THE RULES OF INSURANCE POLICY EXHAUSTION. By Mary E. Borja, Partner, Wiley Rein LLP

THE RULES OF INSURANCE POLICY EXHAUSTION By Mary E. Borja, Partner, Wiley Rein LLP I. INTRODUCTION Excess insurance policies generally attach after exhaustion of underlying insurance. Exhaustion must take into account both allocation of loss from open or pending claims as well as erosion of underlying limits from prior claims under the relevant policies. This paper discusses the rules for apportionment of loss, the order in which policies are attached, proof of erosion of policy limits as a result of earlier losses, and related issues. A. Allocation Methodologies The rate of exhaustion of any policy or tower of coverage will depend on the allocation methodology. 1. Joint and Several or All Sums Liability Policyholders usually argue that each insurer whose policy has been triggered is jointly and severally liable for all damages up to the insurer s policy limit. The argument for asserting joint and several liability under a general liability policy is that it is required by the language in the policy requiring the insurer to pay all sums which the insured becomes legally obligated to pay. Policyholders argue that, once an insurer pays the full extent of the insured s damages, subject to the limit of liability, it is then up to that insurer to seek contribution from other insurers whose policies are triggered by the loss. Under the all sums approach, also known as a vertical spike, the policyholder can target a selected primary policy among the multiple years triggered (typically one with a lower deductible, higher limit or remaining limit, or where the primary insurer is still in business as opposed to other years where the primary insurers are now insolvent). This shifts to the selected primary insurer the obligation to seek contribution and the risk that other insurers have gone out of business or cannot be identified. Notably, New York s highest court, in In re Viking Pump, Inc., 52 N.E.3d 1144 (N.Y. 2016), recently embraced an all sums allocation methodology for certain policies. In Viking Pump, two pump manufacturers faced asbestos-related bodily injury claims with exposure during their predecessor s ownership. The successors sought coverage under predecessor s primary and 2017 Wiley Rein LLP

excess insurance policies in effect from 1972 to 1985. Although New York s highest court had previously adopted pro rata allocation for long tail losses in Consolidated Edison Co. of New York v. Allstate Insurance Co., 774 N.E.2d 687, 693-94 (N.Y. 2002), in Viking Pump, the court answered a certified question by the Delaware Supreme Court by distinguishing Consolidated Edison and explaining that the non-cumulation clauses in the particular policies at issue could not be reconciled with pro rata allocation. The court stated that non-cumulation clauses would be rendered surplusage in a pro rata allocation, a construction that cannot be countenanced under our principles of contract interpretation. Viking Pump, 52 N.E.3d at 1154. Accordingly, the court held that an all sums allocation method applied in light of the particular policy language. 2. Pro Rata Allocation Many jurisdictions allocate liability pro rata across all time periods implicated by the loss. The amount allocated to each period depends on whether the loss is apportioned based upon time on the risk or by limits or by premiums paid. Time on the risk means that, for example, if an insurer issued policies for two out of five triggered policy years, that insurer would have a 40% share of liability. In other words, the insurer was on the risk for 40% of the triggered period. By limits takes into account the total limits of liability of the insurer that issued the triggered policies. Alternatively, using premiums paid recognizes that a high level excess insurer that received relatively lower premium for its policy than an insurer that issued a policy for the same limit of liability but at a lower layer and for greater premium should not be treated equally, even though their limits are the same. Instead, the higher layer excess insurer that received the lower premium would also have allocated to it a lower proportion of the total loss amount. Whichever pro rata allocation method is used, it is often based upon the recognition that occurrence-based liability policies afford coverage only for the damages that take place during the policy period. While policyholders point to the all sums language in some policies to advocate for joint and several liability instead of pro rata allocation, the phrase all sums is expressly qualified by the damages to which this policy applies, if