Icon v Liberty Mutual Insurance & QBE Underwriting and Opal Tower The Annual Contract Works Insurance Program Imbroglio Part 2
Links to the Judgements
Primary Court
Full Court
Disclaimers
The author has not had access to the significant body of evidence presented to the Court, relying on what appears in the judgements and industry knowledge (e.g. that cover limitations on activities/contracts insured apply equally to annual contract works material damage and liability policies).
Absent that detail, no firm conclusions can be drawn on individuals’ positions before or during the Federal Court of Australia hearings. Rather, when making comparisons, comment is focussed on the systems (or lack thereof) and how the involved entities should have acted in market context.
I do observe that no one fully understood how Icon’s insurance program (as written) worked. And, having struggled to follow what went on, I have some sympathy for those having to reconstruct the happenings.
The inordinate complexity of this case means I may have misinterpreted players actions or judges’ approach but, hopefully, still produced a balanced article that gets to the nub of the insurance issues.
What I can more confidently say is that this case should never have got to court and I suspect, if there are many more like attempts to avoid liability, courts/legislatures will have more to say about effective use of court time and taxpayer money, and the responsibility to honour claims in the first instance – so that insureds don’t get caught up in fine policy design and intermediary agency issues.
With my focus being on the insurance program, no conclusions are drawn regarding Icon’s entitlement to an indemnity nor the performance of the Opal Tower design & build contract and the systems used.
Introduction
Conscious of subject matter dryness and the challenges in holding readers attention I have split this article into 3 parts, to be delivered sequentially –
1.??????Setting the Scene – backgrounding the players, exploring the evolution of contract works insurance, and detailing how Icon’s ‘hybrid’ contract works insurance program differed.
2.??????What Went Wrong – listing the panoply of factors that created the slippery slope, starting with that hybrid program and ending in the liability insurer denying indemnity.
3.??????The Products/Completed Operations Conundrum – a sidebar issue for the court but, from a policy drafting perspective, a critical component.
Reminder: Icon went to the Federal Court of Australia for a judicial interpretation of the Liberty Mutual and Lloyds public/products liability policy wordings. They did not seek a ruling on whether their claim(s) arising from the 24 December 2018 Opal Tower damage were covered.
Part 2 - What Went Wrong
The Slippery Slope
In the intervening years between the hybrid decision and the Opal Tower claim, competitive drivers, inept communication, inadequate structures, incomplete processes, knowledge gaps, poor record-keeping and power imbalances all played out, creating a background where ‘commercially sensible’ and ‘objective intentions’ – tests the primary judge was obliged to explore as part of his rectification approach - were hard to find. Those tests, of course, assume the parties, especially the insurer, have the necessary skills to translate that intention into the policy wording. This and other recent cases[1] shows how flimsy that assumption can be.
Rather than a narrative, the following list better paints a picture of the panoply of factors in play which, to a greater or lesser extent, combined to result in two court sittings and 250 judgement pages of discursive complexity and mind-numbing detail that hasn’t really resolved matters.
Involving technical, transmission and transactional issues, they include:
·?????Unaligned cover requirements – Austbrokers seem to have wanted (upon 1st renewal) both contracts transferred and contracts commenced (depending on size/length of contract) under the one policy, with Chase reflecting that in their communications to insurers when contracts were notified despite, at least, the Liberty policy wording saying otherwise.
·?????Quote and placement inconsistencies – using broad terminology and loose expressions Austbrokers asked for one thing, confirmed another to their client, while the quote response and subsequent placement delivered something different again (what was viewed as relatively minor differences and inconsequential cover summaries becoming major issues).
·?????Different policies/insurers - separate contract works (material damage) and public liability (third party liability) policies with different insurers, masked by Austbrokers conflating the covers into one contract works program in its dealings with Icon and Chase, went against best practice and, if unavoidable, the need to be very precise in aligning covers.
·?????Disconnected insuring clauses – policy wordings that strayed from their jurisdictional antecedents and were not fit-for-purpose with operative phrases residing in conditions, extensions and endorsements and/or not connected back to the policy wordings’ main operative clause.
·?????Overlapping conditions and definitions – that define cover scope (physically and temporally) and control turnover/contract declarations captured by that scope, were not clearly separated section by section or from premium calculation requirements.
·?????Unclear cover periods - no clear distinction between policy period (the relevant annual insurance period) and period of insurance[2] (for each contract) in dealings surrounding the policy, imperfect alignment of cover periods between the MD and TPL policies because they were completely de-linked and, as a result, treatment of maintenance period cover in the TPL policy(s) that departed from standard market practice.
