Mobis Parts Australia Pty Ltd v XL Insurance Company SE (No 7)

On 25 April 2015, a severe storm occurred in Sydney. It caused a warehouse owned by the plaintiff, Mobis Parts Australia Pty Ltd, to collapse.

Mobis is a wholly owned Australian subsidiary of Hyundai Mobis, a company incorporated in Korea which I will call “Mobis Korea”. Mobis used the warehouse to store and distribute spare parts for Hyundai and Kia motor vehicles.

The warehouse was vast; equivalent in size to several city blocks. Fortunately, as the storm occurred on a public holiday, no one was in the warehouse when it collapsed.

A little under three months later, on 30 July 2015, the warehouse (and virtually all of its then contents, including a large amount of stock) was destroyed in a fire which broke out during the demolition recovery process.

Mobis seeks indemnity in respect of the loss that it claims to have suffered by reason of the collapse (but not the fire) from the first defendant, XL Insurance Company SE, under a Property Damage and Business Interruption Policy issued by XL for the period 23 June 2014 to 23 June 2015. The parties referred to this as the “Local Policy”.

The Local Policy was issued by XL as part of an International Property Damage and Business Interruption Program.

Under that program, XL (together with the second defendant, AIG Europe Ltd, and the third defendant, UNIQA Versicherungs AG) issued a Property Damage and Business Interruption Policy (covering the period 23 June 2014 to 23 June 2016) in the name of another wholly owned subsidiary of Mobis Korea, Mobis Slovakia s.r.o. The parties referred to this as the “Master Policy”.

Under the Master Policy, XL was the lead insurer (with a 50% share of risk); with AIG and UNIQA having a 30% and 20% share of the risk, respectively.

Mobis’s primary claim is against XL under the Local Policy. Mobis only seeks indemnity under the 2014 Master Policy if, contrary to its case, the Local Policy does not respond to its claim.

Mobis seeks indemnity in the order of $62 million (less some $14.4 million already paid to it by XL, discussed below) in respect of:

The hearing of this matter proceeded over 21 hearing days between 29 May and 30 June 2017. There was further argument thereafter on a number of points, resulting in further submissions delivered during July and August 2017.

I have been greatly assisted by the commendably competent and efficient manner in which the case was prepared by the solicitors for the parties and presented by counsel throughout that period. I have been particularly assisted by the comprehensive written and oral submissions I received following the conclusion of evidence. Much of what follows is taken, with gratitude, from those written submissions, especially as to uncontroversial and background matters.

The Local Policy has a limit for “storm” damage of $72,105,000 (equivalent to EUR 50 million).

The Master Policy has a corresponding limit for storm damage (EUR 50 million).

The Master Policy also has a limit of EUR 10 million for “hail”. The parties referred to this as the “Hail Limit”.

XL contends that the agreement of the parties was that there should be a corresponding hail limit in the Local Policy.

No such limit appeared in the terms of the Local Policy as originally written with effect from 1 January 2011, or as renewed annually from 23 June each year until, and including, the relevant policy period (23 June 2014 to 23 June 2015).

A number of issues arise as a result of this omission.

On 5 June 2015, shortly after the warehouse collapsed, XL agreed to make a payment to Mobis in the amount of the asserted Hail Limit (hence the payment of some $14.4 million referred to at [10]; being the equivalent of EUR 10 million).

The first issue is whether a letter that XL wrote to Mobis on 5 June 2015, Mobis’s acceptance of the offer said to have been made in that letter and the subsequent payment by XL to Mobis of the $14.4 million referred to at [10] and [19], had the effect of foreclosing debate about XL’s obligations under the Local Policy, other than as to whether the Local Policy has the Hail Limit and as to quantum. The parties referred to this as the “Acceptance of Liability” issue.

The second issue (which is relevant assuming that the Local Policy has the Hail Limit) is whether the cause of the collapse was hail or storm. This involves consideration of what meteorological event caused the collapse of the warehouse.

The third issue (which is relevant assuming the collapse was caused by hail) is whether, by reason of cl 1.9 of the Local Policy, the Hail Limit in the Master Policy is, in effect, imported into the Local Policy.

The fourth issue is whether the Local Policy should be rectified so as to incorporate the Hail Limit.

The fifth issue is whether the damage suffered by Mobis was caused by the “faulty” or “defective” design of the warehouse so as to enliven exclusion cl 3.2.1 of the Local Policy (the “Faulty Design Exclusion”).

Finally, there are issues as to quantum; particularly concerning Mobis’s claim for the loss of and damage to stock in the warehouse on the date of collapse.

Assuming Mobis is not able to recover under the Local Policy, and must resort to the Master Policy, the following issues arise under the Master Policy.

The first is whether, in the events that have happened, the Hail Limit in the Master Policy is engaged (this is the same issue that arises under the Local Policy, assuming it has a hail limit).

The second issue is whether, by reason of Art 3.1.8 of the Master Policy (which the parties called the “Overlap Clause”), and assuming the damage was caused by “storm” as well as by hail, Mobis is entitled to indemnity up to the storm limit of EUR 50 million, even if the Hail Limit was engaged.

The third issue is whether the damage suffered by Mobis is the result of “faulty building construction” of the warehouse so as to enliven exclusion Art 6.3 of the Master Policy (the “Faulty Construction Exclusion”). This raises questions similar to, but not quite the same as, those which arise under the Faulty Design Exclusion in the Local Policy.

Again, issues of quantum arise.

UNIQA (one of the three insurers under the Master Policy) raises the following issues (which are not raised by either XL or AIG).

The first is whether Mobis is an “Insured Company” under the Master Policy.

The second is whether, assuming Mobis is an Insured Company under the Master Policy, it is entitled to make a claim under that policy (or whether only Mobis Slovakia is entitled to bring a claim).

The third is whether, on the proper construction of the Master Policy, UNIQA is liable for risks in Australia.

The fourth is whether UNIQA’s participation in the Master Policy would constitute a breach of s 10 of the Insurance Act 1973 (Cth) and thus not be “permitted” for the purpose of Art 5.1 of the Master Policy.

Subject to a number of qualifications concerning the quantum of its claim, Mobis is entitled to indemnity under the Local Policy.

Accordingly, the question of whether the Master Policy would respond to a claim by Mobis does not arise.

The answers to the particular questions set out above are:

By the insuring clause under the Local Policy, XL agreed to:

There is no hail limit in the “Schedule” which is set out in cl 1.10 of the Local Policy as follows:

XL did not exercise its option to indemnify Mobis “by replacement or repair”.

Thus, the basis of indemnity is the “cost of reinstatement”. “Reinstatement” is defined in cl 4.2, to which I will return (see [843] below).

I will set out below other clauses in the Local Policy as necessary to understand the issues which arise.

The Master Policy had an insuring clause which, although expressed somewhat differently to that in the Local Policy, was to the same effect.

The Master Policy had limits for “storm”, “wind storm” and for “natural peril events” (including a hail limit).

The table setting out the various liability limits was in the following form and headed “Insured Perils/Limits of Liability”:

The Hail Limit formed part of the limit:

Somewhat repetitively, Art 13.2 of the Master Policy defined “Natural Peril Events (natural peril cover)” as follows:

Thus, the limits of liability under the Master Policy were significantly different in nature to those in the Local Policy.

Again, I will set out below the other relevant terms of the Master Policy as necessary to understand the issues that arise.

The Acceptance of Liability issue is centred around a letter XL sent to Mobis’s solicitors on 5 June 2015.

On 26 May 2015, Mobis’s solicitor wrote to XL:

On 5 June 2015, XL replied:

The investigations referred to in the 5 June 2015 letter included whether the warehouse complied with the Australian Standards for buildings of its type. Those investigations culminated in a report produced on 2 June 2015 by a firm of consulting engineers, Costin Roe Consulting. The Costin Roe report concluded that the warehouse's structural design complied with the requirements of the relevant Australian Standards.

On 12 June 2015, Mobis’s solicitors replied:

On 23 June 2015, XL paid Mobis the Australian dollar equivalent of EUR 10 million (being the $14.4 million referred to at [10] and [19] above).

Mobis put its case on the Acceptance of Liability issue in three ways:

Mobis also submitted that the 5 June 2015 letter contained an admission by XL of its liability under the Local Policy. It clearly does and XL made no submission to the contrary. However, in closing submissions, Mobis did not submit that this fact alone was decisive of any question in the case; in particular, the question of whether the Faulty Design Exclusion was enlivened. This was no doubt because Mobis recognised that as the admission was made otherwise than in the course of a formal court process, it was merely an item of evidence to be weighed up against all other relevant evidence: see for example The Nominal Defendant v Gabriel (2007) 71 NSWLR 150; NSWCA 52 at [113] (Campbell JA).

Mobis submitted that “the 5 June letter constitutes a settlement (subject to the applicable limit of liability) of Mobis’s claim under the Local Policy, in relation to the issue of the warehouse’s compliance with the applicable Australian Standards and the application of [the Faulty Design Exclusion]”.

Mobis submitted:

Mobis contended that this case is relevantly indistinguishable from that considered by the Court of Appeal in Baulderstone Hornibrook Engineering Pty Ltd v Gordian Runoff Ltd [2008] NSWCA 243.

In that case, Allsop P (as his Honour then was) with whom Beazley JA (as her Honour then was) and Campbell JA agreed, considered a letter sent by an insurer to the insured stating that “we…confirm that indemnity is granted” pursuant to the policy and stating that:

However, the only conclusion to which Allsop P came was that the letter constituted an admission by the insurer of its liability. Thus, his Honour said (at [290]):

His Honour did not conclude that the letter constituted a settlement contract.

Indeed his Honour continued (at [291]):

Allsop P referred to the observations of Chesterman J in Thiess Pty Ltd v ERC Frankona Reinsurance Limited [2007] QSC 4.

In that case, Chesterman J said (at [40] and [41]):

In reaching the conclusion that acceptance of a claim will “normally” amount to a legally binding contract, Chesterman J referred to Newton, Bellamy & Wolfe v State Government Insurance Office (Qld) [1986] 1 Qd R 431.

In that case, Andrews ACJ and Derrington J said (at 437):

McPherson J said (at 444):

These authorities establish that a legally binding contract of settlement may be created between an insurer and an insured where the insurer states that it accepts liability to indemnify the insured under the policy in question, and where the insured can be shown to have given consideration for that acceptance, constituted, for example, by a forbearance to sue the insurer, following from the insurer’s request (to be implied from its acceptance of liability) that it not do so.

But each case must depend on its own facts.

Here, it is plain from the terms of the 5 June 2015 letter, and the reply by Mobis’s solicitor of 12 June 2015, that both parties accepted that Mobis reserved its right to contend that, contrary to XL’s assertion, the limit of XL’s liability under the Local Policy exceeded EUR 10 million. If any implication is to be drawn from the correspondence, it is that XL expected, or at least contemplated the possibility that Mobis would bring proceedings to establish the correctness of that proposition; hardly surprising, bearing in mind that the Local Policy did not, in terms, contain any limit of liability of the kind asserted by XL.

I see no basis to conclude that, by the 5 June 2015 letter, XL was impliedly requesting Mobis to forbear from bringing action against it. XL was simply stating its position, and accepting that Mobis reserved its rights about the correctness of that position.

Nor do I see any other consideration given by Mobis for the qualified acceptance of liability given by XL in the 5 June 2015 letter.

The letter contains an admission. But it does not evidence a contract.

If, contrary to my finding, there was a contract of settlement, it was a term of that contract that XL’s acceptance of liability under the Local Policy was “on the basis of known facts and circumstances”.

Mobis submitted that, assuming the 5 June 2015 letter bespoke a settlement contract, there was no relevant change in the facts and circumstances known to XL such as would, notwithstanding its acceptance of liability, justify it now relying on the Faulty Design Exclusion.

In view of the conclusion to which I have come concerning the contract of settlement issue, there is no need for me to express any opinion about this matter.

However, had it been relevant for me to do so, I would have accepted Mobis’s submission.

In my opinion, a reasonable business person in the position of the parties would have understood that XL’s reference to “known facts and circumstances” was a reference to facts and circumstances relating to what actually caused the warehouse to collapse.

The facts and circumstances known to XL, as at 5 June 2015, were that it had:

XL does not contend that Costin Roe’s opinion was expressed upon the basis of incorrect information about the design of the warehouse, or that any new material facts about that design have come to light since the Costin Roe report.

XL seeks to justify its change of position solely on the basis that it received a second opinion from a structural engineer, Mr Paul Summers, to the effect that the warehouse did not comply with the relevant Australian Standards and was, in fact, defectively designed.

I do not accept that the receipt of that second opinion is a further “fact and circumstances” that would, assuming there was a contract of settlement, have justified XL from departing from its terms.

I accept Mobis’s submission that an opinion expressed by an expert after the loss is not a “fact or circumstance” relating to the loss but is rather, a means chosen by XL to evaluate the significance of the “known facts and circumstances”.

XL’s case seeks to read the words “on the basis of known facts and circumstances” as if they meant “as presently advised”, or something to that effect.

Had it been relevant for me to consider this question, I would not have accepted XL’s submission.

Mobis submitted that the 5 June 2015 letter constituted an election by XL to accept liability under the Local Policy and not to rely on the Faulty Design Exclusion.

Mobis submitted that by accepting liability under the Local Policy, and making the payment of $14.4 million referred to at [10] and [19] above, XL acted inconsistently with the maintenance of a defence based on the Faulty Design Exclusion and has thus made an unequivocal election between inconsistent rights.

I do not accept this submission.

In final submissions, Mobis accepted that the election by an insurer to accept or deny liability under a policy does not, without more, constitute an irrevocable election.

So much is established, for example, in the decision of the Court of Appeal in National Vulcan Engineering Insurance Group Ltd v Transfield Pty Ltd; National Vulcan Engineering Insurance Group Ltd v Connell Wagner Pty Ltd; National Vulcan Engineering Insurance Group Ltd v Coffey Partners International Pty Ltd (2003) 59 NSWLR 119; NSWCA 327 in which Santow JA (with whom Ipp JA and Young CJ in Eq agreed) said (at [64]):

I do not think the fact that XL’s acceptance of liability was limited to EUR 10 million affects this conclusion.

In any event, in order for there to be a waiver by election, the election must be made with knowledge of the relevant facts (for example, see Moore v The National Mutual Life Association of Australasia Limited [2011] NSWSC 416 at [73] (Ball J)).

The knowledge that XL would have had to have had in order to make an election in this case is the (alleged) fact that the warehouse was defectively designed.

If that is the true position (and I deal with this below) there is no suggestion that XL knew this to be so when it wrote the 5 June 2015 letter.

On the contrary, at that time it had the advice of Costin Roe that the design of the building complied with the relevant Australian Standards.

As I have concluded that the 5 June 2015 letter neither bespeaks a settlement contract nor a waiver by election, I see no basis upon which I could conclude that XL’s reliance on the Faulty Design Exclusion amounts to conduct otherwise than in accordance with its obligation to act with utmost good faith.

Whether XL is in fact entitled to rely upon the Faulty Design Exclusion is a matter which I consider later in these reasons.

A vital question in this case is what, as a matter of fact, caused the collapse of the warehouse; in particular, whether it was caused by the hail that fell on the day.

As the answer to many of the legal questions which arise depends on the answer to that question, I shall deal with it before addressing those legal questions.

The matter for consideration is the proximate cause, or causes, of the warehouse collapse.

Was that proximate cause hail alone, as XL contends, or storm (comprising a combination of rain and hail), as Mobis contends?

If the correct conclusion is that the proximate cause of the collapse was hail, and if the cover available to Mobis (whether under the Local Policy or the Master Policy) is confined by a hail limit, then Mobis’s case fails, as XL has already paid it the amount of that limit.

“Storm” is defined (somewhat unhelpfully) in the Local Policy as “storm, tempest, windstorm, hurricane, tornado, cyclone and typhoon”.

There is no suggestion is this case that what occurred on 25 April 2015 was anything other than a “storm” for the purpose of that definition (rather than a “tempest” and so on). I see no reason not to give the word “storm” its ordinary and natural meaning.

“Storm” is defined in the Australian Oxford Dictionary (online) as:

The definition in the Macquarie Dictionary (online) is:

As those definitions and common experience suggest, hail is usually accompanied by rain. As Einstein J said in Caine v Lumley General Insurance Ltd (2006) 14 ANZ Ins Cas 61-698; NSWSC 337 (at [99]):

In Caine v Lumley caravans were damaged in a storm that included hail and torrential rain. The relevant insurance policy contained a provision which capped liability under the policy for damage to caravans caused by hail to $100,000.

Einstein J held:

Einstein J’s decision was upheld in the Court of Appeal in Caine v Lumley General Insurance Ltd [2008] NSWCA 4.

McColl JA (with whom Mason P and McClellan CJ at CL agreed) said (at [76]):

Her Honour then set out, with approval, the passages from the judgment of Einstein J to which I have referred and concluded (at [82]):

A video shot by an independent witness, Ms Susan Bree, on her smart phone showed that the warehouse collapsed at, or more likely very slightly before, 3:49 pm.

Ms Bree was standing in or near a building immediately adjacent to the warehouse. At the start of the video one can hear what Mobis described in its submissions as a “rumbling sound” together with someone (presumably Ms Bree) exclaiming that “the whole Mobis building’s gone down”; suggesting the collapse was either then underway, or had concluded. The collapse itself occurred over a very short time; some 15 seconds.

Each of Mobis and XL called meteorological experts: Mr Barry Cook for Mobis and Prof Michael Manton for XL.

The evidence of those experts established the following sequence of meteorological events:

The evidence pointed very strongly to the conclusion that the warehouse would not have collapsed from rain alone on 25 April 2015.

I did not understand Mobis to make a submission to the contrary.

Each party adduced evidence from an expert hydrologist: Dr Ian Joliffe for Mobis and Mr Drew Bewsher for XL.

In their joint report, Dr Joliffe and Mr Bewsher agreed that from their perspective, without the presence of hail, the roof would not have collapsed from rain alone had it been properly designed and constructed.

Further, the evidence established that since the warehouse was constructed in 2008, there had been a number of days when heavier rain fell than occurred on 25 April 2015, without causing the warehouse to collapse.

As recently as 21 March 2015 (only a month before the date of collapse) a rain storm of considerably greater severity than that of 25 April 2015 occurred over the warehouse without causing any damage to it.

It must follow that, on those occasions, notwithstanding the considerable amount of rain that fell, the roof of the warehouse, and its related drainage systems, were adequate to cope with the downfall.

Those facts point, strongly in my opinion, to the probability that on 25 April 2015, the hail was the proximate cause of the collapse.

In their joint report, Dr Joliffe and Mr Bewsher also agreed that:

Dr Joliffe and Mr Bewsher also agreed that:

The structural engineer who gave evidence on behalf of Mobis, Mr David Carolan, expressed the opinion that the combination of hail and rain on the roof would have contributed to the imposed load on the roof structure. Mr Carolan opined that the combined weight of hail and rain caused the beams supporting the roof to deflect, creating reservoirs in which falling rain could pond, and that the combined weight of the ponded rain and hail led to the collapse.

Mobis submitted:

Mr Carolan prepared a computer model which sought to simulate this hypothesis. That model assumed a somewhat simplified sequence of weather events compared to those that actually occurred on the day. Nonetheless, it correctly predicted the timing of the roof collapse.

Mr Carolan explained that, on his hypothesis, it was a combination of the rain and hail that, in these circumstances, caused the roof to deflect. Mr Carolan hypothesised that the hail on the roof would have impeded the flow of water off the roof and caused water to leak through the “laps” in the roof sheeting and enter the building. This was consistent with images from CCTV footage within the warehouse which appeared to show water accumulating on the floor (evidently having leaked through the roof) some minutes before collapse.

The structural engineer retained by XL, Mr Summers, made some criticisms of Mr Carolan’s hypothesis.

However, assuming Mr Carolan’s hypothesis should be accepted, it suggests that it was the accumulation of hail, not rain, that was the real cause of the build-up of the load on the roof and the cause of rain water build up in reservoirs that Mr Carolan opined had been caused by such deflection. It seems probable that it was because of the hail on the roof that the rain water falling on the roof could not drain away. After all, the warehouse had successfully withstood heavier rain storms without the beams suffering deflection and without the warehouse collapsing.

That further suggests that it was the hail that caused the collapse.

Mr Cook and Prof Manton agreed that:

Based on these figures, Mr Carolan and Mr Summers performed calculations to assess whether the hail load thereby caused exceeded the collapse load (that is, the load needed to cause the warehouse to collapse) of the building.

If the hail load (that is, leaving aside the effect of rain) exceeded the collapse load, that would suggest that the hail, alone, caused the collapse.

Mr Summers calculated a collapse load of 0.33 kPa. Mr Carolan calculated a collapse load of 0.37 kPa. I was not invited to make any assessment of which of these calculations was correct.

Assuming that the hail stone density and the packing density were at the midpoint of the figures agreed by Mr Cook and Prof Manton (that is, 850 kg/m3 and 66% respectively), the evidence showed that the hail load would exceed Mr Summers’s collapse load if there was 6 cm of hail and would exceed Mr Carolan’s collapse load if there was 7 cm of hail.

