In Re Silicone Implant Insurance Coverage Litigation

652 N.W.2d 46, 2002 Minn. App. LEXIS 1093, 2002 WL 31109708

• Court: Court of Appeals of Minnesota
• Decided: September 24, 2002
• Docket: C5-01-1546, C5-01-1742, C4-01-1778, C1-01-1821, C8-01-1850, C7-01-1869, C1-01-1897, C3-01-1920, C3-01-1738, C4-01-1747, C9-01-1811, C4-01-1828, C2-01-1861, C6-01-1894, C9-01-1906, C7-01-1922, C3-01-1965, C3-01-1741, C6-01-1748, C0-01-1812, C6-01-1829, C4-01-1862, CX-01-1896, C3-01-1917 and C2-01-1925
• Status: Published

Opinion

OPINION

TOUSSAINT, Chief Judge.

In this dispute, the parties seek resolution of insurance coverage and other issues. 3M had been sued by women who claimed that their silicone-gel breast implants, manufactured by 3M, caused them injury. 3M settled the suits and then turned to its liability insurers for coverage. After extensive proceedings, the district court resolved the claims. 3M and the insurers brought separate appeals from the district court’s posttrial orders and final partial judgment, and this court consolidated the appeals.

3M conténds that (a) the policies in effect at the time of the implantations provided full coverage to the policy limits; (b) all of the excess policies provided coverage for defense costs; (c) the district court erred in denying its motion to amend its counterclaim to add a claim for punitive damages; and (d) the district court abused its discretion in denying 3M additional equitable relief.

The insurers jointly argue that (a) the policies were triggered shortly before the underlying plaintiffs exhibited overt symptoms of their diseases rather than at the time of implantations; (b) the damages were continuous and should be allocated through the time the underlying plaintiffs filed their lawsuits or died; (c) the judgment-reduction undertaking ought to be enforced; (d) where 3M waived coverage for implant ruptures as excludable expected-and-intended damages, supplemental payments to plaintiffs with both disease and ruptured implants should not be covered; (e) the district court erred in awarding 3M its coverage-action fees and costs; and (f) the district court abused its discretion in instructing the jury on risk of loss and waiver in connection with the insurers’ misrepresentation claim.

Finally, individual insurers argue that as to their particular policies: (a) defense costs were excluded from their excess policies; (b) defense costs were in addition to policy limits; (c) the $5,000 deductible in the relevant underlying layer applied to each implant claim and applied to the excess layers; (d) claims against 3M’s subsidiary that manufactured the implants were not covered because 3M did not give the notice necessary to make the subsidiary a named insured; and (e) the prejudgment interest awarded 3M should be calculated from the date the insurers were presented with a claim for reimbursement.

Because (a) the policies were triggered by injuries occurring both around the time of the implantations and afterwards; (b) damages were continuous, requiring allocation pro rata by time on the risk; (c) the allocation period should end at the time the underlying plaintiffs filed their lawsuits or died; (d) in light of the allocation ruling, this court need not reach the judgment-reduction issue; (e) the district court properly interpreted the defense-cost pro visions of the excess policies; (f) the $5,000 deductible applies per claim but does not apply to the excess policies; (g) 3M complied with the named insured provisions; (h) the prejudgment interest should be calculated from the earlier of the dates on which the individual insurers’ policies were reached by 3M’s payment of covered costs or its obligation to pay such costs; (i) 3M waived coverage for excludable expeeted- and-intended damages for implant ruptures and may not recover for supplemental payments it made for mixed disease and rupture claims; (j) attorney fees are not recoverable absent statutory authorization or breach of a contractual duty to defend; (k) the district court did not err in denying 3M’s motion to add a punitive-damages claim; (1) the district court did not abuse its discretion in denying 3M’s request for additional equitable relief; and (m) the district courts instruction on risk of loss and waiver was harmless error, we affirm in part and reverse in part.

FACTS

In 1977, 3M Company acquired McGhan Medical Corporation, which manufactured silicone-gel breast implants. Seven years later, in 1984, 3M sold the business. It has not manufactured or sold breast implants since then, although some products 3M had manufactured were sold by the purchaser of the business in 1985 and perhaps later.

Beginning in 1992, plaintiffs began bringing claims against 3M, alleging that the implants caused various diseases. It is undisputed that many women who received the implants between 1977 and 1985 later became ill with some sort of autoimmune or other systemic disease. They suffered from a wide variety of symptoms, including joint pain and stiffness, muscle weakness, numbness, fatigue, irritable-bowel symptoms, flu-like symptoms, burning, itching, sleep disturbance, and memory loss. Although 3M asserted there was no reliable evidence of medical causation, several plaintiffs obtained large verdicts. 3M, which settled many individual cases and participated in a global settlement, asserts that its defense and settlement costs exceed $1 billion.

From 1977 to 1985, 3M purchased occurrence insurance with coverage of approximately $1.48 billion. Under the occurrence policies, claims can be made after the policy period but coverage is triggered only if bodily injury or property damage occurred during the policy period.

In 1977, Travelers Insurance Co. provided 3M’s primary occurrence policy. From 1978 through 1985, Lakeside Insurance Limited, a “captive” insurance company owned by 3M, provided 3M’s primary occurrence insurance policy. 3M also had layers of excess policies over the primary layer each year. Typically, several different insurers issued policies within the same excess layer, with each providing a percentage or “quota share” of the coverage limit in that layer. These excess occurrence policies were triggered only after the underlying primary policy and lower-level policies in a particular year had been exhausted. They generally adopt the same provisions, or “follow form,” of the relevant primary layer, absent specific exclusions to the contrary. Generally, the excess occurrence policies issued to 3M from 1977 to 1985 do not contain a duty to defend, but they contain differing provisions as to the duty to pay defense costs.