·?????Unrefined administration protocols –
o?the delivery and handling of underwriting information was not properly established, e.g. a standard framework for advising individual contract details,
o?administration matters like premium charging should have been kept clearly separate from policy coverage matters,
o?record-keeping was poor, details of conversations not recorded or confirmed in writing, and a history of why the insurance is what it is, the advice given in formulating the program, and the reason for complexity vs simplicity or variations from the standard, was patchy.
·?????Disjointed declaration process – instead of ‘declaring’ contracts at end of the policy period, Icon/Austbrokers adopted a combined, contract-by-contract, ‘notification’ approach to Chase despite the separate insurers for MD and TPL, to which Liberty/Chase responded by way of (probably unnecessary) policy endorsements; a practice that was never fully explained.
·?????Ambiguity in labels and differing terminology –
o?construction terms used interchangeably, e.g. contracts v projects, practical completion v contract completion (handover), defects period v maintenance period (noting construction contracts often conflate the two even though liability for defects extends well beyond any contractual maintenance period),
o?Insurance terms expressed inconsistently or without basis, e.g. period of insurance, policy period, period of cover and (estimated) contract period, commencing rather than commenced (yes, there is a difference in the participles, incl. when it comes to timing of declarations),[3] and mixed up with insurance industry jargon (e.g. ‘run-off’ for contracts commenced, ‘turnover’ for contracts transferred, and ‘turnover’ also used to represent ‘contract value’) – leaving open the opportunity to “repurpose the meaning” of those terms.
·?????Confused endorsements –
o?often used cover periods, scope references and terminology (e.g. ‘floater’), inconsistent with the policy, construction sector practices and standard contract requirements,
o?used words interchangeably, e.g. ‘amended’ and ‘endorsed’, and ‘lazy’ language (e.g. by not stating which part/fully restating the part of the policy being changed, and why),
o?made unnecessary cover changes more the province of certificates of insurance (e.g. added Insureds who were already covered by general definition),
o?mixed (necessary) cover changes, e.g. undefined Insureds and contracts outside policy parameters, with underwriting matters such as location/risk/exposure information (which should have had a separate mechanism for advising the insurers),
o?compromised the operation of Condition 15 and, particularly, the requirement that terms and conditions existing upon expiry of the current policy period apply to any contracts declared as requiring ‘run-off’ cover,
and, critically from Liberty’s perspective,
o?added contracts falling outside policy cover parameters into the current annual policy without clearly separating policy amendments, e.g. changes to the cover parameters from the administrative requirements of Condition 15 to engage (‘run-off’) cover.[4]
·?????Tortuous certificate of insurance (CoI) constructs – for each declared contract a CoI was issued by Austbrokers[5] to Icon combining both MD and TPL covers and containing estimated project periods, maintenance period (sometimes), values, etc. Like the endorsements, some terms and phrases were inconsistent with the policy wordings. It was as though Chase/Liberty felt a contract-specific endorsement to the policy was necessary before Austbrokers could issue a contract-specific CoI, whereas, in normal use, a CoI is simply a truncated document catering for a particular contract and evidencing the existence of a policy that covers it.
·?????Imprecise cover transfer - within/between policies, especially in relation to the products/completed operations transition in the Liberty and Lloyds (and QBE/Lloyds) TPL policies, but also in relation to what work/contract activity was picked up by the following policy, e.g. “clean-off” for Liberty and “dirty-on” for Lloyds creating cover duplication, and a combined construction/maintenance insurance period for both Liberty and Lloyds creating a cover gap.
Along with unclear cover periods, what set the scene for the imprecise cover transfer point was Liberty and Lloyds TPL policies purporting to be North American Commercial General Liability (CGL) wordings but missing at least two key elements of that framework –
1.??????a clear temporal definition that separated construction operations (pre-handover) from that of the maintenance period (post-handover) and the cover scope for each – something that may have been avoided if the MD and TPL cover components remained within one policy wording,[6]
and
2.??????product and completed operations definitions that created a physical separation of completed buildings (or, to use the North American term, “real property”) from other products - as the TPL cover, in lockstep with the MD cover, transitioned from construction operations to completed operations (with completed operations including any maintenance period activity).
Note: The policies were also missing the physical separation of business operations (general activity that enveloped on-site construction work, incl. pre-construction preparation) from construction operations.
What all the parties involved wound up with - to coin the words of Liberty when responding in court to Austbrokers view of the CW insurance program – something “so confused as to be almost meaningless”, characterised by “imprecision”, “looseness”, “incoherence” and a “muddled interpretation”. [7]
That none of the parties undertook a sweep-up exercise to confirm the cover required by Icon had been achieved, where any differences lay, and how it was to be administered, reflects how far market discipline has been eroded.
