THE STATE OF CALIFORNIA, Plaintiff, Cross-defendant and Appellant, v. CONTINENTAL INSURANCE COMPANY et al., Defendants, Cross-complainants and Appellants; EMPLOYERS INSURANCE OF WAUSAU, Defendant, Cross-complainant and Respondent, 55 Cal. 4th 186

Opinion CHIN, J. This case considers complex questions of insurance policy coverage

interpretation in connection with a federal court-ordered cleanup of the state’s Stringfellow Acid Pits waste site. We initially address the “ ‘continuous injury’ trigger of coverage,” as that principle was explained in Montrose Chemical Corp. v. Admiral Ins. Co. (1995) 10 Cal.4th 645, 655 [42 Cal.Rptr.2d 324, 913 P.2d 878] (Montrose) and the “all sums” rule adopted in Aerojet-General Corp. v.

Transport Indemnity Co. (1997) 17 Cal.4th 38, 55-57 [70 Cal.Rptr.2d 118, 948 P.2d 909] (Aerojet), and conclude that the principles announced in those cases apply to the insurers’ indemnity obligations in this case, so long as the insurers insured the State of California during the property damage itself. Because we conclude that the continuous injury trigger and all sums rule apply to the duty to indemnify here, we must also determine how best to allocate the indemnity duty among the insurers responsible for covering the State of California’s liability.

As we explain, we conclude that the Court of Appeal below correctly applied the “all-sums-with-stacking” allocation rule. We therefore affirm the judgment of the Court of Appeal. Factual and Procedural Background The State of California (State) seeks indemnity from several of its insurers in connection with a federal court-ordered cleanup of the State’s Stringfellow Acid Pits waste site. The site was an industrial waste disposal facility that the State designed and operated from 1956 to 1972.

Each insurer that is party to this appeal issued one or more excess commercial (also known as comprehensive) general liability (CGL) insurance policies to the State between 1964 and 1976. The State was uninsured before 1963, and after 1978. In 1955, a state geologist determined that a Riverside County quarry was a suitable location for the disposal of industrial waste. According to the geologist’s report, the site was a canyon lined on its bottom with impermeable rock. The geologist advised the State to build a concrete barrier dam to close a 250-foot gap in the canyon’s natural walls.

He claimed that, once the dam was in place, “the operation of the site for industrial wastes [would] not constitute a threat of pollution.” The State subsequently developed the facility, which went into operation in 1956, and eventually received more than 30 million gallons of industrial waste. In reality, the site suffered from three major flaws that made it ill suited to serve as an industrial waste facility. First, the state geologist had failed to identify an underground aquifer located 70 feet below the canyon floor that facilitated the movement of groundwater into and out of the site.

Second, the rock underlying the canyon floor was fractured, so it allowed waste to leak into the groundwater system and escape the facility. Third, the barrier dam proved ineffective. It permitted contaminants to escape the facility during heavy rains in 1969 and again in 1978. The severity of the latter event forced the State to conduct a “controlled discharge” of contaminants into Pyrite Channel. The ensuing plume of waste extended for miles. The State closed the facility in 1972 after discovering the groundwater contamination.

In 1998, a federal court found the State liable for, inter alia, negligence in investigating, choosing, and designing the site, overseeing its construction, failing to correct conditions at it, and delaying its remediation. The State was held liable for all past and future cleanup costs. The State claims costs associated with the Stringfellow site remediation could reach $700 million. The insurers stipulate that the State is liable for at least $50 million. The State filed an action against several of its insurers in September 1993, seeking indemnification for its liability in the federal action.

The pertinent language of all the policies at issue is essentially identical.

Under the heading “Insuring Agreement,” insurers agreed “[t]o pay on behalf of the Insured all sums which the Insured shall become obligated to pay by reason of liability imposed by law ... for damages . . . because of injury to or destruction of property, including loss of use thereof.” Limits on liability in the agreements were stated as a specified dollar amount of the “ultimate net loss [of] each occurrence.” “Occurrence” was defined as meaning “an accident or a continuous or repeated exposure to conditions which result in . . . damage to property during the policy period ...” In addition, “ ‘ultimate net loss’ [was] understood to mean the amount payable in settlement of the liability of the Insured arising only from the hazards covered by this policy after making deductions for all recoveries and for other valid and collectible insurances . . . .” The trial was conducted in multiple phases.

