CPC International, Inc. v. Aerojet-General Corp.

OPINION HILLMAN, Senior District Judge. On August 27, 1991, this court found CPC International, Inc., (“CPC”) directly liable under the Comprehensive Environmental Response, Compensation,. and Liability Act of 1980 (“CERCLA”), § 107(a)(2), as amended, 42 U.S.C. § 9607(a)(2), for response and remediation costs of cleaning up groundwater contamination surrounding the Ott/Story/Cordova site in Muskegon County, Michigan. 1 In the third-party action presently before the court, CPC seeks coverage and indemnification for these response and remediation costs under nineteen separate comprehensive general liability (“CGL”) policies in effect from. 1971-1977. CPC bought these policies from the fourteen third-party defendant-insurance carriers (collectively, the “insurers”). 2 Presently before the court are all insurers’ motions for summary judgment and CPC’s motion for partial summary judgment on the issues of choice of law and duty to defend. The principal point of contention between opposing parties is the timing of the “occurrences” that caused the groundwater contamination resulting in the environmental damage for which CPC seeks coverage. All third-party defendant insurers argue in their motions for summary judgment that these occurrences took place long before 1971-77, when they insured CPC. ‘ Insurers claim that the groundwater in the area of the site was contaminated in the 1960’s by the routine, intentional, and often careless waste disposal practices of both CPC’s subsidiary Ott Chemical Company (“Ott II”) and the former owner of the site (“Ott I”). CPC argues that subsequent owner Story Chemical Company’s actions in turning off the on-site purge well system in 1974 was the occurrence that permitted the spread of groundwater contamination which polluted residential wells near the site in 1974 and the Little Bear Creek .and its unnamed tributary by 1976. All insurers claim that they are entitled to summary judgment on the independent grounds of either the “known-risk” or “loss-in-progress” doctrines. All insurers claim that they are entitled to summary judgment because either the pollution exclusion clauses or the contractual definitions of “occurrence” bar coverage under their policies. Three insurers also offer proofs that coverage is barred because of CPC’s late notice to them of a possible claim. In addition, Commercial Union Insurance Company claims that “cove *799 nant-not-to-sue” language contained' in a “Hold Harmless Agreement” it had with CPC also bars coverage, CPC opposes the insurers’ motions and requests that this .court order Commercial Union Insurance Company (“CU”) to assume its contractual obligations to defend CPC in its underlying litigation with the United States Environmental Protection Agency (“USEPA”). CPC also requests that this court order CU to reimburse all CPC’s costs in defending against the USEPA’s claim. CPC claims that the record does not sufficiently establish parent company CPC’s imputed knowledge of the waste disposal practices of its wholly-owned subsidiary Ott Chemical Company for CPC to be denied coverage under its insurance policies. CU denies that it had any duty to defend CPC in the underlying claim. CPC further asks for an order of this court determining that New Jersey substantive law applies in interpreting all insurance contracts at issue in this case. The parties have submitted voluminous Fed.R.Civ.P. 56(c) materials to the court to support their claims, including copies of relevant portions of trial transcripts, exhibits, and depositions. I have carefully reviewed all these materials, as well as my findings of fact and conclusions of law in the underlying, case. In the discussion that follows, I first analyze the conflicts-of-law problem to determine what state law must apply in interpreting the insurers’ contracts. I then dispose of the insurers’ motions for summary judgment on the grounds of the “known-risk” and “loss-in-progress” doctrines. In the alternative, I conclude that three insurers are entitled to summary judgment on the grounds of CPC’s late notice to them of a possible claim. I therefore do not reach the other grounds upon which the insurers claim that they are entitled to summary judgment. I further conclude that CU had no duty to defend CPC in the underlying litigation. For the benefit of this record, I reiterate below the findings of fact relevant to this discussion that I made after a 15-day trial in the CERCLA liability phase of this case. I. Factual Background The history of the contaminated site is as follows. ■ From 1957-65, the site was owned and operated by Ott Chemical 'Company, a Michigan corporation (“Ott I”). On September 29, 1965, Ott I was purchased by a wholly-owned subsidiary of CPC, which continued operations at the site after October 1, 1965, under the same name, that of the Ott Chemical Company (“Ott II”). From 1965-72, the contaminated site was owned and operated by Ott II, CPC’s wholly-owned subsidiary. From 1972-1977, the site was owned and operated by Story Chemical Company (“Story”). Since 1977, it has been owned by the Cordova Chemical Company. 3 "In the underlying case, this court found CPC liable as an “operator” while the site was technically owned by CPC’s subsidiary Ott II, from 1965-72. In that action, I held CPC directly liable under CERCLA section 107(a) because parent corporation CPC actively participated in and exerted significant control over its wholly-owned subsidiary Ott II’s business and decision-making, including specific policy matters such as hazardous waste disposal. In this regard, I found particularly probative the fact that CPC officials actively participated in and controlled Ott II’s environmental matters, particularly through CPC’s environmental director G.R.D. Williams. CPC’s Williams helped formulate Ott II’s policies, participated in regulatory meetings, and issued directives regarding Ott IPs responses to regulatory inquiries. CPC Int’l, 777 F.Supp. at 574-75. Moreover, CPC’s Williams specifically instructed Ott II officials to limit cooperation with state and federal regulators regarding waste disposal and to consult with CPC before responding to regulatory questionnaires or other inquiries. CPC Int’l, 777 F.Supp. at 561. I based my finding that CPC was directly liable under CERCLA § 107(a) for response and remediation costs for the contaminated site on the following findings of fact concerning the releases of hazardous substances at the site: The principal source of contamination at the site was using engineered, unlined lagoons at the northwestern edge of the site *800 for chemical waste disposal. CPC Int’l, 777 F.Supp. at 556. From 1959 to at least 1968, during both Ott I’s and Ott IFs periods of ownership, wastewaters and other chemical waste used in the manufacturing process were discharged into these lagoons, from which much of the contaminants seeped into the ground and water. Id. During both Ott I’s and Ott IFs periods of ownership, chemical waste also entered the ground because workers buried and slit hundreds of drums in a sandy pit; spilled hundreds of gallons of chemicals from train ears onto railroad tracks; overflowed chemical waste at a cement-lined equalization basin; and dumped into the woods buckets of hazardous chemicals that had spilled during the manufacturing process. Id. Contamination entering the ground from disposal in the lagoons or through these other disposals or spills then seeped into the ground and migrated away from the site via the aquifer to the southeast, ultimately reaching two waterways, Little Bear Creek and the unnamed tributary. Id. Beginning in 1965 during Ott I’s period of ownership and continuing during Ott II’s period of ownership, purge wells were used intermittently at the site in an attempt to treat the groundwater contamination and to retard the spread of contamination away from the site. CPC Int’l, 777 F.Supp. at 556. II. Summary Judgment Standard Summary judgment is appropriate when there is no genuine issue as to any material fact, and the moving party is entitled to judgment as a matter of law. Fed.R.Civ.P. 56(c). The crucial issue in deciding, a motion for summary judgment is determining “whether the evidence presents a sufficient disagreement to require submission to a jury or whether it is so one-sided that one party must prevail as a matter of law.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 251-52, 106 S.Ct. 2505, 2511-12, 91. L.Ed.2d 202 (1986). In making this determination, the court must examine the record as a whole by reviewing all pleadings, depositions, affidavits, and admissions on file, drawing all justifiable inferences in favor of the party opposing the motion. Fed.R.Civ.P. 56(c); Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 587, 106 S.Ct. 1348, 1356, 89 L.Ed.2d 538 (1986) (citing United States v. Diebold, Inc., 369 U.S. 654, 655, 82 S.Ct. 993, 993, 8 L.Ed.2d 176 (1962)). The moving party bears the initial burden of showing the absence of a genuine issue of material fact. Celotex Corp. v. Catrett, 477 U.S. 317, 323, 106 S.Ct. 2548, 2552-53, 91 L.Ed.2d 265 (1986); Kramer v. Bochan Aerospace Corp., 912 F.2d 151, 153-54 (6th Cir.1990). If the moving party demonstrates that there is an absence of evidence supporting the non-moving party’s case, the party opposing the motion must come forward with specific facts showing that there is a genuine issue for trial. Matsushita, 475 U.S. at 587, 106 S.Ct. at 1356; Kramer, 912 F.2d at 153-54. To sustain this burden, the non-moving party cannot rest on the mere allegations of the pleadings. Celotex, 477 U.S. at 324, 106 S.Ct. at 2553; Fed.R.Civ.P. 56(e). Rather, the non-moving party must come forward with specific facts to support its claims and must show that there is a genuine issue for trial. Fed.R.Civ.P. 56(e). In recent years, the Supreme Court has encouraged the use of summary judgment where appropriate to ensure just, speedy and efficient determinations of cases. Celotex, 477 U.S. at 327, 106 S.Ct. at 2554-55 (1986); see Daughenbaugh v. Bethlehem Steel Corp., 891 F.2d 1199, 1205 (6th Cir.1989). As a result, mere allegations in support of the non-moving party are inadequate to resist a motion for summary judgment: “The mere existence of a scintilla of evidence in support of the position will be insufficient; there must be evidence on which the jury could reasonably find” for the non-moving party. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 252, 106 S.Ct. 2505, 2512, 91 L.Ed.2d 202 (1986); Cloverdale Equip. Co. v. Simon Aerials, Inc., 869 F.2d 934, 937 (6th Cir.1989). Moreover, “[o]nly disputes over facts that might affect the outcome of the suit will properly preclude the entry of summary judgment. Factual disputes that are irrelevant or unnecessary will not be counted.” Anderson, 477 U.S. at 247-48, 106 S.Ct. at 2509-10. III. Threshold Choice-of-Law Issue Normally, the parties themselves designate in their contracts which state’s law they wish *801 to have applied to future disputes that may arise under the contracts. Barring public policy considerations to the contrary, courts usually honor such explicit choices, since they demonstrate the intent of the parties at the time of contracting. However, the parties presently before me did not include choice-of-law provisions in their contracts. Therefore, the court must make the choice-of-law determination in this ease without knowing which state’s law the parties themselves would contemporaneously have chosen to govern future disputes. CPC asks for an order of this court determining that New Jersey substantive law applies in interpreting all insurance contracts at issue in CPC’s claims against-its former CGL insurers. CPC argues that applying Michigan’s choice-of-law rules will lead the court to apply New Jersey substantive law. Because N.J.StatAnn. 17:22-6.15 required all policies at issue to be countersigned by a New Jersey licensed agent, 4 CPC claims that this court should presume that the New Jersey countersignature endorsements on the policies were executed in New Jersey and were the last acts necessary to make the contracts binding. Therefore CPC reasons that, according to Michigan choice-of-law rules, if the last act necessary to make a contract a binding agreement was performed in New Jersey, the law of New Jersey should be applied in construing the contracts. In addition, because CPC maintains its corporate headquarters in New Jersey, CPC claims that applying New Jersey law would be consistent with CPC’s reasonable expectations that its comprehensive coverage program would be subject to one body of law. By a September 30, 1992, affidavit of Alex Moreau, Esq., 5 CPC offers proof that several of the insurers, including the Insurance Com-. pany of the State of Pennsylvania, the Insurance Company of North America, and Home Insurance Company, substantially complied with New Jersey’s statutory countersignature requirement in the endorsements on their policies at issue here. CPC submitted copies of the insurers’ policies as exhibits to a July 29, 1992, affidavit of Attorney Moreau. CPC claims that its insurers uniformly included a countersignature endorsement in their policies, which required on its face that the agent of the insurer countersign the policy in order for it to become effective. However, CPC admits that the place of countersignature is shown on only three of the nineteen policies and that none of these three policies was actually countersigned in New Jersey. 6 CPC admits that no countersigned copy of the Fireman’s Fund policy has been produced. CPC further admits that the countersignature endorsement on each of the nineteen policies postdates their respective effective dates, sometimes by as much as several months: “The countersignatures uniformly were executed after the parties were bound and were performing pursuant to these policies.' In many instances, the actual countersigning occurred months after the effective date of policies.” CPC’s Dec. 12, 1992, Memo, of Law at 5 n. 2 (citing July 29, 1992, Moreau Aff.). In different fact-specific settings, other insurers argue that this court should apply the law of other jurisdictions such as New York, California, or Bermuda in interpreting their policies. • Several insurers submitted affidavits of representatives or former employees who stated that their, policies were actually countersigned at a place other than New Jersey. • International Insurance Company and North River Insurance Company took no position on the choice-of-law issue. Since *802 these insurers have their principal places of business in New Jersey, CPC argues that their policies were presumably countersigned in New Jersey. Affiliated FM Insurance Co. submitted an affidavit' in which a representative of the company stated that the place of countersignature for its policy is unknown. The Insurance Company of North America submitted a letter stating that it cannot find a signed copy o'f its policy. Consequently, it cannot say where this policy was countersigned. Under these circumstances, the company believes that its policy became effective upon delivery of the policy and payment of the premium, both of which took place in New York. However, CPC has attached a signed and countersigned copy of this policy to the July 29, 1992, Moreau affidavit. This copy states on its face that the President and Secretary-Treasurer signed the policy at Philadelphia, Pennsylvania; it does not give a place of countersignature. Insurers such as Commercial Union Insurance Company (“CU”) and Home Insurance Company argue that CPC is not entitled to a choice-of-law ruling at this point .in the litigation. CU notes the general rule that the substantive law of the forum (Michigan) will apply where there is no conflict between the law of the forum and foreign law, and asserts that no such conflict exists here. CPC states that it seeks a choice-of-law ruling from this court at this time because Michigan, New York, California, and New Jersey law have developed conflicting theories in constructing and interpreting CGL policy language, particularly pollution exclusion clauses. CPC also asserts that New York and New Jersey law conflict with Michigan decisions as to the applicability of the “known-risk, “loss-in-progress,” and “late-notice” doctrines. Although Michigan currently uses the “place-of-contracting” choice-of-law test, insurer Newington requests this court to adopt instead the “most-significant-relationship” test in making its choice-of-law determination in this case. A. Relevant Case Law on Michigan Choice-of-Law Rules In a diversity action, the district court must apply the choice-of-law rules of the state in which it sits. Klaxon Co. v. Stentor Electric Mfg. Co., 313 U.S. 487, 491, 61 S.Ct. 1020, 1021, 85 L.Ed. 1477 (1941); National Union Fire Insurance Co. v. Watts, 963 F.2d 148, 150 (6th Cir.1992). In this case, that of course is Michigan; Michigan presently follows the Restatement (First) of Conflict of Laws rule that the nature and effect of a contract are determined by the law of the place where the contract was made. Fireman’s Fund Insurance Cos. v. Ex-Cell-O Corp., 702 F.Supp. 1317, 1328 (E.D.Mich.1988) (citations omitted). In determining where in fact a contract was made, Michigan looks to the last act necessary to make the contract a binding agreement and considers construction 'of a contract to be governed by the law of the state in which this last act occurs. State of Ohio v. Eubank, 295 Mich. 230, 233-34, 294 N.W. 166 (1940); Bonelli v. Volkswagen of America, Inc., 166 Mich.App. 483, 507, 421 N.W.2d 213, appeal denied, 430 Mich. 896 (1988). (citations omitted). However, in specifically interpreting insurance contracts, the Michigan Supreme Court has never decided what is the last act necessary to make an insurance contract a binding agreement. More than twenty years ago, a federal district court sitting in Michigan predicted that the Michigan Supreme Court would adopt the rule that, if an insurance contract required a countersignature before it could be binding on the parties, the state of countersigning will be deemed the place where the contract was made and the law of that -state will be applied in construing the contract. Chrysler Corp. v. Ins. Co. of North America, 328 F.Supp. 445, 449 (E.D.Mich.1971). Accord Ins. Co. of North America v. Forty-Eight Insulations, Inc., 633 F.2d 1212, 1219 (6th Cir.1980), reh’g granted in part on other grounds, 657 F.2d 814 (6th Cir.), cert. denied, 454 U.S. 1109, 102 S.Ct. 686, 70 L.Ed.2d 650 (1981). However, both Chrysler and Forty-Eight Insulations were personal injury, not hazardous waste, cases. Moreover, in neither case was the court asked to construe contracts, for which the countersignatures were admittedly not the last acts necessary to make binding agreements. In Chrysler and Forty-Eight Insulations, the *803 countersignatures apparently were the last acts necessaryto make binding agreements. Morbark Industries, Inc. v. Western Employers Ins. Co., 170 Mich.App. 603, 615, 429 N.W.2d 213 (1988), appeal denied, 432 Mich. 896 (1989), was the first Michigan appellate opinion to address choice-of-law' questions while construing contracts with countersignature endorsements. In Morbark; the parties asked the court to construe two different insurance