And, that neither intermediary, Austbrokers or Chase, were aware enough to recognise the limitations, nor confident enough to stand firm with their respective principals, meant they left themselves exposed to the almost genetic disposition of insurers to look for claim ‘outs’ - especially when they realise they’ve made a mistake and/or big bucks are involved. ??
Icon v Liberty Mutual & QBE/Lloyds – the case as presented and determined
For the plaintiff, Icon, the key declarations sought from the court were –
A.???The Opal Tower contract had been properly declared to Liberty under Condition 15.
B.???The 24/12/2018 damage was an Occurrence within the “period of cover” of the Liberty TPL policy, i.e. the 20/9/2015-16 period,
or in the alternative,
C.???The Lloyds TPL policy, i.e. the 20/9/2018 – 31/12/2018 period, responded to the Occurrence (Opal Tower having become a ‘product’ at that point).
Icon did not ask the court to rule on the indemnity, i.e. what they could claim, so, the Federal Court of Australia action should be seen as just the opening skirmish.
However, Icon’s battle plan suffers from a serious flaw – an assumption that the form of both Liberty and Lloyds policies, and the way they were managed, effected an unbroken transition of cover between policies for the activity surrounding Opal Tower, and the transition of cover between policy sections happened seamlessly. As will be seen, this was not the case.
The primary judge made the right decision when favouring Icon with all 3 declarations; holding that mistakes of omission or commission should not deprive an insured of a cover all parties involved bar Liberty and Lloyds thought they had. While he viewed Condition 15 as ambiguous, he was less than enamoured with Liberty’s defence (feeling it was a repurposing of cover to suit their stance and finding their decision not to call a Chase witness to support their position, telling) and decided policy rectification was the answer – Liberty having, in all practical respects (incl. via its agent, Chase) treated the cover as “contracts commencing”. Note: The Lloyds policy decision is covered later in this article.
The judge’s frustration was also evident from the considerable overreach (incl. tapping into obiter comments in other judgements and using S54 of the Insurance Contracts Act as a defence) in Liberty’s arguments[8] and it was obvious he viewed much of it as nonsensical.
The full court, in determining that Condition 15 did apply to the Opal Tower contract, i.e. it was a “contracts commencing” cover, found policy construction was the way to go; the Opal Tower contract held to be covered by the Liberty 2015/16 policy, not the Lloyds policy (because the contract was still within the “Construction Period” and, therefore, did not engage the Lloyds policy product/completed operations sub-section ) – thus knocking out any meaningful cover for Icon; particularly as the construction period for the MD section of the CW insurance program had expired.
Having not been asked for such a ruling, the judges did not canvass in depth the temporal cover transition rules, i.e. the transition into the maintenance period switching cover from the MD policy to the TPL policy general liability section, and from TPL general liability to the product/completed operations section).
Where insurers wish to place time or value limits on the definition of the insured interest, they normally do so by placing those limits in the policy schedule (and referring to those limits in the definition, e.g. of Insured Contracts). That was undoubtedly what the policy schedule section headed ‘Construction Period’ in Lloyds endorsement TPL008 was meant to be but the full bench treated it as a definition in its own right, which had the wholly unintended consequence of altering the insurance periods for contract from –
·????????date contract commenced until date of handover, to
·????????date contract commenced until expiry of the maintenance period.
Also, the judgements did not address how the hybrid arrangement had effectively separated contracts under and over $50m[9] - the latter covered at Liberty’s discretion and only then by the specific terms of the related endorsement to the annual policy (rather than at expiring terms as per Condition 15).
Assuming the above contract split did occur, cover for the Opal Tower contract could not have been invoked via Condition 15 as, due to the policy parameters, Opal Tower was not automatically covered. When Chase/Liberty were asked to cover the Opal Tower contract it wasn’t a declaration or even a notification, it was a request.
As there was not enough focus on identifying and resolving the operative clauses/definition issues, the Conditions 8 and 15 inconsistencies relative to market practice, and sum insured/maximum contract value missteps arising from the original decision of the parties to exercise a hybrid model, Icon lost the opportunity for a broader and deeper consideration. As judges can only deal with what is in front of them, one has to ask if law courts are the best forum for complex policy wording issues?
While the hybrid problem was ultimately recognised and (partially) fixed by Chase/Lloyds adopting a contracts commenced approach in the?2017/18 TPL policy, even that wording has its issues, especially the ‘Construction Period’ definition and silence on excluding contracts entered into prior to policy commencement, i.e. being ‘clean on’.