Further, if there was 7 cm of hail, the hail load would have exceeded Mr Summers’s collapse load no matter what (within the range agreed by Mr Cook and Prof Manton) the hail stone density and packing density.

Even assuming the correctness of Mr Carolan’s (higher) collapse load calculation, very little rain (over and above the hail) would have been needed to cause the combined rain/hail load to exceed the collapse load.

Again, those matters point to the conclusion that it was the hail that caused the collapse.

Finally, hail was falling onto and accumulating on the roof of the warehouse until a very short time before the collapse.

As I have described, the warehouse collapsed at, or probably very slightly before, 3:49 pm. Heavy hail had been falling onto the roof of the warehouse until very shortly before then; at most, 3 minutes, but perhaps only 1 minute: see [119] above).

There was thus only a brief period prior to the collapse when rain alone was falling onto the roof of the warehouse. During that brief period it seems very likely that most, if not all, of the hail that had fallen over the previous 5, 6 or 7 minutes remained on the roof, impeding the runoff of the rain and, on Mr Carolan’s hypothesis, causing deflection of the beams, and creation of reservoirs for the rain.

When these factors are considered together, they persuade me that it was the hail that fell on 25 April 2015 that was the proximate cause of the warehouse collapse.

To adopt the language of McColl JA in Caine, my opinion is that when considering these matters, and applying a common sense standard, the man or woman in the street would reach the conclusion that the “reality, predominance and efficiency” of the impact of the hail in the storm that occurred shows that it was the proximate cause of the collapse.

A critical issue in the proceedings is whether the Local Policy has a hail limit, notwithstanding the fact that the policy document issued by XL on 1 March 2011, and renewed annually thereafter, specifies no such limit.

XL contends that the Local Policy has such a limit for two reasons.

The first is the asserted effect of cl 1.9b of the Local Policy (although this argument was not maintained with great enthusiasm in final submissions).

The second is XL’s contention that the Local Policy should be rectified to include the Hail Limit.

Clause 1.9 of the Local Policy provided:

It is common ground that the reference to “programme limits of indemnity” in the clause is a reference to the aggregate limits of indemnity in the Master Policy.

In my opinion, the effect of cl 1.9b is that if payment of an earlier claim under the Master Policy eroded the aggregate limit available under the Master Policy for, say, a storm claim, then the amount available to Mobis under the Local Policy for a storm claim would be reduced accordingly; even if the amount of that claim did not exceed the storm limit in the Local Policy.

I accept Mobis’s submission that cl 1.9b requires no more than a matching of limits of liability by reference to specific perils appearing in the Local Policy to determine whether the aggregate limits of liability under the Master Policy have been exceeded.

This is consistent with the third paragraph of cl 3.1.8 of the Master Policy which provides:

There is no suggestion that, so construed, cl 1.9b has been enlivened.

Contrary to XL’s submission, I do not accept that the effect of this clause is to import the Hail Limit in the Master Policy into the Local Policy.

If the parties to the Local Policy had intended such a result, they would have used language more clearly directed to that result.

XL contends that the Local Policy should be rectified so as to include the Hail Limit.

In opening submissions, XL put its case succinctly:

The relevant principles were recently restated by the High Court in Simic v New South Wales Land and Housing Corporation [2016] HCA 47; as follows:

Further, each of these matters must be established by clear and convincing evidence. As Kiefel J (as the Chief Justice then was) stated in Simic at [41]:

Rectification is concerned with the subjective and actual state of mind of the parties, objectively ascertained.

As Kiefel J explained in Simic (at [42]):

The Local Policy was incepted on 1 January 2011, part way through the 23 June 2010 to 23 June 2011 period of the then current (and first) Master Policy.

The Local Policy wording was not provided by XL to Mobis’s Australian broker, Marsh Australia, until 1 March 2011. It contained no hail limit.

Thus, the issue to be determined on XL’s rectification case is whether, as at 1 January 2011, it was the agreement of XL and Mobis, or their common intention, that the Local Policy have the Hail Limit and that, by reason of common mistake, it did not.

The only witness called by XL to give evidence of having made a mistake of any kind was Mr Timothy Jones. Mr Jones was XL’s Underwriting Manager in Australia and the person responsible for signing the Local Policy on behalf of XL.

In his affidavit, Mr Jones said:

The XL officer responsible for drafting the Local Policy wording issued on 1 March 2011 was Ms Denise Stanley. XL did not call Ms Stanley. Mr Jones said that Ms Stanley left XL “several years ago”. XL offered no evidence as to any steps taken to locate and take a statement from Ms Stanley. Mr Jones did not suggest that Ms Stanley left XL’s employ on bad terms. Although Mobis did not suggest that a Jones v Dunkel (1959) 101 CLR 298 inference should be drawn by reason of her absence from the witness box, the fact remains that there is no explanation from her as to how the Local Policy came to be worded as it is.

In his affidavit, Mr Jones said:

The implications of that evidence were not developed by Mr Jones in his affidavit, or by XL in its submissions.

However, as is revealed below, there is no suggestion in the written communications between the parties in January and February 2011 that Ms Stanley utilised the services of XL India to prepare the wording of the Local Policy.

As I explain below, the event which led to the issue of the Local Policy without the Hail Limit was an email that Ms Ladislava Hurtajova sent on 22 December 2010 to a colleague, Mr Patrick Hofmann.

Ms Hurtajova was, at the relevant time, XL’s Upper Middle Markets Property Underwriter for the Austrian and Central and Eastern European markets. Ms Hurtajova held this position from 2009 until mid 2013; she was XL’s program underwriter for the policy years 2010/11, 2011/12 and 2012/13. Ms Hurtajova was responsible for the writing of the original Master Policy and its renewal in those years.

In the circumstances I discuss below, it was Ms Hurtajova who determined that the Hail Limit should be included in the Master Policy. Ms Hurtajova gave evidence that she was not involved in the drafting of the Local Policy (although that assertion requires some qualification), and that, so far as she knew, at no time did anyone on behalf of Mobis Korea or Mobis request that a policy be issued without the Hail Limit.

Ms Hurtajova said she assumed and intended that the Local Policy would have the same limits as the Master Policy.

Ms Hurtajova said that the only exception to this was “that a local policy might have a lower limit if the declared values in that country were lower than the relevant limit in the Master Policy, or if a different coverage was specifically requested”. The latter circumstance did occur in relation to the Local Policy in the 2012/2013 policy year after Mobis sought to have its newly acquired Perth premises covered. That led to a lower limit for earthquake loss specified in the case of Perth than for Eastern Creek (see [40] above).

Although in her affidavit Ms Hurtajova said the Local Policy “contained a mistake” in that it did not include the Hail Limit, she did not say that she made a mistake in her 22 December 2010 email. Indeed she did not mention that email at all in her affidavit.

XL also called evidence from Mr Stefan Rossa. Mr Rossa assumed the responsibilities formerly undertaken by Ms Hurtajova for the 2014/15 policy period.

Mr Rossa played no role in the issue and renewal of the Local Policy. However, like Ms Hurtajova, Mr Rossa gave evidence that he assumed, expected and intended that the Local Policy would have the same sub limits as the Master Policy.

The officer at Mobis most directly involved in the circumstances leading to the issue of the Local Policy was Mr Sam Papa, who was then Mobis’s General Manager, Finance and Administration. Mobis did not call Mr Papa. Mr Papa has left Mobis’s employ, evidently in less than happy circumstances. XL made no Jones v Dunkel submission concerning his absence. The result is that assessment of his state of mind at the critical time depends on his email communications (particularly with Mr Roman Kalocai, a Senior Client Executive from Mobis Korea’s insurance broker, Marsh Europe).

In December 2009, Mobis Korea appointed Marsh Europe as its worldwide insurance broker for property damage and business interruption insurance.

Thus, on 15 December 2009, Mobis Korea wrote a letter addressed “[t]o whom it may concern” stating:

A few days earlier, on 8 December 2009, Mr Kalocai wrote to a colleague:

On 1 April 2010, Marsh Europe sent XL (for the attention of Ms Hurtajova and others) a proposal for property damage and business interruption insurance for “Mobis subsidiaries” in the countries referred to in Mr Kalocai’s email of 8 December 2009 above (including Australia).

The proposal specified an inception date of 1 January 2011 in the case of the “Mobis subsidiary” in Australia.

One of the sub limits that Marsh Europe proposed was EUR 50 million for “flood, windstorm, hail, earthquake”.

At around the time Marsh Europe made this proposal, Ms Hurtajova told Mr Kalocai that:

So far as concerns Marsh Europe’s proposal that cover be granted with a EUR 50 million sub limit for hail, Ms Hurtajova said in her affidavit:

Ms Hurtajova was not challenged about this evidence.

Ms Hurtajova prepared draft wording for the proposed cover. In her affidavit she said:

Ms Hurtajova and Mr Kalocai met on 18 June 2010 at XL’s Vienna office. Ms Hurtajova had sent her draft wording to Mr Kalocai the day before. It contained, and highlighted, the EUR 10 million limit for:

Evidently, during the 18 June 2010 meeting, Ms Hurtajova and Mr Kalocai discussed the lower limits on which Ms Hurtajova was insisting. Thus, following the meeting, Mr Kalocai wrote to Ms Hurtajova:

Ultimately, XL agreed to a lower deductible for machinery breakdown, evidently in exchange for Mr Kalocai’s agreement that “you can have the lower limits for the other perils we were talking about”.

After the 18 June 2010 meeting, neither Mr Kalocai nor anyone else from Marsh Europe or Mobis requested insurance for natural peril events (in particular, for hail) in an amount greater than EUR 10 million.

On 27 July 2010, the first Master Policy was issued with effect for the period 23 June 2010 to 23 June 2011.

It had a limit of EUR 10 million for:

The Master Policy recorded that the inception date for Australia would be 1 January 2011.

On 26 July 2010, Ms Hurtajova made a file note in which she described the “Program Structure” as follows:

In her affidavit, Ms Hurtajova explained that “Freedom of Services” policies could be issued for all Mobis subsidiaries in the European Union but that for non-European countries “it would be necessary for XL to issue separate ‘local policies’ as different regulatory and legal requirements apply”.

Evidently, no steps were taken to effect a Local Policy for Australia until September 2010.

On 28 September 2010, Mr Kalocai sent an email to a colleague at Marsh Australia (Ms Carmel Adelson):

Mr Kalocai attached to that email a “Summary of Global Property Insurance” for Mobis for the period 23 June 2010 to 23 June 2011 (that is, the terms of the then current Master Policy).

That summary specified the “Main Sub-limits” as follows:

The summary thus included the Hail Limit of EUR 10 million.

It also stated:

This was a reference to Art 5.2 of the Master Policy. That article is headed “Difference in Limits (DIL)/Difference in Conditions (DIC)”. The parties referred to the article as the “DIL/DIC clause”. It provides:

I will return to this article when discussing a later email exchange between Mr Kalocai and Mr Papa.

The summary stated that for Australia (that is, for Mobis) there was to be a Local Policy with an inception date of 1 January 2011.

On 27 October 2010, Ms Adelson sent an email to Mr Papa. This appears to be the first communication to Mobis following the issue of the Master Policy about the possibility of it becoming a party to a Local Policy with XL.

Ms Adelson’s email stated:

Ms Adelson attached to that email the Summary of Global Property Insurance that she had received from Mr Kalocai.

On 29 October 2010, Mr Papa sent an internal email to Mr Chung at Mobis as follows:

On 4 November 2010 Mr Papa replied to Ms Adelson’s 27 October 2010 email:

Ms Adelson replied stating:

Mr Papa’s somewhat robust response to Ms Adelson’s enquiries evidently prompted Mr Kalocai to intervene.

Mr Kalocai rang Mr Papa from Europe and, on 15 November 2010, wrote:

Mr Papa replied on 25 November 2010 stating that “insurances have not been cancelled yet” and making enquiries about premium and applicable excess.

On 26 November 2010, Mr Kalocai sent Mr Papa a further copy of the Summary of Global Property Insurance document.

On 30 November 2010, Mr Kalocai sent an email to Mr Papa stating:

On 7 December 2010, Mr Kalocai, in the course of responding to an enquiry by Mr Papa concerning premium and like matters, asked Ms Adelson to inform Mr Papa that:

Later, on 16 December 2010, Mr Kalocai sent Mr Papa an email requesting that he “please provide us with your final decision”.

Mr Papa replied on 17 December 2010 saying that he had been “waiting for the indication on pricing”, that Mobis’s “CEO wants to know total cost before I commit” and that “the sooner I get a response the sooner I can decide”.

Ms Adelson replied the same day providing details of what the premium was to be.

On the same day Mr Papa wrote to Ms Adelson, and her colleague, Ms Cory Ormsby:

Ms Adelson replied that this was correct.

Later on 17 December 2010, Mr Papa sent an email to Mr Shaun Ryu (a colleague within Mobis):

The “brief summary” of the “[i]nsurance policy implemented by HQ” contained details of declared values and premium (but made no reference to policy limits).

On 20 December 2010, Mr Papa sent an email to, amongst others, Mr Kalocai and Ms Adelson:

Mr Papa was referring to a statement in the Summary Document that I set out at [210] above which summarised the effect of Art 5.2 of the Master Policy (the DIL/DIC clause).

Mr Kalocai replied on 21 December 2010 but did not, in terms, respond to Mr Papa’s enquiry as to whether Mobis would have “locally defined…coverage” under the proposed local policy.

Rather, he said:

From this Mr Papa must have understood that XL Australia would issue a Local Policy that would accommodate any (local) “legislative” requirements and would be the “best local standard wording”.

The fact that Mr Kalocai put quotation marks around the expression “best local standard wording” may have suggested to Mr Papa that the expression had some particular significance from XL’s point of view. Earlier, Ms Hurtajova had explained to Mr Kalocai that “good local standard” wording meant wording drafted to meet local regulatory and legal requirements (see [192] above).

In his email, Mr Kalocai also explained to Mr Papa the effect of the DIL/DIC clause, and that by reason of that clause, Mobis might have more cover under the Master Policy than under the proposed Local Policy.

Mr Kalocai did not suggest, in terms, that the opposite might be the case: namely, that cover under the proposed Local Policy might be more generous than under the Master Policy.

But the descriptors “best” and “local” were apt to convey to Mr Papa that the wording of the proposed Local Policy would be as generous as XL was able to be in the local context. And the fact that Mr Kalocai did not disabuse Mr Papa from the notion that there might be “locally defined…coverage” was apt to leave Mr Papa with the impression that this might come to pass; that is, that there might be locally defined coverage.

The fact that, as I describe below, Mr Papa was, as soon as the Local Policy incepted on 1 January 2011, anxious to obtain the Local Policy wording, suggests he did have that impression.

In his 20 December 2010 email, Mr Papa also made an enquiry about a EUR 5 million sub limit for what he described as “building destruction”. He asked:

Mr Papa was evidently referring to the EUR 5 million sub limit for “Erection and Construction” referred to in the Summary of Global Property Insurance that he had received from Ms Adelson and from Mr Kalocai (see [208] above). His enquiry was whether cover for what he described as “building destruction” would be limited to EUR 5 million if the building became “unusable” for reasons other than the perils that had a EUR 50 million limit (earthquake, volcano and so on).

On 21 December 2010, in response to that enquiry, and in the same email in which he dealt with the DIL/DIC clause, Mr Kalocai replied:

Although neither Mr Papa nor Mr Kalocai in this exchange referred to the EUR 10 million hail sub limit referred to in the Summary Document, it shows that Mr Papa was conscious of the various sub limits specified in that document and that issues might arise as to how they interplayed.

Later on 21 December 2010, Mr Papa replied to Mr Kalocai:

Again, this shows that Mr Papa had applied his mind to the sub limits in the Summary Document (and thus in the Master Policy).

A short time later Mr Papa sent a further email to Mr Kalocai which included:

This email suggests that Mr Papa had discussed the Summary Document with Mobis’s existing broker and that, following that discussion with that broker, he was “reasonably comfortable” with its terms.

Nonetheless, the email records that Mr Papa wanted to see the terms and conditions of the Local Policy before committing Mobis.

Mr Kalocai replied:

Mr Papa must have understood that the only document that Mr Kalocai was then “able to send” him was the Master Policy. There is no suggestion in the evidence that the wording of the Local Policy was available at this time.

On 21 December 2010, Mr Papa wrote to Mr Kalocai stating:

The next day, Mr Kalocai responded to Mr Papa’s request for the “wording” by sending a copy of the Master Policy.

On 21 December 2010, Mr Kalocai sent an email to Ms Hurtajova (in Slovak):

Mr Kalocai also provided Ms Hurtajova with the address of Mobis’s warehouse.

Ms Hurtajova explained in her affidavit that Mr Kalocai’s email “meant that he confirmed the interest of Mobis Australia was to be covered from 1 January 2011 and gave the location and values”.

As reproduced in the Court Book, this email bears the date 22 December 2010. In her affidavit, Ms Hurtajova described the email as having the date 21 December 2010. The date ascribed to the email by Ms Hurtajova is more likely, as it appears to have led to an email Ms Hurtajova sent to her colleague, Mr Patrick Hofmann, to which I will return. Evidently, the date has altered from European to Australian time in the course of reproduction.

On 31 December 2010, XL issued an endorsement of the Master Policy showing as the “Reason for Issuance” the inclusion of, amongst other countries, Australia, “with effect of 01.01.2011”.

The endorsement recorded revised declared values for Mobis’s building, machinery and stock, and for business interruption cover. All these figures were greater than appeared in the Master Policy when issued earlier in 2010 (see [208] above). At least some were the result of instructions from Mr Papa.

The endorsement included the statement:

Also on 31 December 2010, Mr Kalocai sent Mr Papa a copy of that endorsement under cover of an email:

The endorsement to the Master Policy had the effect that Mobis became an “Insured Company” under the Master Policy with effect from 1 January 2011. This was consistent with the proposed inception date for Australia specified in the Master Policy when it was issued on 27 July 2010 (see [200] above).

A Local Policy was to be issued, also with effect from 1 January 2011.

In final oral submissions, there was some debate as to what would have occurred had a storm (such as that of 25 April 2015) occurred on say 1 January 2011 itself. Mobis would then have had cover under the Master Policy. That cover would have been subject to the EUR 10 million Hail Limit in the Master Policy. But Mobis would also have had cover under the Local Policy, albeit on terms not yet published by XL. To ascertain Mobis’s rights under the Local Policy, it would have been necessary for XL to publish those terms. I see no reason to assume those terms would have differed from those published in March 2011.

On 12 January 2011, Mr Papa wrote to Mr Kalocai saying that he had not been contacted by “your local offices” and that “as far as I know I don’t have a copy of a local policy”. Mr Papa asked for “a Certificate of Currency for our Financier (ANZ Bank) as they hold security over our building”.

On 14 January 2011 Ms Adelson sent Mr Papa a “Certificate of Currency” for a total of $64.6 million. The certificate bore the policy number ultimately ascribed to the Local Policy.

As XL contends that the Hail Limit was omitted from the Local Policy wording issued under cover of Ms Stanley’s letter of 1 March 2011 by common mistake, it is necessary to examine with some care how that circumstance came about.

XL’s procedures at the time required that a “Local Policy Instruction” or “LPI” be prepared in order to generate the wording of a local policy.

Evidently, the person responsible for creating the relevant LPI was an officer of XL stationed in Zurich, Mr Patrick Hofmann.

As I have mentioned (see [256] above) on 21 December 2010, Mr Kalocai sent an email to Ms Hurtajova in Slovak which, Ms Hurtajova said, “meant that [Mr Kalocai] confirmed the interest of Mobis Australia was to be covered from 1 January 2011 and gave the location and values”.

The translation of that email in evidence (which I have set out at [256] above) does not reveal that Mr Kalocai set out any “values”. The translation simply reads that Mr Kalocai confirmed coverage “in the sense of [sic] the reported insurance sums”.

On 22 December 2010, Ms Hurtajova sent Mr Hofmann an email. I infer Ms Hurtajova sent this email as a result of receiving Mr Kalocai’s email of 21 December 2010. In her email, she gave instructions to Mr Hofmann as to the sub limits to be included in the Local Policy.

I see this email as being critical to XL’s rectification case. XL did not refer to it in opening submissions. It was referred to, albeit obliquely, in Mobis’s opening submissions and given careful attention during the cross-examination of Ms Hurtajova. It requires careful examination.

The email read:

It will be recalled (see [47] and [48] above) that the relevant limit in the Master Policy of EUR 10 million was expressed to be in respect of:

In her email, Ms Hurtajova specified a EUR 40 million sub limit for “storm” and a EUR 10 million sub limit only for:

The first two of the perils corresponded with perils identified in the Master Policy as being subject to a EUR 10 million limit: “avalanche” and “landslide”. The third (“snow pressure”) corresponded, approximately, to “weight of snow and ice” (also with a EUR 10 million limit).

But Ms Hurtajova specified no limit for hail (or for any other peril the subject of the EUR 10 million limit in the Master Policy: inundation, windstorm in excess of 75 km/hour, tsunami, rockslide or falling stones).

