In Allstate, the reinsured sought to maximize its recovery against the reinsurer by abandoning, in an environmental pollution clean up case, the one occurrence per polluted site allocation directed by a district court ruling and the multi-occurrence position taken by both sides in the underlying litigation and settlement negotiations.Summary of this case from United States Fid. & Guar. Co. v. American Re-Ins. Co.
June 12, 2007.
APPEAL from an order of the Supreme Court, New York County (Karla Moskowitz, J.), entered February 22, 2006. The order denied plaintiff reinsurer's motion for partial summary judgment and granted defendant reinsured's cross motion for partial summary judgment declaring that its postsettlement allocation did not violate the terms of the parties' facultative reinsurance certificates and was reasonable and made in good faith.
LeBoeuf, Lamb, Greene MacRae LLP, New York City ( John M. Nonna, Eridania Perez and Franklin Monsour of counsel), for appellant.
Lord, Bissell Brook, LLP, Chicago ( Mark A. Kreger and Hugh Balsam, of the Illinois bar, admitted pro hac vice, of counsel), and Lord, Bissell Brook, LLP, New York City ( R. James DeRose, III, of counsel), for respondent.
Before: TOM, J.P., SAXE, BUCKLEY and McGUIRE, JJ., concur.
OPINION OF THE COURT
In this declaratory judgment action, we are asked whether we must give deference under the "follow-the-fortunes" doctrine to an insurer's reinsurance loss allocation on a one-occurrence-per-site basis that allowed the insurer to exceed the facultative reinsurance agreement's $1 million-per-occurrence deductible with respect to four sites. In the underlying litigation between the insurer and its insured as to one of the sites, a United States District Court ruled that there were seven occurrences requiring the application of seven deductibles, and as to the remaining three sites, the insurer in both the litigation and settlement negotiations and eventual settlement took the position that there were multiple occurrences at each site. As is undisputed, had the insurer applied the same multiple occurrence allocation in the loss allocation bill presented to its reinsurer, it would have no reinsurance claim.
Facultative reinsurance, as opposed to treaty reinsurance, is policy-specific.
Defendant, a member of American International Group (AIG), issued commercial property insurance policies to United Technologies Corporation (UTC) for lessor damage to property for the period 1975-1978 and 1978-1981. Both policies had liability limits, for any "one loss, disaster, or casualty," of $6 million under the 1975 policy and $10 million under the 1978 policy. Under both policies, UTC had a self-insured retention or deductible of $200,000 for "any one occurrence."
In the insurance context, the terms retention and deductible, the self-insured portion of a property or liability loss retained by the policyholder, are used interchangeably.
Defendant sought and obtained reinsurance for the UTC policies as follows: reinsurance of the $6 million limit of the 1975 UTC policy was structured in two layers — a $1 million primary layer and a $5 million excess of the $1 million layer. Reinsurance of the $10 million limit of the 1978 UTC policy was structured in three layers — a $1 million primary layer, a $5 million excess of the $1 million layer, and $4 million excess over the other two layers.
On or about November 17, 1975, plaintiff and defendant entered into a reinsurance agreement, effective for the same period as the 1975 policy, whereby plaintiff reinsured 22% of the $5 million excess of $1 million layer. Under the 1975 certificate, plaintiff's 22% participation is not triggered until defendant's payment under the 1975 policy exceeds a deductible of $1 million, the primary layer, of any single occurrence loss.
When the 1975 certificate expired, plaintiff issued another to cover the term of the 1978 policy, reinsuring 25% of the $5 million excess of $1 million layer. As with the 1975 certificate, plaintiff's liability under the 1978 certificate is not triggered until defendant's payment under the 1978 policy exceeds a retention or deductible of $1 million for any single occurrence loss.