Whether created by design or neglect, the artificiality of the Icon CW insurance program’s insured interest definitions and cover transition arrangements should have been apparent. And, whether or not the hybrid approach was a pre- or post-loss construct by Liberty, the fact that Liberty and Lloyds did not tell the others that the TPL part of the CW insurance program was not the hybrid arrangement Icon asked for, counted heavily against their arguments - in equity, equality, fairness and just plain good sense.
But, because Icon effectively got a wording rectification to the Liberty TPL policy that introduced a “Construction Period” definition which, against accepted market practice, included the maintenance period, both Lloyds and Liberty policies were infected with the same transitional virus that precluded products/completed operations cover engaging because the maintenance period had not expired.
Had Icon not tried to ‘innovate’ their CW insurance program in the first place, and had they not sought to change the Liberty TPL policy operative clause, the outcome would probably have been quite different. ?
There has to be a way to ensure that a cover circumstance never intended can be corrected. Perhaps what’s needed for policy interpretation matters is a binding arbitration process involving insurance experts that precedes high bar access to the law courts and, then, only on questions of law?
So, although the full court has found that only the Liberty 2015/16 policy period, not the Lloyds 20/09/2018 – 31/12/2018 policy period, applies, the available indemnity remains hanging.
Unfortunately, I expect this will involve more angst as both the general and products liability insurance cover sections only cover damage caused by, not to, the defective product (Opal Tower) – the MD cover or the principal’s/strata corporation’s building insurance only responding to damage flowing from a defect; the defect itself the province of guarantee/warranty insurance.
At the very least, especially if the majority of Opal Tower’s cracked building components were faulty (Exclusion 5 - Defects) or under-engineered (Exclusion 6.2 – Representation or Warranty), a big proportion of the $17m repair costs still won’t be indemnifiable by either section of the Liberty policy.
And it would be too long a bow to draw to suggest any meaningful amount of the repair cost, loss of use, alternative accommodation and loss of amenity/value claims arising out of the building insurer’s subrogation recovery or the class action against the developer and builder, will ultimately be indemnified by the Liberty TPL policy – other exclusions likely to be vigorously explored, e.g. loss of use, products guarantee, breach of professional duty, etc.
Whether Icon has other insurance policies that could pick up some of the slack remains to be seen.
Stay tuned for Part 3.
---------------------------------------------------------------------------------------------------------------
Footnotes
[1] e.g. insurers response to business interruption claims from the CoViD-19 pandemic.
[2] In connection with the Icon CW program, ‘period of insurance’ was used to describe the policy period (at least until Icon proffered a rectification re-draft that introduced a disconnected “construction period” definition. This was to have material consequences for how Icon’s TPL policy wordings were interpreted when the primary judge’s decision went to appeal.
[3] Contracts commenced cover applies only to those contracts where work actually starts in a particular policy period, not contracts likely to commence (“contracts commencing”) in that period. The fine difference lies in the declaration requirement – either worded to require declaration during the policy period (commencing) or at the end (commenced). While the Icon Liberty TPL policy Condt. 15 was ambiguous (in that it required declarations to occur during the policy period rather than at the end), it retained all the hallmarks of a contracts transferred cover (with a ‘run-off’ option). Note: See the judges’ reference to the use of past participles in the definition of a product – central to the timing of cover transition between work being done and work completed in liability insurance.
[4] Which, in any event (on the true reading of the policy wording), only applied to contracts ≤$50m. Had Liberty done this, or Chase/Liberty followed good practice, clear endorsements or separate policies would have been issued and some of the mess avoided.
[5] In neither judgement was it made clear whether Austbrokers were agents of Chase/Liberty for this purpose.
[6] ‘Construction Period’ being a defined term solely for Section 1 Material Damage cover – Section 2 Third Party Liability being a cover for all activity irrespective of any contract’s construction period (the only reference to that period in the liability section being the end point at which cover transitions between the general liability and products liability sub-sections.
[7] Para 217 of the primary judgement.
[8] As there was, also, in QBE Underwriting’s arguments against the 2nd limb of Icon’s request, the application of products (incl. completed operations) liability to the occurrence. The extent to which QBE’s approach muddied things was summarised by the primary judge in the 2nd sentence of para. 313 – worth a read to reinforce this point and further referenced later on in this article.
[9] See paras 175 - 177 of the appeal judgement bearing in mind the judges were constrained by the pleadings and evidence and that the author has assumed that a maximum contract value parameter existed.