A comparison between the limits Ms Hurtajova specified in her 22 December 2010 email with those then in the Master Policy is illustrated by this table (with the Hail Limit in the Master Policy emphasised):

It can be seen that Ms Hurtajova specified a number limits that had no equivalent in the Master Policy (accidental damage/all risk, SCCR (strike, civil commotion, riot), malicious acts/vandalism, sprinkler leakage, vehicle impact, volcanic eruption and water damage).

On the other hand, Ms Hurtajova did not specify a number of sub limits that did appear in the Master Policy; most notably hail, but also high water/inundation/windstorm greater than 75 km/hour, rockslide, falling stones and landslide (which were grouped with hail in the Master Policy with a EUR 10 million limit) as well as contract works and/or property under construction, machinery breakdown and electronic equipment breakdown.

Ms Hurtajova also expressed lower sub limits for perils that were named in the Master Policy (FLEXA (fire, lightning, explosion, aircraft impact), storm, flood and earthquake). This was consistent with the existence in the Master Policy of the DIL/DIC clause (see [211] above) and Art 3.1.8 (the Overlap Clause; see below at [623]), as was the fact that Ms Hurtajova did not express any higher sub limits from the proposed local policy than existed in the Master Policy.

This suggests that Ms Hurtajova took great care with her email. She descended to a great deal of detail. But she did not, anywhere in her email, mention “hail”.

As I have mentioned, Ms Hurtajova’s evidence was that she was not “involved in drafting the local policy for Australia” but that she assumed and intended that the Local Policy would have the same limits as in the Master Policy.

But Ms Hurtajova’s email shows that she was “involved” in drafting the Local Policy in that she started the process whereby the Local Policy came to be created. She did that by sending her email of 22 December 2010 to Mr Hofmann. That email is expressed to contain a “summary of data” for “Mobis Slovakia expansion per 01.01.2011 Australia” that Mr Hofmann would need. That “data” comprised a lengthy and evidently comprehensive list of the limits and sub limits to go in the Local Policy.

In those circumstances, one might have expected that if Ms Hurtajova’s position was that she mistakenly failed to mention any limit for hail in her email, she would have said so. But at no time has Ms Hurtajova said this. She did not say so in her affidavit. She did not mention her email in her affidavit (which, I should add, comprised 97 paragraphs and was 18 pages long).

In cross-examination, Ms Hurtajova agreed that her email represented her instructions to Mr Hofmann about what was to go in the LPI (and it must follow, thereby into the Local Policy itself).

Thus:

The impression I had of Ms Hurtajova was that she was attending carefully to each question put and taking care to answer only the precise question. Although her command of English was of a very high order, it is not her first language. Indeed, she had available an interpreter to translate into her native Slovak language (although she needed only very occasionally to have the interpreter’s assistance).

The proposition put to Ms Hurtajova, and accepted by her, was that her email represented her “instructions” to Mr Hofmann as to the limits to be inserted in the LPI. Her affirmative response was, literally, correct. Her email did represent her instructions to Mr Hofmann about what limits were to go in the LPI.

That does beg the question of whether she was mistaken in giving those instructions; in particular, by not instructing Mr Hofmann to include a EUR 10 million limit for hail or, indeed, making any reference to hail in her email.

It was not put to Ms Hurtajova that she had made no mistake in giving those instructions, or that she deliberately refrained from instructing Mr Hofmann to include a EUR 10 million limit for hail.

That was no doubt because, although Ms Hurtajova said she expected and intended that the Local Policy would have the same limits as the Master Policy and that there was a “mistake” in the Local Policy, Ms Hurtajova has never asserted that she had made any mistake, whether by sending this email, or at all.

As Mobis submitted:

On 28 December 2010, Mr Hofmann replied by email to Ms Hurtajova attaching an “updated” LPI. The LPI contained details for local policies in the United Arab Emirates and Russia, as well as Australia.

That LPI stated that the limits for the Australian Local Policy were to include:

Unsurprisingly, bearing in mind the contents of Ms Hurtajova’s 22 December 2010 email, and the absence from that email of any reference to “hail”, the LPI did not include any limit for “hail”.

It did include a EUR 40 million limit for “Windstorm/Hail” (corresponding to Ms Hurtajova’s proposed EUR 40 million limit for “Storm”). This, evidently, reflected Mr Hofmann’s understanding of the effect of Ms Hurtajova’s instructions of 22 December 2010.

A comparison between the limits and sub limits Ms Hurtajova specified in her 22 December 2010 email, and those set out in Mr Hofmann’s LPI, is illustrated by this table (with the reference “Windstorm/Hail” highlighted):

The table shows that Mr Hofmann matched each of Ms Hurtajova’s proposed sub limits virtually word for word; but with “Windstorm/Hail” substituted for “Storm”. As I have mentioned, Mr Hofmann did not give evidence. I therefore have no evidence before me as to why Mr Hofmann decided to make that change.

In his 28 December 2010 email, Mr Hofmann asked Ms Hurtajova to “please quickly check and confirm it so that I can then send it to the Australian PAM”. In his affidavit, Mr Jones explained that “PAM” was an acronym used by XL for local policy administrators. At this time, the relevant Australian policy administrator was Ms Stanley (to whom I have referred at [173] and [176] above).

In cross-examination, Ms Hurtajova said that she was out of the office in the days following Christmas in 2010 and that she did not know how she “answered his request for checking”.

However, as emerges below, Ms Hurtajova later checked and approved a version of the LPI which was, so far as concerned the sub limit descriptions, the same as that Mr Hofmann forwarded to her on 28 December 2010.

In cross-examination, Ms Hurtajova was taken to the entries in the LPI attached to Mr Hofmann’s email concerning the UAE and Russia (which were relevantly in the same format as those for Australia, albeit with some differing monetary limits) and gave this evidence:

Ms Hurtajova answered “No” to the two questions suggesting that the “Windstorm/Hail” limit of EUR 50 million in the LPI applied to both windstorm and hailstorm. But she then agreed that “that’s what [the LPI] says” and that this was the only instruction given in the LPI concerning hail.

The implications of Ms Hurtajova’s negative answers to these two questions were not explored in re-examination (although an attempt was made, by questions that I rejected, to deal generally with the “Windstorm/Hail” reference in a later version of the LPI to which I refer below).

Ms Hurtajova was then taken to the page in the LPI dealing with the Australian Local Policy, and gave this evidence:

Once again, Ms Hurtajova gave a literally correct answer to precisely crafted questions.

The following day, 29 December 2010, Mr Hofmann sent an email to Ms Stanley (at XL Australia) (with a copy to Ms Hurtajova):

The details for “snow pressure/weight of snow and ice” and “wind storm/hail” were now expressed to be in Australian dollars. The figure for “snow pressure/weight of snow and ice” cover was stated to be $13,397,000 (equivalent to EUR 10 million). The limit for “wind storm/hail” was stated to be $64,975,450. This was equivalent, for a reason not explained in the evidence, to EUR 48.5 million rather than EUR 40 million as set out in Mr Hofmann’s 28 December 2010 draft LPI. Neither party contended this was significant. Ultimately, the “storm” limit in the Local Policy, as issued on 1 March 2011, was the Australian dollar equivalent of EUR 50 million. There was no limit for hail.

Ms Hurtajova gave the following evidence in cross-examination in relation to the LPI that Mr Hofmann sent Ms Stanley on 28 December 2010:

Again, Ms Hurtajova attended closely to the question put and gave a literally correct answer. The proposition with which she agreed was that the instructions in the LPI sent by Mr Hofmann to Ms Stanley were “implemented” (by XL in Australia) in the Local Policy. That was true (save for the irrelevant difference that the “storm” limit which ended up being slightly higher).

By 8 February 2011, a final version of the LPI was created.

Ms Hurtajova signed this final version of the LPI on that day. This version of the LPI had the sub limits in Euros; namely EUR 10 million for “Snow Pressure/Weight of Snow and Ice” and EUR 50 million for “Wind storm/Hail” (a slight increase from the EUR 48.5 million limit referred to in earlier versions of the Local Policy Instruction). Again, there was no other, or lower, limit for “hail”.

In cross-examination, Ms Hurtajova was taken to this document and gave this evidence:

It is hard to see how Ms Hurtajova could have “checked and approved” this document without noticing that it contained a EUR 50 million limit for “Windstorm/Hail”, and no separate limit for hail.

Ms Hurtajova did not explain this in her affidavit. Indeed, she made no mention in her affidavit of signing this document.

Ms Hurtajova did not assert, in response to the questions put in cross-examination, that she was mistaken to have approved such a limit. It is true that such an assertion would not have been responsive to the careful questions put. But the fact remains that I have no explanation of how this came about.

Also on 8 February 2011, Ms Stanley wrote to Ms Ormsby at Marsh asking for “the address of the local client so that I can invoice”.

Ms Ormsby replied on 14 February 2011, attaching the Summary of Global Property Insurance that Mr Kalocai had sent Ms Adelson on 28 September 2010 (see [207] above) and Mr Papa on 26 November 2010 (see [223] above) and was the subject of the email exchange between Mr Papa and Mr Kalocai in December 2010 that I have set out above (see [225] to [235] above). That document contained “the address of the local client” as requested by Ms Stanley. It also stated that the “Global” insurance was subject to the “Main Sublimits” set out above, including the EUR 10 million Hail Limit.

On 18 February 2011, Ms Ormsby wrote to Ms Stanley asking her to “send the policy document to Mobis” and stating that “the client requires a copy of the schedule and wording asap”.

Ms Ormsby followed up on the matter on 22 February 2011 when she wrote to Ms Stanley saying that “the client has been chasing their policy document” and enquiring “can you please advise when to expect”.

Ms Stanley replied within minutes:

On 23 February 2011, Mr Papa wrote to Marsh Australia:

Ms Ormsby from Marsh Australia replied the same day:

The “version that had been sent to [Ms Ormsby] last week” was not in evidence before me. Nor was there any evidence of what revisions were made to the wording at this time.

On 1 March 2011, Ms Stanley sent Marsh Australia the policy wording for the Local Policy.

It had the following limits:

Those limits differ slightly, but not relevantly, from those in the Local Policy at the time of the collapse.

A comparison of the limits on the 8 February 2011 LPI, and those in the Local Policy as issued, is illustrated in this table:

The AUD figures correspond to the EUR figures at the prevailing exchange rate.

Although the LPI specified 15 separate perils and sub limits, the Local Policy specified only 6; 5 of which reflected the overall policy limit.

Mr Jones (to whom I have referred above: see [171] to [175]) said in his affidavit that “the process within XL by which local policies are prepared so as to be presented to the local underwriter for final review and signature” involved the creation of LPIs and the use of a “precedent or template wording for each of the countries in which a local policy is required as part of a global program of a particular insured known as Good Local Standard policy wording”.

I infer from this evidence that Ms Stanley used the LPI certified as correct by Ms Hurtajova on 8 February 2011, as well as the template to which Mr Jones referred, to create the document issued on 1 March 2011.

Otherwise, there is no evidence to explain by what process Ms Stanley selected the limits which appeared in the Local Policy, and to explain, in particular, why the limits differ from those in the 8 February 2011 LPI and from the Summary of Global Property Insurance document sent by Ms Ormsby to Ms Stanley on 14 February 2011.

Under cover of a letter dated 1 March 2011, Ms Stanley sent Ms Ormsby the “Policy Wording” for the Local Policy (with no hail limit).

Evidently, on 3 March 2011 Ms Ormsby had not received that wording as, on that day, she sent an email to Ms Stanley:

Ms Stanley replied saying that “the wording was mailed yesterday” and attached a “soft copy in the interim”.

On 22 March 2011, Ms Ormsby wrote to Mr Papa:

The cover page of that wording described Mobis’s address as “77 Brock Drive, Eastern Creek”.

Mr Papa replied to Marsh Australia pointing out that Mobis’s address was in fact 77 “Peter” Brock Drive.

Mr Papa’s reply was sent within an hour of his receipt of Ms Ormsby’s email. However, the fact that Mr Papa was able to detect the slight error in Mobis’s address as recorded on the cover sheet of the Local Policy, when considered in the context of the numerous enquiries he made in January and February 2011 concerning provision of the Local Policy wording (including a “copy of the schedule”: see [323] above), suggests that it is probable that he read the wording when he received it and, at least, saw the limits stated in it.

XL does not suggest that Mr Papa saw that there was no hail limit in the policy, understood its absence to be inconsistent with what had been explained to him by Mr Kalocai in December 2010 and remained silent, knowing that XL had made an error. XL’s case is not that there was a unilateral mistake (by it) known to Mobis in the sense considered in Taylor v Johnson (1983) 151 CLR 422. XL’s case is that there was a common mistake.

I would infer that Mr Papa was expecting a policy which, although issued in the context of the global coverage of the Mobis group as a whole, and which thus involved the Master Policy, was to have “best local standard wording” (see [237] above), and which may well have “locally defined excesses and coverage” (see [233] and [235] to [236] above).

I infer, from the manner of Mr Papa’s response when he finally received the Local Policy wording, that he saw nothing in it inconsistent with his discussions with Mr Kalocai in December 2010.

The first Local Policy was for the period 1 January 2011 to 23 June 2012.

Thereafter, it was renewed annually on terms which, relevantly, were the same as those of 1 March 2011 and did not include a hail limit.

For the policy year 23 June 2013 to 23 June 2014, Marsh Europe went to the market with a fresh request for proposal which sought insurance for a program, including a local policy for Australia, which had a hail limit of EUR 10 million. XL won that business and continued as insurer for the Mobis group. As XL submitted:

On 27 April 2015, (two days after the collapse) Mr Kalocai sent an email to Mr Gerardo Pierro (at Marsh Sydney) stating:

Mr Kalocai was referring to the Local Policy. His email suggests that he assumed that the Local Policy had the Hail Limit.

Mr Pierro’s reply was that “the loss was caused by strong winds, hail and resulted water damage”.

On 30 April 2015, Mr Nicholas Kerr (from Marsh Sydney) sent Mr Kalocai a copy of the wording of the Local Policy.

On the same day, Mr Kalocai replied, attaching a copy of the current Summary of Global Property Insurance under the Master Policy and continued:

Again, this email suggests that Mr Kalocai assumed there to be a hail limit in the Local Policy.

I will return to the significance (if any) of this fact.

In order to succeed on its rectification case, XL must prove, to the relevant standard, that as a result of a mistake common to both XL and Mobis, the Local Policy does not reflect their common intention that it include the Hail Limit.

XL must prove these matters on the probabilities, but “to a high standard” and “in the clearest and most satisfactory manner” (see [165] above).

I have set out, in some detail, the circumstances that appear to have led to the absence of the Hail Limit in the Local Policy.

The trail of events started with Ms Hurtajova’s email to Mr Hofmann of 22 December 2010.

As I have said, the detail in Ms Hurtajova’s email of 22 December 2010, and the differences between the limits specified in it, compared to those then in the Master Policy, both suggest that Ms Hurtajova applied her mind carefully to the task of spelling out the precise limits that she wanted Mr Hofmann to include in the LPI.

The email does not have the appearance of being prepared by reference to a template. It appears to be a bespoke document, prepared by Ms Hurtajova for the purpose of giving Mr Hofmann instructions as to the sub limits he should include in the LPI.

Ms Hurtajova did not propose a hail limit in that email.

On 28 December 2010, Mr Hofmann prepared the “updated” LPI on that basis: without a hail limit. As Ms Hurtajova’s email of 22 December 2010 made no reference to a hail limit, it is unsurprising that Mr Hofmann did not include any such limit in the LPI. How it was that he decided to add “hail” into the “Windstorm/Hail” limit of EUR 40 million is not explained on the evidence.

On 8 February 2011 Ms Hurtajova certified the final iteration of the LPI on that basis: without a hail limit. By then, that “Windstorm/Hail” limit had been increased to EUR 48.5 million. The LPI contained no other reference to “hail” and yet, Ms Hurtajova certified that iteration of the LPI as being correct.

When taken to the “Windstorm/Hail” references in the LPI, Ms Hurtajova rejected the proposition that the limit applied to both windstorm and hailstorm, despite the fact that is clearly what the document says.

However, as I have said, the implications of that answer were not explored in re‑examination.

Ms Stanley prepared the wording of the Local Policy on the same basis: without a hail limit. Evidently, Ms Stanley prepared the limits in the Local Policy by reference to the 8 February 2011 LPI and, consistently with that document, did not include a hail limit.

There is no evidence to explain why Ms Stanley decided to change the “Windstorm/Hail” limit to a “Storm” limit.

Thus, many matters are unexplained.

What is clear, however, is that the reason there was no hail limit in the wording of the Local Policy as published by Ms Stanley in March 2011 was the absence of any reference to a hail limit in Ms Hurtajova’s email of 22 December 2010, and the fact that on 8 February 2011, Ms Hurtajova certified as correct the final iteration of the LPI which (consistently with her 22 December 2010 email) contained no hail limit.

Ms Hurtajova’s evidence was that:

It might be thought to be implicit in that evidence, which was not challenged, that Ms Hurtajova made a mistake in her email of 22 December 2010 (by not specifying the Hail Limit) and in her 8 February 2011 certification of the LPI (by certifying, as correct, the LPI absent the Hail Limit).

But Ms Hurtajova did not say so.

And, in the absence of that evidence, I am not persuaded that I should find she was mistaken.

Although Ms Hurtajova was not challenged about her evidence as set out at [373], and although she determined that the Master Policy have the Hail Limit, I think the most reliable guide as to what she intended so far as concerns the Local Policy to be her contemporaneous, internal business communications (her 22 December 2010 email and her 8 February 2011 certification).

Ms Hurtajova created those documents at a time when there was no reason to think there would be any controversy about this matter.

Those communications suggest there was no mistake. At the very least, they are inconsistent with and cannot, in my opinion, be reconciled with a conclusion that there was a mistake.

I see this as being fatal to XL’s case on rectification.

In oral closing submissions, Mobis put the matter this way:

The reference in that submission to the acceptance by XL’s witnesses that local policies might have different limits from those in the Master Policy and local policies comprised, first, this evidence from the cross-examination of Ms Hurtajova:

Similarly, Mr Rossa gave this evidence:

Again, these questions were carefully framed.

But the answers given suggest that Ms Hurtajova and Mr Rossa contemplated that, in some circumstances, local policies might have different limits to the Master Policy.

That, too, is not consistent with the conclusion that there has been a mistake of the kind for which XL contends.

In any event, I am not persuaded that Mobis intended the Local Policy to have the Hail Limit.

The email exchange between Mr Papa and Mr Kalocai in December 2010 concluded on the basis that:

As I have said, although Mr Papa asked for a copy of the wording in the Local Policy, he must have seen that the document with which he was provided was the Master Policy.

Mr Papa made numerous enquiries throughout January and February seeking a copy of the wording of the Local Policy, including “a copy of the schedule and wording” (see [323] above).

Mr Papa, finally, saw the policy wording on 22 March 2010. He, as I have found, reviewed it, noted a correction that needed to be made to Mobis’s address but did not express, and I find did not detect, any inconsistency between the wording of the Local Policy and his discussions with Mr Kalocai in December 2010.

In those circumstances, I am not persuaded that XL has shown, to the requisite standard, that Mobis intended that the Local Policy have the Hail Limit.

The emails set out (at [352] to [357] above) show Mr Kalocai had assumed that the Local Policy would have a hail limit corresponding to that in the Master Policy. But the question here is what Mobis, subjectively and actually, intended (with such intention to be ascertained objectively). I do not see what relevance Mr Kalocai’s assumption about the Local Policy, not based on any communication or information obtained from XL concerning the process whereby the Local Policy wording was created, and not communicated to anyone at Mobis, could have to that question.

When a contract is negotiated by a duly authorised agent, that agent’s intention may be relevant for the purposes of rectification: Metlife Insurance Ltd v Visy Board Pty Ltd [2007] NSWSC 1481 at [32] (Brereton J citing the High Court in Australian Gypsum Ltd v Hume Steel Ltd (1930) 45 CLR 54; HCA 38 at 67).

Mr Kalocai negotiated the Master Policy (with a hail limit).

But he did not negotiate the terms of the Local Policy.

Mr Kalocai appears to have assumed that it, too, would have the Hail Limit.

But all Mr Kalocai knew was that:

There is no evidence that Mr Kalocai knew what process XL followed to cause local policy wording to be prepared, nor what communications had occurred within XL that led to the Local Policy having wording which did not include the Hail Limit, let alone that he had any intention about those matters.

The conclusion to which I have come is that XL has not established its case for rectification to the relevant standard.

In those circumstances, it is not necessary for me to deal with UNIQA’s submission that, in light of the effect that rectification of the Local Policy would or might have on its position as an insurer under the Master Policy, rectification should in any event, and as a matter of discretion, be denied to XL.

In oral submissions in reply, Mobis submitted (for the first time) that, assuming it was the common intention of the parties as at 1 January 2011 that the Local Policy have the Hail Limit, any agreement to that effect was varied by the issue of the Local Policy wording on 1 March 2011 (without the Hail Limit).

In view of the conclusion to which I have come, I do not need to deal with this submission.

In further answer to Mobis’s claim, XL relies on the Faulty Design Exclusion in the Local Policy.

That exclusion is in cl 3.2.1(a) which provided that the Local Policy does not cover:

XL contends that the design of the warehouse was “faulty or defective”.