At the conclusion of a June 1999 bench trial, the court ruled that the policy limits under policies with multiple-year periods applied “per occurrence” and not annually. Following this, in April 2002, the trial court held that the State’s failure to remediate and its delay in remediating the site was not a breach of any duty to mitigate the insurers’ damages. In September 2002, the State brought a second suit, asserting related claims against additional insurers, including those which are parties to this appeal.

This case was consolidated with the first action, and defendant insurers in the second suit agreed to be bound by all prior rulings in the original action. All parties stipulated that the property damage that the Stringfellow site’s selection, design, and construction caused took place continuously throughout the defendant insurers’ multiple consecutive policy periods from 1964 to 1976. The trial court held that each insurer was liable for damages, subject to its particular policy limits for the total amount of the loss.

The court based this ruling on the “all sums” language in the insuring agreements. (Ante, at p. 193.) It also held that the State could not recover the policy limits in effect for every policy period, and could not “stack,” or combine, policy periods to recover more than one policy’s limits for covered occurrences. The court then concluded that the State had to choose a single policy period for the entire liability coverage, and it could recover only up to the total policy limits in effect during that policy period. The court based its ruling on the decision in FMC Corp. v.

Plaisted & Companies (1998) 61 Cal.App.4th 1132 [72 Cal.Rptr.2d 467] (FMC), which prevented an insured from stacking multiple consecutive policies in a case in which the insured had caused toxic contamination “over a period of many years” (id. at p. 1142). In May 2005, a jury in phase three of the trial rendered special verdicts finding the insurers had breached their policies. By that time, the State had already entered into settlement agreements totaling approximately $120 million with several other insurers. The trial court required that these settlements reduce the insurers’ liability as setoffs.

Therefore, “[u]nder the trial court’s one-occurrence, no-annualization and no-stacking rulings, the most the State could recover [from all insurers] was $48 million.” Because the State had already recovered $120 million, the court entered judgment nominally in the State’s favor, but in the amount of “$0.” The State filed an appeal and, with the exception of Wausau, all of the insurers filed cross-appeals. The Court of Appeal affirmed in part and reversed in part the trial court’s ruling.

The Court of Appeal, like the trial court, rejected the insurers’ contention that they could not be liable for property damage occurring outside their respective policy periods. It held that once coverage was triggered, all of the insurers had to indemnify the insured for the loss. However, the Court of Appeal reversed the trial court’s ruling that prohibited the State from stacking the total policy limits in effect during all policy periods.

In doing so, the Court of Appeal rejected the holding of FMC, supra, 61 Cal.App.4th 1132, characterizing that antistacking decision as “flawed and unconvincing.” Our grant of review followed the insurers’ petitions for review. Discussion A. Background 1. Standard of Review and Insurance Law Principles In general,

interpretation of an insurance policy is a question of law that is decided under settled rules of contract

interpretation. (E.M.M.I. Inc. v. Zurich American Ins. Co. (2004) 32 Cal.4th 465, 470 [9 Cal.Rptr.3d 701, 84 P.3d 385]; Waller v. Truck Ins. Exchange, Inc. (1995) 11 Cal.4th 1, 18 [44 Cal.Rptr.2d 370, 900 P.2d 619].) “ ‘While insurance contracts have special features, they are still contracts to which the ordinary rules of contractual

interpretation apply.’ (Bank of the West v. Superior Court (1992) 2 Cal.4th 1254, 1264 [10 Cal.Rptr.2d 538, 833 P.2d 545]; see AIU [Ins. Co. v. Superior Court (1990)] 51 Cal.3d [807,] 821-822 [274 Cal.Rptr. 820, 799 P.2d 1253].)” (Foster-Gardner, Inc. v. National Union Fire Ins. Co. (1998) 18 Cal.4th 857, 868 [77 Cal.Rptr.2d 107, 959 P.2d 265].) “The fundamental goal of contractual

interpretation is to give effect to the mutual intention of the parties.” (Bank of the West v. Superior Court, supra, 2 Cal.4th at p. 1264; see Civ. Code, § 1636.) “Such intent is to be inferred, if possible, solely from the written provisions of the contract.” (AIU, supra, 51 Cal.3d at p. 822; see Civ. Code, § 1639.) “If contractual language is clear and explicit, it governs.” (Bank of the West v. Superior Court, supra, 2 Cal.4th at p. 1264.) “ ‘The “clear and explicit” meaning of these provisions, interpreted in their “ordinary and popular sense,” unless “used by the parties in a technical sense or a special meaning is given to them by usage” ([Civ. Code,] § 1644), controls judicial