contracts for umbrella or excess liability coverage. Plaintiffs insurers under its general liability policies had become insolvent. Plaintiff, a Michigan manufacturer, therefore demanded that its excess insurers provide coverage and defend the products liability suits brought against it. The first contract was not countersigned, although the parties stipulated that it was a valid contrac~t. This policy was with Western Employers Insurance Company, a California insurer. It contained a countersignature endorsement that read as follows: It is agreed that the signature appearing on this endorsement is the signature of a person duly authorized to countersign on behalf of the Company in the state designated above [Michigan] and which is appended hereto in cortfor'inity with the insurance laws of that state [Michigan]. Morbark, 170 Mich.App. at 615, 429 N.W.2d 213 (emphasis added). Plaintiff insured (the Michigan manufacturer) had urged the trial court to apply the law of California, the insm~er's home state, in construing this contract. However, the trial judge concluded that the parties intended the countersignature endorsement (quoted above) to be part of thO agreement and therefore intended the contract to be a Michigan contract. Consequently, the trial judge concluded that Michigan law applied in interpreting this contract, then granted summary judgment to the defendant insurer. On appeal, the Michigan appellate court affirmed. It stated that a requirement for countersignature endorsements existed, in Michigan at least, to assure that the law of the state "would govern contracts of insurance entered into by the citizens thereof." Morbark, 170 Mich.App. at 616, 429 N.W.2d 213. Although Michigan's countersignature requirement was repealed in 1972, the court concluded that the countersignature endorsement conforming to Michigan's former statutory requirement could only have been intended to meet the statutory purpose of that former law. The court therefore: held that the Western Employers contract was a Michigan contract. Morbark, 170 Mich.App. at 616, 429 N.W.2d 213. The second policy construed by the Mor-bark courts was between plaintiff and First State Insurance Company, a Massachusetts insurer. The Michigan appellate court stated that this contract was countersigned in Massachusetts and that the parties agreed the law of Massachusetts was controlling. Id. The court cqncluded that this contract was governed by Massachusetts law, discussed recent Massachusetts decisions, then affirmed the trial court's judgment in favor of the defendant insurer; B. Discussion CPC requests that 11 adopt the Mor-bark court's conclusion dealing with the first contract it construed, the one that was not in fact countersigned. The Morbark court considered the last act binding the parties to this contract to be their presumed compliance with Michigan's former statutory coun-tersignatüre requirement. In other words, although the contract was actually never countersigned, the court nevertheless treated it as if it was countersigned in Michigan. The court therefore applied Michigan law in construing this contract. CPC requests that I follow the same tack and consider the last acts binding the parties to their agreements in the contracts before me to be their presumed compliance with New Jersey's former statuto~y countersignature requirement. GPO therefore requests that I apply New Jersey law in construing these contracta 7 *804 For four reasons, I find no merit in this argument and have concluded that it would be inappropriate to apply New Jersey law. First, the relevant case law does not support applying New Jersey law. Second, the parties did not have any reasonable expectations as to which body of law would apply other than the law of the forum state, which is Michigan. Third, since on this record the countersignatures were admittedly not the last acts necessary to make binding agreements, the facts of the case do not support my applying New Jersey law. Fourth, I am satisfied that the Michigan Supreme Court, faced today with deciding this choice-of-law issue in a hazardous waste case in which the contaminated site is located in Michigan, would apply Michigan law. For these reasons, which I discuss in turn below, I will apply Michigan law to resolve the issues in dispute between the parties before me now. 1. Relevant Case Laid on Neiv Jersey Countersignature Statute For three reasons, I conclude the relevant case law does not support applying New Jersey law in interpreting the contracts at issue in this case. First, I am not convinced by CPC’s analysis that the Michigan Mor-bark court’s discussion about the legislative purpose of the former Michigan statutory countersignature requirement has any bearing whatsoever upon determining the legislative purpose of the former New Jersey statutory countersignature requirement. The Morbark court stated that countersignature requirements existed in Michigan to assure that the law of the state “would govern contracts of insurance entered into by the citizens thereof.” Morbark, 170 Mich.App. at 616, 429 N.W.2d 213. However, the Morbark court did not hold that this was the purpose of New Jersey’s former statutory countersignature requirement. Morbark cannot be read as making a finding on the purpose of the New Jersey countersignature statute. That was never before the Michigan court. Quite the contrary, a New Jersey Superior Court has recently stated, that this was not the purpose of New Jersey’s former statutory countersignature requirement. CPC International, Inc., et al. v. Hartford Accident and Indemnity Co. et al., 262 N.J.Super. 191, 620 A.2d 462 (1992). In Hartford, the court concluded that the New Jersey legislature did not intend New Jersey law to apply every time there was a countersignature by a New Jersey resident agent. Historically, the two principal objectives for requiring countersignature by New Jersey, resident agents of insurance contracts were “to protect the interest of all resident agents and to assure payment of premium taxes.” Id. at 36. I accept this conclusion , of the New Jersey Superior Court regarding the legislative purpose of New Jersey’s countersignature requirement. Id. I therefore conclude that New Jersey’s former statutory countersignature requirement was not adopted to assure that its law governed insurance contracts entered into by its citizens. Second, even assuming, arguendo, that the purposes of New Jersey’s and Michigan’s former statutory countersignature requirements were similar, the rationale used by the Morbark court to construe a contract by presuming compliance with á former statutory countersignature requirement could possibly apply only to one of the nineteen contracts at issue here. The Fireman’s Fund policy is the only one of CPC’s CGL contracts with the defendant insurers for'which a countersigned copy was not produced. In Morbark, when the policy was countersigned, the court applied the law of the state in which the contract was in fact countersigned. CPC admits that eighteen of the nineteen policies at issue in this case were countersigned. For these eighteen contracts, Morbark would therefore provide authority to apply the law of the states in which these contracts were in fact countersigned, not the law of New Jersey, the state with whose former statutory countersignature requirement CPC claims the parties intended to comply. Moreover, insurer Fireman’s Fund has submitted an affidavit of its employee Carol Fatzer that offers proof of countersignature in the state of New York: It was the customary practice of FFIC’s New York branch office to type the insurance contracts in carbon copy sets. The actual countersignature would only be affixed on one copy. While the counter *805 signed copy has not been produced, the countersignature date, as well as other in-dicia on the insurance contract, confirm that it would have been countersigned in FFIC's New York branch office. Affidavit of Carol Fatzer at 3. Drawing all justifiable inferences in favor of the party opposing CPC's motion for summary judgment, as I must, I conclude that this affidavit satisfies Fireman's Fund's, burden to come forward with specific facts to support its claim that its contract was countersigned in New York. Fed.R.Civ.P. 56(e). Morbcvrk does not therefore entitle CPC to a ruling from this court that the Fireman Fund's policy must be deemed countersigned in New Jersey. Third, the Third Circuit case law that CPC cites in support of its claim that New Jerse~r law should be applied to the contracts at issue in this case does not actually support this position. , See Wootton Hotel Corp. v. Northern Assur. Co., 155 F.2d 988 (3d Cir), cert. denied, 329 U.S. 758, 67 S.Ct. 111, 91 L.Ed. 654 (1946). In Wootton, defendant insurer offered parol evidence that, notwithstanding the requirements of the countersignature ~tatute, the insurer's agent was not actually in New Jersey when he countersigned the policy, but had in fact countersigned it in Pennsylvania. Nevertheless, the court held that the countersignature was deemed to have taken place in New Jersey, as indicated in the policy itself and as required by the New Jersey statute. Id. at 990. 