] Gibbs J agreed.[9] While Menzies J dissented on this point, he too considered that the word “series” would “normally carry with it the notion of a sequence with some connexion between the items in the sequence”.[10]
[7]Distillers Co (Bio-Chemicals) (Australia) Pty Ltd v Ajax Insurance Co Ltd, [1974] HCA 3, (1974) 130 CLR 1.
[8]At 21.
[9]At 10.
[10]At 6.
To similar effect, is the New York Court of Appeals’ observation in Travelers Casualty and Surety Co v Certain Underwriters at Lloyd’s of London:[11]
The word “series” is commonly defined as “a group of [usually] three or more things or events standing or succeeding in order and having a like relationship to each other: a spatial or temporal succession of persons or things”.
(Citation omitted.)
[11]Travelers Casualty and Surety Co v Certain Underwriters at Lloyd’s of London 760 N E 2d 319 (NY 2001) at [8].
Paying due regard to the temporal aspect suggested by the word “series” also enables the provision to be read as a coherent whole and in a manner consistent with the primary rule it expresses. The primary rule is that the limit of cover for any loss (here, sudden accidental physical loss or physical damage) is the sum insured. The subordinate rule, indicated by its placement in parentheses in clause C2, covers the closely comparable position where the same insured fortuity causes a related and proximate sequence of occurrences of sudden accidental physical loss or physical damage.
We need to consider first the type of circumstances envisaged where two separate occurrences of “sudden” physical damage caused by a single event would form a “series” for the purposes of the aggregation clause. Read in combination, the words “series of” and “sudden” indicate an unbroken series of sudden accidental occurrences of physical loss or physical damage, following one after the other in temporal proximity and caused by the same event. It can be readily understood why such closely related occurrences, in both time and cause, would be grouped together when determining the limit of cover. In such circumstances, it could be quite artificial to attempt to do otherwise.
The further qualification to the primary rule, through the extended definition of “one event” to include a series of events which have the same cause, is in turn subordinate to the subordinate rule. We do not consider that this part of the clause, incorporated only by reference, should be interpreted in a way that substantially undermines the primary rule. On the Judge’s interpretation, the primary rule, which reinstates the limit of cover after each occurrence of sudden accidental physical damage from an insured event, is overtaken by an extended definition of a word in a subordinate rule so that all instances of sudden accidental physical damage occurring as a result of any number of insured events, no matter how unrelated or distant in time during the 12‑month policy period, will be grouped for the purposes of applying the policy limit so long as the events have a common underlying cause. If that were correct, a claim like Mr Moore’s for sudden accidental physical damage caused by an earthquake could not be settled until the sum insured was reached or the policy period had expired.
We conclude that two separate and distinct occurrences of sudden accidental physical damage occurring four months apart cannot properly be regarded as a series of such losses in terms of the aggregation clause.

Series of events

An event in this context means something “which happens at a particular time, at a particular place, in a particular way”.[12] An event refers to something specific having happened rather than the reason or underlying cause for what happened.[13] So, for example, a war would not of itself be considered an event but acts taking place during a war would be.[14] The February and June 2011 earthquakes would clearly be regarded as separate events for the purposes of the present clause.[15]
[12]AXA Reinsurance (UK) plc v Field [1996] 1 WLR 1026 (HL) at 1035.
[13]Countryside Assured Group plc v Marshall [2002] EWHC 2082 (Comm), [2003] 1 All ER (Comm) 237.
[14]Robert Merkin and Chris Nicoll (eds) Colinvaux’s Law of Insurance in New Zealand (2nd ed, Thomson Reuters, Wellington, 2017) at 633.
[15]See Crystal Imports Ltd v Certain Underwriters at Lloyds of London [2013] NZHC 3513, (2013) 18 ANZ Insurance Cases 61-997 at [52].
Much of the analysis above is equally applicable here. The aggregation clause expressly contemplates that it will apply in each of three scenarios:
(a)The most that IAG will pay for any (single) loss (caused by one event) is the sum insured (the primary rule).
(b)The most that IAG will pay for any series of losses caused by one event is the sum insured (the subordinate rule).
(c)The most that IAG will pay for any series of losses caused by a series of events which have the same cause is the sum insured (subordinate to the subordinate rule).
The first of these scenarios — the primary rule — is self-explanatory.
The second scenario — the subordinate rule — requires consideration of what must have been intended by the words “any series of losses caused by one event” where “one event” means “a single event”. The following example illustrates how the clause would operate in those circumstances. A tree branch falls on the roof of the house. The sudden accidental physical damage to the house is the damage to the roof. A short time later, it starts raining. Water enters through the damaged part of the roof and causes further sudden accidental physical damage to the internal walls and flooring in the upstairs bedroom immediately below. In this example, we have two sets of sudden accidental physical damage (the damage to the roof followed by the water damage in the bedroom). Unlike the example in paragraph [17] above, these losses could readily be described as a “series of losses” caused by a single event (the tree branch falling on the roof). The sudden accidental losses occur proximately in time, one after the other and are all caused by one event.
The following example illustrates the third scenario — “a series of losses caused by a series of events which have the same cause”. A tropical storm causes a series of events which in turn cause a series of losses. As the storm approaches, high winds cause a tree to topple onto the roof causing sudden accidental physical damage to the roof (as in the previous example). The following morning, as the storm continues to rage, lightning strikes the house causing sudden accidental fire damage in the kitchen. Later that day, as the storm passes over, the high winds and lightning are followed by a period of torrential rain. This causes sudden accidental flood damage to the ground floor of the house in addition to the water damage in the upstairs bedroom caused by rain coming through the damaged part of the roof (as in the previous example). In this example, we have a series of occurrences of sudden accidental physical damage, proximate in time, one following the other — to the roof, upstairs bedroom, kitchen and ground floor. This series of losses were caused by a series of events — the high winds causing the tree to fall onto the roof, the lightning causing the fire in the kitchen and the torrential rain causing the ground floor flooding. These “events” “have the same cause” — the tropical storm, an insured fortuity.

