New York’s Highest Court Interprets Specific Excess Insurance Policy Language to Require “All Sums” Allocation and Vertical Exhaustion of Underlying Policies

On May 3, 2016, the New York Court of Appeals responded to two questions certified by the Delaware Supreme Court and interpreted the language of certain excess insurance policies as requiring “all sums” allocation, and vertical, rather than horizontal, exhaustion of underlying policies at issue in asbestos exposure litigation pending in Delaware. In rendering its opinion in In the Matter of Viking Pump, Inc.1 the Court of Appeals noted that its decision applying the pro rata method of allocation in Consolidated Edison Co. v. Allstate Ins. Co. (“Con Ed”)2 did not adopt a blanket rule that pro rata is always the appropriate method of dividing indemnity among successive insurance policies.

The Viking Pump litigation was rooted in liabilities for asbestos exposure claims for which Viking Pumps, Inc. and Warren Pumps, LLC became liable after acquiring pump manufacturing businesses from Houdaille Industries in the 1980s.  Houdaille had extensive multi-layer insurance coverage, which included coverage under successive primary and umbrella excess policies issued by Liberty Mutual totaling approximately $17.5 million and $42 million, respectively.  Also in place were additional layers of excess policies providing a total of approximately $400 million in coverage.  As the Liberty Mutual policies neared exhaustion, the insureds commenced a declaratory judgment action to determine whether they were entitled to coverage under the additional excess policies and, if so, how indemnity should be allocated across the triggered policy periods.

The insuring language of the Liberty umbrella policies provided that the insurer “will pay on behalf of the insured all sums in excess of the retained limit which the insured shall become legally obligated to pay … as damages … because of … personal injury … with respect to which this policy applies and caused by an occurrence.” The policies defined personal injury as personal or bodily injury “which occurs during the policy period.”  The excess policies either followed form or provided for what the court deemed to be substantively identical coverage.

Most of the excess policies also followed form to a non-cumulation of liability (anti-stacking) provision in the Liberty umbrella policies, which provided:

If the same occurrence gives rise to personal injury … or damage which occurs partly before and partly within any annual period of this policy, the each occurrence limit and the applicable aggregate limit or limits of this policy shall be reduced by the amount of each payment made by [Liberty Mutual] with respect to such occurrence, either under a previous policy or policies of which this is a replacement, or under this policy with respect to previous annual periods thereof.

Those excess policies that did not follow form contained a similar two-part prior insurance and non-cumulation of liability provision.

It is agreed that if any loss covered hereunder is also covered in whole or in part under any other excess Policy issued to [the insured] prior to the inception date hereof the limit of liability hereon … shall be reduced by any amounts due to [the insured] on account of such loss under such prior insurance.

Subject to the foregoing paragraph and to all the other terms and conditions of this Policy in the event that personal injury … arising out of an occurrence covered hereunder is continuing at the time of termination of this Policy the Company will continue to protect [the insured] for liability in respect of such personal injury … without payment of additional premium.

In the underlying litigation the Delaware Chancery Court granted the insureds summary judgment, holding that New York law applied to the dispute, that the insureds were each entitled to coverage under the excess policies and that the proper method of allocation was the all sums approach. In response to the insurers’ argument that pro rata was the appropriate allocation method under New York law, the Chancery Court distinguished the policy language in this case from the language at issue in Con Ed, stating that the non-cumulation and prior insurance provisions in the Viking policies evinced a clear and unambiguous intent to use all sums allocation and that utilizing a pro rata allocation would render such provisions surplusage, contravening New York’s principles of contract interpretation.  Furthermore, even if the provisions were deemed to be ambiguous, the Chancery Court noted that the only substantive extrinsic evidence offered by the parties was that Liberty Mutual routinely allocated its liability under the policies under the all sums method in the past.  In addition, any ambiguity must be resolved in favor of the insured.