As I have said, the warehouse was vast. It was equivalent in size to several city blocks. It was some 246 m long and 123 m wide.

The structure of the steel frame of the warehouse was depicted in a diagram in one of Mr Carolan’s reports:

The north side of the warehouse is shown by the line indicated by “A”. “D” represents the southern edge of the warehouse. The red boxes depict office buildings on the north side of the warehouse. There were two internal rows of columns, which are shown as “B” and “C”. The connections in the north to south direction (“A” to “D”) were referred to by the parties as beams or rafters. Each was a little over 40 m in length.

A section from north to south was referred to by the parties as a frame. The primary components of the frame were the two end columns, the two middle columns and the three connecting beams.

There were 26 internal frames in addition to the west wall and the east wall. Each frame included 3 beams. Each of the beams in a frame is a “structural element” for the purpose of the relevant Australian Standards.

Purlins, running east to west were connected between each of the beams. There were some 20 purlins attached to each beam, and thus some 60 purlins attached to each frame as a whole.

A typical frame is shown in the following diagram:

There were three spans making up each frame. There were 78 spans (excluding the east and west wall).

As can be seen from the frame section, the ridge line was off-centre, being closer to the row of internal columns depicted as “C”.

The roof had an average slope in the order of 2°. A roof of that slope would, to the naked eye, appear almost flat. However, water could easily drain off it. Indeed, as I have mentioned, the warehouse was, prior to 25 April 2015, subjected to rain storms far more severe than that which occurred on 25 April 2015. On those occasions the slope of the roof was adequate to cause rain to run off and be dealt with in the warehouse’s drainage system.

Clause 3.2.1(a) directs attention to “its own” faulty or defective design. A question arises as to the significance of these words.

The answer to that question is revealed once the exclusion is read incorporating defined terms and excluding unnecessary words, thus:

This makes clear, in my opinion, that the word “its” in the expression “its own faulty or defective design” refers to the property the subject of the loss, and not to Mobis.

Thus, the “faulty or defective design” to which the exclusion is directed is of the “property covered by the Policy” which is the subject of the “physical loss or destruction…or damage” (that is, of the warehouse itself) and that what is excluded is a claim for the loss of that property if that loss be caused by the faulty or defective design of that property.

Accordingly, if it be the case that the collapse of the warehouse was caused by its faulty or defective design, Mobis’s claim for damage to the warehouse itself would be excluded. But Mobis’s claim for loss or damage to contents and stock would not be excluded as that loss would not have been caused by the faulty or defective design of the contents and stock.

As to business interruption, what is excluded is “business interruption…caused by or consisting of its own faulty or defective design”. The wording is awkward as it is hard to see how “business interruption” could have a design, let alone one which was faulty or defective. The better reading of the exclusion is, probably, simply to ignore the words “business interruption” as it appears “clear that something has gone wrong with the language”: see Chartbrook Ltd v Persimmons Homes Ltd [2009] 1 AC 1101 at [25] (Lord Hoffmann).

In any event, the proviso to the exclusion makes clear that, assuming the design of the warehouse was faulty or defective, only damage to the warehouse itself is excluded.

That proviso states that the exclusion does not apply to “subsequent” loss.

The effect of the proviso was that loss or business interruption which results from a cause “not otherwise excluded” is not excluded by cl 3.2.1(a).

In Prime Infrastructure (DBCT) Management Pty Ltd v Vero Insurance Ltd [2005] QCA 369, the Queensland Court of Appeal (at [33] and [38]) considered a proviso similar to that in cl 3.2.1(a) to mean damage “occurring or coming later or after; following in order or succession” or “causing or occurring later or after or following in order”. It is not necessary that the “subsequent damage” be separate and distinct from the initial damage.

In the same case, the words “a peril (not otherwise excluded)” was held to mean a peril that is not excluded by any other provision in the policy. That is, the “proviso peril” could be a peril excluded by the clause containing the proviso, but not any other exclusion in the policy (Prime at [22]).

As all damage (other than to the building itself) is, in this sense, “subsequent” to the actual collapse of the warehouse and is not “otherwise excluded”, it is not caught by cl 3.2.1(a).

In the result, all that is capable of exclusion by reason of cl 3.2.1(a) is Mobis’s claim for damage to the building itself, not its claim for loss of stock, contents or for business interruption.

This argument was developed for the first time in UNIQA’s opening submissions. In final submissions, it was embraced by Mobis, and barely disputed by XL.

In final submissions, Mobis accepted that the phrase “faulty or defective design” does not require XL to establish negligence or blameworthiness in order to engage the exclusion (for example, see Manufacturers' Mutual Insurance Ltd v Queensland Government Railways (1968) 118 CLR 314 at 323).

In its opening submissions, XL submitted that the mere fact that the warehouse collapsed showed that it was defective in design. XL accepted, at that point, that it was making a submission “close to” a res ipsa loquitur submission.

Although that proposition was barely (if at all) developed in final submissions, I should make clear that I do not accept it.

The Local Policy expressly provides for destruction of an insured building “caused by its own collapse” if the damage results from a “named peril” (which includes “storm”: see cll 2.7 and 3.2.1(i)).

I accept Mobis’s submission that when cll 3.2.1(a) and 3.2.1(i) are considered together, it is clear that the fact that the warehouse collapsed during a storm will not, itself, engage the Faulty Design Exclusion. As Mobis pointed out, a building might collapse even though its design complied with all applicable design standards. In those circumstances, it is not open to XL to resort to the res ipsa loquitur principle (or anything close to it). XL must establish that the warehouse was, objectively, defective in design.

In Manufacturers' Mutual, the High Court was concerned with the proper construction of an exclusion in an insurance policy for loss “arising from faulty design”.

The plurality (Barwick CJ, McTiernan, Kitto and Menzies JJ) said (at 321):

Windeyer J said (at 322):

In Chalmers Leask Underwriting Agencies v Mayne Nickless Ltd (1982) 2 ANZ Ins Cas 60-463 Hutley JA said (at 77, 581):

Chalmers Leask was upheld on appeal (Chalmers Leask Underwriting Agencies v Mayne Nickless Ltd (1983) 155 CLR 279; 2 ANZ Ins Cas 60-514) and cited with approval by the Court of Appeal in State of New South Wales v AXA Insurance Australia Ltd (2002) 54 NSWLR 409; NSWCA 63 by Ipp AJA (with whom Mason P and Giles JA agreed).

His Honour said (at [39]):

I accept Mobis’s submission that while it is implicit in the High Court’s judgment in Manufacturers' Mutual that the “faulty design” in question fell short of an objective standard, the plurality did not need to articulate precisely what standard was being applied because it concluded that the factual findings made by the arbitrator (whose decision was in question) were sufficient to establish a departure. As I have set out, Windeyer J referred expressly to the design not being “up to a required standard”.

In this case, the question arises in the context of the ability of the warehouse to resist loads.

That question is the subject of particular provisions in the Building Code of Australia and in the relevant Australian Standards.

As the warehouse was constructed in 2008, and in the events that have happened, the relevant Australian Standards are:

As I discuss below, much of the dispute between the parties can be resolved by reference to the proper construction of the Note to Table 3.2 in AS/NZS 1170.1: 2002.

The parties adopted very different positions on this question.

Mobis contended that if the design of the warehouse complied with the Standards, it must follow that its design was not faulty.

On the other hand, XL submitted that if the design was not the product of “prudent design and engineering judgment” it was thereby defective, whether or not it complied with the Standards.

What was common ground was that if the design did not comply with the Standards, it was faulty or defective.

Accordingly, I will turn first to that question.

Before turning to the Standards, I will refer to the Building Code of Australia (“BCA”).

At the relevant time, reg 98(1)(a) of the Environmental Planning and Assessment Regulation 2000 (NSW) stated that, as a condition of development consent for any development involving building work “the work must be carried out in accordance with the requirements of the [BCA]”.

The BCA provided that, so far as concerns the structure of the building, the structure:

The BCA contained “Deemed-to-Satisfy Provisions” the net effect of which is that if a building like the warehouse complied with, relevantly, AS/NZS 1170.1 and AS/NZS 1170.2, it was deemed to comply with the BCA (Pt B1.0, B1.1 and B1.2).

Section 2 of this Standard dealt with “Structural Design Procedure”.

Clause 2.1 provided that structural design for “ultimate limit states” must be carried out using the procedure in cl 2.2.

“Limit states” were defined in the Standard as “states beyond which the structure no longer satisfies the design criteria”. “Ultimate limit states” were defined as “states associated with collapse, or with other similar forms of structural failure”. Relevantly, this meant the strength of the structure.

Clause 2.2 of this Standard set out the procedure to be followed to “design for ultimate limit states”.

Relevantly, those procedures included:

The Standard required that when calculating strength, the “design capacity” of the building as built must be equal to or exceed its “design action effect” (the load the Standard states that it must bear).

There was little difference between Mr Carolan and Mr Summers as to the first of these elements: the “design capacity” of the warehouse. What divided Mr Carolan and Mr Summers on this question was the second of those elements: the “design action effect”: the load that the Standard requires the as built structure to bear.

The parties referred to the relationship between these two concepts as the “Utilisation Ratio”.

It was common ground that, in order to comply with the Standard, the Utilisation Ratio must be less than or equal to 1.

The manner in which imposed loads on roof structures such as that at the warehouse were at the relevant time to be calculated was set out in Pt 3.5 of AS/NZS 1170.1; in particular, by reference to Table 3.2 in that part.

Table 3.2 was in the following form:

The relevant part of Table 3.2 is that dealing with “other roofs (i) Structural elements”.

The “structural elements” in question are the 78 beams forming the 26 internal frames of the warehouse as I described above.

Each of those beams supported more than 200 m2 of roof area.

Accordingly, Note 1 to Table 3.2 was engaged and required that those structural elements:

Table 3.2 thus sets out the minimum “imposed action” that the “structural elements”, (that is, the beams) must bear.

It was common ground that the “supported area that gives the worst effect” was the weakest part of the beam, namely, the mid-point. Mr Carolan and Mr Summers both adopted a rectangular area of 9.1 m by 22 m as the 200 m2 “worst effect” area.

The commentary to AS/NZS 1170.1 stated (at Pt C3.5.1):

The matter that divided the parties was the proper construction of Table 3.2 and, in particular, Note 1.

XL contended that, on its proper construction, Table 3.2 required that all structural elements in the building of the kind referred to in the Table, relevantly the beams, be able simultaneously to sustain application of the 0.25 kPa pressure at their points of worst effect. That is, looking at the warehouse, that each beam be able simultaneously to withstand 0.25 kPa pressure, as depicted in the following diagram (in which the pink highlighting represents an example of the points at which pressure was to be applied):

If the construction of Note 1 for which XL contended is correct, the warehouse had a Utilisation Ratio of more than 1 (Mr Summers calculated it to be 1.03) and did not comply with the Standard.

Mobis contended that XL should not be permitted to advance this argument because, it was said, XL had not made clear (through the evidence of Mr Summers or otherwise) until final submissions that it wished to do so. In view of the conclusion to which I have come on the question of the proper construction of Note 1, it is not necessary for me to deal with this argument.

On the other hand, Mobis contended that, on its proper construction, Note 1 required only that each structural element (that is, relevantly, each beam) separately, be able to withstand that imposed action; as illustrated in the following diagram:

If Mobis’s construction of Note 1 is correct, and subject to the “load shedding” point that I discuss below at [512] to [532], the warehouse had a Utilisation Ratio of less than 1 and complied with the Standard.

Mobis submitted that, properly understood, Mr Carolan’s evidence was that, at the point where the collapse occurred, the Utilisation Ratio of the warehouse was well under 1: namely 0.86.

In his report of 21 October 2016, Mr Carolan said:

The figure that followed was in the following form:

As Mobis explained in its final submissions, this diagram shows that Mr Carolan performed six compliance checks on structural elements at different locations. The Utilisation Ratio of 0.99 was a result of a check performed on a beam on the second frame from the end of the warehouse: described as “Check 1” in the diagram. Mr Carolan calculated a lower Utilisation Ratio (0.86) from a check he did of a beam in the middle of the warehouse: at “Check 3”.

Mr Carolan gave this evidence:

Mr Carolan was not asked to identify which area of the structure his Utilisation Ratio of 0.99 related to. However, it is apparent from Mr Carolan’s diagram that, as I have said, he calculated that Utilisation Ratio in relation to the second frame from the end (Mr Carolan’s “Check 1”).

It is common ground that the three end frames on both the western and eastern ends of the warehouse are not “typical” frames. They are designed differently from the rest of the warehouse. They are more heavily braced to withstand lateral wind loads and include thicker and heavier rafter beams.

Mr Carolan and Mr Summers agreed that the end frames were not the cause of the collapse. They agreed, and this was in any event apparent from the CCTV footage shown during evidence, that the collapse initiated near the loading dock in the northern span of one or more of the middle frames of the warehouse. The collapse then spread out towards the eastern and western edges of the structure.

Mr Carolan illustrated the location of the actual collapse in a diagram he included in his report of 19 February 2016.

In final submissions, Mobis produced a marked up version of that diagram showing (in yellow) the frame tested in Mr Carolan’s “Check 1” that produced the Utilisation Ratio of 0.99.

That diagram follows:

The Utilisation Ratio of 0.99 for “Check 1” was the highest Utilisation Ratio that Mr Carolan calculated for the structure. However, as Mobis submitted, the relevant Utilisation Ratio for present purposes is the one Mr Carolan produced at “Check 3” – that is, 0.86, which is the highest Utilisation Ratio for the relevant part of the structure.

I am persuaded that Mobis’s construction of Note 1 is correct. This is for a number of reasons.

First, as a matter of language, Note 1 to Table 3.2 directs attention to the pressure that each of the “structural elements” in a roof must withstand and identifies the pressure that that element must bear. The Table does not, in terms, call for multiple spans of the roof to be simultaneously loaded.

“Structural element” is defined in cl 1.4.20 of AS/NZS 1170.0 as a “physically distinguishable part of a structure, for example, wall, column, beam, connection”.

As Mobis submitted, the term does not refer to types of elements collectively, but to each physically distinguishable part individually.

In this case, the “structural elements” are the rafters or beams depicted in “Figure 4” at [408] above and described at [468] above.

Note 2 to Table 3.2 (which explains the meaning of “A” in the formula referred to in the body of Table 3.2 (1.8/A + 0.12)) refers to “the” member under analysis, suggesting (as Mobis submitted) that Table 3.2 contemplates that the 0.25 kPa is to be applied to one member (or structural element) at a time.

Note 1 calls for each structural element to be designed to support “the” 200 m2 of “the” supported area that gives the worst effect; suggesting again, that one structural element at a time is to be considered.

Those factors point to the correctness of Mobis’s construction.

There was controversy between the parties as to what light an earlier provision of the Standard cast on the proper construction of Note 1.

That earlier provision dealt with “Partial Load” and was in the following terms:

It is common ground that although the balance of cl 3.3 concerns imposed actions on floors (that is, not roofs), the first sentence is of general application, including to roofs.

The first sentence states that imposed actions shall be considered “to be absent” (that is, they are not to be taken into account) from any parts of the structure if such “absence” would cause “more adverse effects” on “that or any other part”.

The wording is awkward, but I read it to mean that if there would be a “more adverse” effect on any part of a structure were the imposed actions not to be applied to a particular part of the structure, then consideration of the building’s ability to withstand the imposed action (that is, its strength) must be considered in that context.

This is how Mr Carolan, who is deeply experienced in this area, understood the provision. His evidence was that this sentence was concerned with the “absence” of imposed actions and simply required the designer to “apply either the live load or not [apply] the live load, whichever produced the worst effect”.

There is no suggestion in the evidence that, for the beams in the Mobis warehouse, the absence of a live load would cause a worse effect than the presence of a live load.

The position may be different in a differently designed structure. For example, Mobis suggested that on a cantilevered rafter, the absence of the live load could have a worse effect on the rafter than the presence of a live load, because the live load could act as a counterweight to the dead load of the cantilevered section. However that may be, I do not see anything in the first sentence of cl 3.3 which casts light on the proper construction of Note 1 to Table 3.2.

I accept Mobis’s submission that nothing in cl 3.3 purports to:

Further, as Mobis pointed out, in the case of floors, the Standard states in terms that loading of multiple spans is required: see the third paragraph in cl 3.3 (commencing with “[f]or partial loading on continuous beams”).

As Mobis submitted, the specific stipulation that spans supporting floors be subjected to multiple loading, and the absence of such a specific stipulation in the case of roofs, suggests that the drafters of the Standard did not intend such a requirement in the case of roofs.

For all those reasons, my conclusion is that the construction of Note 1 for which Mobis contended is to be preferred. What is required, in my opinion, is that each beam, separately, be able to support 0.25 kPa on the 200 m2 of supported area that gives the worst effect.

XL submitted that even if the Standard was to be construed this way, the warehouse’s compliance with the Standard depended on:

The point arises this way.

In applying the load specified in the Note to Table 3.2, Mr Carolan attributed approximately 75% of the load to the beam in question and assumed that the balance of the load was shed, equally, to the adjacent beams through the purlins (which, as I have said, connected the beams from east to west throughout the warehouse: see [412] above).

Thus, Mr Carolan applied the load as depicted below:

XL accepted that, if that was the correct way to proceed, the beams complied with the Standard.

On the other hand, Mr Summers applied the whole of the load specified in the Note to Table 3.2 (as a “line load”) to the beam itself, as illustrated in this diagram:

If that is the correct way to proceed, the beams do not comply with the Standard.

However, I do not accept that the Standard requires the course taken by Mr Summers.

Note 1 to Table 3.2 requires that the beam support the 0.25 kPa pressure on the 200 m2 of the roof area that which gives the worst effect. That is, the Note requires that the beam support such a load over the whole of that area and thus calls for the force to be uniformly distributed over that area.

In my opinion, as there are purlins between each of the beams, which purlins span some 20 spans, the course adopted by Mr Carolan responds to the requirements of Note 1 to Table 3.2, as it has the effect of simulating uniform distribution of the load as required by the Note.

In his report of 21 October 2016, Mr Carolan included Fig A9.2 as follows:

This figure was included in an appendix to Mr Carolan’s report headed “Structural Calculations to Check Compliance with Australian Standards”.

In cross-examination, Mr Carolan said that “we’ve done our own calculation to verify the 75 per cent on the centre rafter”.

Mr Carolan described that what he had done was an “elastic analysis” using “Strand7” (a computer program). Mr Carolan said:

Although this may not be obvious to a lay observer, Figure A9.2 shows that in his modelling, Mr Carolan applied 75.24% of the 0.25 kPa load called for by Note 1 to the beam in question, and 12.32% to each adjacent rafter.

In his report of 18 June 2017, Mr Carolan said:

In cross-examination it was put to Mr Carolan that:

Mr Carolan’s evidence in response was that:

Mr Summers’s evidence was that he did not model the purlins.

The only evidence before me as to what proportion of the load on the beam would be shared to adjacent purlins is that of Mr Carolan. Such challenge as was made to Mr Carolan’s calculations in the cross-examination to which I have referred, does not persuade me that Mr Carolan’s calculations are not correct.

Accordingly, my conclusion is that, in checking for compliance with AS/NZS 1170.1, it is appropriate to take account of the shedding of load through the purlins to the adjacent beams, and I accept Mr Carolan’s opinion that some 25% of the load should be so shed.

As I have discussed, cl 2.2 of AS/NZS 1170.0 provided that “ultimate limit states” for imposed loads (such as hail) are to be determined in accordance with AS/NZS 1170.1 (relevantly, in accordance with Table 3.2: see the discussion above) and for wind loads, in accordance with AS/NZS 1170.2 (see below from [552] to [592]).

Clause 2.2(i) also provided that:

“Structural robustness” was defined at cl 1.4.21 of the AS/NZS 1170.0 as the:

Part 6.1 and 6.2 of the Standard provided, under the heading “Structural Robustness”:

Thus, the notion of “robustness” in the Standard was concerned with the “detailing” of the structure (including the provision of “load paths”) so as to ensure that the damage sustained to the structure following the event that causes that damage (in this case, hail) is not “disproportionate” to that event.

Clause 2.2(i) required that both the “design and detail” of the structure achieve this object.

The question of “robustness” was dealt with further in the Supplement (or Commentary) to AS/NZS 1170.0 (AS/NZS 1170.0 Supp 1:2002).

It stated (at C1.1):

Section C6.1 of the Commentary dealt further with “Structural Robustness” and states:

I accept Mobis’s submission that the Supplement or Commentary to AS/NZS 1170.0:2002 cannot enlarge the requirements of that Standard. But the Supplement emphasised the need for a structure to be designed so that it did not respond disproportionately to an event causing damage and gave examples of how that might be achieved: for example, a design that caused the structure to collapse progressively (C1.1); and not collapse without warning (C6.1(d)); and not be sensitive to the “hazards considered” (C6.1(b)).

Mr Carolan agreed that the warehouse was sensitive to hail damage. Thus, he gave this evidence (in response to questions from me):

Mr Carolan made no separate analysis in his reports of the “robustness” of the building. His analysis was directed to whether the building could withstand imposed loads (see above) and wind loads (see below).