8 The court in Wootton thus concluded that the insurer's failure to abide by New Jersey's statutory countersignature requirement would not defeat the applicability of New Jersey law under the circumstances at issue in that case. This is the result that CPC asks this court to reach in the present case. However, in a much more recent case, the Third Circuit affirmed the district court's holding that Pennsylvania, not Ne* Jersey, law would govern interpreting the insurance policies at issue in the case because the policies were actually countersigned in Phila-deiphia. Armotek Industries, Inc. v. Employers Ins. of Wausau, 952 F.2d 756 (3d Cir.1991). While the New Jersey countersignature statute "required that the policies be countersigned by an agent licensed in New Jersey, the statute did not state that its agent's signature had to be affixed in New Jersey." Armotek, 952 F.2d at 760-61. In Avinotek, the Third Circuit explicitly explained that it had not held otherwise in the earlier Wootton. Id. at 761 n. 6. According to the Third Circuit's recent interpretation of New Jersey's former statutory counte~signa-ture requirement, what matters is the place at which an insurance policy was in fact countersigned, not, as CPC claims, the place at which it should have been countersigned. Armotek, 952 F.2d at 760-761. I accept the Third Circuit's recent interpretation of this statute and conclude that it does not require my presuming that all the policies at issue in this case were countersigned in New Jersey. Accord CPC International Inc., et al. Hartford Accident and Indemnity Co., et al., 262 N.J.Super. 191, 620 A.2d 462 (1992). 2. Reasonable Expectations of the Parties Both CPC and the insurers conduct business on a national and international basis, dealing with insurance policies covering risks that are located in numerous different locations in numerous different forums. Any of tlvise parties can therefore reasonably expect to sue or be sued over insurance coverage disputes in many different jurisdictions. The only certain way for parties to state their reasonable expectations of the substantive law that will apply in such disputes is to negotiate a choice-of-law clause into their contracts. The parties presently before me did not do this. 1 conclude that none of the parties had any reasonable expectations as to which body of law would apply in resolving the disputes at issue here other than the law of the forum state, which is Michigan. 3. Practical Difficulties in Determining Contract Formation Neither the federal cases nor the Michigan appellate case cited above that discussed *806 countersigning as the last acts necessary to make ■ a binding agreement did' so in the context of deciding the very practical difficulties in determining contract formation in such complex factual situations as are before me here. None of these previous eases contained an admission, as there is by CPC in this case, that the countersignatures on the contracts at issue were uniformly executed after, sometimes long after, the parties were bound and were performing pursuant to those contracts. By definition, if. the parties were already bound and performing pursuant to the contracts before the contracts were countersigned, the countersignatures could not have been the last acts necessary to, make binding agreements. The papers filed by the parties did not claim that these contracts of insurance did not exist for the period of time from the effective dates from which premiums had been paid to procure insurance to the dates upon which countersignature endorsements were finally executed for the contracts. If valid contracts of insurance existed on the respective effective dates given in the contracts, some other acts constituting acceptance — whether they were oral or written, or whether they were payment for or issuance and delivery of the contracts — must have been the last acts necessary to make these contracts binding agreements. The parties have not provided the court with the kind of detailed chronology of facts relating to contract formation that would be necessary to decide such issues, presumably because such evidence no longer exists. 9 On this record, I conclude that the countersignatures on the contracts at issue in this case were not in fact the last acts necessary to make -binding agreements. The facts of the case therefore do not support my applying New Jersey law to interpret contracts for which acts other than presumed countersignature in New Jersey were the last acts necessary to make binding agreements. 4. Likely Action of the Michigan Supreme Court if Deciding This Issue Furthermore, even if such evidence was freely available, I am satisfied that the Michigan Supreme Court, faced today with deciding this choice-of-law issue in a case of such factual complexity, would not mechanically apply the “plaee-of-contracting” rule that it articulated in the 1940’s and has not revisited since. See State of Ohio v. Eubank, 295 Mich. 230, 233-34, 294 N.W. 166 (1940). As discussed above, there are no Michigan Supreme Court cases that define the last act necessary to make an insurance contract. Moreover, neither the federal cases nor the Michigan appellate case cited above that discussed countersigning as this last act were hazardous waste cases, in which the contaminated site was located in the forum state. In addition, the factually complex case presently before me involves a hazardous waste dump sité at which exists “some of the worst groundwater contamination [Michigan] state regulators have encountered to date.” CPC Int’l, 777 F.Supp. at 563. The United States Environmental Protection Agency (“USEPA”) placed this site on the federal government’s National Priorities List of locations in need of a long-term remedial response in 1982. I have previously documented the public health and safety threats posed by the conditions at this site: The site’s environmental problems were legion. The toll exacted by decades of hazardous waste contamination was abundantly clear. Groundwater pumped to the surface contained foam and a brownish color like root beer. The stench of chemicals permeated the air. Soil excavation revealed the taint of toxic pollution, showing purplish colors. Hundreds of chemical drums, many piled atop each other, lay around the site, randomly strewn among trees, across pavement and into sandy pits. Many of the drums and barrels were crushed, corroded and leaking, with their contents seeping into the ground. Chemical waste by-products, including thousands of broken bottles, littered the land. Among the chemicals found at the site were potentially deadly toxics. Tanks of explosive phosgene gas, potentially deadly *807 if released into the air, presented' a serious risk to neighboring residents in the event of vandalism or an accident. Other chemicals known as probable human carcinogens were found scattered around the site, including benzene, phenol, methylene -chloride, and methyl isocyanate. CPC Int'l 777 F.Supp. at 554-555. In diversity cases, if the forum state’s highest court has not addressed an issue, the federal court must determine from all available data, including “the decisional law of the state’s lower courts, restatements of law, law review commentaries, and decisions from other jurisdictions on the Majority’ rule,” what the state’s highest court would do if deciding the issue. Grantham and Mann, Inc. v. American Safety Products, Inc., 831 F.2d 596 (6th Cir.1987). Neither the federal cases nor the Michigan appellate case discussed above applied the First Restatement ’s “place-of-contracting” choice-of-law rule to countersignatures that were admittedly not the last acts necessary to make the contracts at issue binding agreements. None of these previous cases therefore addressed the very practical difficulties in determining contract formation in the complex factual situations before me now. Moreover, a recent article surveying choice-of-law rules applied in contract disputes states that “the modern approach” to resolving conflicts of law in this context involves “determining which state has the most significant relationship to the insurance litigation.” Edward J. Zulkey, Resolving Choice of Law Questions in Insurance Disputes, THE BRIEF, Summer 1990, at 43. The Second Restatement formulated the “most-significant-relationship” test as follows: (1) In the absence of an effective choice of law by the parties, the contacts to be taken into account to determine the law applicable to an issue include: (a) the place of contracting, (b) the place of negotiation of the contract, (c) the place of performance, (d) the location of the subject matter of the contract, (e) the domicile, residence, nationality, place of incorporation and place' of business of the parties. Restatement (Second) of Conflict of Laws § 188 (1971). In addition, the trend in many jurisdictions that' have recently abandoned the “place-of-contracting” choice-of-law rule is to adopt the Second Restatement’s “most-significant-relationship” choice-of-law rule. Zulkey, THE BRIEF' at 44. 10 A New Jersey appellate couft, for example, recently chose to adopt the “most-significant-relationship” rule in a hazardous waste case similar to the one now before me. That court applied the law of the forum state (New Jersey) in an action brought by an insured against its liability insurers for declarations of potential coverage. Johnson Matthey, Inc. v. Pennsylvania Mfrs’ Ass’n Ins. Co., 250 N.J.Super. 