After 1985, due to a change in the worldwide insurance market, occurrence insurance was no longer widely available, and 3M purchased claims-made insurance. Unlike occurrence policies, the insured may make a claim on a claims-made policy only during the policy period. Coverage is further limited on a claims-made policy by the use of a retroactive date, which defines the period during which injury must take place in order to be covered. 3M’s elaims-made policies contained retroactive dates of November 27, 1985, May 1, 1986, and January 1, 1986, and were intended to dovetail with the occurrence policies.

In 1992, 3M retained coverage counsel and designated its principal insurance broker as its litigation consultant to help map a strategy for maximizing insurance coverage. In April 1992, 3M reported the implant occurrence to its claims-made insurers. Sixteen months later, in July 1993, 3M notified its occurrence insurers.

In September 1994, three excess insurers initiated this action against. 3M and other excess insurers, seeking a declaratory judgment on coverage defenses and trigger, scope, and allocation issues related to the implant cases. The defendant insurers soon became aligned with the plaintiff insurers. 3M counterclaimed for declaratory relief on the same issues and asserted a variety of claims against the insurers, including breach of contract and breach of the implied covenant of good faith and fair dealing.

In October 1995, 3M and the insurers moved for summary judgment on the issue of when actual injury occurred. 3M asserted that injuries began around the time of implantation, while the insurers asserted they began when the underlying plaintiffs first experienced overt symptoms, usually some years after the implant. The district court denied the motions, ruling that the issues of when the injury took place and whether it took place in more than one policy period were questions of fact triable to the court.

In June and July 1996, the court held a “medical trigger” trial to determine when injury occurred for purposes of triggering the occurrence policies. The court found that the injury took place on or about the time of implantation. It adopted a “continuous-trigger theory” and held that the policies would be triggered continuously from date of implant to date of claim or death and that 3M’s losses would be allocated “pro rata by time on the risk” over the period of injury. This ruling meant that the policies that were “on the risk” or in effect over the relevant period of time would each pay a pro rata share of 3M’s losses.

In July 1997, the district court sua sponte reversed its continuous-trigger ruling based on its interpretation of Domtar, Inc. v. Niagara Fire Ins. Co., 563 N.W.2d 724 (Minn.1997), a then-recent case that addressed allocation issues. The insurers moved to reconsider, providing the court with the transcript from the medical-trigger trial and citing to the underlying evidence. In November 1997, the court reinstated its prior trigger and allocation decisions, finding the record demonstrated continuing injuries.

On January 6, 1998, the predecessor district court judge was appointed to the court of. appeals and a successor judge was appointed to preside. 3M then moved to clarify the trigger and allocation rulings. The court ultimately revised the ruling to provide that allocation ended on December 31, 1985, the last date that any occurrence policies were in effect, rather than the date the underlying plaintiffs filed their lawsuits or died.

Early in the case, 3M had stated that it was not seeking coverage for expected- and-intended damages relating to mechanical failures of the implant, specifically, rupture of the implant. In 1997, the parties addressed whether 3M was entitled to reimbursement for supplemental payments to global-settlement claimants whose implants had ruptured and who suffered disease. The court held that 3M had disa vowed such claims when it “voluntarily narrowed the issue * * * to exclude mechanical claims.” 3M later asked the court to reconsider the ruling. The court interpreted the prior order as not resolving this specific reimbursement issue and held that 3M had not disclaimed coverage for specific payments for rupture made to women who also asserted disease claims.

A four-month jury trial on 3M’s breaeh-of-contract claims and certain insurer coverage defenses, including misrepresentation, began in October 1999. The first phase addressed coverage defenses, which the jury ultimately rejected. The second phase addressed 3M’s breach claims. At the end of the trial, the court granted the insurers’ directed-verdict motions, finding, among other things, that although 3M had established a prima facie case for breach, it had not presented a viable damages theory to support a breach claim. The court then discharged the jury over 3M’s objection.

The insurers then asked the court to enforce 3M’s judgment-reduction undertaking. Early in the litigation, ACE and XL, two of the claims-made insurers who provided coverage to 3M after 1985, moved to dismiss, citing policy provisions requiring 3M, in relevant part, to obtain their dismissal from any action filed by another insurer by filing a judgment-reduction undertaking. 3M filed that undertaking, and the trial court dismissed ACE and XL on that basis. The court denied the insurers’ request to enforce the judgment-reduction undertaking, finding that the claims-made insurers were not liable for any implanta-tions performed before the policies’ retroactive dates.

In April 2000, 3M resurrected its breach claims by filing a postverdiet motion for “further declaratory relief.” 3M requested an award of its coverage-action fees and costs based on the same implied-covenant claim that the court had just resolved against it. It also requested further equitable relief.

In an order issued in September 2000, the court made general findings about the insurers’ conduct as a group, particularly before the coverage action commenced in 1994, and concluded that the insurers had repudiated their contracts and breached the implied covenant of good faith and fair dealing. The court awarded 3M attorney fees and costs as an equitable remedy for the insurers’ breach of the implied covenant. The court apportioned 3M’s fees and costs among all the insurers in the program, including insurers that had settled or become insolvent. But the court declined to grant 3M any substantive monetary or equitable relief. The court had also previously declined to allow 3M to amend its pleadings to seek punitive damages.

At issue at different stages of the proceedings was whether, under the occurrence policies, the insurers were hable to 3M for defense costs incurred in the underlying actions. Various insurers sought rulings that their policies excluded coverage of defense costs or that their coverage of defense costs should be included within policy limits, rather than in addition to limits. Most of these motions were ultimately denied.

In May 2001, the court entered its order for partial final judgment, which included monetary judgments against most of the remaining insurers. The $169,340,679 in judgments against the insurers remaining in the case break down as follows: indemnity, $123,030,541; defense costs, $22,456,112; and prejudgment interest, $23,854,026. These appeals followed.

ISSUES

I. Did the district court err in ruling that coverage was triggered shortly after implantation, based on a finding injury first occurred at that time?