The matter was transferred to the Delaware Superior Court, where a trial was held and on appeal the Delaware Supreme Court concluded that resolution of the allocation and exhaustion issues depended on significant and unsettled questions of New York law. It thus certified two questions to the New York Court of Appeals:

  1. Under New York law, is the proper method of allocation to be used all sums or pro rata when there are non-cumulation and prior insurance provisions?
  2. Given the Court’s answer to Question #1, under New York law and based on the policy language at issue here, when the underlying primary umbrella insurance in the same policy period has been exhausted, does vertical or horizontal exhaustion apply to determine when a policyholder may access its excess insurance?

The New York Court of Appeals noted that it did not adopt a blanket rule in Con Ed that pro rata allocation was always the appropriate method for dividing indemnity among successive insurance policies.  Instead, it pointed out that the Con Ed court relied on general principles of contract interpretation and made clear that the contract language controls the question of allocation.  The policies at issue in Con Ed contained an agreement to indemnify the insured for “all sums” for which the insured was liable and which were caused by or arose out of an “occurrence” and that the policies provided for indemnity “for liability incurred as a result of an accident or occurrence during the policy period, not outside that period.”3  In the view of the Con Ed court, pro rata allocation, “while not explicitly mandated by the policies,” was consistent with the language used therein and that the use of the term “all sums” was insufficient to establish a contrary interpretation.4  Acknowledging that, in Con Ed, the court “suggested that, in the absence of language weighing in favor of a different conclusion, pro rata was the preferable method of allocation in long-tail claims,” the court recognized that different policy language might compel all sums allocation.

With regard to the policies at issue in Viking Pump, the court stated that by inclusion of the non-cumulation provisions and the two-part non-cumulation and prior insurance provisions, the language is “substantively distinguishable” from that interpreted in Con Ed and the arguments by the parties were likewise different.  In the view of the court, the excess policies at issue in Viking Pump “present the very type of language that we signaled might compel all sums allocation” in Con Ed. In the view of the court, it would be inconsistent with the language of the non-cumulation clauses to apply pro rata allocation because such policy provisions “plainly contemplate that multiple successive insurance policies can indemnify the insured for the same loss or occurrence….”  By contrast, “the very essence of pro rata allocation” is that the policy language limits indemnity to losses and occurrences during the policy period, meaning that no two successive policies would indemnify the same loss or occurrence.  The court explained that pro rata allocation is a legal fiction designed to treat continuous and indivisible injuries as distinct in each policy period, despite the fact that the injuries may not actually be capable of being confined to specific time periods.  The non-cumulation provision negates that premise by presupposing that two policies may be called upon to indemnify the insured for the same loss or occurrence.  Thus, pro rata allocation would render the non-cumulation provisions surplusage, thereby violating basic precepts of contract construction.

With regard to exhaustion, the court rejected the argument that the insureds were required to exhaust all triggered primary and umbrella policies “horizontally” before tapping into any of the additional excess insurance policies and instead held that the insureds could access each excess policy once the immediately underlying policies’ limits were exhausted (“vertical” exhaustion), even if the limits of other primary or umbrella policies issued for other policy periods had not yet been reached. The court was of the view that vertical exhaustion is more consistent with the language tying attachment of the excess policies specifically to identified policies that span the same policy period.  Furthermore, vertical exhaustion is more consistent with all sums allocation, permitting insureds to seek coverage through the layers of insurance available for a specific year.  In addition, the “other insurance” provisions of the excess policies did not support horizontal exhaustion because such provisions only apply when multiple policies cover the same policy period, not to situations involving coverage under policies issued for successive policy periods.

As a result of this decision, pro rata will continue to be the presumptive method of allocating indemnity obligations for long-tail claims under New York law, but that presumption can be rebutted by specific policy language.

2016 WL 1735790 (N.Y. May 3, 2016).

2 774 N.E.2d 687 (N.Y. 2002).

3 Con Ed, 774 N.E.2d at 695.

4 Id.