51, 593 A.2d 367 (N.J.Super.Ct.App.Div.1991). The case involved determining whether pollutants that leaked onto the land constituted a “sudden, and accidental” “occurrence.” The New Jersey appellate court explained as follows its rationale for' rejecting the “place-of-contracting” test: “In these days of multistate insurers, multistate insureds, and instantaneous interstate transmission of voice and document, it. is not easy to identify a state of contracting.” Johnson Matthey, 593 A.2d at 372. “A Delaware company, for example, secures a casualty insurance policy for a New Jersey site, among others, through a Philadelphia agent from an insurer with a Hartford home office that retains final underwriting approval on large policies. The handshake deal for the insurance is made over lunch in Manhattan.” Id. “Choosing a locus contractu in such a case would be a difficult and perhaps pointless exercise. Pointless, because there is nothing about the choice that tells very'much about the insur- *808 anee transaction involved.” Johnson Matthey, 593 A.2d at 372. The New Jersey appellate court then discussed its rationale both for adopting the “most-signifieant-relationship” choice-of-law rule and for judging the place of the insured contaminated site as the most important of the many factors that it considered in applying this rule. “The existence or absence of insurance proceeds can very well determine whether'or not a waste site is remediated or a toxic tort victim is compensated.” Johnson Matthey, 593 A.2d at 370. “New Jersey’s paramount interest in the remediation of toxic waste sites, and in the fair compensation of victims of pollution, extends to assuring that casualty insurance companies fairly recognize the legal liabilities of their insureds.” Id. “New Jersey’s urgent concern is for the health and safety of its citizens.... The gravity of the problem is out of doubt; the necessity of finding solutions is manifest.” Johnson Matthey, 593 A.2d at 371. 11 Having discussed the relatively strong interest of the forum state (New Jersey) 'in having the contaminated site remediated, the New Jersey appellate court in Johnson Mat-they then discussed the “relatively remote” interest of the place of contracting (Pennsylvania) in having the contract interpreted according to its law. The court stated , that Pennsylvania had no legitimate interest “in preventing the fair and predictable application of foreign law to Pennsylvania insurance contracts” written to cover insureds’ liabilities in other states. Johnson Matthey, 593 A.2d at 371. As the place of contracting, the court stated that Pennsylvania might justifiably assert an interest in the “uniform interpretation” of Pennsylvania insurance contracts in other states. Id. However, the court concluded that this goal was finally “illusory” and “must give way to New Jersey’s paramount interest in the health and safety of its people.” Johnson Matthey, 593 A.2d at 371. In deciding the coverage issue before it in a similar hazardous waste case, the New Jersey appellate court thus chose not to apply the First Restatement’s “place-of-contracting” choice-of-law rule. The court chose instead to follow what it considered to be the more appropriate Second Restatement’s “most-significant-relationship” choice-of-law rule. This choice led the court to apply the law of the .forum state, the state in which the contaminated site was located, rather than the law of the state where the contract was made. The appellate court therefore reversed the trial court, which had decided that Pennsylvania law applied to the action. The appellate court held instead that New Jersey law governed the meaning of the terms “sudden and accidental” used in the policies at issue in the case. Johnson Matthey, 593 A.2d at 367. Michigan’s “place-of-cóntraeting” and “last-act-necessary” choice of-law rules were formulated long before our present era of hazardous waste litigation was opened by the Comprehensive, Environmental Response, Compensation, and Liability Act, 42 U.S.C. § 9601 et seq. See Rubin v. Gallagher, 294 Mich. 124, 128, 292 N.W. 584 (1940); State of Ohio v. Eubank, 295 Mich. 230, 233-234, 294 N.W. 166 (1940). Today, those states that have adopted the Second Restatement’s “most-significant-relationship” choice-of-law rule have considered a “critical factor” in hazardous waste cases to be the location of the dump site. Zulkey, THE BRIEF at 45. I conclude that, if presented with the choice-of-law issue presently before me, the Michigan Supreme Court would also consider a “critical factor” in this hazardous waste case to be. the location of the -dump site. In this situation, I conclude that the Michigan Supreme Court would find similar “paramount interest” in the health and safety of its citizens and would employ a similar means as the New Jersey appellate court in Johnson Matthey did in attempting to effect this interest. *809 I therefore conclude that the Michigan Supreme Court would not mechanically apply a “place-of-contracting” test that would result in the ■ pointless application of the law of several different forums to the numerous insurance transactions at issue here. I further conclude that the Michigan Supreme Court would in this case adopt the “most-significant-relationship” choice-of-law test, judge the place of the insured contaminated site to be the most important of the many factors that it considered in applying the test, and apply the law of the forum state to resolve the issues disputed between the parties here. I will therefore apply Michigan law to resolve the disputed issues presently before me in this third-party-diversity action. IV. Insurers’ Motions for Summary Judgment A. Relevant Case Law on the “Known-Risk” and “Loss-in-Progress” Doctrines According to Michigan law, insurance provides coverage for fortuitous events only. No coverage exists for events that must happen or that have happened. See Nash v. New York Life Ins. Co., 272 Mich. 680, 682, 262 N.W. 441 (1935), in which the court stated, “Insurance policies are contracts to indemnify against contingent losses.” See also Harper v. Michigan Mut. T., C. & W. Ins. Co., 173 Mich. 459, 139 N.W. 27 (1912); Gauntlett v. Sea Ins. Co., 127 Mich. 504, 86 N.W. 1047 (1901); Wigle v. Aetna Casualty & Surety Co., 177 F.Supp. 932, 933 (E.D.Mich.1959). In a pollution setting, several Michigan courts have held that a loss was uninsurable when it was a “known risk,” that is, when the insured knew or should have known that there was a substantial probability of loss before the contract period began. See Mitchell v. Ins. Co. of North America, No. 86-51880-CK (Kent Cty. Cir.Ct. Nov. 11, 1988), in which the court specifically embraced the “known-risk” doctrine as part of Michigan law. See also County of Kent v. Home Ins. Co., No. 85-46740-CK (Kent Cty. Cir.Ct. Apr. 3, 1990); Central Quality Services Corp. v. Ins. Co. of North America, No. 87-CV-74473-DT, slip. op. (E.D.Mich. Sept. 6, 1989) (Central Quality I), aff'd, 977 F.2d 580 (6th Cir.1992) (Central Quality II); Inland Waters Pollution Control, Inc. v. National Union Fire Ins. Co., 783 F.Supp. 325, 329 (E.D.Mich.1992). The court in Inland Waters also adopted the “loss-in-progress” doctrine. According to this doctrine, no part of a loss may be insured when it was already in progress before the contract period, that is, when the damage for which coverage is claimed began to occur before the contract period and continues into the contract period. Inland Waters, 783 F.Supp. at 328. The facts in several of the cases cited above that adopted the “known-risk”'doctrine are similar to the facts in the casé presently before me. In County of Kent, for instance, numerous insurance companies whose coverage of the county began in 1980 brought motions for summary judgment. Based upon several reports made by the Michigan Department of Natural Resources (“MDNR”) and the Department of Health, including a 1979 report that pinpointed the 'county’s landfill as the 'source of the groundwater contamination, insurers argued that the county knew of the groundwater contamination before the policies took effect. The court noted that the remedial action undertaken at the direction of the MDNR had determined that the county was the source of the contamination. Based upon the information before it, the court accepted and applied the “known-risk” doctrine, granting the insurers’ motions for summary judgment: “This court is convinced that where the insured has knowledge prior to the beginning of a policy period of a substantial probability that an event has occurred which will in the near future produce a monetary loss, that the insurance company has no liability.” County of Kent v. Home Ins. Co., No. 85-46740-CK at 7 (Kent Cty. Cir.Ct. Apr. 3, 1990). . In other words, to allow coverage in such a situation would be akin to insuring a ship after it has sunk. Id. A similar result was reached in Central Quality I, where plaintiff insured, who operated an industrial dry cleaning business from 1968 to 1987, routinely discharged wastewa-ter containing perchloroethylene into unlined lagoons, a dry well, and a stream. In 1974, *810 the MDNR noted elevated levels of the chemical in the wastewater discharges to the lagoons. In 1978, the MDNR concluded that plaintiff insured’s discharges had contaminated the groundwater and wells surrounding its plant. Despite these facts, the company took out insurance between 1980-87. When sued for coverage, insurers requested summary judgment. The court noted, “For the known risk doctrine, the relevant question is whether the insured- knew or reasonably should have known there was a substantial probability of a loss before the policy period began. If such probability exists, then the risk is known and cannot be insured against.” Central Quality Services Corp. v. Ins. Co. of North America, No. 87-CV-74473-DT, slip. op. at 29 n. 13 (E.D.Mich. Sept. 6, 1989) (Central Quality I), aff'd, 977 F.2d 580 (6th Cir.1992) (Central Quality II) (citations omitted). The court in Central Quality I adopted the “known-risk” doctrine because it was a “commonly accepted” premise of insurance law throughout the United States. Id. The court also decided that, when the risk was known before the policy of insurance took effect, “there is no need to reach the question of whether an ‘occurrence’ took