II. Did the district court err in ruling that damages were continuous from time of implantation and should be allocated pro rata by time on the risk through December 31, 1985, the last date of the occurrence policies?

III. Did the district court err in denying the occurrence insurers’ motion to enforce 3M’s judgment-reduction undertaking?

IV. Did the district court err in its rulings on coverage of defense costs?

V. Did the district court err in ruling that the $5,000 deductible in the primary Lakeside policy applied as a single deductible each policy year and did not survive the exhaustion of the primary policy?

VT. Did the district court err in rejecting First State Insurance Co.’s claim that its policies did not provide coverage for 3M’s subsidiary that manufactured the implants because 3M failed to comply with the notification requirement as to named insureds?

VII. When the district court awarded 3M prejudgment interest, was it clearly erroneous in determining the dates on which 3M gave the insurers sufficient notice of the amount due and in determining that further demands for payment would have been futile?

VIII. Did the district court err in ruling that 3M did not waive coverage for supplemental payments for rupture given to the underlying plaintiffs who also suffered disease?

IX. Absent statutory authorization or breach of a contractual duty to defend, did the district court err in awarding 3M its coverage-action fees and costs on a finding that the excess insurers breached the implied covenant of good faith and fair dealing?

X. Did the district court err in denying 3M’s request to amend its counterclaim to add a claim for punitive damages?

XI. Did the district court abuse its discretion in denying 3M additional equitable relief?

XII. Did the district court abuse its discretion in instructing the jury on risk of loss and waiver in connection with the insurers’ misrepresentation claim?

ANALYSIS

In reviewing an appeal from a declaratory judgment, findings of fact will not be reversed unless clearly erroneous, while issues of law will be reviewed de novo. Minn. Ctr. for Envtl. Advocacy v. Big Stone County Bd. of Comm’rs., 638 N.W.2d 198, 202 (Minn.App.2002), review denied (Minn. Mar. 27, 2002). “Insurance coverage issues are questions of law for the court.” State Farm Ins. Cos. v. Seefeld, 481 N.W.2d 62, 64 (Minn.1992).

I.

The underlying plaintiffs — women who received silicone-gel breast implants between 1977 and 1985 that were manufactured by 3M — sued 3M on the theory that the implants caused various diseases. 3M, which denied that there was any causal connection, nonetheless settled many such lawsuits after several plaintiffs in implant litigation obtained large verdicts.

3M then turned to its liability insurers, whose occurrence policies were in effect between 1977 and 1985. It asserted that for purposes of coverage the injuries occurred around the time the women received their implants. The insurers, however, claimed that injuries occurred shortly before the women experienced overt symptoms of their diseases, which was often years after the implant. Because the implants took place while the occurrence insurers policies were in effect, but most of the overt symptoms arose after the policies expired, resolution of this issue was determinative of coverage in most cases. 1 After a trial on this “trigger” issue, the district court ruled in favor of 3M, holding that injuries first occurred around the time of the implant.

The insurers contend that the court erred as a matter of law in holding that injuries occurred, for purposes of coverage under the occurrence policies, shortly after implantation. They assert that this ruling is inconsistent with the scientific fact-accepted by both the insurers and 3M — that implants do not cause disease. They argue the court should have held that injuries occurred shortly before the underlying plaintiffs experienced overt symptoms.

The occurrence policies provide indemnity for “all sums” that 3M becomes legally obligated to pay as damages as a result of an “occurrence,” typically defined as “an accident, event or happening, including injurious exposure to conditions, which results, during the policy period, in bodily injury” that is “neither expected nor intended” by the insured. To establish or trigger coverage, “an insured must demonstrate that damage ‘occurred’ while the policy was in effect.” N. States Power Co. v. Fid. and Cas. Co., 523 N.W.2d 657, 659-60 n. 3 (Minn.1994) (quotation omitted). An occurrence takes place not when the policyholder engages in the wrongful act, “but the time the complaining party was actually damaged.” Singsaas v. Diederick, 307 Minn. 153, 156, 238 N.W.2d 878, 880-81 (1976). Consequently, if the damage occurs outside of the policy period, the policy does not provide coverage. See id. at 155, 238 N.W.2d at 880-81 (holding that policy did not provide coverage where negligent repair of lift in elevator occurred during policy period, while resulting injury occurred after policy was canceled).

As a preliminary matter, it should be noted that the question of when injury first occurred was not previously decided in the litigation between 3M and the underlying plaintiffs. In some situations, critical coverage issues may not be resolved in the actions underlying the insureds liability for a judgment. See, e.g., Brown v. State Auto. & Cas. Underwriters, 293 N.W.2d 822, 825 (Minn.1980) (holding that where issue of intent to injure was not necessary or essential issue in previous action it was nevertheless material fact for litigation in coverage action). Such is the case here, where the insured first settled with the plaintiffs and then sought coverage from its liability insurers. Whether the “occurrence” is covered by the policies remained to be resolved. Jostens, Inc. v. CNA Ins./Cont’l Cas. Co., 403 N.W.2d 625, 629-30 (Minn.1987), overruled on other grounds, NSP, 523 N.W.2d at 664. The issue of coverage “is distinct from the issue of whether a settlement is reasonable.” Id. In such coverage disputes, the courts may have to “determine what injuries arose under the policies and are thus indemnified.” Id. at 630.

The district court held a seven-day bench trial on the sharply disputed issue of when actual injury occurred to the underlying plaintiffs. The court issued an order on July 11, 1996, containing extensive findings. It first noted the difficulty of resolving the factual issue of when the injury occurred, in light of the uniform opinion of the testifying experts “that silicone gel breast implants do not cause systemic disease.” Nonetheless, the court found it undisputed that the underlying plaintiffs suffered symptoms of systemic disease, that these diseases were bodily injuries covered by 3M’s liability policies, and that legal causation for purposes of the policies was assumed.