Plaintiff's certificates provide that subject to the terms and conditions of the certificates and the UTC policies, its liability follows that of defendant under the UTC policies. The 1975 certificate provides: "This Certificate is issued as reinsurance and is subject to the same risks, valuations, privileges, clauses and conditions, endorsements (except changes in location), assignments[,] adjustments and mode of settlement, as are, or were, or may be assumed or adopted by the Reinsured." Similarly, the 1978 certificate provides: "The liability of Allstate shall follow that of the Company [American Home] and, except as otherwise specifically provided in this certificate, shall be subject to the terms and conditions of the policy(ies) reinsured."
In April 1992, UTC commenced an action against defendant in the United States District Court for the District of Connecticut for indemnification under the UTC policies for physical loss and damage allegedly sustained due to environmental pollution at various sites, including plants in West Palm Beach, Florida, Stratford and Windsor Locks, Connecticut, and Santa Clara County, California (also known as Coyote).
During the UTC litigation, both defendant and UTC consistently argued that there were multiple occurrences (losses) at each of these sites; neither ever took the position that each site constituted one occurrence. Their only disagreement concerned the number of occurrences at each site, a key issue that affected defendant's $200,000-per-occurrence deductible under the UTC policies. The number of occurrences dictated the number of deductibles that would be applied. More deductibles meant less covered loss under the policies. Due to the number of sites involved, the District Court determined to handle the discovery and trial in phases.
On March 2, 1998, after trial, a jury found that UTC had shown loss or damage at seven different areas at the Windsor Locks site during the period covered by the 1975 and 1978 policies, allocating an amount to each of the seven areas at which damage or loss occurred. After the verdict, the parties sought summary judgment on the number of occurrences involved so they could determine the number of $200,000 deductibles to be applied, UTC claiming seven and defendant claiming at least 18. The District Court found "for purposes of calculating deductible amounts . . . seven occurrences of damage or loss took place at the Windsor Locks site consistent with the jury's . . . verdict." The court applied one deductible per occurrence and an additional deductible for each policy period during which the jury found UTC suffered a covered loss.
Defendant moved for a new trial and petitioned for an interlocutory appeal, arguing that the court had applied the wrong standard for determining the number of occurrences at the Windsor Locks site. The petition for an interlocutory appeal was denied, and defendant and UTC settled the litigation before the motion for a new trial was decided. In settling the matter, neither party conceded its position as to the number of occurrences at the Windsor Locks site.
Meanwhile, the UTC litigation continued on the coverage question with respect to the 16 sites, including the West Palm Beach, Stratford and Coyote sites, that were the focus of the second trial. As with their positions on Windsor Locks, both UTC and defendant, in settlement discussions, discovery and trial preparation, took the position that multiple occurrences had taken place at these sites. Each of the exposure analyses prepared by defendant demonstrates a consistent approach, i.e., assessing its exposure at the 16 sites on a multiple-occurrence-per-site basis and estimating the cost of each occurrence at those sites. During discovery and trial preparation, defendant identified and estimated the cost of 11 occurrences at West Palm Beach, 4 at Stratford and 6 at Coyote. UTC also identified multiple occurrences at each of these sites. Thomas Carey, AIG's vice-president of domestic property operations, confirmed that its multiple-occurrence-per-site analysis represented good faith estimates of the cost of each occurrence at these sites.
In January 2002, two months before the trial regarding the 16 sites, defendant settled with UTC, agreeing to pay a lump sum of $112 million for all of the sites at issue. In the settlement negotiations, defendant advanced the position that there had been multiple occurrences at each site, 95 in all at the 16 sites. UTC, on the other hand, maintained that there had been only 44 occurrences. During these negotiations, UTC, referring to the District Court's postverdict ruling at the first trial, took the position that it had a "verdict in hand" for Windsor Locks and that there should be no discount in settlement for that site. While defendant insisted that it was looking for a global settlement, no concession was made to limit the District Court's ruling that there were seven loss occurrences at the Windsor Locks site.
After the settlement, at defendant's request, William Hassler, its lead counsel in the UTC litigation, who had never before prepared an allocation of an environmental coverage dispute settlement for reinsurance purposes, prepared a fourfold analysis for allocating the settlement to the UTC policies and to defendant's reinsurers. In his analysis, Hassler, in stark contrast to defendant's position before the Connecticut Federal District Court and the court's ruling and the position of defendant and UTC throughout the UTC litigation, treated each site as one occurrence. Hassler evaluated 74 sites, separated into categories based on whether the claimed loss was already tried or was pending, dropped or had no value. Hassler's allocation analysis was fourfold.