the all sums language even appears in the policy at issue at all. By way of example, in the Order Regarding Allocation, Radiator Specialty Co. v. Fireman s Fund Insurance Co., No. 13 CVS 2271 (N.C. Super. Ct. Jan. 28, 2016), the court allocated both defense costs and indemnity for benzene-related disease claims and asbestos claims using a pro rata time-on-the-risk allocation method. The court held that the pro rata allocation most reasonably interprets the policy language in the subject policies, requiring that sums paid by an insurer apply to those injuries occurring during that insurer s policy period. The court further held that the policyholder was responsible for its pro rata share of defense and indemnity costs where there has been settled, insolvent or lost policies, as well as periods where [the policyholder] was uninsured, underinsured or self-insured. Similarly, in Arceneaux v. Amstar Corp., 200 So.3d 277 (La. 2016), the Louisiana Supreme Court recently held that in long latency disease cases, the defense cost amounts, like indemnity, 2017 Wiley Rein LLP 2

should be prorated between the insurers and the insured where the policy provided coverage for only a portion of the time during which the exposure occurred. The policy at issue provided that the insurer had the right and duty to defend any suit against the insured seeking damages on account of... bodily injury. The policy defined bodily injury as bodily injury, sickness or disease sustained by any person which occurs during the policy period, including death at any time resulting therefrom. The Arceneaux court had previously held that [F]or long-latency occupational claims involving long-term exposure, liability should be allocated on a pro rata basis over all periods in which the exposure took place, including years in which the insured was uninsured. For uninsured periods, the insured is treated as being self-insured and is assigned a pro rata share. In its 2016 decision, the court further held that the insured would be required to pay for its defense costs allocable to those years when it did not obtain a policy that would afford a defense for the underlying claims. As suggested by the decisions in Radiator Specialty and Arceneaux, the rate of exhaustion of triggered policies will further depend upon whether and to what extent the insured bears a share of liability for periods when it was self-insured, had inadequate limits, or where it purchased insurance from a now-insolvent insurer. In cases where the insured is found to be self-insured for part of the relevant period, the weight of the authority is that the insured must bear a pro rata share of the defense costs. In both Radiator Specialty and Arceneaux, discussed above, the court apportioned loss, including defense costs, to the policyholder for uninsured periods. But at least one court has warned that an insurer who aggressively pursues allocation of loss to the insured where the insurer s policy has a duty to defend runs the risk of a bad faith claim. See, e.g., Ray Indus., Inc. v. Liberty Mut. Ins. Co., 974 F.2d 754 (6th Cir. 1992). 3. Equal Shares Allocation A minority of jurisdictions allocate costs in equal shares among all insurers on the risk. For primary policies with a duty to defend, courts may adopt this approach because each insurer assumed an obligation to defend and therefore arguably should bear an equal share of the defense costs. Some courts also have used equal shares allocation of defense costs as an equitable approach when a pro rata approach is deemed unworkable. See, e.g., Travelers Indem. Co. of America v. AAA Waterproofing, Inc., No. 10-CV-02826, 2014 WL 201726, at *3 (D. Colo. Jan. 17, 2014) ( While equal shares allocation has been applied in a minority of jurisdictions, the circumstances of this case in which each subcontractor/insurer had a joint and several duty to provide a complete defense to DRH, and no reliable method exists to tie allocation to liability or policy limits militates in favor of a holding that those subcontractors/insurers responsibility for equitable contribution to DRH s defense costs should be apportioned evenly. ). However, equal shares allocation is typically used, if at all, only for defense costs. Where defense costs are outside of limits, this method of apportionment would not affect the rate of exhaustion and attachment of the excess policies. 2017 Wiley Rein LLP 3