The court’s findings describe the body’s formation of a capsule around the implant and the two hypotheses on what occurs thereafter. All of the experts at trial agreed that after certain normal immune responses to the implant, including acute and chronic inflammatory response and capsule formation around the implant, the body was then immunologically normalized and no systemic disease resulted. The other hypothesis, which the underlying plaintiffs espoused and which is held by other physicians and scientists, was that an abnormal autoimmune response occurred and that systemic disease symptoms appeared. The court noted that even though the experts at trial rejected the latter hypothesis, some of them provided sufficient scientific information from which the plausibility of the hypothesis could be inferred. For purposes of the coverage issue, therefore, the court accepted the second hypothesis of an abnormal autoimmune response as more likely true than not.

While 3M and the insurers differed on when the injury occurred, they all based their arguments on the premise that “the immune system reacts immediately to the presence of a foreign substance in the body.” The insurers’ experts argued that this rapid-response injury occurred shortly before overt symptoms appeared. 3M’s experts asserted that the often insidious immune-system diseases might not become symptomatic for many years after the injury.

The court made specific findings, based on the record, that the injury occurred at about the time of the implant and described the mechanism by which it occurred:

Leaked silicone is in contact with body tissues from the time of implant until the formation of the protective capsule, a period of several weeks. Silicone is bioreactive during that period and more likely than not that is the period during which cellular abnormality is produced. Thus, “bodily injury” within the purview of the trigger language occurs at or about the time of implant.

The greater weight of the evidence, in the context of the undisputed fact of systemic disease symptoms and the assumed fact of legal causation and the necessary inference of the occurrence of an abnormality, supports the conclusion that the leaking silicone gel is the cause, the cellular damage is the injury, and the disease symptoms are the effects. Such cellular damage is determinable, constitutes the underlying bodily harm without which there would be no manifestation in the form of disease symptoms, and satisfies the “actual injury” legal standard for trigger.

(Emphasis added.)

The insurers explicitly state that they do not challenge these findings as clearly erroneous but instead argue that the court erred as a matter of law. They contend that the district court’s error lay in assuming a causal link between the implants and disease when, in fact, no such relationship exists. Because there is no causal relationship, they argue, the cellular-injury mechanism used to explain that causal relationship similarly does not exist.

The insurers do not contest that there was actual injury to the implant plaintiffs. They reject the district court’s cellular-injury theory, however, and cite evidence indicating that once autoimmune disease begins (often years after the implantation), it typically produces overt symptoms in a matter of weeks or months. They assert that it was therefore possible, with reasonable medical certainty, to identify these first disease symptoms by examining the plaintiffs’ medical records and to determine when the cellular-level injury that triggered coverage occurred. The injury would have occurred, the insurers assert, shortly before the overt disease symptoms were observed.

We re-emphasize that we are not addressing the issue of liability between the implant plaintiffs and 3M. Instead, we are addressing the issue of coverage for the liability 3M incurred as a result of the lawsuits. Further, although neither 3M nor the insurers agreed with the plaintiffs’ theory that silicone implants caused disease, that was the theory on which the plaintiffs obtained their settlements. And, as 3M points out, as a result of this theory, 3M incurred more than $1 billion in implant liabilities. The district court, finding the issue of trigger hotly contested and unresolved by the underlying litigation, properly held a trial on the issue. See id. Based on conflicting testimony, the court found as fact that injury occurs on a cellular basis shortly after implant. We hold that the district court properly applied the law and was not clearly erroneous in its findings.

II.

After the resolution of the trigger issue, the question of allocation of damages among insurers arose. The district court ultimately ruled that the damages were continuous from the time of implant and should be allocated pro rata by time on the risk, with the allocation period ending on December 31, 1985, the last date that any of the occurrence policies was in effect.

3M challenges the determination that the damages were continuous; it contends that the damages arose from the discrete event of implantation and argues that any policies in effect at that time should pay up to the policy limits, under an “all sums” theory. The insurers agree with the district court that the damages were continuous, but they argue that the allocation period should end not on December 31, 1985, but on the date that the underlying plaintiffs filed their lawsuits or died, whichever occurred earlier.

If the damages are the result of a single discrete event, only the policies on the risk at that time are triggered and it is not necessary to allocate damages. SCSC Corp. v. Allied Mut Ins. Co., 536 N.W.2d 305, 318. But when damages occur over more than one policy period, courts will apply an “actual injury” or “injury-in-fact” theory in determining which policies are “on the risk.” NSP, 523 N.W.2d at 662. The essence of this theory “is that each insurer is held liable for only those damages which occurred during its policy period; no insurer is held liable for damages outside its policy period.” Id. Further, the court will presume the damages are continuous, unless an insurer can show that damages did not occur during its policy period. NSP, 523 N.W.2d at 664. “Each triggered policy therefore bears a share of the total damages proportionate to the number of years it was on the risk relative to the total number of years of coverage triggered.” Id. at 663 (citation omitted).

“The proper scope of coverage also will depend on the facts of the case.” Domtar Inc., v. Niagara Fire Ins. Co., 563 N.W.2d 724, 733 (Minn.1997). In environmental contamination cases, when the origins of the damages “are uncertain or indivisible and the property damage is continuous,” the insurers are consecutively liable. Id. at 732 (citation omitted). This is limited to “those difficult cases in which property damage is both continuous and so intermingled as to be practically indivisible that NSP properly applies.” Id. at 733. It provides a “judicially manageable way” for trial courts to adjudicate certain pollution-coverage disputes when it is difficult to determine the event, occurrence, or damage giving rise to legal liability. Id. In contrast, if environmental contamination “arises from discrete and identifiable events,” only the policies on the risk provide coverage. Id. In SCSC, the jury found the damage arose from the single spill, although the leaching of chemicals from the spill into the groundwater in subsequent years caused damage. 536 N.W.2d at 318. Only the policies at risk at the time of the spill were triggered. Id.