In summary, we accept Mr Campbell’s submission that the phrase “series of losses” in the aggregation clause does not mean any two or more losses no matter how unrelated or distant in time so long as they have an underlying common cause. However, we do not agree with his further proposition that the series of events, acting together, must cause the series of losses. Each event may cause its own loss or losses. However, both the series of losses and the series of events causing those losses must occur in a proximate temporal sequence before they can be grouped for the purposes of applying the policy limit. They must also have a common cause. We turn now to that issue.

Series of events which have the same cause

A series of losses caused by a series of events which have the same cause. What does “cause” mean in this context? The Judge accepted IAG’s submission that it means cause in the “but for” sense.[16] In other words, so long as the series of events causing the series of losses would not have occurred “but for” some other operative factor, that would suffice. An example might be global warming. Assume that, but for global warming, it is unlikely that two severe storms would have occurred during a 12‑month policy period, as they did, eight months apart. Assume that each storm caused sudden accidental physical damage. In the first, the damage was to the roof. In the second, the damage was caused by flooding to the downstairs area. Although these were unrelated events occurring eight months apart and causing quite separate losses, on the Judge’s analysis the losses would be grouped for the purposes of the aggregation clause because it is unlikely that both storms would have happened during the policy period but for global warming. Expressed another way, on the balance of probabilities, the two storms would not have occurred in the policy period but for global warming.
[16]High Court judgment, above n 1, at [52].
Ms Meechan QC, for the respondent, submits that the Judge was correct to apply a “but for” approach in the present case. However, we do not consider the aggregation clause, read as a whole and in context, anticipates that such a remote “but for” test would suffice. The general rule that the sum insured is reinstated after each loss would be stretched considerably, if not departed from, if a “but for” test for causation was to apply. It does not sit well with our analysis of what constitutes a series of losses for the purposes of the clause. We consider that a series of events causing a series of losses (temporally proximate and linked) will have the same cause only if they have the same proximate cause in the usual sense.
The doctrine of proximate causation holds that an insurer is liable only for any loss proximately caused by an insured peril.[17] The proximate cause of any given loss is a question of fact to be determined in accordance with common sense principles.[18] The test to be applied is the same as in tort law and has been expressed in a variety of ways, including as the direct cause,[19] the immediate cause from which the loss arose as a natural consequence,[20] the dominant cause,[21] or the real efficient cause.[22] Proximate cause is to be contrasted with a more remote cause, such as one that simply sets the scene.[23]

Application of the aggregation clause to the facts of this case

[17]Codified in the Marine Insurance Act 1908, s 55(1) but equally applicable in non-marine insurance law. See Techni-Chemicals Products Co Ltd v South British Insurance Co Ltd [1977] 1 NZLR 311 (SC) and Groves v AMP Fire & General Insurance Co (NZ) Ltd [1990] 2 NZLR 408 (CA) at 411.

[18]Groves v AMP Fire & General Insurance Co (NZ) Ltd, above n 17, at 412, referring to Yorkshire Dale Steamship Co v Minister of War Transport [1942] AC 691 (HL) at 706.

[19]Leyland Shipping Co Ltd v Norwich Union Fire Insurance Society Ltd [1918] AC 350 (HL) at 358 per Lord Finlay.

[20]At 362 per Viscount Haldane.

[21]At 363 per Lord Dunedin.

[22]At 369 per Lord Shaw.

[23]Groves v AMP Fire & General Insurance Co (NZ) Ltd, above n 17, at 412.