B. Attachment of the Excess Layers The important divide, with respect to attachment of excess policies, is essentially between two camps. Some courts hold that all lower layers must be exhausted before to proceeding to any higher excess layer (horizontal allocation). Other courts hold that an excess insurance policy can be attached as soon as the policy immediately below it has been exhausted, regardless of whether primary policies or lower level excess from other periods have been exhausted (vertical allocation). Put another way, with horizontal allocation an insured typically spreads a loss across all triggered policy years at the same rate; with vertical allocation, an insured works its way up a single tower of coverage before moving to another tower. 1. Horizontal Allocation Horizontal allocation in its most basic form requires that every triggered primary or lower level excess layer be exhausted before any secondary or higher layer can be attached. See, e.g., Pacific Coast Bldg. Prods., Inc. v. AIU Ins. Co., 300 F. App x 546, 548 (9th Cir. 2008) ( Horizontal exhaustion applies where an excess policy, by its terms, is excess to all underlying insurance. ). The argument for horizontal allocation is typically founded in the excess insurance language of the policies. Excess insurance policies that state that they are excess of all other collectable insurance, as opposed to only a specifically identified lower-layer policy or policies, may be interpreted as requiring exhaustion not only of the lower level policies in a particular tower, but all other collectable insurance, and thus support depleting all available lower layers of insurance before attaching that policy. For example, in Community Redevelopment Agency v. Aetna Casualty & Surety Co., 50 Cal. App. 4th 329 (Cal. Ct. App. 1996), as modified (Nov. 13, 1996), the excess policy stated that it was excess not only to a specified primary policy but was also excess to the applicable limits of any other underlying insurance collectible by the [insured parties]. Id. at 338 (emphasis omitted). The court held that [a]bsent a provision in the excess policy specifically describing and limiting the underlying insurance, a horizontal exhaustion rule should be applied in continuous loss cases. Id. at 340; see also Mayor & City Council of Balt. v. Utica Mut. Ins. Co., 802 A.2d 1070, 1105 (Md. Ct. Spec. App. 2002) (holding that horizontal exhaustion would apply unless the language of the excess policy states that (1) it is excess insurance over a particular, specific, primary policy, and (2) will be triggered when that discrete policy is exhausted ). Horizontal allocation may not spread costs in an even bathtub fashion in all cases. Lower layer policies may deplete at different rates depending on the specific allocation method as well as the application of aggregate limits of liability. In National Union Fire Insurance Co. of Pittsburgh, Pa. v. Porter Hayden Co., Civil Nos. CCB 03 3414, CCB 03 3408, 2014 WL 43506 (D. Md. Jan. 2, 2014), the court granted the insurers motion seeking application of horizontal exhaustion, but denied the insurers request to find that excess coverage had not been triggered. In this case, not all of the primary policies contained an aggregate limit of liability. The court found that certain excess policies, therefore, had not attached based upon lack of exhaustion of primary policies that lacked aggregate limits. Thus, the pro rata allocation and 2017 Wiley Rein LLP 4

horizontal exhaustion under these circumstances resulted in certain excess layers attaching before the exhaustion of all the triggered primary policies. The policy year towers depleted at different rates, depending on the presence or absence of aggregate limits. Additionally, even when horizontal allocation applies to the primary policies, it does not necessarily apply to the excess layers. So, while a court might require that all primary policies must be exhausted prior to attachment of any excess policies, it might not require that all firstlayer excess policies have to be exhausted prior to attachment of second layer policies. In In re Viking Pump, Inc., the Court of Appeals of New York adopted vertical exhaustion for excess layers even though horizontal exhaustion had been applied to primary layers. See Viking Pump, 52 N.E.3d at 1148-49. 