Here, the district court made specific findings of fact regarding the continuous nature of the damages. The court emphasized that mere leakage of silicone was not the injury. It found that new bodily injury in the form of new cellular abnormalities occurred on an ongoing basis as cells came in contact with leaked microdroplets of silicone.

Unlike SCSC where there was one spillage of a contaminant that for some time afterward leaked into and damaged the soil, the mere leakage of the silicone into the system is not the injury. Rather, it is an autoimmune bioreaction with the silicone in the system that leads to a cell distortion that constitutes the injury. If such bioreaction occurs, it occurs at the time of implant but it also occurs at various times thereafter. The initial autoimmune response begins the injury process. .As cells later come into contact with the silicone and provoke an autoimmune response, new cell distortions, and hence new injuries, occur. It is impossible to tell when any particular injury occurs. This is precisely the situation that requires the application of a continuous trigger.

The district court specifically found that putting an implant in the body was not the injury or the injury-causing event. Instead, the court found that the injury-causing event was the continuous leakage of silicone that comes into contact with the body’s cells, causing incremental cellular damage and eventually producing disease. Even though the successor judge seemed to be leaning toward an interpretation similar to SCSC, the finding of continuous damage remained. We agree that, under the district court’s explicit findings as to the mechanism by which the injury occurred, this is a continuous-damage case requiring allocation.

3M next analogizes this injury to that in a typical personal injury case. In such an action, if the claimant seeks damages for future pain and suffering, the insurer cannot avoid or reduce coverage on the theory that future injuries are outside of its policy period. The insurers respond that this was the type of scenario arguably the subject of the predecessor judge’s decision, finding that implantation was the discrete event causing injury, which decision was later withdrawn. However, the predecessor judge ruled in a subsequent order, which found damages to be continuous, that this was not the case, because new contact with the leaking silicone produced new bodily injury over time.

Relying on the NSP reasoning, 3M argues that when other relevant factors— policy language, the parties intent or reasonable expectations, canons of construction, and public policy — are considered, it is entitled to full coverage under the occurrence policies in effect at the time of the implantation or, in the alternative, under each such policy in effect from the date of implantation to the date of claim. See NSP, 523 N.W.2d at 661 (citing these factors). 3M also asserts that the policy language does not specifically authorize allocation, which is a judicially created concept. It cites evidence that insurance-industry drafters rejected proposals to include pro rata provisions in the policies and argues that the court in NSP did not consider such information. Finally, 3M asserts that its reasonable expectations are supported by the fact that most courts in other jurisdictions have concluded that occurrence policies in effect during any part of the period in which continuing damages have occurred must provide full coverage.

The supreme court in NSP considered most of these factors 3M cites in reaching its decision and ruling that continuous damage must be allocated pro rata by time on the risk. While allocation is a creature of judicial construction, we are bound to follow the supreme court’s interpretation of law. 3M — apparently contrary to the holdings of NSP and Domtar — asks this court to hold that when policies are triggered, the insurers must pay all costs under each policy up to the policy limits, rather than as allocated pro rata based on time on the risk. In light of the supreme court’s definitive rulings on this issue, 3M cannot prevail.

3M next argues that once triggered by injury during the policy period, 3M’s policies promise to indemnify 3M for “all sums” that 3M “becomes legally obligated to pay as damages.” 3M contends that, at a minimum, it is reasonable for insureds such as 3M to expect coverage on that basis under the policies in effect on the dates of implantations. But the supreme court has specifically rejected this approach in cases involving continuing damages. Domtar, 563 N.W.2d at 731-32; NSP, 523 N.W.2d at 662.

Having affirmed the district court’s decision that damages were continuous, requiring allocation pro rata by time on the risk, we address the insurers’ challenge to the district court’s decision that the end date of the allocation period is December 31,1985, the last date the occurrence policies were in effect. When damages occur over multiple policy periods, it is deemed “a continuous process in which the property damage is evenly distributed over the period of time from the first contamination to the end of the last triggered policy (or self-insured) period.” NSP, 523 N.W.2d at 664. This period is not limited only to times when insurance policies are in effect; in NSP, the allocation period included times when NSP had insurance, had no insurance, and had insurance with different limits or self-insured retentions. Id. at 659, 662-63 n. 7. While the length of the allocation period was not at issue in NSP, the supreme court referred to the end of the allocation period as being the time of “discovery, cleanup or whenever the last triggered policy period ended.” Id. at 663-64.

Later, the supreme court addressed more specifically the time over which damages should be allocated. Domtar, 563 N.W.2d at 731. In Domtar, the damages began in 1933 and cleanup efforts began in the 1990s, but the company could produce proof of insurance only between 1956 and 1970. Domtar argued that the losses should be allocated only over the 15 years during which insurance coverage was triggered and not for the entire period over which damages occurred. Id. The supreme court rejected the argument, holding that the insurer was liable “for that period of time it was on the risk compared to the entire period during which damages occurred.” Id. at 732.

Applying this rule of law to the facts as found by the district court — -that damages began upon implantation and were continuous — the damages should be allocated from implantation through the date the underlying plaintiffs’ lawsuits were filed or the plaintiffs died. The district court, however, basing its decisions on equitable considerations, held that the allocation period should end December 31, 1985, the last date the occurrence policies were in effect. It reasoned that 3M had transferred risk to the insurers by purchasing the insurance and that 3M had the reasonable expectation that damages occurring during the policy periods would be covered. The district court concluded that allocating after 1985 would ignore insurance marketplace realities, establish a disincentive to purchase insurance by injecting uncertainty of claim resolution, and construe policies in such a way that would lessen prospects for compensation of injured victims.