In the first step, Hassler used the amount UTC claimed for each site, as provided by UTC's counsel, except that the "[f]igures for the Windsor Locks site were taken from the 1998 jury verdict for that site, as modified by the Court's Ruling on Deductibles/Number of Occurrences." In the second step, Hassler modified the UTC amounts per site by applying one $200,000 deductible per site for each of the 1975 and 1978 policies. If a site was owned or used during the 1975 and 1978 policy periods, the amount deducted was $400,000. For Windsor Locks, however, Hassler relied on the District Court's determination of the number of occurrences at that site. Thus, instead of applying a deductible of $400,000 for the 1975 and 1978 UTC policy periods for Windsor Locks, defendant deducted $2,194,707, which was computed on the basis of the amount of deductibles for the number of occurrences the District Court found at the site, with additional application of $400,000 for the 1975 and 1978 policy periods.
In the third step, Hassler calculated for each site the cost per year of insurance coverage. He did so by dividing the cost by the number of years that UTC owned or used a site. For most of the sites, Hassler assigned 11 years of coverage for the period covered by all three policies defendant issued to UTC. For Windsor Locks, Hassler accepted the jury's finding that damage at the site had occurred only over the six years that the 1975 and 1978 policies were in effect. In the fourth step of this analysis, Hassler allocated a settlement cost to each policy year, broken down on a site-by-site basis, thereby treating each site as one occurrence. For Windsor Locks, the settlement cost was allocated evenly over six years based on the District Court's determination in that proceeding that damage at that site had occurred during the period of the 1975 and 1978 policies. When Hassler determined the actual per-site-per-year settlement cost for Windsor Locks, however, he did not reduce the cost by the deductibles for seven occurrences, as had been judicially determined. Thus, he did not apply the $1 million self-retention for each of the Windsor Locks occurrences under the 1975 and 1978 certificates in the same way he applied the $200,000 deductible under the 1975 and 1978 policies in the second step of his analysis. Similarly, for West Palm Beach, Stratford and Coyote, Hassler did not break down the per-site-per-year settlement cost by the number of occurrences that defendant, and even UTC, had claimed in the UTC litigation at those sites.
The 11 years of coverage include the period 1981-1986 covered by a third policy that is not at issue here.
Defendant utilized the step four analysis to bill its reinsurers on a one-occurrence-per-site-per-year basis. By treating each site as one occurrence, Windsor Locks, West Palm Beach, Stratford and Coyote were each allocated settlement costs per year in excess of $1 million, thus triggering plaintiff's reinsurance obligations under the 1975 and 1978 certificates. In considering the $1 million-per-occurrence deductible and its effect on the second layer reinsurers like plaintiff ($5 million excess of $1 million), Hassler testified that he did not break down his one-occurrence-per-site allocation to actual occurrences because it would be too subjective, as he could not do so fairly where the per-site-per-year amount was in excess of $1 million. He claimed that this approach was consistent with defendant's past practices. Hassler applied this rationale as well to Windsor Locks.
When defendant demanded payment of $2,578,638.21 from plaintiff for its reinsurance share of defendant's liability under the 1975 and 1978 UTC policies for the Windsor Locks, West Palm Beach, Stratford and Coyote sites, Patrick Parrington, plaintiff's reinsurance claims analyst, noticed the inconsistency in "the manner in which the claim was evaluated by AIG prior to settlement and the subsequent methodology used in ceding the loss to the reinsurer," and asked for an explanation for, inter alia, AIG's "disregard of its own multiple occurrence positions taken" at the West Palm Beach, Coyote and Stratford sites. As a result, Parrington audited defendant's handling of the UTC claims and litigation.