2. Vertical Allocation Vertical allocation allows the attachment of an excess policy as soon as the policy immediately below it has been exhausted. Some courts that have applied a vertical allocation have concluded that it is dictated by the excess policy language at issue. If the excess policy specifies that it is excess to an identified and particular underlying policy then a court may be more likely to conclude that once that identified policy has been exhausted, the excess policy attaches. For example, in Westport Insurance Corp. v. Appleton Papers, Inc., 787 N.W.2d 894 (Wis. Ct. App. 2010), the Wisconsin Court of Appeals held that the excess policies were subject to vertical exhaustion. The court reasoned that [t]he excess insurers wrote their policies based on other policies providing coverage beneath them in that particular policy year; the excess policy thus required exhaustion only of the policies below them in that particular policy year before the attachment point for which each had been compensated was reached. Horizontal exhaustion is thus not consistent with policy provisions, but vertical exhaustion is consistent. Id. at 918. New York s highest court has concluded that vertical exhaustion is more consistent than horizontal exhaustion with language tying attachment of the excess policies specifically to identified policies that span the same policy period. Viking Pump, 52 N.E.3d at 1156. The excess insurers in that case argued that other insurance provisions of the excess policies dictated horizontal exhaustion. Id. at 1156-57. Declining to apply horizontal exhaustion to the excess policies, the Court explained that under New York law other insurance clauses are not implicated in situations involving successive as opposed to concurrent insurance policies. Id. at 1157. The court stated that, in light of the language in the excess policies tying their attachment only to specific underlying policies in effect during the same policy period as the applicable excess policy we conclude that the excess policies are triggered by vertical exhaustion of the underlying available coverage within the same policy period. Id. The Viking Pump court also concluded that vertical exhaustion is conceptually consistent with an all sums allocation.... Id. at 1156. In contrast, in U.S. Gypsum Co. v. Admiral Insurance Co., 643 N.E.2d 1226, 1261 (Ill. App. Ct. 1994), an Illinois Appellate Court interpreted an other insurance clause as support[ing] an interpretation that this policy serves as an excess policy to all triggered primary policies, regardless of whether they extend over multiple policy periods or only one. 2017 Wiley Rein LLP 5

C. Additional Issues 1. Proof of exhaustion or erosion of policy limits Of course, exhaustion of the underlying insurance policies limits of liability may be demonstrated through proof of payment of the policies limits of liability for covered claims. Typically, this evidence is adduced through proof of claim payments and/or final settlements of prior claims. See, e.g., R.T. Vanderbilt Co., Inc v. Hartford Accident & Indem. Co., No. X02UWYCV075016321, 2014 WL 1647135, at *21 (Conn. Super. Ct. Mar. 28, 2014) ( [T]he court finds... the Hartford... and the CNA primary policies... to be exhausted [because] both Hartford and CNA have paid the limits of their primary policies.... ). Even where actual payments have been made, a party challenging exhaustion under the purportedly exhausted policies may dispute whether the payments were made in good faith or possibly to prematurely end the duty to defend. A party may also challenge whether the correct allocation methodology was used. See, e.g., Royal Indem. Co. v. C.H. Robinson Worldwide, Inc., No. A08-0996, 2009 WL 2149637 (Minn. Ct. App. July 21, 2009) (holding that an excess insurer may challenge whether underlying policies were exhausted through payment of covered loss, including reasonable defense costs); but see Continental Cas. Co. v. BorgWarner Inc., 04 CH 01708, at 15 (Ill. Cir. Ct., Cook Cnty., Oct. 8, 2015) (holding that the upper level excess insurers could not second-guess the lower level umbrella and excess insurer s settlement decisions, explaining that an excess insurer does not have the right to challenge exhaustion of underlying insurance by substituting its own analysis or policy requirements for an underlying insurers settlement decisions ). 2. Can a settlement of less than full limits fully exhaust a policy? A frequent issue with respect to attachment of excess policies is whether the underlying policies can be exhausted by payment of something less than the full underlying limits of liability. Absent actual payment of the limits of liability of the underlying by policies, an excess insurer may argue that its obligations under the excess policy are not triggered. In response, the policyholder would assert that actual payment is not required, nor even that payment be for covered loss under the underlying policies, and that, therefore, it should be able to fill in any gaps on the underlying insurance. The particular exhaustion language in the excess policies may lead to different results. A policyholder is more likely to be permitted to fill the gap through its own payments where the excess insurer