NSP and Domtar, however, achieve equity by spreading the risk consistent with the “actual-injury” trigger. Further, NSP and Domtar specifically rejected arguments that the timing of the triggering injuries should be based on the amount of insurance purchased; both declined to exclude periods in which there was no insurance from the pro rata allocation period. Domtar, 563 N.W.2d at 733; NSP, 523 N.W.2d at 664. Applying the law1 clearly set out in NSP and Domtar, the insurers are liable for their time on the risk in proportion to the entire period over which damages occurred; there is no precedent for the district court’s equitable considerations on this point. Under the findings in this case, the damages must be allocated from the time of implantation through the time the underlying plaintiffs filed their lawsuits or died, whichever occurred earlier. Therefore, the district court’s decision to end the allocation period on December 31, 1985, is incorrect as a matter of law and is reversed.

TIG Insurance, an occurrence insurer, separately briefed the issue of the allocation order, arguing against the district court allocation scheme. Our decision reversing the district court decision on allocation resolves TIG’s challenge.

III.

The occurrence insurers challenge the district courts denial of their motion to enforce 3M’s judgment-reduction obligation. After 1985, occurrence policies were generally unavailable. 3M then obtained claims-made policies from four companies: Seaside Insurance Company (Seaside), a “captive” insurer that was wholly owned by 3M; X.L. Insurance Company (XL); A.C.E. Insurance Company (ACE); and Bowring Excess Slip (BES). Seaside, the primary insurer, settled with 3M. XL and ACE, both named parties in this action, also settled with 3M. BES was not made a party to the action.

Both XL and ACE made summary-judgment motions early in the case to obtain dismissal based on judgment-reduction provisions of their policies. These provisions required 3M to undertake to the court to reduce judgment against other insurers to the extent the court determined that XL or ACE would have been liable for indemnity or contribution. 3M filed a judgment-reduction undertaking on XL’s and ACE’s behalf and joined in their motions. The district court granted summary judgment and dismissed the claims against XL and ACE.

The occurrence insurers did not initially seek to enforce the judgment-reduction undertaking under the allocation formula because they believed that the allocation was spread fairly across the period of damages and that they were liable only for the period their policies were on the risk. But after the successor judge imposed the allocation period end date of December 31, 1985, the occurrence insurers contend that they became responsible for damages that accrued after December 31,1985, when the claims-made policies were in effect, for which they believed claims-made insurers should be liable.

In April 2000, they moved for an order to enforce 3M’s judgment-reduction obligation, so that any judgment against the occurrence insurers would be reduced by the amount that would have been allocated to the claims-made carriers had they remained in the case. The court denied the motion and the joint insurers challenge that decision. Because we have determined that the court erred as a matter of law in ending the allocation period on December 31, 1985, and have held that the allocation period should instead extend through the date the underlying plaintiffs filed their lawsuits or died, the insurers are no longer responsible for damages accruing after their policies ended, and this issue is now moot. This court addresses only actual controversies and will not decide moot questions. In re Schmidt, 443 N.W.2d 824, 826 (Minn.1989).

IV.

3M and four of the insurers challenge rulings regarding payment of defense costs under the excess policies. Unless otherwise specified, the excess layers “followed form” of the relevant underlying layer, in this case either the primary policy or the first excess policy. These underlying layers provided for payment of defense costs. For the years 1977 and 1985, the payment of defense costs applied toward the exhaustion of the policy limits, i.e., defense costs were within the policy limits; for the years 1978 through 1984, the payment of defense costs did not apply toward exhaustion, i.e., defense costs were in addition to policy limits. Over the course of the declaratory judgment proceedings, there were numerous challenges regarding the applicability of these defense-cost provisions to particular insurers.

1. SM’s Challenges to Defense Costs

The district court granted summary declaratory judgment to Columbia Casualty Co., Federal Insurance Co., United States Fire Insurance Co., Employers Mutual Casualty Co., Utica Mutual Insurance Co., The Home Insurance Co., and Mt. McKinley Insurance Co., ruling that these insurers’ policies did not impose the duty to pay defense costs.

3M argues that under its multi-layered insurance program, the policies were uniform throughout the layers, requiring payment of defense costs by the occurrence insurers. This' argument, however, is not based on evidence that was before the district court at the time the summary-judgment motions were made. 3M proffers primarily affidavits and related material that it submitted in 1999 in support of its request for leave to file a motion for reconsideration. When the court denied the motion, 3M submitted an offer of proof. 2 3M also contends that, even considering the language of each particular insurers’ policy, the rulings are erroneous. While 3M addressed policy language only generally in its opening brief, it discussed it more specifically in its reply brief, still relying in large part on the additional material contained in its offer of proof.

■ In connection with a summary-judgment motion, the district court will consider materials on file with the court and any affidavits submitted by the parties. Minn. R. Civ. P. 56.03. In reviewing the decision, this court may only consider whether there are genuine issues of mate rial fact at issue and whether the district court erred as a matter of law. State by Cooper v. French, 460 N.W.2d 2, 4 (Minn.1990). An appellate court may not consider documents that were filed in an offer of proof after the district court denied summary judgment, which the district court never considered. Sullivan v. Spot Weld, Inc., 560 N.W.2d 712, 716 (Minn.App.1997), review denied (Minn. Apr. 24, 1997). Further, this court will not consider arguments made for the first time in a reply brief. Albert Lea Ice Fuel Co. v. U.S. Fire Ins. Co., 239 Minn. 198, 205, 58 N.W.2d 614, 619 (1953); see Minn. R. Civ.App. P. 128.02, subd. 3 (stating that, generally, reply brief is limited to addressing new points made by respondent).

3M does not assert that there were genuine issues of material fact based on the record as it existed at the time of the summary judgment. Because the district court ruled on the policies as a matter of law and this court agrees with the reasoning of the district court in its summary judgment decisions, we affirm its decisions that these policies exclude payment of defense costs.