Using the documents obtained during the audit, Parrington performed his own per-occurrence-per-site evaluation of the Windsor Locks, West Palm Beach, Stratford and Coyote sites. In his analysis, Parrington relied on the number of occurrences and cost allocated to each occurrence by the jury and District Court in the Windsor Locks trial. For West Palm Beach, Stratford and Coyote, Parrington relied on defendant's own multiple-occurrences-per-site analysis utilized during the UTC litigation in which it identified and allocated a cost for 21 occurrences — 11 at West Palm Beach, 4 at Stratford and 6 at Coyote.
Parrington's allocation analysis confirmed that no single occurrence at these sites would have exceeded the $1 million retention. Thus, plaintiff's reinsurance obligations would not have been triggered under the 1975 or 1978 certificate. Plaintiff notified defendant that it would not honor the payment request "because the retentions required under the various Allstate facultative certificates have not been [reached]. Even under the most lenient accounting of occurrences as put forth in this matter by [UTC], no single occurrence impacts the excess of [$1,000,000] reinsurance layer in any given year." Plaintiff thereupon commenced this declaratory judgment action.
After joinder of issue and discovery, plaintiff moved for partial summary judgment seeking a declaration that it was not bound to the follow-the-fortunes doctrine because defendant's reinsurance loss allocation of the Windsor Locks, West Palm Beach, Stratford and Coyote sites was unreasonable, and therefore, it was not liable to defendant under the facultative reinsurance certificates. Defendant cross-moved for partial summary judgment on the same issue, arguing that its allocation was reasonable. The motion court denied the motion and granted the cross motion, relying on North Riv. Ins. Co. v ACE Am. Reins. Co. ( 361 F3d 134 [2d Cir 2004]) and Travelers Cas. Sur. Co. v Gerling Global Reins. Corp. of Am. ( 419 F3d 181 [2d Cir 2005]) for the proposition that the follow-the-fortunes doctrine is applicable regardless of any inconsistency between the reinsured's presettlement and postsettlement allocation positions. Otherwise, the court held, citing North River ( 361 F3d at 141), it would be engaging in an "intrusive factual inquiry" into defendant's settlement process. The court rejected the effect of the Federal District Court's earlier determination as to the number of occurrences at Windsor Locks because the decision was nonfinal ( see 28 USC § 1291) and because the parties to that proceeding (UTC and defendant) had eventually settled the dispute. Finally, the court found that defendant's one-occurrence-per-site allocation was reasonable because it was consistent with Hassler's manner of applying "bright line tests" rather than "his own subjective judgment."
We reject such reasoning, which reflects, on this record, nothing more than Hassler's subjective judgment, and in effect lends the court's imprimatur to defendant's playing by two sets of rules: one, applied at the insured's claim level where the occurrence deductible is used as often as possible to minimize the amount of the insurer's exposure and loss, and later, in the same loss setting, another, where the occurrence deductible is used as sparingly as possible to maximize the reinsured's recovery against the reinsurer. The follow-the-fortunes doctrine was intended to foster consistency in the treatment of losses at both levels, insured and reinsured, not to allow an insurer to use a different set of rules at each level. We soundly reject the notion that the follow-the-fortunes doctrine requires that courts turn a blind eye to such manifest manipulation of the allocation process in total disregard of the reinsured's obligation to act in good faith.
The follow-the-fortunes doctrine requires a reinsurer to indemnify for payments reasonably within the terms of the original policy, even if technically not covered by it ( Travelers Cos. Sur. Co. v Certain Underwriters at Lloyd's of London, 96 NY2d 583, 596). Pursuant to the doctrine, a reinsurer is bound "to accept the cedent's good faith decisions on all things concerning the underlying insurance terms and claims against the underlying insured: coverage, tactics, lawsuits, compromise, resistance or capitulation" ( British Intl. Ins. Co. Ltd. v Seguros La Republica, S.A., 342 F3d 78, 85 [2d Cir 2003]). While, obviously, the doctrine is not so expansive as to "give reinsureds carte blanche to impose whatever settlement decisions they make on their reinsurers" ( Affiliated F.M. Ins. Co. v Employers Reins. Corp., 369 F Supp 2d 217, 227 [D RI 2005]), it is intended to avoid relitigating coverage issues under the insured's policies and to promote good faith settlement of the insured's claims ( see id.). It "insulates a reinsured's liability determinations from challenge by a reinsurer unless they are fraudulent, in bad faith, or the payments are `clearly beyond the scope of the original policy' or `in excess of [the reinsurer's] agreed-to exposure'" ( North Riv., 361 F3d at 140, quoting Christiania Gen. Ins. Corp. of N.Y. v Great Am. Ins. Co., 979 F2d 268, 280 [2d Cir 1992]).