The warehouse did collapse very suddenly. Rain and hail fell during the periods to which I have referred. Within a number of minutes after the hail ceased, the load on the warehouse reached the collapse load. Thereafter, the warehouse collapsed in approximately 15 seconds. Mr Carolan described what happened as an “un-zippering” effect. One beam failed and then, virtually immediately, many others followed.

This was revealed in the vision from CCTV within the warehouse. Mr Carolan said that, had he been in the warehouse at the time, he (with his expert eye) would have been able to detect some deflection in the beams before the moment of collapse.

However, after the relevant vision was displayed in Court, Mr Carolan gave this evidence in response to questions from me:

Thus, the warehouse did not collapse progressively; and it did collapse without any, or any great, warning.

However, in final submissions, my attention was not directed to any evidence which showed that the design or “detailing” of the warehouse was what led to the “un-zippering” phenomenon; nor to any respect in which there was a shortcoming in the manner in which the structure of the warehouse was “tied together both in horizontal and vertical planes” (see [536] above); nor to any shortcoming in the provision of “load paths to the foundations” of the warehouse (see [536] above), such as would suggest that the design of the warehouse was not “robust”.

Nor was my attention directed to any evidence showing what damage to the warehouse would have been “proportionate” to the hail load; or to contrast the damage which did occur with the damage that would have occurred had the detailing of its design and the provision of load paths been different.

In those circumstances, I am not persuaded that the design of the warehouse failed to achieve the requisite “robustness” or that the collapse of the warehouse was caused by any such short coming in the design.

Clause 2.2(c) of AS/NZS 1170.0:2002 requires that the ultimate limit states for a structural design to comply with wind loads are as specified in AS/NZS 1170.2 (which I call “the Wind Standard”).

The Design Exclusion, construed as I have set out above, is enlivened only if the damage to the warehouse was caused by its faulty design.

There is no suggestion that wind caused the warehouse collapse. There was very little wind on the day of collapse.

There is thus no suggestion that any failure of the warehouse to comply with the Wind Standard itself led to the collapse.

However, XL contended that had the warehouse been designed to comply with three particular requirements of the Wind Standard, it would have been sufficiently strong to withstand the weather events of 25 April 2015.

Mr Summers identified those three requirements as relating to:

However, it does not necessarily follow that, had the warehouse been designed to satisfy what Mr Summers said were the wind standard requirements in these three respects, it would not have collapsed.

As Mobis submitted:

In any event, for the reasons that follow, I am not satisfied that XL has proved that the warehouse failed to comply with the Wind Standard.

The warehouse had ridge vents along the ridge line.

Ridge vents have throats to allow the passage of air and thus contribute to the equalisation of air pressure within and without the warehouse.

What divided Mr Carolan and Mr Summers was the assumption that should be made as to the width of the ridge vent throat in the warehouse as constructed.

Mr Carolan assumed such width to be 1000 mm and that the ridge vent had an efficiency of 35%.

On the other hand, Mr Summers, by reference to a photograph of the ridge vent taken from the floor of the warehouse, calculated the width of the vent to be 900 mm.

I do not consider I need to come to any conclusion as to which of these was correct.

That is because Mobis’s General Manager, Warehouse and Logistics, Mr Malcolm Stoddart (to whom I refer later when discussing quantum) gave evidence that, in addition to the roof vents (and no matter how wide their throats were), there were venting devices in the southern wall of the warehouse.

Neither Mr Carolan nor Mr Summers took these vents into account.

In cross-examination, Mr Summers accepted that the Wind Standard permitted the use of other design alternatives, such as a venting device, to make a structure comply. He said that “adding vents changes everything” and that adding vents “would change the whole functionality of the building”.

It was for XL to show that the warehouse design did not comply with the Wind Standard. As Mr Summers did not address what effect the vents in the southern wall had on the capacity of the warehouse to bear wind loads, I am not persuaded that the ridge vent width bespoke a want of compliance with the Standard.

The terrain category in the Wind Standard is relevant to assumptions made by the designer as to the force of wind on the structure.

There was no dispute between Mr Carolan and Mr Summers as to the kind of terrain categories that the warehouse designer should take into account.

The matter that divided Mr Carolan and Mr Summers was as to what terrain the warehouse designer should assume for the purpose of considering winds coming from the south.

In cross-examination, Mr Summers said that “the terrain category is the most subjective issue that one faces as a designer and when designing for wind loads”.

Ultimately he gave this evidence in response to questions from me:

In light of that evidence, I accept the following submission from Mobis:

In those circumstances, I am not satisfied that XL has established that the warehouse design did not comply with this aspect of the Wind Standard.

The issue here is as to the assumption a designer of a structure should make as to which doors would be open when a wind event occurred.

Mr Carolan contended that a reasonable assumption for a designer to make was that 50% of the doors would be open when a wind event occurred.

In his first report, Mr Summers performed his calculations assuming that no doors were left open.

In his final report, Mr Summers assumed that all doors were left open.

Argument before me proceeded upon the basis that, all other things being equal, adoption of Mr Carolan’s assumption would have caused the design of the warehouse to have a Utilisation Ratio which was (just) under 1.

It was for XL to show that the assumption that Mr Carolan made was not reasonable.

The Commentary to the Wind Standard is set out at C5.3.1 and provides:

In one of his reports, Mr Carolan stated:

Later in his report, he continued:

In a later report Mr Carolan said:

I accept Mobis’s submissions that the assumption Mr Carolan made was, in all the circumstances, conservative. It is true that the warehouse had a design life of many decades and that Mobis might in the future be replaced by another occupier. Nonetheless, the warehouse was purpose designed, and unlikely to be used for anything else other than as a warehouse. That points to the reasonableness of Mr Carolan’s assumption.

On the other hand, Mr Summers’s assumption was that all doors would be open, even during a windstorm. He justified this assumption by saying that:

That appears to me to be a very unlikely circumstance and one that would not be reasonable for a designer to contemplate.

Overall, I am not persuaded that the assumption that Mr Carolan said a reasonable designer should adopt was not, itself, reasonable.

For those reasons, my conclusions are that XL has failed to show that: any failure of the design of the warehouse to comply with the Wind Standard caused the warehouse collapse and thus enlivened the Design Exclusion; or that the design of the warehouse did not comply with the Wind Standard.

For those reasons, I am not satisfied that the design of the warehouse failed to comply with the relevant Standards or that the design of the warehouse was for that reason “faulty”.

The question which then arises is whether XL has shown that the design of the warehouse was, nonetheless, faulty.

The authorities to which I have referred have held that a design is faulty if it:

Further, the appropriateness of the design must be measured against the purpose for which it was intended to be used (AXA at [441] above).

Manufacturers’ Mutual was concerned with the design of a railway bridge, Chalmers Leask with the design of a cofferdam and AXA with a child’s toy.

None of those cases addressed a circumstances where, as here, there were detailed standards specifying what the design must achieve.

The parties to the Local Policy agreed that Mobis would have no cover for the warehouse if it was damaged by reason of “its own faulty or defective design”.

In my opinion, reasonable business people in the position of the parties would have understood the words “faulty or defective design” to refer to a design which was (to adopt the language of Windeyer J in Manufactures’ Mutual) “not up to a required standard”.

In this case, that standard was that required by the BCA and thus AS/NZS 1170.0, 1170.1 and 1170.2.

XL has failed to show that the design of the warehouse did not comply with those Standards.

Mr Summers expressed the opinion that the design of the warehouse was not “prudent” for various reasons.

Whether or not the design was or was not “prudent”, it has not been shown to be “not up to [the] required standard”.

For those reasons, my conclusion is that XL has failed to establish that the Faulty Design Exclusion has been enlivened.

For these reasons, and subject to questions of quantum, Mobis is entitled to indemnity under the Local Policy.

Accordingly, the question of whether Mobis would be entitled to indemnity under the Master Policy does not arise and a number of issues, which occupied time during the hearing, are now moot.

However, in deference to the detailed submissions of the parties, I shall deal with a number of them.

The first of those issues is that arising under the Faulty Construction Exclusion.

Article 6.3(j) of the Master Policy provided that the policy did not cover loss or damage:

The article raises issues similar to, but not the same as those which arise under the Faulty Design Exclusion.

Article 6.3(j) of the Master Policy directs attention to the “construction” of the building and not its design.

I accept Mobis’s submission that the expression “faulty building construction” is apt to encompass a process by which a building is constructed – essentially workmanship – but not apt to cover the design itself.

There is no suggestion that there was any deficiency in the manner in which the warehouse was constructed.

If the parties to the Master Policy had intended that Art 6.3(j) apply in cases of faulty design, it would have been very easy for them to say so (as did the parties to the Local Policy).

I see no reason to give the word “construction” in this article anything other than its ordinary meaning.

It follows that assuming Mobis had a claim under the Master Policy, it would not have been excluded by this article.

Had I been persuaded to rectify the Local Policy to include the Hail Limit, it would have followed from that conclusion and from my conclusion that the proximate cause of the collapse was hail that Mobis had no further entitlement to indemnity under the Local Policy (as XL has paid Mobis an amount equal to what the Hail Limit would have been).

The question would then have arisen as to whether, by reason of the DIL/DIC clause in the Master Policy, Mobis could have claimed indemnity under the Master Policy, notwithstanding the existence of the Hail Limit in that policy.

In that event, a question would have arisen as to the proper construction of the Overlap Clause: Art 3.1.8.

Mobis submitted that Art 3.1.8 had the effect that had it been necessary for it to make a claim under the Master Policy, it would have been entitled to indemnity up to the storm limit (EUR 50 million), notwithstanding the EUR 10 million Hail Limit.

For the following reasons, I do not accept that submission.

Article 3.1.8 of the Master Policy provided:

The first sentence of the article refers to Arts 2 and 10.4 of the Master Policy.

Article 2 set out the “Limits of Liability and Deductibles” (including the limits as set out at [46] above).

Article 10.4 was in the following terms:

Article 3.1.8 provided a mechanism for resolving any “overlap” between different limits of liability and different deductibles and provided that, in the event of such an overlap, the highest limit applied.

The issue on which the parties were divided was as to whether the mechanism was directed to overlaps within the Master Policy (as Mobis contended) or overlaps between the Master Policy and the Local Policy (as XL and UNIQA contended).

Article 5.2 of the Master Policy (the “DIL/DIC clause” (see [211] above)) contemplated that there may be wider cover under the Master Policy than under the Local Policy for a particular peril and that in that event, the Master Policy “shall provide additional cover over and above the indemnities under the local policies”.

The Master Policy thus contemplated that Mobis might make a claim under the Master Policy for a peril which was also insured under the Local Policy (to take advantage of the wider cover offered under the Master Policy for that peril).

In those circumstances, there might have been an overlap between the deductibles and limits of liability in each policy.

The first paragraph in Art 3.1.8 of the Master Policy stated that, what followed, was:

The third paragraph of Art 10.4 dealt with the circumstances in which there was an overlap between deductibles under the Local Policy and the Master Policy.

That paragraph had the effect that if Mobis made a claim under the Master Policy for a peril also covered by the Local Policy:

Thus, that paragraph of Art 10.4 set out the rules to be applied when there was a claim made under both the Local Policy and the Master Policy and there was an overlap in deductibles.

That suggests that the parties intended that Art 3.1.8 deal with overlaps in deductibles between the two policies. Why else would Art 3.1.8 state that its contents were “in accordance with” Art 10.4?

The first paragraph of Art 3.1.8 also stated that that paragraph is to be “in accordance with” Art 2 of the Master Policy.

Article 2 of the Master Policy set out limits of liability, but also the deductibles under the Master Policy.

For “property damage” those deductibles were as follows:

Thus, there was a deductible of EUR 25 million as a “general deductible for all locations” and a deductible of EUR 10 million for theft and vandalism.

As UNIQA pointed out in final submissions, if, as Mobis contends, the second paragraph of Art 3.1.8 was directed to overlapping deductibles within the Master Policy (and directed that the higher deductible apply) then, as the “general deductible for all locations” was EUR 25 million, and was thus higher than the EUR 10 million deductible for theft and vandalism, the latter would never apply. It is unlikely that the parties intended this result.

This suggests, strongly in my view, that the parties intended that the second paragraph of Art 3.1.8 deal with any overlap in deductibles between the Master Policy and the Local Policy (and not any overlap of deductibles within the Master Policy).

If the parties intended that this paragraph deal with overlaps in deductibles between the Master Policy and the Local Policy, they must also have intended that the clause govern an overlap between limits of liability between those two policies.

It is true, as Mobis pointed out, that the third paragraph of Art 3.1.8 (which deals with aggregate limits of liability) made particular reference to “local policies”, whereas the second paragraph of Art 3.1.8 did not.

However, in my opinion, the second paragraph of Art 3.1.8 must be read in light of the first paragraph which, as I have said, made particular reference to Art 10.4 and states that what follows is to be “in accordance” with that paragraph.

For those reasons, I am persuaded that XL’s construction of Art 3.1.8 is correct.

It follows that Mobis’s position under the Master Policy would be no better than under the Local Policy, assuming it were rectified to include the Hail Limit.

UNIQA contended that, in any event, it had no liability to Mobis under the Master Policy.

In light of my conclusions thus far, the question of what UNIQA’s liability to Mobis is under the Master Policy, all other things being equal, does not arise.

Nonetheless, in deference to the detailed submissions put by Mobis and UNIQA on the question, I shall deal with it.

UNIQA is a Lichtenstein insurer.

When the Master Policy was first written in July 2010, XL was the only insurer. XL continued as the only insurer for the years ended 23 June 2012, 2013 and 2014.

UNIQA and AIG became insurers under the Master Policy for the year ended 23 June 2015, with participations of 20% and 30% respectively.

At all relevant times the “Policyholder” named in the Master Policy was Mobis Slovakia.

Mobis Slovakia is a sister company to Mobis; they are both wholly owned subsidiaries of Mobis Korea.

Mobis did not execute the Master Policy although, as discussed below, it is included in a “List of Insured Companies”.

UNIQA contends that, on the proper construction of the Master Policy, Mobis is not an “Insured Company”.

I do not agree. In my opinion, Mobis is an “Insured Company” under the Master Policy.

There is no insuring clause in the Master Policy of the kind found in the Local Policy.

The Master Policy refers to both the “Policyholder” (Mobis Slovakia) and to “Insured Companies”, one of which is Mobis.

The critical provisions were Arts 4.1 and 5.1 which, relevantly, were in the following terms:

Article 5.1 dealt with the “Function of this Insurance Programme” and the “Cover Principle” of the policy.

The first paragraph of Art 5.1 recited that the Master Policy was issued as part an “international property and business interruption insurance programme”.

The second paragraph of Art 5.1 provided that cover under the Master Policy “is provided subject to the registration of the Insured Companies…with and acceptance by the Leading Insurer [that is, XL]”.

It is not clear what “registration” was contemplated by this provision. What it does make clear, however, is that cover under the Master Policy is extended to “the Insured Companies” once they are “accepted” (as Insured Companies) by XL.

Thus, once such registration has occurred (and there is no suggestion that it has not) then cover “is provided” to, amongst others, the Insured Companies.

Reference is then made to Art 4 (“see also Art 4. Insured Companies”) suggesting that that article is relevant to the question of whether the leading insurer (XL) has “accepted” companies as Insured Companies.

Article 4 deals with “Insured Companies”. There was no other definition of “Insured Companies” in the Master Policy or any other provision directed, in terms, to the question of what entities are “Insured Companies”.

A striking feature of Art 4 is that Art 4.2 contains a “List of Insured Companies”. That list includes Mobis.

The first three sentences of the first paragraph of Art 4.1 might, absent context, suggest that the parties intended that “Insured Companies” be subsidiaries of Mobis Slovakia (being companies in which Mobis Slovakia “directly or indirectly holds at least 50% of the voting share capital”) or companies in which Mobis Slovakia “holds participations” (companies in which Mobis Slovakia has “management responsibility”).

Thus, the first sentence of Art 4.1 stated that the Master Policy covered the Policyholder (Mobis Slovakia) and its subsidiaries and companies “in which it holds participations”. Although the wording is awkward, the third sentence of Art 4.1 appears to be intended to cast light on the meaning of the expression “in which it holds participations” in the first sentence and stated that this requirement was satisfied if Mobis Slovakia held 50% or more of the voting share capital or had “management responsibility”. The function of the intervening sentence is not clear but may constitute a further requirement for a company to be an “Insured Company”; namely, that it constitute “a financial entity with central management”.

Mobis does not qualify as an Insured Company if that expression is to be read as the first paragraph of Art 4.1, seen in isolation, might suggest. It is not a subsidiary of Mobis Slovakia. Mobis Slovakia holds no “voting share capital” in it. Nor, so far as the evidence reveals, does Mobis Slovakia have any “management responsibility” for it.

But Art 4.1 must be seen in the context in which is appears; particularly Art 4.2.

Article 4.2 is headed “List of Insured Companies”. Included in the list is Mobis, notwithstanding the fact that Mobis does not answer to the descriptions in Art 4.1.

The listing of the companies in Art 4.2 is consistent with the provision in Art 5.1 that cover is extended to Insured Companies once they are “accepted” as Insured Companies by XL.

I cannot see why the parties to the Master Policy included Mobis as one of the “Insured Companies” if they did not intend that Mobis be an Insured Company. Indeed, the inclusion of Mobis in Art 4.2 made clear that XL had “accepted” it as an Insured Company.

Indeed, XL has admitted this. In its amended commercial list response, XL admitted that Mobis “is an insured under the 2014 Master Policy subject to its terms and conditions”.

Ultimately, in final submissions, UNIQA submitted that Art 4.2 was merely a list of “potential candidates” as “Insured Companies”. To attribute to the parties the intention that Art 4.2 have this effect carries with it the implication that the parties contemplated that although Mobis was not currently an Insured Company (because it was not a subsidiary of Mobis Slovakia or a company in which Mobis Slovakia held “participations”), that circumstance might change, and that Mobis might one day become a subsidiary of Mobis Slovakia or a company in which Mobis Slovakia held “participations” and thereby be transformed into an Insured Company. There is no suggestion that the parties had any such contemplation. I do not accept that this is what a reasonable business person in the position of the parties would have thought the words to mean or that the contract would have this effect.

UNIQA also drew attention to the final sentence of the first paragraph of Art 4.1 which provided that, subject to any “special agreement”:

I do not see that this takes the matter any further.

The sentence was expressed in prospective terms and contemplated a circumstance where a company once complied with the provisions of the first paragraph of Art 4.1, but no longer did so. That provision did not apply to Mobis as it was never a subsidiary of Mobis Slovakia or a company for which Mobis Slovakia had management responsibility. The effect of the sentence is merely that a company that qualified as an Insured Company by reason of compliance with the requirements of Art 4.1 ceased to do so if it ceases to comply with those requirements.

Looking more widely in the Master Policy there are other indicators that the parties intended that Mobis be an Insured Company.

Article 1 of the Master Policy provided that the insurance under the Policy “shall cover”, amongst other countries, Australia (with particular limits being specified for property damage and for business interruption).

Further, in Art 2, one of the limits of liability was for “Earthquake Australia (named area Perth)” with a lower limit of liability, EUR 10 million, than the general earthquake limit of EUR 50 million.

Further, the second paragraph of Art 5.1 states that cover was subject to “acceptance by the Leading Insurer”; that is, XL.

Mobis also submitted that there was a “special agreement” between it and XL for the purposes of the last sentence of the first paragraph of Art 4.1 and that UNIQA was estopped from denying that Mobis is an Insured Company.

In view of the conclusions to which I have come, I do not need to deal with these matters.

For those reasons, my conclusion is that Mobis is an “Insured Company” under the Policy.

UNIQA submitted that, even if Mobis is an Insured Company under the Master Policy, its rights under the Master Policy would only be enlivened at the suit of Mobis Slovakia.

I accept that submission.

That submission directs attention to the question of whether Mobis was a party to the Master Policy.

The fact that Mobis is an “Insured Company” for the purpose of the Master Policy does not mean that it is a party to the policy.

As the Full Court of the Federal Court said in ABN AMRO Bank NV v Bathurst Regional Council (2014) 224 FCR 1; FCAFC 65 (Jacobson, Gilmour and Gordon JJ) at [1629]:

Mobis is not identified as a party to the Master Policy. The Master Policy was executed by Mobis Slovakia, as the (only) Policyholder on the one hand and the three insurers, XL, AIG and UNIQA on the other.

There is no term in the Master Policy that, in terms, gave Mobis an entitlement to make a claim under it.

Indeed, there were terms of the Master Policy which suggest that Mobis had no such right.

For example, the second paragraph of Art 5.2 (the DIL/DIC clause – see [211] above) provides:

Mobis submitted this constituted no more than a direction (by Mobis) to pay. I do not accept that submission. It does not take into account the fact that Mobis Slovakia, and not Mobis, was the Policyholder. I see the provision as a fundamental obligation imposed on the insurers to pay any amount payable under the DIL/DIC clause to Mobis Slovakia. There was no provision in the Master Policy obliging the insurers to make a payment to any of the Insured Companies (other than Mobis Slovakia). I see nothing in the Master Policy which contemplated the possibility that this might happen.