place.” Id. The court then concluded that plaintiff insured knew or should have known that there was a substantial probability of loss due to perchloroethylene by 1978, before insurers’ policies- took effect, and granted summary judgment for defendant insurers. The Sixth Circuit affirmed this result, specifically agreeing with the district court’s conclusion that the “known-risk” doctrine applied under the circumstances 'of the case. Central Quality II, 977 F.2d at 580. The Michigan Supreme Court has not yet ruled on the “known-risk” and “loss-in-progress” doctrines. However, the court in Inland Waters adopted both of these doctrines for the following reasons: the weight of áu-thority of other jurisdictions; the indications within the state of Michigan that these doctrines would be acceptable; and the court’s belief that the Michigan Supreme Court, if presented with this question, would adopt these doctrines into Michigan law. Inland Waters Pollution Control, Inc. v. National Union Fire Ins. Co., 783 F.Supp. 325, 329 (E.D.Mich.1992) (Inland Waters III) 12 (citations omitted). The factual background of the Inland Waters cases is as follows. Plaintiff insured, who was in the business of cleaning up and transporting waste materials, disposed of and crushed several hundred drunis containing liquid waste on the “Strieker” site in January and February 1981. The first liability policy under which plaintiff sought coverage took effect in August 1981. Several years later, after the MDNR instructed Strieker Corp. to conduct a hydrogeological investigation of its property, Strieker’s environmental consultant noted contamination in the area in which plaintiff had disposed of the drums. Strieker filed suit against plaintiff, whose insurer refused to defend it in this lawsuit or to indemnify it for any claims made in this matter. When plaintiff then brought suit against its insurer for breach of the insurer’s obligations under its “occurrence”-based and “claims-made” policies of insurance, the district court granted summary judgment in favor of the insurer. Inland Waters Pollution Control, Inc. v. National Union Fire Ins. Co., No. 89-70584, slip op. (E.D.Mich. Oct. 29, 1990) (Inland Waters I), aff'd in part, rev’d and remanded in part, 943 F.2d 52 (6th Cir.1991) (Inland Waters II). In Inland Waters II, the Sixth Circuit concluded that the Michigan Supreme Court would adopt the “injury-in-fact” trigger of coverage if that court was deciding the issues presented in Inland Waters II. This approach provides that coverage is triggered when real personal injury or actual property damage first occurs. The Sixth Circuit concluded that the soil contamination had first occurred in January or February 1981, before the policy at issue had taken effect. Consequently, the'court held that there was no coverage under the policy for the soil contamination- and affirmed the district court’s grant of summary judgment to the insurer on this issue. *811 However, the Sixth Circuit could not determine from the record before it when the groundwater contamination first occurred. “If evidence establishes that the groundwater was first contaminated after August 1981, there could be coverage under the occurrence policy, and [insurer] National Union could be liable for breach of the duty to defend.” Inland Waters Pollution Control, Inc. v. National Union Fire Ins. Co., 943 F.2d 52 (6th Cir.1991) (Inland Waters II). “On the other hand, if evidence establishes that the groundwater first became contaminated before August 1981, there would be no coverage, and, hence, no duty to defend.” Id. The dispositive issue for the Sixth Circuit in Inland Waters II was therefore determining whether groundwater contamination had occurred before the policies at issue took effect. If it had, the court concluded -that there could be no coverage. Because the court could not determine from the record whether it had or not, the court reversed the district court’s grant of summary judgment on the issue of the insurer’s obligations to indemnify or defend in a suit seeking damages for groundwater1 contamination. The court then remanded the case for the district court to resolve the factual question as to when the groundwater first became contaminated. Upon remand, defendant insurer renewed part of its motion for summary judgment that had not previously been reached by the court, claiming that the “known-risk” and the “loss-in-progress” doctrines barred plaintiffs suit. Inland Waters Pollution Control, Inc. v. National Union Fire Ins. Co., 783 F.Supp. 325 (E.D.Mich.1992) (Inland Waters III). In response, plaintiff insured introduced the affidavit of an expert who stated that plaintiff could not have known with any certainty the future risks to the water table that would ultimately result from the spill in question. After adopting both the “known-risk” and the “loss-in-progress” doctrines, the court concluded that this expert’s affidavit raised a genuine issue of material fact as to when plaintiff “reasonably should have known” that groundwater contamination would ultimately result from this spill. 13 Id. at 329. The court therefore denied summary judgment to defendant insurer under the “known-risk” doctrine. However, the court found no factual dispute under the “loss-in-progress” doctrine. “Plaintiff spilled the contaminants on the ground. Despite its efforts at mitigating the damage, the water table became contaminated as a result of the spill. The damage may have taken years to manifest itself, but it is a direct result of the, spill caused by plaintiff.” Id. Because the spill occurred before the effective date of the insurance contract, the court concluded that defendant insurer was not liable to plaintiff for insuring against any injury flowing from that occurrence. The court therefore granted summary judgment to the defendant insurer under the “loss-in-progress” doctrine. B. Discussion I agree with the court in Inland Waters III that, if presented with the question, the Michigan Supreme Court would adopt the doctrines of “known risk” and “loss in progress” into Michigan law. I therefore adopt these doctrines. Inland Waters Pollution Control, Inc. v. National Union Fire Ms. Co., 783 F.Supp. 325, 329 (E.D.Mich.1992). See also Central Quality Services Corp. v. Ins. Co. of North America, No. 87-CV-74473-DT, slip. op. at. 29, n. 13 (E.D.Mich. Sept. 6, 1989) (Central Quality I), aff'd, 977 F.2d 580 (6th Cir.1992) (Central Quality II), in which the courts accepted and applied the “known-risk” doctrine. According to Michigan law, knowledge acquired by employees within the scope of their employment is imputed to the corporation: “In consequence, a corporation cannot plead innocence by asserting that the information obtained by several employees was not acquired by any one individual employee who then would have comprehended its full import.” Upjohn Co. v. New Hampshire Ins. Co, 438 Mich. 197, 476 N.W.2d 392, *812 400-401 (1991). The corporation is thus considered to have the collective knowledge of its employees and is held responsible for. their failure to act appropriately. Id. In the underlying case, I held that CPC is directly liable under CERCLA section 107(a) as an operator because parent corporation CPC actively participated in and exerted significant control Over its wholly-owned subsidiary Ott II’s business and decision-making, including specific policy matters such as hazardous waste disposal. CPC Int’l, 777 F.Supp. at 574-75. I reached this conclusion only after engaging in a highly fact-specific inquiry into the history of CPC’s involvement with Ott II. Id. As I specifically found in the underlying case, several individuals served simultaneously as top-ranking Ott II and CPC officials. These CPC officials served within Ott II’s management in such positions as president and chief executive officer. CPC Int’l, 777 F.Supp. at 559. During CPC’s entire period of ownership of Ott II, the chairman of Ott II’s board was always a top-ranking CPC official. No fewer than three CPC-affiliated directors served on Ott II’s board at all times, and CPC, as 100-percent shareholder, controlled the selection of board members. CPC Int’l, 777 F.Supp. at 558. Moreover, discussions of waste disposal problems and potential solutions were major topics of discussion within both Ott II’s management structure and its board. CPC Int’l, 777 F.Supp. at 561. In basing my holding that CPC was directly liable under CERCLA section 107(a) as an operator on the fact that parent corporation CPC actively participated in and exerted significant control over its wholly-owned subsidiary Ott II’s business and decision-making, I found particularly probative the role played by CPC’s environmental director G.R.D. Williams. CPC Int’l, 777 F.Supp. at 561, 575. CPC’s Williams coordinated all pollution activities for CPC and its divisions and subsidiaries and became heavily involved in environmental issues at Ott II. CPC Int’l, 777 F.Supp. at 561. As more fully discussed in the opinion in the underlying case, CPC’s Williams actively participated in and’ exerted control over a variety of Ott II’s environmental matters. Id. On this record, I find no ■merit , to CPC’s claim that the record does not sufficiently establish parent company CPC’s imputed knowledge of the-waste disposal practices of its wholly-owned subsidiary Ott II. On this record, I conclude that CPC knew of and was responsible for the waste disposal practices of its wholly-owned subsidiary Ott II. I further conclude that CPC’s environmental director Williams, as well as other CPC officials who served as Ott II’s top managers or as members of its board, of directors, either knew or reasonably should have known that there was a substantial probability of a loss caused by groundwater contamination in the area surrounding the Ott/Story/Cordova site before the policy periods at issue, here began. See CPC Int’l, 777 F.Supp. at 555-56, particularly the following finding of fact: “Beginning in 1965 during the Ott I period of ownership and continuing during the Ott II period, purge wells were used intermittently at the site in an attempt to treat the groundwater contamination and retard the spread of contamination away from the site” (emphasis added). See also the March 16,1965, letter, from the Michigan Water Resources Commission (“MWRC”) to CPC’s subsidiary Ott II, which stated that the MWRC was increasingly concerned about the method of disposal of Ott II’s waste, that the quality of the groundwater in the area had bqen significantly altered, and that there was no reason to believe that the affected area was or would be confined to the company’s property. 