interpretation. (Id., § 1638.)’ [Citations.]” (Waller v. Truck Ins. Exchange, Inc., supra, 11 Cal.4th at p. 18.) “A policy provision will be considered ambiguous when it is capable of two or more constructions, both of which are reasonable.” (Waller v. Truck Ins. Exchange, Inc., supra, 11 Cal.4th at p. 18, citing Bay Cities Paving & Grading, Inc. v. Lawyers’ Mutual Ins. Co. (1993) 5 Cal.4th 854, 867 [21 Cal.Rptr.2d 691, 855 P.2d 1263].) A term is not ambiguous merely because the policies do not define it. (Bay Cities Paving, supra, 5 Cal.4th at p. 866; Bank of the West v.

Superior Court, supra, 2 Cal.4th at pp. 1264-1265; Castro v. Fireman’s Fund American Life Ins. Co. (1988) 206 Cal.App.3d 1114, 1120 [53 Cal.Rptr. 833].) Nor is it ambiguous because of “[disagreement concerning the meaning of a phrase,” or “ ‘the fact that a word or phrase isolated from its context is susceptible of more than one meaning.’ ” (Castro v. Fireman’s Fund American Life Ins.

Co., supra, 206 Cal.App.3d at p. 1120.) “ ‘[L]anguage in a contract must be construed in the context of that instrument as a whole, and in the circumstances of that case, and cannot be found to be ambiguous in the abstract.’ ” (Bank of the West v. Superior Court, supra, 2 Cal.4th at p. 1265, italics omitted, quoting Producers Dairy Delivery Co. v. Sentry Ins.

Co. (1986) 41 Cal.3d 903, 916, fn. 7 [226 Cal.Rptr. 558, 718 P.2d 920].) “If an asserted ambiguity is not eliminated by the language and context of the policy, courts then invoke the principle that ambiguities are generally construed against the party who caused the uncertainty to exist (i.e., the insurer) in order to protect the insured’s reasonable expectation of coverage.” (La Jolla Beach & Tennis Club, Inc. v.

Industrial Indemnity Co. (1994) 9 Cal.4th 27, 37 [36 Cal.Rptr.2d 100, 884 P.2d 1048].) We now apply these principles to the present case. 2. “Long-tail” Claims Disputes like the one here frequently occur in the context of environmental damage and toxic exposure litigation. The kind of property damage associated with the Stringfellow site, often termed a “long-tail” injury, is characterized as a series of indivisible injuries attributable to continuing events without a single unambiguous “cause.” Long-tail injuries produce progressive damage that takes place slowly over years or even decades.

Traditional CGL insurance policies, including those drafted before such environmental suits were common, are typically silent as to this type of injury. (Hickman & DeYoung, Allocation of Environmental Cleanup Liability Between Successive Insurers (1990) 17 N.Ky. L.Rev. 291, 292 (Hickman Sc DeYoung).) Because of this circumstance, many insurers are unwilling to indemnify insureds for long-tail claims. Their refusal to indemnify often causes insureds to sue for coverage. As the present case highlights, these suits tend to be complex.

Typically they involve dozens of litigants and even larger numbers of insurance policies covering multiple time periods that stretch back over many years.

It is often “virtually impossible” for an insured to prove what specific damage occurred during each of the multiple consecutive policy periods in a progressive property damage case. (Hickman & DeYoung, supra, at p. 292.) If such evidence were required, an insured who had procured insurance coverage for each year during which a long-tail injury occurred likely would be unable to recover. “While CGL policies [such as the ones at issue here] limit coverage to their policy period, the policies . . . require only that some damage occur during the policy period. . . .

Unfortunately, CGL policies leave unanswered the crucial question for long-tail injuries: when does a continuous condition become an ‘occurrence’ for the purposes of [triggering] insurance coverage?” (Bratspies, Splitting the Baby: Apportioning Environmental Liability Among Triggered Insurance Policies (1999) 1999 B.Y.U. L.Rev. 1215, 1228-1229, fn. omitted (Bratspies).) B.

Montrose and Aerojet While the term “trigger of coverage” does not appear in the language of the CGL insurance policies here, it is a term of “convenience used to describe that which, under the specific terms of an insurance policy, must happen in the policy period in order for the potential of coverage to arise.