Article 5.3 provided for “Financial Interest Cover”. No claim is made for such cover in these proceedings. However, it casts light on the question to hand. The article provided that if an Insured Company suffered loss which cannot be recovered because of being “legally prohibited” for the purposes of the third paragraph of Art 5.1:

Further, the last paragraph of Art 3.1.8 (which I dealt with earlier in a different context: see [623] above) provided that:

Article 12.5 (the “Agent/Intermediary/Broker Clause”) provides:

Article 10.5 dealt with “Payment of Indemnity” and provides (in its last two paragraphs):

Although Art 10.5 provides that “in principle” a payment under the policy must be made in the currency at the domicile of the Insured Company, they do not provide that the payment must be made “to” the Insured Company. The second quoted paragraph makes clear that payments must be made to the Policyholder.

The provisions set out in the preceding paragraphs point to the conclusion that the parties to the Master Policy intended that only the Policyholder, Mobis Slovakia, could bring proceedings to recover the loss suffered by any Insured Company (such as Mobis).

Mobis sought to meet the implication of these provisions in the Master Policy by submitting that:

In support of that proposition, Mobis cited the following passage from the United Kingdom text book, Colinvaux’s Law of Insurance (11th UK ed, Sweet & Maxwell), at [15-003]:

However, the authors of Colinvaux went on to say (at [15-006]):

In any event, the first passage quoted from Colinvaux does not appear to reflect the position in Australia.

In Federation Insurance Ltd v Wasson (1987) 163 CLR 303; HCA 34 Mason CJ, Wilson and Dawson JJ said (at 303 and 314) that:

Thus, as UNIQA submitted, a composite policy is one contract with multiple insureds, not multiple contracts with individual insureds.

The plurality in Federation Insurance quoted with approval (at 310) the description of a composite policy by Sir Wilfrid Greene MR in General Accident Fire & Life Assurance Corporation Ltd v Midland Bank Ltd [1940] 2 KB 388 at 404-405:

This description (“one piece of paper”, “one policy”), also, is inconsistent with the proposition that a composite policy is in reality a series of individual contracts. A similar point was made by Gibbs ACJ in Deaves v CML Fire & General Insurance Co Ltd (1979) 143 CLR 24 at 41:

In any event, looking at the terms of the Master Policy, I cannot see how it can be construed as being other than one contract that confers benefits and liabilities on Mobis Slovakia.

In my opinion, on its proper construction, the Master Policy was one contract of insurance that provided Mobis Slovakia with direct cover for losses Mobis Slovakia itself suffered, and also constituted an agreement whereby the insurers agreed to pay to Mobis Slovakia the value of specifically identified property owned by the various Insured Companies (whether or not they be subsidiaries of Mobis Slovakia) when property was lost or damaged in the circumstances set out in the Master Policy.

Section 48 of the Insurance Contracts Act provides that:

However, the Insurance Contracts Act only applies if the proper law of the insurance contract in question is a State or Territory of Australia (s 8(1)).

Mobis submitted that the proper law of the Master Policy was Australian law because the subject of the proceedings was a warehouse located in Australia and because Art 10.5.1 of the Master Policy provided that any indemnity to which Mobis was entitled would be paid in Australian dollars and in Australia.

However the Master Policy:

The only connection with Australia is the fact that Mobis and some of the insured property is in Australia. As UNIQA submitted, that fact alone does not distinguish Australia from any of the other 10 jurisdictions in which an Insured Company under the Master Policy was domicile and held property.

In my opinion, the proper law of the Master Policy is likely to be that of Slovakia. It certainly is not Australia.

For those reasons, my conclusion is that although Mobis is an Insured Company for the purposes of the Master Policy, it has no standing to bring proceedings under that Policy.

Any proceedings must be brought by Mobis Slovakia.

Contrary to Mobis’s submissions, it does not follow from that conclusion that the Master Policy could never respond to losses suffered by the Insured Companies named in Art 4.2 of the Master Policy or that those entities would be “effectively uninsured”.

The conclusion means only that it is Mobis Slovakia, the Policyholder named in the Master Policy, that must bring proceedings to seek the benefit of the Master Policy for those Insured Companies.

UNIQA submitted that, in any event, on the proper construction of the Master Policy, it is not liable for any risks in Australia.

In that regard, UNIQA relied upon a Note in Art 3.1.3 of the Master Policy. To place that Note in context it is necessary to have regard to the whole of the article:

Article 3.1.1 defined the share of risk to be undertaken by each of XL, AIG and UNIQA under the Master Policy: 50% to XL, 30% to AIG and 20% to UNIQA.

The statement in the Note operated as a qualification to that risk sharing regime. So far as concerns “local policies” in Australia, the Note stated that UNIQA “is not taking any share” and XL’s share increases (correspondingly) from 50% to 70%.

Although the Note referred to a share of “local” policies, it is clear that the Note addressed UNIQA’s exposure under the Master Policy to claims which also arose under the relevant local policy: here, the Local Policy. Thus, it stated that XL’s liability in such circumstances was 70%: 20% more than what would otherwise be its share of liability.

Thus, the effect of the Note was that, in the circumstances to which the Note referred, XL would bear, in addition to its 50% share of liability, UNIQA’s 20% share.

Mobis submitted that the words in the Note are directed solely to UNIQA’s position under the Local Policies.

But that cannot be right. If that were so, there would be no reason to make any reference to an increase in XL’s “share”.

Further, were the Note to be construed as being directed solely to UNIQA’s position under the Local Policies, it would be superfluous and included in the Master Policy without any apparent purpose.

Mobis submitted that its construction of the Note was consistent with email correspondence between XL and UNIQA in June 2014 (shortly before UNIQA became an insured under the Master Policy on 23 June 2014) which, Mobis submitted, showed that UNIQA’s non-participation in the Australian and UAE local policies was based upon it being rejected as a reinsurer of those local policies.

Thus, as an example, Mobis drew attention to an email of 13 June 2014 from XL to UNIQA stating:

This material was not, in my opinion, admissible on the question of the proper construction of the Master Policy.

However, to the extent that it is relevant, it points in UNIQA’s favour.

When the full context of the email exchange is examined, it is clear that XL’s request that UNIQA confirm that it would be interested in “taking a share in the programme excluding AU and UAE” was directed to UNIQA’s potential participation as an insurer accepting 20% share of the risk under the Master Policy; and that the reason Australia and the UAE were to be excluded was because XL was not able to provide UNIQA with “risk engineering information” for those countries.

That exchange indicates that the construction of the Note for which UNIQA contends, and which I accept, is not one which is commercially absurd. Indeed, Mobis did not submit that it was.

Accordingly, my conclusion is that, for this reason alone, on the proper construction of the Master Policy, UNIQA has no liability for Mobis’s claim.

The third paragraph of Art 5.1 of the Master Policy (set out at [661] above) provides relevantly that:

UNIQA submitted that the effect of Art 5.1 would be to exclude any liability it had to extend cover to Mobis because the affording of such cover would be “impermissible” under Australian law.

That is because, UNIQA submitted, for it to afford cover to Mobis in such circumstances would be to carry on insurance business in Australia contrary to s 10 of the Insurance Act.

Section 10 of the Insurance Act relevantly provides that a body corporate commits an offence if (a) the body corporate “carries on insurance business in Australia” and (b) the body corporate is not a “general insurer” and (c) there is no determination in force under s 7(1) that s 10 does not apply.

Subsections (b) and (c) of s 10 are satisfied in the present case. UNIQA is not and never has been a “general insurer”. That term is defined in s 11 to mean a body corporate that is authorised under s 12 to carry on business in Australia. UNIQA has never been authorised under s 12. Further, there has never been any determination in force under s 7(1) that s 10 does not apply.

The question then is, on these assumptions, would UNIQA be carrying on insurance business in Australia within the meaning of s 10(a) so as to deprive Mobis of what would otherwise have been its entitlement to cover under Art 5.1 (and substitute an entitlement to Financial Interest Cover under Art 5.3 in favour of Mobis Slovakia).

UNIQA has no place of business in Australia. Its participation in the Master Policy is the only occasion upon which it has done any business having any connection with Australia.

I do not consider that, in those circumstances, it can be said that UNIQA was “carrying on business” in Australia.

The notion of carrying on a business involves the repetition of acts. Entry into an isolated transaction of a business nature does not constitute “carrying on a business” (see, for example, Thiel v Federal Commissioner of Taxation (1990) 171 CLR 338; HCA 37 at 347 (Dawson J) and 359 (McHugh J)).

By notice of motion filed in Court on 29 June 2017, Mobis sought leave to amend its claim against, relevantly, UNIQA to include the underlined words as follows:

In view of the conclusions to which I have come it is not necessary for me to deal with this application to amend.

In any event, I would not be prepared to allow the amendment. Any such order could only be made at the suit of Mobis Slovakia. In due course I will order that Mobis’s notice of motion be dismissed.

Mobis claims to be entitled to indemnity under the Local Policy in the sum of $62,190,665 calculated as follows:

This amount does not take into account the amount already paid to Mobis by XL, referred to at [10] above. It is common ground that this amount must be brought to account in due course.

Numerous issues arose in respect of quantum. I have endeavoured to deal with them all. I have not, however, attempted to calculate what final figure XL must pay Mobis. Indeed, as I explain below (see [795] – [797], [936] – [937], [1076] – [1080] and [1086] – [1090]), there are a number of topics in respect of which I have not been able to come to a conclusion and in respect of which I will invite further submissions. In the meantime, I invite the parties to confer and agree on the arithmetic consequences of the various conclusions to which I have come.

The warehouse has been replaced.

Mobis claims $17,247,234 for the cost of replacing the warehouse.

That claim is made up as follows:

The starting point of the claim is the cost of rebuilding the warehouse.

Mobis engaged FDC Construction and Fitout Pty Ltd to build the new warehouse.

It is common ground that the total amount paid by Mobis to FDC to construct the new warehouse was $19,717,498 comprising:

There is also now no controversy about the last two items claimed by Mobis in the above schedule, namely removal of debris ($1,181,451) and professional fees (which, Mobis accepts, are subject to a policy limit of $100,000).

Mobis seeks to add to the costs of construction of the new warehouse an amount it described as “Separate Building Costs” of $289,833.

Mobis’s quantum expert, Mr Paul Johnson, stated in his report of 1 June 2016:

Mr Johnson did not identify what building costs constitute the figure of $289,833.

However, his evidence about this matter was not challenged by XL’s quantum expert, Ms Kimberley Daley, or in cross-examination.

In its submissions, XL stated that “there is a real risk of such costs being duplicated by items of expenditure specifically claimed by reference to the general ledger”.

However, Ms Daley did not identify any such duplication.

In those circumstances, I accept Mobis’s submission that Mr Johnson’s unchallenged evidence should be accepted and that this figure should be added to Mobis’s building claim.

The slab on which the warehouse was constructed was not damaged by the collapse. It was, however, damaged by the fire.

The cost of replacing the slab is included in the amount of the FDC contract and, thus, must be deducted from Mobis’s claim.

Mobis accepts that $3,165,869 must be deducted from the amount of the FDC building cost on this account. That is the amount that Mobis claimed in a Notice of Dispute it wrote on 19 October 2015 to the demolition contractor it had engaged to demolish the warehouse, Metropolitan Demolitions Pty Ltd (see [867] below).

However, as XL pointed out, in a demand sent to Metropolitan (on 18 January 2016: see [873] – [874] below), Mobis’s solicitor contended that the cost of replacing the slab was $3,865,000; $699,161 more. That figure is very close to provisional sum 15 ($670,309.77) which, as the above table reveals, Mobis has already deducted from its building claim. Provisional sum 15 relates to the costs of remediating the foundations underpinning the slab. In those circumstances, I accept Mobis’s submissions that I should draw the inference that the figure of $3,165,869 represents the net cost of remediating the foundations of the warehouse and that, as provisional sum 15 is to be deducted from Mobis’s claim, no further deduction is called for on account of the replacement of the slab.

XL contends that certain items that Mobis has included in its building claim (for items such as air-conditioning units, dock levellers, lifts, switchboards, carpets, signage, exhaust fans, security systems, switchboard generators, roller doors and smoke alarms) should be deducted from Mobis’s building claim, and included as part of its contents claim (and thus be subject to average under cl 4.2.1(e) of the Local Policy).

For the reasons I set out below at [1035] to [1055], I accept that submission.

There is otherwise no dispute as to the amount in respect of which Mobis is entitled to indemnity for these items.

The warehouse that Mobis has constructed to replace that which collapsed is more robust than its predecessor.

Mobis incurred some $793,000 costs in relation to extra steel work that was incorporated into the reconstructed building, and makes no claim on XL for that cost.

However, XL submitted (and Mobis did not in its final submissions dispute) that the evidence revealed that the additional steel columns that were installed throughout the reconstructed warehouse resulted in the need for additional piles, pile capping and pad footings.

XL submitted that the cost incurred by Mobis in carrying out this work should also be deducted from the building claim.

There is, however, no evidence which identifies the cost of this work.

Mobis and XL disagreed as to which of them had the onus to prove this cost.

The question is one of betterment.

Both parties relied on the decision of the Court of Appeal in General Accident Insurance Asia Ltd v Sakr [2001] NSWCA 402.

As I read Sakr, it is authority for the propositions that:

Thus, at [71] Giles JA said:

As I read this passage, his Honour held that the insured must establish what the insurer was obliged to pay under the policy but that there was no onus on the insured to prove the absence of betterment.

Hodgson JA substantially agreed with Giles JA and stated (at [77]) that the onus was on the insured to prove their damages but stated (at [78]) that so far as concerns betterment “an evidentiary onus” was cast on the [insurer] to prove some betterment for which a reduction should be made.

Sperling J said (at [88] and [89]):

In my opinion, XL has sustained its evidentiary onus of showing betterment; namely, that the reconstructed warehouse, as well as having additional steel columns, has additional piles, pile capping and pad footings.

Mobis has not proved, and allowed as a deduction of its claim, the value of that betterment.

But some allowance must be made.

In Sakr, Hodgson JA said at [79] that in such a circumstances:

Here the party “responsible for the deficiency of evidence” is Mobis.

Even doing the best I can, on the basis of the “inadequate material” before me, I can only guess what part of the amount claimed by Mobis for the reinstatement of the warehouse is referrable to the extra piles, pile capping and pad footings.

In those circumstances, one alternative is simply to reject Mobis’s building case in its entirety on the basis that it had not proved the damage in respect of which it is entitled to indemnity.

However, I will invite submissions as to whether there is an alternative course available.

XL contends that the following further amounts should be deducted from Mobis’s claim. These relate to “provisional sums” and “variations” under the FDC building contract. I will deal with these in turn.

Mobis incurred this sum to restore landscaping around the warehouse.

The question is whether these costs are attributable to the warehouse collapse or the demolition process.

The invoice for this amount from FDC is dated 8 June 2016.

The process of demolition did not start until early July 2015.

This suggests that the cost of landscaping work is attributable to the warehouse collapse, rather than its later demolition.

Accordingly, my conclusion is that this amount should not be deducted from Mobis’s claim.

This is an expense incurred by Mobis to reinstate “IT requirements”.

In an appendix to Mobis’s submissions, it accepted that $419,622.49 of this sum amounted to “betterment”.

I am not satisfied that XL has sustained its evidentiary onus of showing that the balance of provisional sum 7 amounts to betterment and I propose to allow it.

This sum represents the cost of installing sprinklers on an additional mezzanine level in the reconstructed warehouse.

Mr Stoddart gave evidence that, prior to its collapse, the warehouse had a sprinkler system suitable for what the parties described as a “third level of binning”.

I am not satisfied that XL has sustained its evidentiary onus of showing that this cost amounts to betterment.

Accordingly I do not propose to deduct this amount from Mobis’s claim.

As I have indicated above, Mobis accepts that this amount should be deducted from its claim. I have concluded that, for that reason, no further amount should be deducted from Mobis’s claim on account of the replacement of the slab, beyond that set out in the table at [753] above.

This amount represents the cost to Mobis of designing and installing an additional thickness to the concrete slab.

According to an email from FDC to Mr Stoddart this work was needed:

Mr Stoddart said in cross-examination that the cause of these “additional loads” was an error made by FDC, and that this expense was incurred to correct that error.

In those circumstances, I do not see this expense as betterment.

This item concerns glazing costs incurred by Mobis due to a directive from Mobis Korea.

In final submissions, Mobis accepted that $141,400 of the claim constituted betterment.

XL made no submissions as to the balance of $10,859, which should be allowed as part of Mobis’s claim.

This cost was in respect of lights that Mr Stoddart accepted were not present in the warehouse at the date of collapse.

They should not be allowed as part of Mobis’s claim.

This item relates to glazed balustrades installed in the rebuilt warehouse as a result of a directive from Mobis Korea.

In final submissions Mobis accepted that $108,700 of this amount constituted betterment and did not form part of Mobis’s claim.

No challenge was made to the balance of Mobis’s claim for this item ($30,517) and I propose to allow it.

This expense related to joinery upgrades. The relevant invoice stated:

XL accepted that Mobis is entitled to claim $52,885 on account of this item but submitted that the balance (of $74,498) should be rejected because:

Mr Stoddart gave this evidence in cross-examination:

In light of this evidence, I am not satisfied that any of the cost of the joinery amounts to betterment and propose to allow the item without deduction.

This expense relates to upgrades in IT and comprised 13 individual expense items.

Mr Stoddart gave this evidence about this item:

Although Mr Stoddart agreed that some of these items were “upgrades to meet new IT requirements” he also added “[w]hat exactly is betterment, I'm not sure”.

XL had the evidentiary onus to show that these items amounted to betterment. I am not satisfied that, by this cross-examination, XL has sustained that onus.

Accordingly, my conclusion is that Mobis is entitled to this amount.

Mobis claims that stock to the value of $30,264,587 was either physically lost, destroyed or damaged when the warehouse collapsed.

The policy sub limit for stock is $27,573,108. Accordingly, Mobis makes a claim for the full amount of that sub limit.

The historic cost of stock in the warehouse when it collapsed was $26,911,372.

Mobis has replaced some, but not all, of the stock that was in the warehouse on the date of collapse.

The replacement cost of the stock that Mobis has replaced was $20,283,301. The historic cost of stock that was in the warehouse on the date of collapse but which had not been replaced was, as at December 2016, $9,797,372. The sum of those two figures is $29,682,339, which exceeds the stock sub limit.

The insuring clause under the Local Policy provides for indemnity under section 1 (“Property Damage”) “resulting directly from any Damage during the Period of Insurance”.

“Damage” is defined in cl 2.6 of the Local Policy to mean “physical loss or destruction of or damage to the Property Insured caused by an Occurrence”.

“Property Insured” includes stock (cl 2.23). “Occurrence” is defined to mean “one loss or a series of losses arising out of and directly resulting from one original cause” (cl 2.22).

Clause 4.2 provides that the basis of settlement is to be “reinstatement”.

Clause 4.2.1(g) provides that reinstatement shall be “the cost of replacement of lost, destroyed or damaged Stock…by similar property as new”.

Mobis summarised its case on stock as follows:

Following the collapse, Mobis engaged loss adjusters who, in turn, engaged Costin Roe.

On 27 April 2015, Costin Roe sent an email to the loss adjusters stating:

On 11 June 2015, Dexion, the manufacturer of the racking in the warehouse, carried out an inspection and reported that:

At about this time, Mr Stoddart, who, as I have mentioned, was Mobis’s General Manager Warehouse and Logistics, prepared a document called “Demolition and Recovery Process”. That document stated that:

This process was followed. The value of stock so recovered in saleable condition was nominal; some $123,590.

Any stock that was in a package that was damaged (whether by water or otherwise) was regarded as unsaleable and destroyed. Mobis’s Senior Manager of After Sales and Marketing, Mr Anthony Lasan, explained that this was because of concerns about consumer safety, an apprehension that Hyundai’s dealers would not purchase stock in damaged packaging, concern about damage to the “Hyundai brand” and Mobis’s experience that its dealers regarded stock in damaged packaging as creating a warranty risk.

No doubt because of an apprehension of further possible collapse and difficulties in gaining access to stock within the damaged warehouse, Mobis investigated, appointing a contractor to perform the demolition and stock recovery process.

Mobis also appointed a project manager to manage the demolition and reconstruction of the warehouse. That project manager obtained quotations from three demolition contractors and reported that one of those contractors, Metropolitan (referred to at [772] above), had a “methodology for lifting sections of the roof which is best suited to maximising salvage of stock” and was “an obvious choice”.

On 2 July 2015, Mobis entered a “Demolition and Stock Recovery Contract” with Metropolitan.

Pursuant to that contract, Metropolitan agreed that it would:

Metropolitan took possession of the site on 6 July 2015. Demolition commenced that day.

The demolition process involved cutting away sections of the roof with oxyacetylene torches. This generated sparks. It also involved demolition workers hosing down the stock with water while oxyacetylene cutting was being performed (to prevent sparks from the oxyacetylene torches igniting the stock). Nonetheless, at least one spot fire broke out during the demolition process.

Mr Stoddart explained that it was physically impossible to remove some stock. Many pallets of stock were not full and were unbalanced, which meant stock could not be retrieved without dropping the packages. Further, some stock was stored in narrow aisles, which required specialist equipment not available to the demolition contractors.