14 See also the January 4, 1966, letter from the MWRC to Ott II, which stated that Ott II’s groundwater pumpage to Little and Big Bear.Creeks resulted in the presence of waste constituents “of much more consequence than chlorides, including compounds which result in tainting of fish flesh.” 15 See also > the January 17, 1966, memorandum of an Ott II employee, which reported the MWRC’s conclusion that Ott IPs’ waste discharges were contaminating Little Bear Creek and the main stream of *813 Bear Creek, and that continuing such discharges to the stream would similarly contaminate Bear Lake, a popular fishing spot. In this memorandum, the Ott II employee divided the problem into two categories, the second of which was “the purge of the present ground waters to clean up the situation which has grown over the time that the plant has been in operation.” 16 .On this record, I conclude as a matter of law that CPC knew or reasonably should have known that there was a substantial probability of loss caused by groundwater contamination in the area surrounding the Ott/Story/Cordova site by at least the mid to late 1960’s, long before the effective dates of any of the insurance policies at issue here. 17 See Upjohn, 438 Mich. at 400-01, 476 N.W.2d 392. Since such probability existed, the risk was known and could not be insured against. Central Quality Services Corp. v. Ins. Co. of North America, No. 87-CV-74473-DT,. slip, op. at 29 n. 13 (E.D.Mich. Sept. 6, 1989) (Central Quality I), aff'd, 977 F.2d 580 (6th Cir.1992) (Central Quality 77); County of Kent v. Home Ins. Co., No. 85-46740-CK (Kent Cty. Cir.Ct. Apr. 3, 1990). All defendant insurers are thus entitled to summary judgment under the “known-risk” doctrine. In addition, I conclude as a matter of law that this record establishes that the loss was already in progress before the earliest contract period for insurance began on October 1, 1971. See CPC Int’l, 777 F.Supp. at 555-56. See also the MRWC’s March 16, 1965, and January 4, 1966, letters and the Ott II employee’s January 13, 1966, memorandum cited above. Since the loss was already in progress before the effective dates of the insurance policies at issue.here, no part of this loss may be insured. Inland Waters III, 783 F.Supp. at 329. Cf. Inland Waters II, 943 F.2d 52 (6th Cir.1991), where, in reversing the Inland Waters I court’s grant of summary judgment to defendant insurers and remanding the case for eventual disposition in Inland Waters III, the Sixth Circuit stated: “If evidence establishes that the groundwater first became contaminated before August 1981 [the effective date of the ‘occurrence’-based CGL insurance policies], there would be no coverage and, hence, no duty to defend.”' In the alternative, I conclude that all defendant insurers except Zurich Insurance- Company 18 are entitled to sümmary judgment under the “loss-in-progress”' doctrine. C. Alternative Grounds: The “Late-Notice” Doctrine In the alternative, I conclude that Commercial Union Insurance Company (“CU”), Zurich Insurance Company (“Zurich”), and North Star Reinsurance Corporation (“North Star”) are entitled to summary judgment on the grounds that CPC’s late notice to these insurers of the USEPA’s enforcement activities and pending claim at the contaminated site violated the terms of. CPC’s policies with them. These policies required CPC to give such notice “immediately” or “as' soon as practicable.” 19 Under Michigan law, the purpose of notice provisions in insurance policies is to allow insurers to make timely investigations of accidents in order to evaluate claims and to defend against fraudulent, invalid, or excessive claims. Wendel v. Swanberg, 384 Mich. 468, 477, 185 N.W.2d 348 (1971); Wehner v. Foster, 331 Mich. 113, 119, 49 N.W.2d 87 (1951). In an insurance policy, the phrases “immediately” or “as soon as practicable” are interpreted , to mean within a reasonable time, dependent upon the 'joarticular circumstances of the case. Motor State Insurance Co. v. Benton, 35 Mich.App. 287, 192 N.W.2d 385, 387. (1971); Kravat v. Indemnity Ins. Co., 152 F.2d 336 (6th Cir.1945). *814 Unless an insurer can show that it has suffered prejudice because of a delay, however, late notice to an insurance company of a possible claim does not eliminate an insurer’s obligations under a policy. Wendel, 384 Mich. at 478, 185 N.W.2d 348. Moreover, the insurer bears the burden of showing prejudice. Id. An insurer may establish prejudice by showing that the delay “materially” impaired its ability to contest either its liability to an insured or the insured’s liability to a third party. Wendel, 384. Mich. at 479, 185 N.W.2d 348. Michigan courts generally leave the question of prejudice to the trier of fact. Wendel, 384 Mich. at 478, 185 N.W.2d 348. However, where the facts are so clear that “one conclusion only is reasonably possible,” the question is one of law. Wehner, 331 Mich. at 120, 49 N.W.2d 87; Wendel, 384 Mich. at 479, 49 N.W.2d 87. CPC offers proof that it gave notice to CU in 1976 of possible claims from two adjacent landowners who were experiencing problems with their well water. 20 However, CPC does not dispute that-it did not. notify CU of the USEPA’s pending CERCLA. claim until December 18, 1986. This was six years after CPC notified, carriers on the risk from 1965-72 of the USEPA’s enforcement activities at the site and four years after the USEPA identified CPC as a potentially responsible party (“PRP”) under CERCLA for the 'response and remediation costs of- cleaning up the contaminated site. Drawing all justifiable inferences in favor of CPC, as I must, I conclude as a matter of law that CPC’s proof of timely notice to CU of adjacent landowners’ claims was not sufficient proof of timely notice to CU of the USEPA’s later CERCLA claim. Under ordinary insurance liability principles, it would not have been possible for CU to provide coverage to CPC for claims made before its policies with CPC took effect. The Comprehensive Environmental Response, Compensation, and Liability Act opened our present era of hazardous waste litigation, including strict liability for .former owners and operators of contaminated sites, in 1980, four years after CPC’s 1976 notice to CU of the adjacent landowners’ claims. If CPC expected CU to cover, indemnify, or defend it for an “occurrence” for whose consequences it was subsequently sued by the USEPA under a new statutory scheme of strict liability, it was CPC’s contractual responsibility to give CU timely notice of the USEPA’s claim. CPC was moreover contractually bound to give CU such notice “as soon as practicable.” Considering the facts and circumstances of this case, I conclude that a reasonable time for CPC to notify CU of the USEPA’s pending CERCLA claim would have been December, 1980, the same time that it notified its carriers on the risk from 1965-72 of the USEPA’s enforcement activities at the site. In any case, if CPC expected CU to provide coverage or indemnification for the USEPA’s pending CERCLA claim, I conclude that it would have been unreasonable for CPC to notify CU of this claim at any time later than October 1982, when the USEPA formally identified CPC as a PRP for the contaminated site. However, CPC did not notify CU of this claim until over four years later, in December 1986. I therefore conclude that CPC did not fulfill its contractual obligations to CU to give notice “as soon as practicable” of the USEPA’s pending CERCLA claim. Michigan courts have recognized that, where enough time has passed, “it virtually becomes. impossible to learn what facts, favorable to defendants], could have been ascertained through prompt inquiry.” In such situations, Michigan courts have been “impelled to the conclusion that prejudice must be presumed.” West Bay Exploration Co. v. AIG Specialty Agencies, Inc., 915 F.2d 1030, 1036-37 (6th Cir.1990) (quoting Wehner, 331 Mich. at 120, 49 N.W.2d at 87). I conclude-that CPC’s notice to CU of the USEPA’s pending CERCLA claim concerning this site was so unreasonably late under the terms of its policies at issue here that prejudice must be presumed. I further conclude as a matter of law that CPC therefore may not recover under its policies with CU for the response and remediation costs of *815 cleaning up groundwater contamination at the Ott/Story/Cordova site. Similarly, I conclude that CPC’s notice to Zurich and to North Star of the USEPA’s pending CERCLA claim concerning this site was so unreasonably late under the terms of their policies at issue here that prejudice must be presumed. It is undisputed that CPC did not notify Zurich' of the USEPA’s pending CERCLA claim until April 15, 1986. 