The issue is largely one of timing—what must take place within the policy’s effective dates for the potential of coverage to be ‘triggered’?” (Montrose, supra, 10 Cal.4th at p. 655, fn. 2.) In Montrose, we held that in the context of a third party liability policy “property damage that is continuous or progressively deteriorating throughout several policy periods is potentially covered by all policies in effect during those periods.” (Id. at p. 655.) In that case, the dispute centered on a series of successive liability policies that seven insurers issued covering a 26-year period. (Id. at p. 656.) At issue was whether an insurer whose policy covered only the last four years of this period had a duty to defend suits alleging continuous and progressive property damage and bodily injury that resulted from hazardous chemicals that the insured manufactured beginning before, but continuing during, the insurer’s policy period.

This court held that “ ‘[property damage’ ” was “ ‘physical injury to or destruction of tangible property which occurs during the policy period....’” (Id. at p. 668.) The policy defined “ ‘occurrence’ ” as “ ‘an accident, including continuous or repeated exposure to conditions, which results in . . . property damage ....’” (Id. at p. 669; see id. at pp. 671-673.) Under the insurance policy language at issue in Montrose, we determined that a continuous condition becomes an occurrence for the purposes of triggering insurance coverage when “ ‘property damage’ ” results from a causative event consisting of “the accident or ‘continuous and repeated exposure to conditions.’ ” (Id. at p. 669.) The limitation on potential indemnity was that the damage must “ ‘occur’ during the policy period, and ‘. . . result[]’ from the accident or ‘continuous and repeated exposure to conditions.’ ” (Ibid.) In 1997, this court again was asked to interpret the all sums insurance policy language in determining an insurer’s defense duties under a similar CGL policy.

We noted that “the ‘settled rule’ of the case law” is that “ ‘an insurer on the risk when continuous or progressively deteriorating [property] damage or [bodily] injury first manifests itself remains obligated to indemnify the insured for the entirety of the ensuing damage or injury.'’ ” (Aerojet, supra, 17 Cal.4th at p. 57, fn. 10, italics added by the Aerojet court.) Although Aerojet, like Montrose, principally involved the duty to defend, the issue the court addressed included the question whether the insurers could require the insured to pay any part of the defense costs. (Id. at pp. 55-56.) Aerojet reasoned that the insurers would be hable to indemnify the insured against all claims that resulted from some triggering harm during the respective policy periods, even if the claims arose after the policy period expired. (Id. at p. 71.) Therefore, the insurers were responsible for defending the insured for all claims that involved the triggering damage. (Ibid.) Aerojet understood Montrose as extending insurers’ indemnity obligations beyond the expiration of the policy period where there has been a continuous loss.

In other words, under Aerojet, as long as the policyholder is insured at some point during the continuing damage period, the insurers’ indemnity obligations persist until the loss is complete, or terminates. (Ibid.) As the present Court of Appeal observed, Aerojet s “ah sums” approach to the duty to indemnify was essential to its holding regarding the duty to defend. Similar reasoning applies to the indemnity question presented here. Neither the State nor the insurers dispute that progressive damage to property at the Stringfellow site “occurred” during numerous policy periods.

In addition, the insurers concede that in cases such as this it is impossible to prove precisely what property damage occurred during any specific policy period. The fact that all policies were covering the risk at some point during the property loss is enough to trigger the insurers’ indemnity obligation.

The insurers rely on footnote 19 in Montrose, supra, 10 Cal.4th at page 681, which generally noted that the court could not endorse a holding that insurers are “jointly and severally liable for the full amount” of a long-tail loss. (Italics omitted.) Aerojet explained the Montrose footnote. “In Montrose, we also made plain that ‘successive’ insurers ‘on the risk when continuous or progressively deteriorating [property] damage or [bodily] injury first manifests itself’ are separately and independently ‘obligated to indemnify the insured’: ‘[W]here successive . . . policies have been purchased, bodily injury and property damage that is continuing or progressively deteriorating throughout more than one policy period is potentially covered by all policies in effect during those periods.’ [Citation.] The successive insurers are not ‘jointly and severally liable.’ [Citation.]” (Aerojet, supra, 17 Cal.4th at p. 57, fn. 10, italics added, quoting Montrose, supra, 10 Cal.4th atpp. 686-687, 681, fn. 19.) Rather, as the Court of Appeal observed, each insurer is severally liable on its own policy up to its policy limits.