Applying our reasoning above, the sudden accidental physical damage to the house caused by the February earthquake and the separate sudden accidental loss to the house caused four months later by the June earthquake cannot be regarded as forming a “series” of sudden accidental “losses”. The event that caused the sudden accidental physical damage to the house in February 2011 was the February earthquake following the rupture of three fault lines including the Port Hills fault. The sudden physical damage (or series of such losses) caused by that event then ceased. Mr Moore had an accrued entitlement to receive payment under the policy for that loss. In the normal course, had time permitted, his claim for that loss would have been paid at that time. Four months later, the house sustained a separate sudden accidental physical loss or physical damage (or a series of such losses) in the June earthquake which was caused by the rupture of two other faults to the east of Christchurch. No part of this physical damage was caused by the February earthquake. These separate losses do not in our assessment form a series of losses — the losses were each sudden, but they were separated by four months and were caused by quite separate events. We do not consider they constituted a “series of losses” caused by a “series of events” for the purposes of the aggregation clause.
Having reached this conclusion, it is not strictly necessary for us to consider whether the February and June events had the same cause, namely the September earthquake which in turn was caused by the rupture of yet another fault, the Greendale fault. We nevertheless address this issue for completeness, applying the proximate causation test that we have found to be the correct test. This requires a more direct causal connection than that established simply by applying the “but for” test.
IAG pleaded that the February and June 2011 earthquakes were “part of a series of events which have the same cause”. They were both said to be “aftershocks of the 4 September 2010 earthquake and, as such, causally-linked”.
Two eminent experts provided briefs of evidence — Euan Smith, a professor of geophysics at Victoria University, Wellington (instructed by IAG) and Mark Quigley, an associate professor of earthquake science at the University of Melbourne (instructed by Mr Moore). Their evidence was not challenged, and neither was cross-examined.
Professor Smith was asked by IAG to advise whether the February and June earthquakes “occurred independently of each other, or whether they were linked”. In his summary, he concluded that these two earthquakes “were both significantly more likely to have been aftershocks of the 4 September 2010 earthquake, than to have been unrelated background earthquakes”.
Dr Quigley did not disagree with any of the scientific content of Professor Smith’s detailed report. He said that at a “coarse level of abstraction” all earthquakes in New Zealand have a related common cause, namely relative movement between the Pacific and Australian tectonic plates. Fault rupture occurs when the stresses acting on a fault plane, primarily caused by plate tectonic stresses that accumulate over millennia, exceed the frictional strength of that fault. On that basis, he would agree that all three earthquakes had this common cause. Dr Quigley also accepted that movement on the faults responsible for the September 2010 earthquake and its aftershocks caused additional loading on the adjacent faults that ruptured in February and June 2011. But, importantly, he added:
47. … However, I consider one could only say that the Darfield [September 2010] earthquake ‘caused’ the February event and June event at a more abstract level that ignores more proximate causes. To illustrate, the February fault was already pre-stressed before the Darfield earthquake caused additional loading. The 4 September 2010 earthquake contributed to ‘setting the stage’ for the February 22 event. However, it was not even the straw that broke the camel’s back, given the time delay between these events, and their occurrence on spatially distinct faults. The Darfield earthquake cannot be said to be necessary for the February event to have occurred, nor was it, as a distinct entity, sufficient to cause the February event.
48. … It is statistically highly probable (on the basis of statistical seismicity analysis) that movement on the faults responsible for the 22 February earthquake caused additional loading on the faults that ruptured in the 13 June 2011 earthquake. Again, however, it is more accurate to say that the February event contributed to the June event occurring when it did (i.e., the February 22 earthquake brought the occurrence time of the inevitable June 13 earthquake forward in time).
Thus, both experts provided some support for IAG’s pleading that the “February 2011 and June 2011 earthquakes are part of a series of events which have the same cause”. In a broad sense, all three earthquakes could be regarded as being part of a series of events caused by the accumulation over thousands of years of stresses resulting from the relative movement of the Pacific and Australian plates (equivalent to our global warming example in [32] above). However, we do not consider that being “causally-linked” in this sense and as pleaded meets the proximate causation test we have found to be posited by the aggregation clause. On Dr Quigley’s unchallenged evidence, the September 2010 earthquake did no more than set the stage and was “not even the straw that broke the camel’s back”. It follows that the September 2010 earthquake was not the proximate cause of either the February or the June 2011 earthquakes.

Conclusion

We conclude that the sudden accidental physical damage caused by the February 2011 earthquake and the sudden accidental physical damage caused by the June 2011 earthquake is not a series of losses caused by a series of events which have the same cause. Accordingly, they cannot be aggregated for the purposes of determining the limit of cover for each loss. The answer to the preliminary question is therefore “no”.

Result

The appeal is allowed.
The answer to the preliminary question set out in [8] is “no”.
The respondent must pay costs to the appellant for a standard appeal on a band A basis and usual disbursements. We certify for second counsel.

Solicitors:
Parry Field Lawyers, Christchurch for Appellant
DLA Piper, Wellington for Respondent

[2020] NZCA 319

31 July 2020 at 9.30 am