s policy does not explicitly state that underlying limits must be paid in full and by the underlying insurer. Zeig v. Massachusetts Bonding & Insurance Co., 23 F.2d 665 (2d Cir. 1928), applying New York law, is oft-cited. In Zeig, the insured had three underlying primary burglary insurance policies that provided $15,000 total in coverage, but which paid out only $6,000 in connection with the loss at issue. Id. at 665. The excess policy provided that it shall apply and cover only after all other insurance herein referred to shall have been exhausted in the payment of claims to the full amount of the expressed limits of such other insurance. Id. Because the insured settled the claim under the primary policies for less than the full value of those policies, the excess 2017 Wiley Rein LLP 6

insurer denied coverage. Id. The appellate court held, however, that the excess insurer should have been held liable for the amount of loss in excess of the $15,000 primary limits. Id. at 666. The court explained that the excess insurer had no rational interest in whether the insured collected the full amount of the primary policies, so long as it was only called upon to pay such portion of the loss as was in excess of the limits of those policies. Id. According to the court, [to] require an absolute collection of the primary insurance to its full limit would in many, if not most, cases involve delay, promote litigation, and prevent an adjustment of disputes which is both convenient and commendable. Id. The court pointed to the fact that the excess policy did not require the collection of the full amount of the limits of the primary policies and that payment could mean the satisfaction of a claim by compromise in reaching its decision. Id. However, the Zeig court noted that excess insurers may have exhaustion based defenses to coverage but only... when the terms of the contract demand it. Id.; see also Lexington Ins. Co. v. Tokio Marine & Nichido Fire Ins. Co., No. 11-civ-391(DAB), 2012 WL 1278005, at *4 (S.D.N.Y. Mar. 28, 2012) (applying Zeig and holding that an excess policy provided coverage for amounts in excess of the primary policy s limit even when the primary insurer paid less than its full limit because [i]n the absence of unambiguous language requiring exhaustion via full payment of the underlying policy, no such exhaustion is required ); Plantation Pipe Line Co. v. Highlands Ins. Co., 444 S.W.3d 307 (Tex. App. 2014) (holding that the policyholder s own payments in addition to the payments made by the underlying insurers for the remediation costs, allowed the policyholder to attach the third layer excess insurer s policy). Although Zeig has been widely cited, the decision has since been limited to the first party insurance context. In Ali v. Federal Insurance Co., 719 F.3d 83 (2d Cir. 2013), the Second Circuit rejected the application of Zeig in the context of excess liability policies. Giving effect to the exhaustion language in the excess insurance policies, the court held that the plain meaning of the phrase payment of losses refers to the actual payment of losses suffered by the Directors not the mere accrual of losses in the form of liability. Id. at 94. The court rejected the policyholder s argument that the excess policies were triggered once the total amount of defense costs and indemnity exceeded the underlying limits, regardless whether such amounts were actually paid. Not surprisingly, a New York bankruptcy court recently observed that Zeig s continuing vitality is open to question following the Second Circuit s decision in Ali v. Federal Insurance Co., 719 F.3d 83 (2d Cir. 2013). In re Rapid-American Corp., No. 13-10687 (SMB), 2016 Bankr. LEXIS 2224 (Bankr. S.D.N.Y. June 7, 2016). Consistent with Ali, a Michigan federal court interpreted the phrase payment of losses to mean that actual payment of losses by the insurer was necessary to trigger the excess coverage. Comerica Inc. v. Zurich Am. Ins. Co., 498 F. Supp. 2d 1019 (E.D. Mich. 2007). The court reasoned that settlements that extinguish liability up to the primary insurer's limits, and agreements to give the excess insurer credit against a judgment or settlement up to the primary insurer's liability limit are not the same as actual payment. Comerica, 498 F. Supp. 2d at 1032. Likewise, in Citigroup, Inc. v. Federal Insurance Co., the Fifth Circuit, applying Texas law, found that numerous excess-coverage provisions unambiguously required the underlying insurers to pay the full amount of the underlying limits. 649 F.3d 367, 371 73 (5th Cir. 2011). The excess policies issued by four of the participants in a quota share excess layer had various 2017 Wiley Rein LLP 7