On 21 July 2015, Mr Stoddart reported to Mobis’s CEO, Mr Jay Suh:

On 24 July 2015 Mr Stoddart reported to Mr Suh:

And on 29 July 2015:

Despite his email correspondence with Mr Suh, Mr Stoddart gave this evidence in cross-examination:

A major fire broke out on 30 July 2015. The warehouse was destroyed. All the contents of the warehouse which had not been retrieved by then were (with only minor exceptions) lost.

At the time of the fire, the demolition process was still at an early stage. It had proceeded from the eastern wall up to, in effect, four or five rows in the racking area.

Evidently, Mobis contended that Metropolitan was responsible for the fire.

Thus, in August 2015, following the fire, Mobis’s solicitors asked Mr Stoddart to provide an estimate as to the loss that Mobis had suffered as a result of the fire. Mr Stoddart prepared an estimate and sent it to Mr Suh, who then told Mr Stoddart to send the material Mobis’s solicitors.

The amount of that estimate was $17,545,389.

On 19 October 2015, Mobis sent a Notice of Dispute to Metropolitan in which it stated that its present estimate of its loss in relation to stock was $17,545,389.

In cross-examination, Mr Stoddart said that this estimate was “broad brush”. Mr Stoddart said, in answer to questions from me, that he had arrived at the estimate of $17,545,389 by ascertaining the value of the stock in the warehouse at the time of the collapse and deducting 25% on account of stock that was “probably wet and irretrievable”.

Thus, Mr Stoddart gave this evidence:

Mr Stoddart said that the "first figure" (referring to the $23 million) was accurate, but that the second (referring to the $17.5 million) was "pure estimate".

However, Mr Stoddart’s initial estimate of $17,545,389 was the subject of a process of refinement.

Mr Stoddart said that between 19 October 2015 (the date of the Notice of Dispute) and 18 January 2016 he spent "some time to ascertain with a degree of precision the amount of stock which perished in the fire". When asked about the exact process he undertook in order to derive the precise figure of $18,438,945.49 he said "we extracted the information from the computer, the total stock". He was assisted by Mr David Peterson who was responsible for inventory control. He agreed that Mr Peterson (whom he described as "very good with his inventory analysis") came to the view in January 2016 that approximately $18.5 million of stock suitable for consumption and distribution was destroyed in the fire.

On 18 January 2016, Mobis’s solicitors sent Metropolitan a “Demand for Indemnity” which quantified Mobis’s loss in relation to stock “as a result of the fire” at $18,438,945.49.

The letter stated:

Mr Stoddart agreed that it was likely that he was the person that established the figure of $18,438,945.49 as being the “salvageable property” of Mobis at the site.

In a letter that Mobis’s solicitors sent Metropolitan’s solicitors on 2 February 2016, Mobis’s solicitors stated that Metropolitan had “been provided with extensive documentation and information which substantiate the quantum of our client’s loss”.

During the hearing, XL served a notice to produce on Mobis seeking production of that “extensive documentation”. Mobis claimed that those documents were the subject of legal professional privilege. I upheld that claim for reasons I gave in a judgment on 5 June 2017: Mobis Parts Australia Pty Ltd v XL Insurance Company SE (No 4) (Supreme Court (NSW), Stevenson J, 5 June 2017, unrep).

Nonetheless, the evidence showed that Mobis, through Mr Stoddart, made a precise calculation of the value of the “salvageable property” at the warehouse as at the date of the fire for the purposes of making its claim on Metropolitan, and that calculation was based on the “extensive documentation and information” that Mobis had provided to Metropolitan (and that Mobis objected to being placed into evidence).

Mr Stoddart gave this evidence:

Mobis’s solicitors’ letter of 18 January 2016 referred to “salvageable” property, being the stock stated to have a value of $18,438,945.49.

It is clear from the letter that its author was speaking of property “salvageable” by Mobis, in the sense of being “suitable for sale and distribution” (as Mr Stoddart agreed in the answer referred to in the evidence just quoted).

The figure of $18,438,945.49 was calculated as a result of a request by Mobis’s legal counsel, Ms Kim, to obtain updated estimates in order to brief Mobis’s solicitors to commence proceedings against Metropolitan.

On 1 June 2016, Mr Stoddart swore an affidavit in which he set out what he said was a calculation of the proportion of stock at the warehouse that “might have been salvaged had there been no fire”.

Once again, it is clear from the context, that Mr Stoddart was referring to stock “salvageable” by Mobis.

Mr Stoddart said in that affidavit:

Mobis described the process undertaken by Mr Stoddart as follows:

Mr Stoddart’s conclusion was that there was “a chance of stock with a value of about $4,399,858.72 could have been recovered [from the warehouse] had the salvage process been completed”.

I understood Mr Stoddart to be saying that stock to this value could have been recovered in a saleable condition.

The process by which Mr Stoddart reached this conclusion was summarised in a table headed “Possible Salvage Calculation” which is attached to this judgment [Possible Salvage Calculation table (944 KB, pdf)].

Counsel for XL did not in cross-examination challenge Mr Stoddart about the detail of his process of estimation, but did put that Mr Stoddart had “sought to exaggerate the value of stock which was lost by reason of the collapse”.

One difficulty with Mr Stoddart’s “calculation” is that it assumed that stock to the value of $8,540,902 “would have been damaged during the demolition process”. The basis for that assertion was a statement made by Mr Stoddart earlier in the affidavit which, because of its form, I rejected. I granted leave to Mobis to adduce evidence from Mr Stoddart in proper form, if such evidence was available. That leave was not taken up. Accordingly, the basis for this aspect of Mr Stoddart’s “calculation” fell away.

Thus, I am faced with two opinions expressed by Mr Stoddart as to the value of the stock that, but for the fire, could have been “salvaged” by Mobis; the $18,438,945.49 estimate made in January 2016 and the estimate of $4,399,858.72 made by Mr Stoddart in his affidavit of 1 June 2016 (suffering, as it does, with the difficulty to which I have referred).

Overall, I am persuaded that I should prefer the higher figure to the lower, essentially for the reasons set forth in UNIQA’s submissions as follows:

I accept those submissions and I am persuaded that I should infer that:

As I have set out above (see [840]) “damage” is defined in the Local Policy to include “physical loss” of Mobis’s property (the full expression being “physical loss or destruction of or damage to” that property).

The question is whether, as Mobis contends, there was a “physical loss” of all the stock (that is, the stock was “lost” for the purposes of the policy) as soon as the warehouse collapsed.

The “lost” stock that is relevant, in monetary terms, is that which was otherwise undamaged; there is no controversy that Mobis is entitled to indemnity in respect of damaged stock (save for that which it has not replaced: see below).

But an analysis of what stock has been “lost”, as opposed to “damaged”, directs attention to all the stock, damaged or not, that is said to have been lost by reason of the collapse.

My attention has been drawn to a number of authorities in which the question of whether insured property has been lost has been considered.

It is common ground before me that, as this is a non-marine case, no question of “constructive loss” arises.

Those authorities (and, indeed, common sense) establish the following propositions concerning establishment of loss for these purposes:

The parties devoted many pages of their submissions to a detailed analysis of the factual circumstances considered in these cases. I do not find it necessary to conduct such an analysis in these reasons. The facts in those cases were very different than that before me. The cases are valuable in so far as they lay down the guidelines that I have referred to in the preceding paragraph. Otherwise, this case must be determined on its own facts.

Mobis’s case is that all the stock in the warehouse was lost when the warehouse collapsed because, at that point, its prospects of recovery in an undamaged condition was uncertain because:

There were 88,080 lines of stock and approximately 1.5 million individual items in the warehouse at the time of collapse. Following the collapse, all of those items of stock remained in the warehouse, which Mobis owns. Each of the items remained in the warehouse until Mobis itself retrieved and moved some of the stock and, thereafter, engaged Metropolitan to recover stock and demolish the warehouse pursuant to the 2 July 2015 Demolition and Stock Recovery Contract to which I have referred.

The process came to an end when the fire destroyed what remained on 30 July 2015.

Mobis makes no claim for indemnity in respect of the fire. Its only claim is in respect of the damage caused by the collapse. Mobis made, but has now withdrawn, a claim under the 2015/2016 Local Policy for indemnity in respect of loss or damage arising from the fire.

Nonetheless, Mobis submitted that:

Mobis’s submission was, thus, that:

I do not accept that submission.

A fair reading of the evidence is that the fire was an unexpected event. It was in no sense a consequence of the storm of 25 April 2015. It was not a necessary or direct consequence of the storm or the collapse. It was an intervening cause of damage and destruction.

It is true that Metropolitan used oxyacetylene torches to dismantle parts of the warehouse.

There is in evidence, photographs showing Metropolitan employees using oxyacetylene equipment and of sparks flying while that equipment is in use. There is also in evidence, vision of a spot fire that broke out (and was contained) prior to 30 July 2015.

But I do not accept that it follows from this evidence that, as Mobis has submitted, “attempting salvage itself gave rise to a significant risk of fire”, let alone that fire was an inevitable consequence of the collapse. Indeed, the fact that Metropolitan engaged in a process of pre-wetting stock (which Mobis relies upon to show how some of its stock was damaged) shows that Metropolitan adopted a procedure directed to minimising the risk of fire which it was obviously confident would be effective; hence the wide ranging warranties that it gave Mobis in the Demolition and Stock Recovery Contract (see [854] above).

Unlike the circumstances discussed in many of the authorities to which I have referred, this is not a case where the insured property became missing or was mislaid. Mobis did not lose physical possession of the stock.

I do not accept that, at the time of the collapse, it was “uncertain” that Mobis would recover any stock from the warehouse.

No witness was called by Mobis to say that.

In my opinion, the proper conclusion from the evidence is that, as at the date of the collapse, it was simply too early to say to what extent Mobis would be able to recover stock from the warehouse. The photographs taken by Mr Stoddart on the day after the collapse suggest that there was at least some stock that could be recovered (albeit much of it was wet and otherwise damaged).

In my opinion, as at the date of the collapse, Mobis was not in a position to form any view as to whether it was “uncertain” that it could recover its stock until it took reasonable steps to investigate the situation.

And that is what Mobis did. Mr Stoddart conceived and implemented the Demolition and Recovery Process to which I have referred. Once that process was completed, Mobis engaged a project manager to investigate the engagement of a demolition contractor. Mobis then engaged a demolition contractor whose processes seemed most likely to maximise the recovery of the stock still within the warehouse.

When the fire broke out, all of the stock that had not been removed from the warehouse was still in the warehouse. As I have said, the most reliable guide as to how much of that stock was still in saleable condition when the fire broke out is the analysis that Mr Stoddart undertook in January 2016, in which he concluded that saleable stock to the value of $18,438,945.49 was present in the warehouse at the date of the fire. The remaining stock was, Mr Stoddart opined, in damaged condition. Some of it may not have been able to be retrieved separately from the components of the warehouse structure that Metropolitan was to demolish. But that stock was not “lost”. It was still in Mobis’s possession, albeit doubtless to some extent in an unsaleable condition.

As at the date of the collapse, the only “uncertainty” that existed was as to the extent to which the stock was damaged or destroyed. There was no “uncertainty” as to whether the stock was lost.

However, much of Mobis’s stock was destroyed or damaged in the collapse.

The best evidence available as to what stock was not destroyed or damaged by the collapse (but which was ultimately destroyed in the fire) is the amount that Mr Stoddart calculated in January 2016 and which was the subject of the demand made by Mobis on Metropolitan on 18 January 2016; namely $18,438,945.49.

The historic cost of Mobis’s stock in the warehouse at the date of collapse was $26,911,372 according to Mobis’s database, AMOS.

Thus, as Ms Daley has calculated, some 69% ($18,438,945.49/$26,911,372 x 100) of Mobis’s stock was not damaged or destroyed in the collapse, and, but for the fire, would have been “salvageable” in the sense I have set out above (that is, saleable).

The balance of stock, some 31%, was thus damaged or destroyed by the collapse.

Mobis has replaced stock to the value of $20,283,301. On the face of it, therefore, Mobis is entitled to indemnity for 31% of that amount: namely $6,287.823.

There are, however, two further matters to consider.

The first is that Mobis destroyed some stock, which may not have been damaged, without inspecting it.

The second is that Mobis has not yet replaced stock with an historic cost of $9,797,372. Based on the above analysis, some 31% of that stock was damaged or destroyed: an amount in the order of $3,037,171. The question is whether Mobis is entitled to indemnity for that amount.

Much of the stock in the warehouse at the date of the collapse comprised spare parts in cardboard packaging, which packaging was damaged (most commonly by water) during the collapse. Mr Lasan (from Hyundai) and Mr Murray (from Kia) gave evidence to the effect that it was the policy of Hyundai and Kia, for reasons of consumer safety and brand protection, to destroy stock in damaged packaging, without investigation of whether or not the spare parts themselves were compromised.

Mr Stoddart accepted it was possible that, in some cases, although the packaging of an item of stock was wet or damaged, the stock itself housed in plastic wrapping, was dry and could otherwise be in good condition. However, according to Mr Stoddart, those parts were not capable of inspection because “there are millions of parts” and “we just couldn’t have done it”.

Nonetheless, as I have set out above (see [848]), Mr Stoddart’s Demolition and Recovery Process anticipated that, as part of the "further inspection” of stock that Mobis removed from the Demolition Area to the Quarantine Area, and then to neighbouring warehouses, might be unpacked “for detailed inspection”.

It seems likely that some stock in saleable condition (albeit in damaged packaging) was thereby destroyed. There is no suggestion in the evidence that Mr Stoddart took such stock into account when coming to his estimate as to the amount of saleable stock at the warehouse at the time of the fire.

There is no evidence before me as to how much stock was destroyed for this reason, or what its value was.

Some allowance must be made on this account.

It is for Mobis to prove how much of its stock was damaged by the collapse. As it appears likely that Mobis has, evidently for reasons of practicality, itself destroyed stock which was not damaged in the collapse, unless Mobis proves the value of the stock so destroyed, it cannot make out the total claim that should be the subject of indemnity.

As in the case of the extra piles and footings issue (see [795]-[797] above), one possible course is simply to dismiss Mobis’s claim for damage to stock for the reason that this element (which would constitute a deduction from its claim) is not proven.

However, I will invite submissions as to whether any alternative course is available and preferable.

There was stock at the warehouse on the date of collapse that Mobis has classified in its records as being obsolete.

Mobis’s Senior Manager for Sales and Marketing, Mr Mark Harrington explained that:

Stock is only declared obsolete (by Mobis Korea) after Mobis has considered whether any “local factors” (such as service recalls or other campaigns in Australia) militate against such classification. That suggests that, once stock is declared obsolete, Mobis has formed the view that there will be no further call for the stock. That is no doubt because, as Mobis put it in final submissions, obsolete stock represents parts used in vehicles that are “long out of production”.

Once stock is declared obsolete, it is written off in Mobis’s books.

Stock that has been declared obsolete includes stock no longer manufactured. In that regard, Mr Harrington said:

Thus, stock declared to be obsolete is not necessarily irreplaceable.

Mr Harrington also said that 6,137 of the 88,080 lines of stock in the warehouse at the date of collapse could not be replaced as at October 2016 for a number of reasons, including that the part was no longer available (by which I assume Mr Harrington meant, readily available) from suppliers. Ms Daley’s investigation of Mobis’s records reveals that this is the explanation for almost all of those stock lines.

The Inventory Soundness Report records that as at April 2015 there was present in the warehouse obsolete stock to the value of $29,185.52.

Mobis has made provision in its financial records for obsolete stock at the date of collapse of $245,059.61.

That appears to me to be the most reliable guide as to the value of stock in the warehouse that Mobis believed to be obsolete on the date of the collapse.

Ms Daley has ascertained that 42,591 of the 88,080 line items of stock in the warehouse at the date of collapse were not replaced within the 13 month period following the collapse; and that of those 42,591 line items, no example of 27,058 of the line items had been sold since January 2014 (15 months before the collapse).

Mr Harrington said that as at October 2016 (approximately 18 months after the collapse), Mobis had replaced only 39,242 of the 88,080 lines of stock.

Ms Daley has calculated, by reference to Mobis’s records, that Mobis has not replaced stock to the value of $9,797,372 (based on historical costs) that was in the warehouse on the date of the collapse. Of that sum, $6,425,641 represents lines of stock not replaced at all, and the balance, $3,371,731, partly replaced lines of stock.

According to Mobis’s profit and loss accounts, it had a turnover of some $100 million for the calendar year 2014, some $106 million for the calendar year 2015 and some $29 million for the three months to March 2016. In those circumstances, XL submits, and I accept, that it is reasonable to infer that Mobis has sold in excess of $200 million worth of parts, and has ordered parts to about that value, since the date of the collapse.

Mobis has procedures for classifying certain stock as “non-moving”. Mobis classifies its stock into five classes. It deems stock in one of those classes as “non-moving” if it is not sold for more than 5 years, and stock in another class as not moving if not sold within 12 months.

As at April 2015, Mobis’s records (in the form of an “Inventory Soundness Report”) identified non-moving stock at the warehouse in the order of $1.25 million. Mr Harrington accepted that, were it not for the collapse of the fire, in all likelihood, some of the stock which perished in those events would also have been declared to have been “non-moving”.

Mobis has a policy of retaining spare parts for older model cars to cater for the possibility that they might be needed from time to time.

In that regard, Mr Harrington gave this evidence:

Mr Harrington also gave this evidence:

Mr Harrington said that “we replaced what was required at the time”.

It may be that, from time to time, customers with an older model Hyundai will require a spare part of a kind that was present in the warehouse at the date of collapse but which Mobis has not replaced or cannot readily replace.

However, it seems highly unlikely that any more than a minute fraction of this stock will ever be called for. It is now almost two and a half years since the collapse, and yet, so far as the evidence reveals, the demand for these components is as I have set out.

The question is whether Mobis is entitled to indemnity under the Local Policy for that proportion of this stock (that is, some 31%) as was damaged or destroyed in the collapse.

It is necessary first to consider cl 4.2.1 (d) of the Local Policy.

Mobis is entitled to indemnity for destroyed stock on the basis of the “cost of replacement” of that stock: cl 4.2.1(g).

Clause 4.2.1(d) is in the following terms:

The question arises as to whether the effect of this clause is that Mobis is not entitled to be indemnified for the “cost of replacement” of stock until that cost has been actually incurred.

Mobis submitted that cl 4.2.1(d) only had effect if XL had exercised its option (set out in the insuring clause: see [39] above) to replace or repair the damaged property.

That submission was based on the proposition that earlier subclauses of cl 4.2.1 (especially cll 4.2.1(a) and (b)), by their terms, could only have relevant application in circumstances where XL had exercised its option under the insuring clause, and that the same can be said for cl 4.2.1(d).

Clauses 4.2.1(a) and 4.2.1(b) are in the following terms:

I see force in the argument that these two subclauses could only have logical application had XL exercised its option under the insuring clause.

But I do not see that the same applies to cl 4.2.1(d).

Mobis submitted that, if that were the correct conclusion, then cl 4.2.1(b) operates as a condition precedent to liability and thus engages the provisions of s 54 of the Insurance Contracts Act.

I now turn to that issue.

Section 54(1) (when read with s 54(6)) provides:

Mobis submitted that a failure by Mobis to incur the cost of replacing stock for the purposes of cl 4.2.1(d) was an omission of the kind to which s 54(1) is directed.

In that regard Mobis drew my attention to the decision of the Court of Appeal in East End Real Estate Pty Ltd (t/as City Living) v CE Heath Casualty & General Insurance Ltd (1991) 25 NSWLR 400 (in particular, the observations of Gleeson CJ at 403E and of Mahoney JA at 406E) and of the High Court in Antico v Heath Fielding Australia Pty Ltd (1997) 188 CLR 652.

Mobis submitted that:

In its further submissions in reply, XL did not dispute these matters but simply submitted:

I read those submissions as, in effect, abandoning reliance on cl 4.2.1(d) (while maintaining XL’s submission that Mobis is not entitled to indemnity for stock which it cannot or will not replace).

Ultimately, XL accepted that it was not entitled to refuse to pay Mobis’s claim by reason only of the fact that Mobis had not complied with cl 4.2.1(d).

As I have mentioned, the total value (historical cost) of stock that Mobis has not replaced is $9,797,372, of which a small proportion ($245,059.61: see [948] above) has been classified as being obsolete and, evidently, cannot be readily replaced.

As a result of the collapse, Mobis has suffered a loss in respect of 31% (see [926] to [930] above) of that stock, in that it acquired that stock, presumably at some cost, and now cannot sell it, as it has been damaged and rendered unsaleable.

If, as seems probable, the fact is that Mobis will not in any event be able to sell that stock (because there is and will be no demand for it), that will be a loss it would have suffered in any event; that is, whether or not the warehouse had collapsed.

The question, then, is Mobis’s entitlement under the Local Policy in respect of this stock.

I see this question as depending on the proper construction of the Local Policy.