21 It is undisputed that North Star first became aware of CPC’s claim relating to the site at issue when it received a copy of a complaint in a New Jersey action on or about July 23, 1987. 22 If CPC expected Zurich or North Star to provide coverage or indemnification for the USEPA’s pending CERCLA claim, I conclude that it would have been unreasonable for CPC to notify them of this claim at any time later than October 1982, when the USEPA formally identified CPC as a PRP for the contaminated site. However, CPC’ did not notify these insurers of this claim until over three to foür years later, in April 1986 and July 1987. I therefore conclude that CPC did not fulfill its contractual obligations to Zurich and North Star to give notice “as soon as practicable” or “immediately” of the USE-PA’s. pending CERCLA claim. I further conclude as a matter of law that CPC may not recover under its policies with Zurich and North Star for the response and remediation costs of cleaning up groundwater contamination at the Ott/Story/Cordova site. In the alternative, Commercial Union Insurance Company, Zurich Insurance Company, and North Star Reinsurance Corporation are entitled to summary judgment on the grounds that CPC’s late notice to these insurers of the USEPA’s enforcement activities and pending claim at the contaminated site violated the terms of its policies with them. V. CPC’s and CU’s Cross-Motions for Summary Judgment on Duty to Defend The insurer has a duty to defend the insured as long as the allegations against the insured even arguably come within the policy’s coverage. Detroit Edison Co. v. Michigan Mut. Ins. Co., 102 Mich.App. 136, 301 N.W.2d 832, 835 (1980). The duty to defend is therefore broader than, and not necessarily conclusive of, an insurer’s duty to indemnify. Allstate Ins. Co. v. Freeman, 432 Mich. 656, 443 N.W.2d 734, 737 (1989). The court must resolve any doubt pertaining to the duty to defend in favor of the insured. Id. I concluded above as a matter of law that Commercial Union Insurance Company is entitléd to summary judgment on the independent grounds of the “known-risk,” “loss-in-progress,” and “late-notice” doctrines. Since CPC may not recover under its policies with CU for the' response and remediation costs of cleaning-up groundwater contamination surrounding the Ott/Story/Cordova site, CU had no duty to defend CPC in its underlying litigation with the USEPA. See Inland Waters II, 943 F.2d 52. (6th Cir.1991); Inland Waters III, 783 F.Supp. 325 (E.D.Mich.). Consequently, I deny CPC’s motion for summary judgment on the duty-to-defend issue. I grant CU’s motion for summary ’ judgment on the issue. On this record, CU, not CPC, has proven that the facts are so clear that “one conclusion only is reasonably possible.” West Bay, 915 F.2d at 1037 (citations omitted). ORDER In accordance with the opinion issued this date: IT IS ORDERED that CPC’s motion for partial summary judgment on choice of law and duty to defend is DENIED. *816 IT IS FURTHER ORDERED that Commercial Union Insurance Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that Zurich Insurance Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that North Star Reinsurance Corporation’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that Highlands Insurance Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that New-ington, Ltd.’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that the Insurance Company of the State of Pennsylvania’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that the Insurance Company of North America’s motion for summary judgment is QRANTED. IT IS FURTHER ORDERED that Home Insurance Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED , that the International Insurance Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that North River Insurance Company’s motion for summary judgment is GRANTED. IT IS FURTHER' ORDERED'that Aetna Casualty and Surety Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that the Northwestern National Insurance. Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that Fireman’s Fund.Insurance Company’s motion for summary judgment is GRANTED. IT IS FURTHER ORDERED that Affiliated FM Insurance Company’s motion- for summary judgment is GRANTED. Footnotes [1] The court’s findings of fact and conclusions of law in the .underlying case may be found at CPC Int’l, Inc. v. Aerojet-General Corp., 777 F.Supp. 549 (W.D.Mich.1991). [2] CPC’s third-party complaint originally named seventeen insurers as third-party defendants. CPC subsequently amended its complaint to add a third-party defendant. The claims against four of the insurers have been dismissed. The fourteen remaining third-party defendants are Commercial Union Insurance Company;' Home Insurance Company; Insurance Company of the State of Pennsylvania; Zurich Insurance Company; Affiliated FM Insurance. Company; North River Insurance Company; North Star Reinsurance Corporation; International Insurance Company; Highlands Insurance Company; Aetna Casualty and Surety Company; Newing-ton, Ltd.; Fireman’s Fund Insurance Company; Northwestern National Insurance Company; and Insurance Company of North America. Commercial Union Insurance Company was the primary carrier from July 1, 1973, to July 1, 1976. The other insurers provided excess layers of coverage. [3] The facility has not been in operation since 1986. [4] N.J.Stat.Ann. 17:22-6.15 provided that "All contracts of insurance for or on behalf of any insurance company, covering any property, or insurable interests, or business activities, located within, or transacted within this State ... shall be countersigned by a licensed agent.” A 1951 amendment to this statute allowed a licensed agent to countersign a contract of .insurance by an attorney in fact if a certificate authorizing the agent to do so was on file in the office of the commissioner. N.J.Stat.Ann. 17:22-6.15 was repealed in 1987, but was in effect during 1971-77. [5] Attorney Moreau is associated with the law firm of Lowestein, Sandler, Kohl, Fisher & Boy-lan, national insurance counsel for CPC. [6] CPC admits that Highland’s two policies were countersigned in New York, New York, and that Northwestern National’s policy was countersigned in Milwaukee, Wisconsin. [7] It should be noted that, at one period of time, both Michigan and New Jersey had countersignature requirements. These requirements have since been repealed in both states, more recently in New Jersey. Nevertheless, New Jersey's former statutory countersignature requirement was in effect at the time the parties before me entered into the insurance contracts at issue in this case. [8] Unlike N.J.Stat.Ann. 17:22-6A5 that was in force when the contracts presently before me were countersigned, the New Jersey countersignature statute in force at the time Wootton was decided did not expressly authorize countersignature by an attorney in fact. [9] "[T]he relevant time period is 1973 to 1977. Many of the relevant documents are lost or destroyed. Some of the necessary witnesses are dead.” Newington's Response to CPC's Motion for Summary Disposition at 10. [10] I note that, in Chrysler Corp. v. Skyline Industrial Services, Inc., 199 Mich.App. 366, 502 N.W.2d 715, 718-20 (1993), the court approved of and used Restatement (Second) of Conflict of Laws, § 188, in deciding a choice-of-law issue. The case involved a contract for painting services that included clauses requiring both applying Michigan law and indemnifying damages of injured workers, where the law of the place of injury forbade enforcing the indemnification clause. [11] Without specifically deciding that Michigan law should apply, the Superior Court of New Jersey dismissed the original declaratory judgment action in the case now before me because Michigan had a strong interest in deciding if insurance coverage should be provided. Commercial Union’s Motion for Summary Judgment (citing Transcript of Sept. 21, 1989, N.J.Super.Ct.). [12] This is actually the third of three Inland Waters cases. I discuss below each of these cases in turn. [13] In this case, the insurance policies took effect just seven months after contaminants had been spilled on the soil and cleaned up. [14] See Ex. L attached to Aetna's Brief in Support of Joinder in Motion for Summary Judgment. [15] See Ex. R attached to Aetna’s Brief in Support of Joinder in Motion for Summary Judgment. [16] See Ex. S attached to Aetna’s Brief in Support of Joinder in Motion -for Summary Judgment. [17] The earliest effective date of any of these policies was October 1, 1971. [18] Zurich Insurance Company did not move for summary judgment on the basis of the "loss-in-progress” doctrine. [19] CU’s Brief in Support of Motion for Summary Judgment at 37; Zurich's Memo, in Support of Motion for Summary Judgment at 27; North Star’s Motion for Summary Disposition at 4. [20] See Ex. 1, 3-5 attached to the July 30, 1992, affidavit of William Wells, CPC’s senior corporate counsel. [21] See Ex. N attached to Zurich’s Memo, in Support of Motion for Summary Judgment. Ex. N is a May 29, 1986, letter from Thomas Krysotek of Zurich to Pamela Rosenthal of CPC acknowledging receipt of CPC’s April 15, 1986, first notice of loss for CPC’s claim in its underlying litigation with the USEPA concerning the Ott/Story/Cordova site.- The letter specifically reserves Zurich’s rights based upon-late notice of. loss on this claim. [22] See Ex. 1 to North Star's Motion for Summary Disposition. Ex. 1 is an affidavit of John Datt-ner, Vice-President of General Reinsurance Corporation, who stated that North Star first became aware of this claim when it received a copy of a complaint filed by CPC in a New Jersey action on or about July 23’, 1987.