Contrary to the motion court's holding, Travelers and North River do not require a reinsurer, under the follow-the-fortunes doctrine, to accept the reinsured's postsettlement loss allocation even if that allocation is contrary to the reinsured's preallocation position and treatment of the loss allocation issue with its own insured, i.e., its treatment of deductibles. While the cases unequivocally hold that the doctrine extends to a postsettlement allocation despite "an inconsistency between that allocation and the [reinsured's] pre-settlement assessments of risk," it applies only " as long as the allocation meets the typical follow-the-settlements requirements, i.e., is in good faith, reasonable, and within the applicable policies" (North Riv., 361 F3d at 141 [emphasis added]).
Here, unlike North River, the inconsistency is not between defendant's postsettlement allocation and its presettlement assessments of the risk, but between its presettlement allocation of loss with its insured (UTC) and its postsettlement allocation with its reinsurer (plaintiff). Defendant was obliged to act in good faith in its postsettlement allocation of loss with plaintiff. As is readily apparent from its postsettlement loss allocation analysis, it did not. A reinsurer is not bound by the follow-the-fortunes doctrine where the reinsured's settlement allocation, at odds with its allocation of the loss with its insured, designed to minimize its loss, reflects an effort to maximize unreasonably the amount of collectible reinsurance ( see Hartford Ace. Indem. v Columbia Cas. Co., 98 F Supp 2d 251, 259 [D Conn 2000]).
Defendant's postsettlement allocation was unreasonable because the one-occurrence-per-site allocation of the Windsor Locks site directly contradicts the District Court ruling as to the number of occurrences at that site. Moreover, the reinsurance allocation is internally inconsistent, as reflected by defendant's highly selective use of the District Court ruling. The one-occurrence-per-site allocation of the other sites, including West Palm Beach, Stratford and Coyote, completely contradicts the multiple-occurrence position that both defendant and its insured pursued in the UTC litigation and settlement negotiations, which culminated in a $112 million lump-sum settlement for all of the sites at issue. For defendant to assert aggressively the maximum number of occurrences at each site to minimize its liability to its insured in the UTC litigation, and then completely change its position in allocating its loss to plaintiff under the reinsurance certificates, is neither reasonable nor reflective of good faith. It is disingenuous.
Here, unlike any other reported case involving the follow-the-fortunes doctrine, defendant ignored a court ruling determining the number of occurrences at a covered site in its allocation of loss to plaintiff, and instead imposed a single occurrence at the Windsor Locks site for reinsurance purposes. Based on the jury's verdict, the District Court ruled that there were seven occurrences at that site and, consistent therewith, allocated seven $200,000 deductibles to the covered loss sustained there. Although the claims for the seven occurrences were ultimately settled, neither defendant, which had argued that there had been at least 18 occurrences, nor UTC, which had argued that there had been seven, conceded its position as to the number of occurrences. Defendant, rather than apply the District Court allocation which suited its purposes in reducing its liability to UTC under the 1975 and 1978 policies, used a one-occurrence-per-site calculation in its postsettlement allocation to exaggerate its reinsurance claim against Allstate as to the Windsor Locks site. As Hassler admitted, had defendant broken down the per-site-per-year settlement cost for Windsor Locks by the seven occurrences as determined by the District Court's ruling, no single occurrence would have exceeded the $1 million retention to trigger plaintiff's liability under the 1975 and 1978 certificates. Defendant's inconsistent application of the District Court's occurrence determination is neither reasonable nor based on good faith. And contrary to the motion court's reasoning, this determination requires no "intrusive factual inquiry" into defendant's allocation. Nor, contrary to defendant's argument, is plaintiff "second guessing" its settlement decisions.