The general principle was stated in D Kelly and M Ball, Kelly and Ball Principles of Insurance Law, (2nd ed, 2001, LexisNexis) at [12.0120.25] as follows:

The learned authors referred to the decision of the Court of Appeal in Leppard v Excess Insurance Co Ltd [1979] 2 All ER 668; 1 WLR 512. The issue in that case was whether the insured was entitled to indemnity on the basis of the costs of reinstatement or market value. The Court concluded that the insured was entitled to recover his real loss, but not exceeding the cost of replacement, and that the real loss was the market value of the insured property.

The Court referred to the general principle enunciated in Castellain v Preston (1883) 11 QBD 380 in which Brett LJ said (at 386):

The question was also considered by the High Court in British Traders’ Insurance Co Ltd v Monson (1964) 111 CLR 86. The issue in that case was whether the insured could recover the full insured value of property destroyed by fire or merely a loss of their interest (as lessee) of the property.

The plurality (Kitto, Taylor and Owen JJ) said (at 94):

XL’s promise under the Local Policy was to indemnify Mobis based on the cost of reinstatement for property damage. In the case of stock, the promise was to pay “the cost of replacement of…damaged Stock…by similar property as new”.

The Local Policy is thus a reinstatement policy.

In CIC Insurance Ltd v Bankstown Football Club Ltd (1997) 187 CLR 384 the majority (Brennan CJ and Dawson, Toohey and Gummow JJ) discussed reinstatement policies and said at 395:

I was not able to find in the New Zealand case cited by their Honours (City Realties (Holdings) Ltd v National Insurance Co of NZ Ltd (1986) 4 ANZ Ins Cas 60-695) a statement to the effect set forth by the majority. It seems Palmer AJ came to the same conclusion in General Newspapers Pty Ltd v Cigna Insurance (1991) 6 ANZ Ins Cas 61-086 at 77,321-2.

However, it appears that the majority did not disapprove of the suggestion that they saw as being contained in City Realties.

However, as their Honours made clear, that view was subject to the proviso that there be no requirement in the policy such as is contained in cl 4.2.1(d) of the Local Policy.

XL accepts that, by reason of s 54 of the Insurance Contracts Act, it may not refuse Mobis’s claim by reason only of the fact that Mobis has not satisfied the “condition precedent to liability” set out in cl 4.2.1(d).

However, the fact remains that the wording of the Local Policy included cl 4.2.1(d). That is part of what the parties agreed. Section 54 does not have the effect that cl 4.2.1(d) is statutorily written out of the Local Policy.

I cannot see why it is not available to assist understand what the parties meant when they agreed that Mobis’s indemnity for damaged stock would be the cost “of” replacement of that stock; and not, for example, the “replacement cost” of that stock.

In my opinion, the wording of cl 4.2.1(d) suggests that the parties intended that the “cost of replacement” of damaged stock was the cost actually incurred by Mobis in replacing that stock. Why else would the parties have agreed that a condition precedent to XL’s liability to pay Mobis’s reinstatement costs (which must include the cost of replacing damaged stock) be the actual incurring of those costs by Mobis.

For those reasons, my conclusion is that XL is not liable to indemnify Mobis for the cost of stock it has not replaced.

At the time of the collapse, Mobis was storing in the warehouse stock owned by Kia Motors Australia Pty Ltd.

Mobis claims that the value of that stock was $460,482.

On 2 June 2015, Kia made a claim on Mobis for that sum and described its claim as being in respect “the value of goods owned by [Kia] which was stored at [the warehouse]”.

In an affidavit, Mr Harrington said that:

In closing submissions, Mobis drew my attention to a Memorandum of Understanding made in March 2013 between Mobis and Kia pursuant to which Mobis agreed to “assume ownership and management of [Kia] inventory of A/S parts and accessories”.

However, on 2 June 2015, Kia wrote to Mobis stating:

Kia concluded:

It is plain from this correspondence that the relevant material belonged to Kia, not Mobis.

In those circumstances I do not see how the Local Policy can respond to it.

Mr Harrington seems to have had the same view as, on 2 June 2015, he wrote to Kia:

I am not satisfied that Mobis is entitled to indemnity in respect of any property belonging to Kia which was damaged or destroyed in the collapse.

Indeed, the matter was barely pressed in final submissions.

Mobis claims an amount of $8,545,089 for contents lost when the warehouse collapsed.

Two issues arise. The first is in respect of a number of individual items.

The second is whether the contents claim should be subject to average.

I will first deal with the individual items that are still in contention.

This item concerns “finger scanners” (devices used to record stock) lost in the collapse.

Ms Daley opined that this claim may have been duplicated. However, Mr Stoddart gave evidence that there was no duplication and that these items were lost as a result of the collapse.

I allow this claim.

The claim for this item was withdrawn in final submissions.

Ms Daley was only able to identify $15,877 of the amount claimed in Mobis’s Asset Register.

However, Mr Guangxiang Liang, an accountant employed by Mobis, gave evidence that the items were not in the Asset Register as they were consumable items. He provided a photograph which showed the existence of the racking and binning labels at the time of collapse.

I allow those items.

The claim for this item was withdrawn in final submissions.

XL accepted this claim in final submissions.

This is the cost of the supply and installation of intumescent coating to steel columns and the supply and installation of fire rated materials within the warehouse following the collapse.

XL submitted that the claimed expense duplicated a claim already made.

But that suggestion was not put to Mr Stoddart and is inconsistent with his evidence.

I allow this amount.

Although Ms Daley pointed out that these items are not included in Mobis’s Asset Register, Mr Stoddart gave evidence that these dividers were installed during the original construction of the warehouse and were present at the time of the collapse.

I allow these items.

Mr Stoddart gave evidence that these items were purchased in 2005 and were still in the warehouse and in use at the time of collapse.

I allow this item.

The sum insured under the Local Policy for contents was $9,683,269 (as set out in the Asset Schedule to the policy).

Clause 4.2.1(e) of the Local Policy provided for average in respect of contents (and also, but not relevantly, for buildings) and was in the following terms:

The effect of cl 4.2.1(e) was that if the sum insured for contents was less than 85% of the cost that would have been incurred for reinstatement of all the contents insured, Mobis’s entitlement to indemnity was rateably reduced in that proportion.

The question of whether average applies depends upon whether certain items that are part of Mobis’s claim for damage to the building (see [774] above) should be characterised as “contents”.

Those items comprise air-conditioning units, dock levellers, lifts, switchboards, carpets, signage, exhaust fans, security systems, switchboard generators, roller doors and smoke alarms.

If those items should be characterised as “contents”, the result is that Mobis was underinsured for contents such as to enliven the average provision. The parties are in agreement as to the amount that should be deducted from Mobis’s claim if this is the correct conclusion.

Clause 2.23.1 of the Local Policy defined “Buildings” as, relevantly:

On the other hand, cl 2.23.2 defined “Contents” to mean, relevantly:

At all relevant times it was a mutually known fact that Mobis owned the warehouse. It was not a “tenant”. There was no “landlord”. Hence, the use of those words in the Local Policy was inapt. The reference “[l]andlord’s fixtures and fittings” and “[t]enants’ improvements and betterments” should be read as any fixtures, fittings, improvements or betterments in the warehouse.

On the face of it, the items in question are fixtures, fittings and improvements, and thus part of the building. As Mobis submits, they are all attached to the building and form an intrinsic part of its function as a warehouse. They are “annexed” to the property in the sense described by the High Court in TEC Desert Pty Ltd v Commissioner of State Revenue (WA) (2010) 241 CLR 576 at 585 (French CJ, Gummow, Heydon, Crennan and Kiefel JJ).

However, XL pointed to the fact that the items are listed in Mobis’s “Asset Register” as items of machinery or hardware, as thus as contents, rather than components of the building.

In that regard, XL relied on cl 4.16 of the Local Policy which provided:

By this clause, XL agreed, for the purposes of determining whether to insure Mobis’s property as “contents” rather than “buildings”, to accept the manner in which Mobis described the property in its own records (provided such description was reasonable).

The question may well have been relevant to the level of premium; although there is no evidence about whether XL would have charged a different premium if Mobis had included the value of these items as part of the building.

The amount Mobis nominated to XL as the value of its contents for the purposes of the Local Policy was $9,683,269. This is the figure specified as “General Contents” in the Asset Schedule of the Local Policy. That figure is recorded in Mobis’s Asset Schedule as the value of its contents, and included the items in question.

Although XL did not see Mobis’s Asset Register when it wrote the Local Policy, by cl 4.16 it agreed to accept Mobis’s designation that these items should be included in the Asset Schedule to the Local Policy, and thus be insured as contents. I would infer that XL assumed (correctly) that this was how Mobis characterised the items in its own records.

In my opinion, it is implicit in cl 4.16 that the parties agreed that if XL insured Mobis’s property on this basis, Mobis could not later contend that the property be differently characterised.

As XL submitted, Mobis has taken the benefit of the clause by designating these items as contents and cannot be now heard to say that they are not. I accept the submission that the designation clause should not be treated as being ambulatory and thus permitting Mobis to alter the heading under which its property is claimed.

In my opinion, on the proper construction of cl 4.16, it represents an agreement between Mobis and XL by which:

Accordingly, the parties should be taken to have agreed that the items in question were “contents”.

The average provision is thus enlivened.

Mobis claims various amounts on account of business interruption. The Local Policy provides cover for Mobis’s loss of gross profit and increased cost of working resulting from the interruption to and interference with its business caused by the warehouse collapse.

Thus, cl 7 of the Local Policy provides, relevantly:

Clause 7.1.1(a) provides for the calculation of the loss of gross profit by reference to cll 8.1 and 8.7.

Clause 7.1.1(b), when read in conjunction with the last two lines of cl 7.1.1, defines “Increase in Cost of Working” as follows:

Mr Johnson has assessed Mobis’s loss of gross profit at $318,557; Ms Daley at $34,735.

The matters that divide the parties on this question are:

Mr Johnson, on behalf of Mobis, applied a rate of 28.95% to lost sales. Ms Daley, on behalf of XL, applied a rate of 27.99%.

Neither Mobis nor XL offered any submission as to which of these rates was to be preferred.

In the absence of such submission, I find that the halfway point: 28.47% should be adopted.

Mr Johnson calculated loss over a period of eight months; Ms Daley over three months.

The only direct evidence on this issue was from Mobis’s sales and marketing manager, Mr Harrington. His evidence was that there was (unsurprisingly) a drop in revenue from the date of collapse (25 April 2015) to the end of May 2015, followed by a surge in sales from June to late September 2015 and a dip in sales after September 2015 (compared to budget).

That evidence suggests that the period over which to assess loss of sales lies somewhere between the figures adopted by Mr Johnson and Ms Daley.

I propose to adopt the period from the date of collapse to the end of September 2015 (a little over five months); being the period the subject of Mr Harrington’s evidence.

The question here is what rate of sales growth should be assumed to have occurred but for the warehouse collapse.

Mr Johnson calculated a trend for sales growth of 11.49%; Ms Daley 5.62%.

Ms Daley opined that the best indicator of expected trend was the six month period prior to the collapse: 5.62%.

But the sales growth actually experienced by Mobis between April and September 2014 (which corresponds, approximately, to the period I have adopted for present purposes) was 13.65%. The growth during the period April to September 2015 was 8.26%, and during the period April to September 2016 it was 20.7%. These figures suggest that Ms Daley’s figure is somewhat pessimistic.

Mr Harrington’s evidence of a drop followed by a surge in sales following the warehouse collapse must be considered in the context of his evidence that most of Mobis’s customers are Hyundai and Kia dealers who, typically, hold an average of 45 days’ worth of spare parts in their own inventories and Mr Stoddart’s evidence that temporary warehouses were fully operational by 30 June 2015. Further, there is no evidence to suggest that the demand for Hyundai and Kia spare parts would have changed during the five month period following the collapse (compared to the corresponding period before the collapse).

That evidence suggests that, despite the warehouse collapse, the sales growth actually achieved in the five month period following the collapse was in the same order it would have been but for the collapse.

In those circumstances, I propose to adopt the figure of 9%; being slightly more than that actually experienced from April to September 2015.

Both Mr Johnson and Ms Daley assumed that the fire on 30 July 2015 had no effect on Mobis’s gross profit.

But it must have; some part of Mobis’s loss of profit must be attributable to the extra disruption caused by the fire.

In the absence of evidence on the subject, I can only guess as to what that effect was.

Just as in the case of the extra pilings issue in the building claim (see [795] to [796] above) and the saleable stock destroyed without inspection (see [931] to [939] above) one possible course, is that I simply disallow this aspect of Mobis’s claim for business interruption because of this hiatus in the evidence.

However, I will invite submissions as to whether an alternative course is available and preferable.

Mobis claims an amount of $768,569 for additional labour costs.

XL accepts all but $39,104 of this claim.

The difference between the parties is principally a product of the differing methodology adopted by Mr Johnson and Ms Daley.

Mr Johnson used average wage costs for the period before and after the collapse, whereas Ms Daley adopted the actual hours worked and hourly rates to ascertain the incremental increase in labour costs.

Ms Daley’s methodology appears to me to be more reliable, as it is based on actuality. I propose to adopt it.

However, once again, neither expert made any allowance on account of labour costs caused by the fire.

It seems likely that Mobis incurred some additional labour costs because of the fire.

But, again, I could only guess what those costs were.

And, again, one course I could follow is simply to dismiss this aspect of Mobis’s claim.

However, I will invite submissions as to whether there is any other course I should follow.

These are costs recorded in a ledger created by Mobis to record expenses incurred by reason of the collapse and the fire.

Mobis claims $5,346,507 under this head, whereas XL contends the correct figure is $4,495,010: a difference of $851,497.

Clause 7.1.1(b) of the Local Policy requires that Mobis demonstrate, in each case, that the additional expenditure was:

I will deal with each of the items in dispute in turn.

The bulk of this amount ($18,000) relates to charges from the local council for road blocks around the damaged warehouse. Mr Stoddart gave evidence that Mobis was required by NSW Fire & Rescue to set up road closures and safe access to the damaged warehouse.

I am not satisfied this expenditure was for the sole purpose of avoiding or diminishing reduction in turnover. It thus does not satisfy the “sole purpose” test.

This cost was incurred to re-establish power to the damaged warehouse. Turnover could not be generated without power. I am satisfied the sole purpose test is satisfied.

This is a cost levied by the local council for road closures. It does not satisfy the sole purpose test.

Mr Stoddart and Mr Harrington gave evidence that these costs related to travel and accommodation expenses of key decision makers from Mobis Korea who travelled to Australia immediately following the collapse to facilitate the making of urgent decisions concerning significant items of expenditure, such as leases, fitout, equipment and staffing.

Mr Stoddart said that the presence in Australia of these key decision makers assisted in getting Mobis’s business up and running and thereby diminished any reduction in Mobis’s turnover.

Although there was no evidence before me of any breakdown or itemisation of these figures, I am satisfied that they are incurred and that they satisfy the sole purpose test.

This expense was for security guards for the warehouse.

Mr Stoddart gave evidence that this cost was incurred to protect potentially salvageable items of stock and other assets, and to also protect Mobis from the risk of people entering the premises and sustaining injury.

However, over half of these costs ($295,000) were incurred after the fire.

Mr Stoddart gave evidence that the security was required “to prevent people going inside and doing whatever those sort of people do”.

That evidence makes clear, in my opinion, that the purpose of the security guards was not to avoid or diminish the reduction of turnover. In any event, much of the expenditure was related to the consequences of the fire, and not the collapse.

The sole purpose test is not satisfied.

These costs were incurred for packing and wrapping stock that Mobis consumes in the course of its business. The items were ordered prior to the date of the collapse as spare stock. Because existing stock was lost when the warehouse collapsed, these items were used to replace it. I do not see how this cost satisfies the sole purpose test.

Mr Johnson agreed (in the Joint Report with Ms Daley) that this item is included in Mobis’s contents claim. I reject it as part of this claim.

This represents the cost of replacement of printers and photocopiers destroyed in the collapse. The amount was paid to the supplier of this equipment (which was leased to Mobis). It does not satisfy the sole purpose test.

This is a claim for the cost of catering which, according to Mobis’s submissions, was “incurred to keep staff motivated as they worked very long hours in the weeks following the storm” and was part of Mobis’s “effort to ensure the temporary warehouse was set up as quickly as possible”. It does not satisfy the sole purpose test.

This is an amount incurred to translate the Local Policy from English to Korean. It does not satisfy the sole purpose test.

This amount is said to relate to various items such as a Bureau of Meteorology Report and Work Health Safety costs and accommodation costs (evidently for a Korean lawyer). I am not satisfied these satisfy the sole purpose test.

Following the collapse of the warehouse, Mobis leased two temporary premises nearby. Mobis was not able to negotiate a lease term of less than 15 months from 6 May 2015 for one of those premises, and 18 months from 1 June 2015 for the other.

The issue which arises is whether the rental payments made by Mobis under the leases beyond the expiry of the Indemnity Period (on 25 April 2016: see cl 8.3 and item 1.14 in the Schedule) are covered.

The amount claimed by Mobis totals $1,562,547.

The cover under the Local Policy for increased cost of working was for additional expenditure necessarily and reasonably incurred for the sole purpose of avoiding or diminishing reduction in turnover which turnover would, but for that expenditure, have taken place “during the Indemnity Period”: cl 7.1.1(b).

The clause thus directs attention to expenditure made for the purpose of avoiding reduction of turnover that would otherwise have taken place “during the Indemnity Period”; that is, from the date of the collapse on 25 April 2015 to the end of the Indemnity Period 12 months later on 26 April 2016.

Whether or not Mobis, by executing the leases during the indemnity period, can be said to have “incurred” an obligation to pay rent beyond that period, I cannot see how the incurring of that obligation can be said to have been made for the purpose (let alone “sole purpose”) of avoiding a reduction in turnover within the period of indemnity. It may have been a necessary consequence (because a shorter lease term could not be negotiated) of the incurring of expense which was for that purpose; but it was not itself incurred for that purpose.

In closing submissions, I was informed that Mobis’s reduction in turnover all occurred within the period of indemnity. That may be how things turned out, but there is no evidence to which my attention has been directed that this was known or expected when these leases were entered into.

For that reason, my conclusion is that Mobis is not entitled to indemnity for this amount.

For the same reason, Mobis is not entitled to costs paid by it after 25 April 2016 to move back to its replacement warehouse.

Clause 7.1.1 of the Local Policy provides that there is to be deducted from the amount payable under the policy for loss of gross profit and increased cost of working, any “sum saved” during the period of indemnity:

XL contends that there are sums “saved” for the purpose of this provision in respect of depreciation ($1,449,509) and IT expenses ($287,642).

The issue is whether a depreciation expense in Mobis’s financial records which is no longer booked because the assets in question were destroyed is an “expense…payable out of gross profit” which has been “saved” for the purpose of cl 7.1.1 and should thus be deducted from the quantum of Mobis’s claim.

The amount in question is $1,449,509.

The depreciation expenses that Ms Daley identified in Mobis’s financial records relate to plant and equipment.

Mobis devoted a number of pages of its submissions to authorities dealing with the meaning of “gross profit”.

However, those authorities are of little assistance as the expression “gross profit” is defined in the Local Policy in cl 8.1 as follows:

The definition thus contemplates that, for the purpose of calculating gross profit, provision will be made for depreciation in opening and closing stock and work in progress. It seems to me to be beside the point that the depreciation in question here is in respect of plant and equipment, rather than stock. The point is that for the purposes of the definition of “gross profit” in the Local Policy, depreciation is taken into account.

The question is whether depreciation can be said to be an expense “payable out of” gross profit.

As Mobis has acknowledged, there is authority for the proposition that it is. In Synergy Health (UK) Ltd v CGU Insurance Plc (t/as Norwich Union) [2010] EWHC 2583 (Comm) Flaux J considered a policy which had terms indistinguishable from those under consideration here and concluded that depreciation was a sum “payable out of gross profit” that had been “saved”.

His Lordship stated:

Mobis submitted that Synergy Health should not be followed because:

But here, because of the definition in the Local Policy, gross profit is not profit before depreciation. It is profit which takes into account depreciation in respect of stock.

In any event, I find Flaux J’s decision to be directly on point and I am persuaded that I should follow it.

It follows that there has been a saving which must be taken into account.

If, as Mobis appears to submit, there is some controversy about the manner in which Ms Daley has assessed the amount of the depreciation expense, that matter can be referred out for separate determination.

The issue here is what savings Mobis made concerning IT by reason of the collapse.

An assistant accountant at Mobis, Mr Liang, gave unchallenged evidence that Mobis had five IT accounts, and that only one, the “network account” was directly connected to the warehouse collapse and that only the IT expenses recorded in Mobis’s “network account” were saved.

Ms Daley’s assessment of the amount of IT expenses saved ($287,642) was calculated on the basis that all five IT accounts were affected by the collapse of the warehouse. Mr Liang’s evidence is that this was not so, that the other IT expenses were not affected by the collapse and should not be treated as a saving.

Mr Johnson’s calculation is based upon that evidence. Mr Johnson assessed the savings at $52,185.

I propose to adopt that figure.

The parties raised numerous issues concerning the quantum of Mobis’s claim. I have endeavoured to address and decide all of the issues raised by the parties.

I now invite the parties to confer and agree as to whether any further issues remain for consideration and as to the monetary consequence of the decisions I have made.

I also invite submissions as to how the issues that I have not been able to deal with should be progressed.