The insurers advocate that we adopt an alternative allocation scheme—a pro rata rule for indemnity allocation. Pro rata (or apportionment) allocation “assigns a dual purpose to the phrase ‘during the policy period’ in the CGL policy’s definition of ‘occurrence.’ The phrase serves both as a trigger of coverage and as a limitation on the promised ‘all sums’ coverage [language in the ‘Insuring Agreement’].” (Bratspies, supra, 1999 B.Y.U.

L.Rev. at p. 1234, fn. omitted.) Courts apportioning coverage on a pro rata basis require the allocation of loss to a particular policy to be “proportionate to the damage suffered during that policy’s term.” (Interim 23, Appleman on Insurance 2d (Holmes ed. 2003) § 145.4[A][2][b], p. 25 & fii. 109 [citing cases].) “This approach emphasizes that part of a long-tail injury will occur outside any particular policy period. Rather than requiring any one policy to cover the entire long-tail loss, [pro rata] allocation instead attempts to produce equity across time.” (Bratspies, supra, 1999 B.Y.U.

L.Rev. at p. 1232.) Of states addressing similar questions concerning indemnification for long-tail injuries involving multiple consecutive CGL policies, several have adopted some variation of the pro rata allocation approach. Under the most basic scheme of pro rata allocation, an equal share of the amount of damage is assigned to each year over which a long-tail injury occurred. The amount owed under any one policy is calculated by dividing the number of years an insurer was “on the risk” by the total number of years that the progressive damage took place.

The resulting fraction is the portion of the liability owed by that particular insurer. Some states, most notably New Jersey, utilize more complicated systems of pro rata allocation allowing for the “weighing” of each insurer’s liability to compensate for an insured’s increased perception of risk over time. (See Owens-Illinois, Inc. v. United Ins. Co., supra, 650 A.2d 974.) Significantly, all pro rata allocation methods assign liability to the insureds for those years of the continuous injury that the insureds chose not to purchase insurance.

Although some states have concluded, as the insurers urge in this case, that pro rata coverage would be more fair and equitable when compared to all sums allocation, we are constrained by the language of the applicable policies here (as noted ante, at p. 193), which supports adoption of the all sums coverage principles, as it does not differ in any meaningful way from the Montrose and Aerojet policies. (Aerojet, supra, 17 Cal.4th at p. 49.) Under the CGL policies here, the plain “all sums” language of the agreement compels the insurers to pay “all sums which the Insured shall become obligated to pay ... for damages . . . because of injury to or destruction of property . . . .” (Ante, at p. 193.) As the State observes, “[t]his grant of coverage does not limit the policies’ promise to pay ‘all sums’ of the policyholder’s liability solely to sums or damage ‘during the policy period.’ ” The insurers contend that it would be “objectively unreasonable” to hold them liable for losses that occurred before or after their respective policy periods.

But as the State correctly points out, the “during the policy period” language that the insurers rely on to limit coverage, does not appear in the “Insuring Agreement”

section of the policy and therefore is neither “logically [n]or grammatically related to the ‘all sums’ language in the insuring agreement.” The insurers’ claim that their indemnity responsibility is limited to damage occurring “during the policy period” would unduly restrict their agreement to pay “all sums” the insured is obligated to pay for damages due to “injury to or destruction of property.” The CGL policy language does not contemplate such a limited result once there is a property damage occurrence that triggers the insurers’ indemnity responsibilities for the entirety of the loss, and a growing number of states have similarly adopted this

interpretation of the all sums language. We therefore conclude that the policies at issue obligate the insurers to pay all sums for property damage attributable to the Stringfellow site, up to their policy limits, if applicable, as long as some of the continuous property damage occurred while each policy was “on the loss.” The coverage extends to the entirety of the ensuing damage or injury (Montrose, supra, 10 Cal.4th at p. 686), and best reflects the insurers’ indemnity obligations under the respective policies, the insured’s expectations, and the true character of the damages that flow from a long-tail injury.

C. Stacking Considerations As we have explained, the all sums indemnity coverage that the Court of Appeal below adopted under Montrose and Aerojet envisions that each successive insurer is potentially liable for the entire loss up to its policy limits. When the entire loss is within the limits of one policy, the insured can recover from that insurer, which may then seek contribution from the other insurers on the risk during the same loss.