ways to describe their attachment: after all Underlying Insurance carriers have paid in cash the full amount of their respective liabilities ; [t]he [excess] [i]nsurer shall only be liable to make payment under this policy after the total amount of the Underlying Limit of Liability has been paid in legal currency by the insurers of the Underlying Insurance as covered loss thereunder ; only after any Insurer subscribing to any Underlying Policy shall have agreed to pay or have been held liable to pay the full amount of its respective limits of liability ; [i]n the event of the exhaustion of all of the limit(s) of liability... solely as a result of payment of loss thereunder. Id. at 372-73. The court considered each of these articulations of how the excess policies would attach and concluded that, under all of the formulations, the language unambiguously required that all lower-layer insurers pay, or be held liable to pay, their full limits before the excess insurers had any coverage obligation. 3. Must an excess insurer drop down if primary is insolvent? Where a primary insurer is insolvent, an issue arises whether an excess insurer must drop down and fill in the gap resulting from the insurer s insolvency or whether the insured must bear the liability otherwise allocated to the insolvent insurer s policy. For example, in Canal Insurance Co. v. Montello, Inc., 632 F. App x 448 (10th Cir. 2015) (Okla. law), the policyholder faced bodily injury claims as a result of exposure to asbestos used in one of its products. The primary insurer was declared insolvent before paying for any claims on the policyholder s behalf, and the policyholder turned to its excess insurer for a defense and indemnity in the underlying claims. The Tenth Circuit held that the two excess policies did not provide dropdown coverage because the insolvency of the primary insurer was not an occurrence triggering coverage. The court also concluded that the excess clause of the policies clearly provided that the policies went into effect when the primary insurer s limit of liability was reduced by payment of loss, and the underlying insurer s inability to pay is not payment of loss. Id. at 452. In so holding, the court rejected the policyholder s argument that the umbrella coverage clause in one excess policy created a drop-down duty to defend or indemnify the asbestos claims. According to the court, an umbrella policy provides primary coverage only for risks that an underlying policy does not cover, and coverage was not expanded by the insolvency of a primary insurer whose policies provided coverage. The court also rejected the policyholder s argument that primary insurer s insolvency rendered the underlying insurance invalid and uncollectible, requiring the excess insurers to provide primary insurance under the other insurance provision of their policies. According to the court, [i]t would be inequitable to read an other insurance clause as insuring the solvency of the underlying primary insurer. Id. at 453. The court similarly rejected the policyholder s argument that its reasonable expectations created drop-down coverage. The court held that the excess insurer s duty to defend does not arise as a result of the primary insurer s inability to defend. Id. at 454. Thus, the excess insurers were not required to contribute to the defense of the insured until the primary insurance policy was exhausted. In Polygon Northwest Co. v. American National Fire Insurance Co., 189 P.3d 777 (Wash. Ct. App. 2008), the Washington appellate court likewise held that an excess insurer did not have to drop down to pay the amounts owed by an insolvent primary insurer. However, the court also held 2017 Wiley Rein LLP 8

that the excess insurer was obligated to pay the amount of the settlement that otherwise implicated its excess layer, regardless whether the primary insurer actually paid the loss amounts within its policies. A minority of courts have required insurers to drop down and gap-fill where the primary insurer is insolvent, though. See, e.g., Gulezian v. Lincoln Ins. Co., 399 Mass. 606, 611-612 (Mass. 1987) (interpreting particular excess policy language referring to collectible or recoverable underlying insurance to require the excess insurer to drop down in the event of the primary insurer s insolvency and cover from dollar one of the primary policy); Mission Nat l Ins. Co. v. Duke Transp. Co., 792 F.2d 550, 553 (5th Cir. 1986) (La. law) ( [w]hen an excess insurer uses the term collectible or recoverable it is agreeing to drop down in the event the primary coverage becomes uncollectible or unrecoverable ). 2017 Wiley Rein LLP 9