2. Century’s Challenges to Defense Costs

The district court ruled that Century Indemnity Co. was liable for payment of defense costs. Century contends that its policies exclude payment of defense costs. The Century policies follow form to the primary policy for payment of defense costs, with certain stated exceptions. First, it is undisputed that the policies explicitly exclude any obligation of Century to defend unless it chooses to do so. Second, the policies contain the following provision regarding expenses:

[T]he insurance afforded by this Certificate shall not apply to any expenses for which insurance is provided in the primary insurance [or designated underlying insurance].

Century asserts that this provision excludes payment for defense costs.

The district court ruled that under a reasonable person’s understanding of this language, Century was not obligated to pay defense costs so long as the underlying policy was obligated to do so, but once the underlying policy was no longer obligated, then Century was. See Hartford Accident & Indem. Co. v. Pac. Employers Ins. Co., 862 F.Supp. 160 (S.D.Tex.1994) (interpreting similar language to provide no exclusion for expenses).

Century contends that the unambiguous language of the paragraph excludes any obligation to pay defense costs and argues that the district court read language into the policies that is not there. Further, Century asserts that the court’s interpretation renders it liable for defense costs immediately upon exhaustion of the designated underlying policy rather than when all of the excess underlying layers are exhausted. Finally, it argues that the court’s reading renders the paragraph mere surplusage by simply restating the basic principle of exhaustion of underlying policies.

Terms in an insurance policy should be interpreted “according to what a •reasonable person in the position of the insured would have understood.” Canadian Universal Ins. Co. v. Fire Watch, Inc., 258 N.W.2d 570, 572 (Minn.1977). Restrictive language is construed against the insurer who drafted the language. Wondra v. Am. Family Ins. Group, 432 N.W.2d 455, 459 (Minn.App.1988), review denied (Minn. Jan. 25, 1989). “Exclusions in insurance contracts are read narrowly against the insurer.” Atwater Creamery Co. v. W. Nat’l. Mut. Ins. Co., 366 N.W.2d 271, 276 (Minn.1985). The disputed lan guage does not clearly exclude payment of defense costs and a reasonable person would not understand that it did. We hold that the district court’s interpretation of this language is correct.

3. TIG’s Challenges to Defense Costs

The court ruled that TIG Insurance Co. was liable for payment of defense costs. TIG argues that its policies exclude payment of defense costs. It reasons that the primary policy imposes the duty to pay defense costs only if the primary insurer agrees to defend the insurer, but that there is no independent duty to pay defense costs. TIG contends that because its policies exclude the duty to defend, it has no duty to pay defense costs. TIG cites the following provision in its policies in support of its theory:

This policy is subject to the same warranties, terms and conditions (except as regards the premium, the obligation to investigate and defend the amount and limits of liability and the renewal agreement, if any, and except as otherwise provided herein) as are contained in or as may be added to the Underlying Policy prior to the happening of an occurrence for which claim is made hereunder ⅝ ⅜ ⅝

If policy language is ambiguous, it must be interpreted in favor of finding coverage. Nordby v. Atlantic Mut. Ins. Co., 329 N.W.2d 820, 822 (Minn.1983). A policy exclusion is strictly construed against the insurer. Atwater Creamery, 366 N.W.2d at 276.

The district court, applying these principles, determined that the language was ambiguous and construed it against TIG. The district court noted first that the plain language of the exception in the TIG policies does not address defense costs. Further, the underlying policy and many of the other excess policies treat the duty to defend and the duty to pay defense costs separately. The court distinguished between these two duties, noting that the duty to defend carries obligations such as retaining an attorney and directing and supervising litigation, while the duty to pay defense costs is a singular one. If a policy includes a duty to defend, there is a concomitant duty to pay defense costs. A policy might exclude a duty to defend, however, while including a duty to pay defense costs. The court noted that the exclusion of the “nonexistent” duty to defend could have been intentional or could have been merely unartful language. Finding the language ambiguous, and interpreting the ambiguity in favor of the insured, the court held the TIG policies followed form and TIG had an obligation to pay defense costs. While TIG complains that this result is inconsistent with the court’s interpretation of other similar policies, the language in the other policies is distinguishable and leads to different results. The district court’s analysis of the policy is not erroneous.

A Executive’s Challenges to Defense Costs

The district court determined that Executive Risk Indemnity Inc. had a duty to reimburse 3M’s defense costs in addition to the policy limits of liability. The court ruled that because Executive’s policies followed form to the relevant underlying policy, and because Executive’s policies did not clearly provide for payment of defense costs within the policy limits, Executive was responsible for payment of defense costs in addition to the policy limits. Executive challenges this decision.

The underlying policies obligated Executive to reimburse 3M for its ultimate net loss, defined as “all sums actually paid or which [3M] is legally obligated to pay, as damages, in settlement or satisfaction of claims or suits ⅜ * ⅜.” The policies do not specifically define damages. The district court noted that undefined terms in insurance contracts are given their plain and ordinary meaning. Minn. Mining & Mfg. Co. v. Travelers Indem. Co., 457 N.W.2d 175, 179 (Minn.1990). The court concluded that the ordinary meaning of damages did not include defense costs and that the definition of ultimate net loss did not include the costs of defense. Because its policies follow form to Lakeside and because Lakeside’s definition of ultimate net loss does not include defense costs, the court ruled that defense costs are payable in addition to the policy limits.

Executive first argues that because its policies do not explicitly provide that defense costs will be reimbursed in addition to policy limits, it has no obligation to reimburse 3M for defense costs that exceed the policy limits. This argument ignores the fact that its policies followed form to the relevant underlying policy, which provides for payment of defense costs outside policy limits.

Executive also argues that its policies incorporate the provisions of underlying insurance except with respect to “obligation to investigate, defend, pay costs and expenses incident to same” and other specified matters. It contends that its obligation to pay defense costs is governed in part by the policy’s definition of “ultimate net loss,” which does not specifically include defense costs. Consequently, it contends that defense costs are expenses reimbursed as part of the insured’s losses within the policy limits for ultimate net loss, and that there is no basis for the decision that defense costs are in addition to the policy limits.