The motion court's attempt to trivialize the District Court's interlocutory ruling on the ground of nonfinality under 28 USC § 1291 is frivolous. Section 1291 relates to finality for appealability purposes. An interlocutory decision that is nonappealable may nevertheless be final in the preclusive sense ( see Lummus Co. v Commonwealth Oil Ref. Co., 297 F2d 80, 89 [2d Cir 1961], cert denied sub nom. Dawson v Lummus Co., 368 US 986) as was the Windsor Locks decision, which was binding on defendant. A trial was held, the jury returned a verdict, and the court ruled on the precise question of the number of occurrences at the site. The issue was fully litigated and clearly was "not avowedly tentative" ( id.). Defendant was effectively bound by the District Court decision when it negotiated a settlement with UTC. Defendant selectively utilized the effect of the Windsor Locks decision, but only when it suited its purpose in reducing its liability to UTC.
Nor is there any support for the motion court's conclusion that defendant and UTC settled the Windsor Locks litigation in a manner that rendered the decision inoperative. A settlement does not entail vacatur of a prior decision. Neither party did anything to eliminate or limit the preclusive effect of the decision. The parties neither agreed nor negotiated to seek to have the court vacate the decision as part of their settlement ( see e.g. Mercantile Gen. Reins. Co. v Colonial Assur. Co., 147 Misc 2d 804, 806-807).
Defendant and, for that matter, UTC, never took the position in that litigation and the settlement negotiations that there was only one occurrence at each of the other sites involved (West Palm Beach, Stratford and Coyote). Their only disagreement was over the number of occurrences. During discovery, trial preparation and settlement, defendant vigorously pursued the maximum number of occurrences per site. In its exposure analysis, defendant contended there were 21 occurrences among the West Palm Beach, Stratford and Coyote sites. Defendant allocated a cost, which it conceded was a reasonable, good faith estimate, to each occurrence identified in those sites. Defendant's claim of multiple occurrences was, as AIG's vice-president of domestic property operations confirmed in his testimony, a key issue in settlement negotiations that he used as leverage to obtain better terms from UTC. It is undisputed that if defendant had allocated the UTC settlement to plaintiff consistent with the position it took in the UTC litigation and settlement negotiation that there were 11 occurrences at West Palm Beach, 4 at Stratford and 6 at Coyote, there would be no reinsurance liability under plaintiff's 1975 and 1978 certificates for those sites.
Finally, we note that defendant's argument as to an industry practice of ceding pollution claims to reinsurers on a single-occurrence-per-site basis is not only unsubstantiated, but offset by the same testimony upon which it relies. In any event, its own standard practice is irrelevant. Moreover, a reinsured cannot treat its insured's claim on a per-occurrence-loss basis at each site and then allocate the loss to its reinsurer on a single-occurrence-per-site basis. That kind of inconsistent handling of loss is the very antithesis of the follow-the-fortunes doctrine.
Accordingly, the order of the Supreme Court, New York County (Karla Moskowitz, J.), entered February 22, 2006, which denied plaintiff Allstate's motion for partial summary judgment and granted defendant American Home's cross motion for partial summary judgment declaring that its postsettlement allocation did not violate the terms of the parties' facultative re-insurance certificates and was reasonable and made in good faith, should be reversed, on the law, with costs and disbursements, the motion granted, the cross motion denied, and it is declared that plaintiff is not liable to defendant pursuant to the 1975 and 1978 facultative reinsurance certificates.
Order, Supreme Court, New York County, entered February 22, 2006, reversed, on the law, with costs and disbursements, plaintiff's motion for partial summary judgment granted, defendant's cross motion for partial summary judgment denied and it is declared that plaintiff is not liable to defendant pursuant to the 1975 and 1978 facultative reinsurance certificates.