Recognizing, however, that this method stops short of satisfying the coverage responsibilities of the policies covering a continuous long-tail loss, and potentially leaves the insured vastly uncovered for a significant portion of the loss, the present Court of Appeal allowed the insured to stack the consecutive policies and recover up to the policy limits of the multiple plans. “Stacking” generally refers to the stacking of policy limits across multiple policy periods that were on a particular risk.

In other words, “Stacking policy limits means that when more than one policy is triggered by an occurrence, each policy can be called upon to respond to the claim up to the full limits of the policy.” (Colon, Pay It Forward: Allocating Defense and Indemnity Costs in Environmental Liability Cases in Cal. (Feb. 2002) 24 Ins. Litig. Rep. 43, 53.) “When the policy limits of a given insurer are exhausted, [the insured] is entitled to seek indemnification from any of the remaining insurers [that were] on the risk . . . .” (J.H. France Refractories Co. v. Allstate Ins.

Co., supra, 626 A.2d at p. 509 [adopting all sums allocation and serial stacking of policies in Pa. for continuous bodily injuries caused by asbestos manufacturer]; see Koppers Co. v.

Aetna Casualty & Surety Co. (3d Cir. 1996) 98 F.3d 1440 [adopting all sums and stacking for environmental cleanup liability].) The all-sums-with-stacking indemnity principle properly incorporates the Montrose continuous injury trigger of coverage rule and the Aerojet all sums rule, and “effectively stacks the insurance coverage from different policy periods to form one giant ‘uber-policy’ with a coverage limit equal to the sum of all purchased insurance policies.

Instead of treating a long-tail injury as though it occurred in one policy period, this approach treats all the triggered insurance as though it were purchased in one policy period. The [insured] has access to far more insurance than it would ever be entitled to within any one period.” (Bratspies, supra, 1999 B.Y.U. L.Rev. at p. 1245.) The all-sums-with-stacking rule means that the insured has immediate access to the insurance it purchased. It does not put the insured in the position of receiving less coverage than it bought.

It also acknowledges the uniquely progressive nature of long-tail injuries that cause progressive damage throughout multiple policy periods. (Ibid.) In adopting the all-sums-with-stacking rule, the Court of Appeal rejected the FMC court’s antistacking ruling because it “disregarded the policy language entirely.” The Court of Appeal noted that, as in this case, the policies in FMC did not include antistacking provisions, so the FMC court resorted to “judicial intervention” in order to avoid stacking.

As the Court of Appeal recognized, absent antistacking provisions, statutes that forbid stacking, or judicial intervention, “standard policy language permits stacking.” We agree with the Court of Appeal, and find that the policies at issue here, which do not contain antistacking language, allow for its application. In so holding, we disapprove FMC Corp. v. Plaisted & Companies, supra, 61 Cal.App.4th 1132. An all-sums-with-stacking rule has numerous advantages. It resolves the question of insurance coverage as equitably as possible, given the immeasurable aspects of a long-tail injury.

It also comports with the parties’ reasonable expectations, in that the insurer reasonably expects to pay for property damage occurring during a long-tail loss it covered, but only up to its policy limits, while the insured reasonably expects indemnification for the time periods in which it purchased insurance coverage. All-sums-with-stacking coverage allocation ascertains each insurer’s liability with a comparatively uncomplicated calculation that looks at the long-tail injury as a whole rather than artificially breaking it into distinct periods of injury.

As the Court of Appeal recognized, if an occurrence is continuous across two or more policy periods, the insured has paid two or more premiums and can recover up to the combined total of the policy limits. There is nothing unfair or unexpected in allowing stacking in a continuous long-tail loss. The most significant caveat to all-sums-with-stacking indemnity allocation is that it contemplates that an insurer may avoid stacking by specifically including an “antistacking” provision in its policy.

Of course, in the future, contracting parties can write into their policies whatever language they agree upon, including limitations on indemnity, equitable pro rata coverage allocation rules, and prohibitions on stacking. Conclusion In the present case, consistent with this court’s precedent, principles of equity, and sound insurance policy

interpretation considerations, we conclude that the all sums approach to insurance indemnity allocation applies to the State’s liability for successive or long-tail property damage. In addition, we conclude that allocation of the cost of indemnification under these circumstances should be determined with stacking. Consequently, we affirm the Court of Appeal’s judgment. Cantil-Sakauye, C. J., Kennard, 1, Baxter, J., Werdegar, J., Corrigan, J., and Liu, J., concurred. On September 19, 2012, the opinion was modified to read as printed above.

Insurers are Continental Insurance Company, successor in interest to Harbor Insurance Company; Continental Casualty Company, successor by merger to CNA Casualty Company of California; Yosemite Insurance Company; Stonebridge Life Insurance Company, successor of Beneficial Fire & Casualty Company (see post, fn. 3); Horace Mann Insurance Company; and Employers Insurance of Wausau (Wausau). Excess liability insurance is coverage “whereby, under the terms of the policy, liability attaches only after a predetermined amount of primary insurance has been exhausted.” (2 Cal.

Insurance Law & Practice (Matthew Bender 1986) The Insurance Contract, § 14.02[1], p. 14-4.) Frequently there are several layers of secondary coverage, sometimes referred to as “excess insurance.” (Ibid.; see Ins. Code, § 676.6, subd. (b).) The concurring and dissenting opinion in Aerojet (Aerojet, supra, 17 Cal.4th at pp. 88-92 (conc. & dis. opn. of Chin, J.)) related to the allocation of defense costs to one insurer’s limited cashflow and self-insurance policy at issue in that case, and is not relevant to the present facts or decision. See, e.g., Owens-Illinois, Inc. v. United Ins.

Co. (1994) 138 N.J. 437 [650 A.2d 974] (adopting pro rata approach to continuous loss); see also Public Service Co. of Colorado v. Wallis & Companies (Colo. 1999) 986 P.2d 924, 935; Security Ins. Co. v. Lumbermens Mutual Casualty Co. (2003) 264 Conn. 688 [826 A.2d 107]; Atchison, Topeka & Santa Fe Railway v. Stonewall Ins. Co. (2003) 275 Kan. 698 [71 P.3d 1097]; Aetna Casualty & Surety Co. v. Commonwealth (Ky. 2005) 179 S.W.3d 830, 842; Southern Silica of Louisiana, Inc. v. Louisiana Ins. Guarantee Assn. (La. 2008) 979 So.2d 460; Boston Gas Co. v.

Century Indemnity Co. (2009) 454 Mass. 337 [910 N.E.2d 290]; Domtar, Inc. v. Niagara Fire Ins. Co. (Minn. 1997) 563 N.W.2d 724, 732; EnergyNorth Natural Gas, Inc. v. Certain Underwriters at Lloyd’s (2007) 156 N.H. 333 [934 A.2d 517]; Consolidated Edison Co. of New York v. Allstate Ins. Co. (2002) 98 N.Y.2d 208 [746 N.Y.S.2d 622, 774 N.E.2d 687]; Sharon Steel Corp. v. Aetna Casualty & Surety Co. (Utah 1997) 931 P.2d 127, 140-142; Towns v. Northern Security Ins. Co. (2008) 184 Vt. 322 [964 A.2d 1150, 1167]. See, e.g., Hercules, Inc. v. AIU Ins. Co. (Del. 2001) 784 A.2d 481, 494; Allstate Ins. Co. v.

Dana Corp. (Ind. 2001) 759 N.E.2d 1049, 1058; Goodyear Tire & Rubber Co. v. Aetna Casualty & Surety Co. (2002) 95 Ohio St. 3d 512 [2002 Ohio 2842, 769 N.E.2d 835]; J.H. France Refractories Co. v. Allstate Ins. Co. (1993) 534 Pa. 29 [626 A.2d 502]; American National Fire Ins. Co. v. B & L Trucking & Construction Co. (1998) 134 Wn.2d 413 [951 P.2d 250]; Plastics Engineering Co. v. Liberty Mutual Ins. Co. (2009) 2009 WI 13 [315 Wis.2d 556, 759 N.W.2d 613, 616].

There is precedent in the Court of Appeal for adopting the stacking rule, although the insurers correctly point out that stacking was allowed in the presence of a stipulation only. (See Stonewall Ins. Co. v. City of Palos Verdes Estates (1996) 46 Cal.App.4th 1810, 1853 [54 Cal.Rptr.2d 176] [adopting “horizontal” approach to excess liability coverage, meaning that if limits of liability of each primary insurance policy adequately cover the occurrences, there is no excess coverage expectation].) This case is the first in our court to consider the stacking of excess policies in the continuous property loss scenario.