Based on the district court’s reasoning as to the policy language, the fact that Executive followed form to the underlying policy, and that Executive did not specifically exclude defense costs, the district court’s ruling is affirmed.

5. ACE’s Challenges to Defense Costs

The district court held that ACE Europe, SA NV, was also liable for defense costs in addition to its policy limit. ACE challenges the decision. The parties were unable to locate ACE’s policies and stipulated that certain memos between brokers provided evidence as to indemnity. The court concluded that this evidence indicated that ACE followed form to the relevant underlying policies, making ACE responsible for defense costs in addition to the policy limits.

ACE explains that the district court must have relied on language in the memos referring to coverage “per underlying policy.” This language, ACE contends, does not establish the extent of its defense-cost obligations because there was no specific provision describing the extent to which ACE agreed to reimburse defense costs. See Schneider Natl. Transp. v. Ford Motor Co., 280 F.3d 532, 538 (5th Cir.2002) (holding that absent “follow form” language, provision that excess policy would continue “as underlying policy” did not impose coverage as provided in underlying policy). Finally, ACE contends that the district court erroneously shifted to ACE the burden of establishing that the terms of the policy did not provide for coverage of defense costs in addition to limits.

Unlike the situation in Schneider, the information available as to the ACE policies indicates that they follow form to the relevant underlying policy, which provides for payment of defense costs in addition to policy limits, and ACE’s arguments do not compel a decision to the contrary. The district court ruling is consistent with its other decisions, that unless there is a spe cific exclusion, policies that follow form to the underlying policy and do not limit defense costs, such as ACE, obligate the insurer to pay defense costs in addition to policy limits. We reject the argument that the court shifted the burden of proof.

V.

For the years 1981 to 1985, Lakeside Insurance provided primary coverage to 3M that contained the following deductible: “$5,000-per-occurrence or event including allocated claim expense.” 3 The parties tried the issues relating to the deductible, and the court concluded that the $5,000 per occurrence deductible should be applied as a single deductible in each policy year and that it would not survive the exhaustion of the primary policy. Federal Insurance challenges the district court’s decision that a single deductible applies for each policy year. Federal also argues that after the Lakeside policies have been exhausted, the deductible should apply on a per-claim basis..

As with all contracts, the interpretation of insurance-policy language is a question of law reviewed de novo. Jenoff, Inc. v. New Hampshire Ins. Co., 558 N.W.2d 260, 262 (Minn.1997). If the terms are not defined, “they must be given their plain, ordinary, or popular meaning.” Id. (citation omitted). The court will construe ambiguous terms against the insurer and “in accordance with the reasonable expectations of the insured.” Id. (citation omitted); Atwater Creamery, 366 N.W.2d at 276.

The district court found that the evidence supported virtually any interpretation of the applicability of the deductible. The court noted that limiting terms in policies must be interpreted in such a way as to make the general purpose of the insurance contract effective, citing Struble v. Occidental Life Ins. Co., 265 Minn. 26, 35, 120 N.W.2d 609, 614 (1963). The court found that treating an “occurrence,” as an event, namely the manufacture of the implants, from which multiple claims might arise was a reasonable interpretation that was consistent with the objectives of 3M’s insurance program and deemed it to be the manufacture of the implants. It reasoned that to interpret occurrence to apply to each claim would be contrary to the intent of the parties and defeat the purpose of the insurance contracts. The district court concluded that the $5,000 per-occurrence deductible should be applied as a single deductible each policy year..

Federal argues that in the case of an ambiguous provision, the practical construction of the' terms controls. See Hawkeye Cas. Co. v. Rose, 181 F.2d 157, 159 (8th Cir.1950). It contends that 3M applied the deductible on a per-claim basis to all of its products-liability bodily injury claims. It argues that because each implant claim is a separate occurrence, a separate deductible applies to each implant claim.

3M contends that the policy definitions do not equate “occurrence” with “claim.” It notes that 3M implants were manufactured at the same location and the policy provides that:

All such exposure to substantially the same general conditions existing at or emanating from one premises location shall be deemed one occurrence.

Further, it cites the industry custom of applying “per occurrence” deductibles once in each triggered year to claims of the same type arising out of a common cause.

We must disagree with the district court’s decision that a single deductible applied each year as a matter of law. The primary policies define occurrence as “an accident, event or happening,” including “continuous or repeated exposure to conditions,” or “injurious exposure to conditions.” A dictionary defines “event” as “[s]omething that takes place; an occurrence.” The American Heritage Dictionary 635 (3d ed.1992). Under the policy, an “occurrence” takes place when the complainant suffered .the injury, which occurred here shortly after implantation, not when the implant was manufactured. Singsaas v. Diederich, 307 Minn. 153, 156, 238 N.W.2d 878, 880 (1976). Second, while there was much testimony as to when the deductible applied and how deductibles had been applied in the past, and while the trial court believed that applying the deductible on a per-implant basis would unduly limit the insurance, we cannot avoid the plain language and interpretation of the policy. Further, we note that even parol evidence including a memo from 3M — supports application of the $5,000 deductible per claim as a matter of law.

As to Federal’s argument that the deductible applied to the excess layers after exhaustion of the primary coverage, we note that the excess policies, as discussed above, follow form except, as is relevant here, regarding the limits of liability. Therefore, the excess policies do not follow form as to the deductible endorsement, which is written in terms of liability limits. Consequently, we agree with the district court that the deductible does not apply to the excess policies.

VI.

First State Insurance Co. argues that its policies do not provide coverage because 3M did not comply with the “Named Insured” provisions. The court granted summary judgment in favor of 3M on the merits and, alternatively, on waiver of the issue by First State.

At the time 3M acquired McGhan in May 1977, 3M was already insured by two First State policies that took effect on March 31, 1977, and that described the named insured as 3M, its existing subsidiaries, and any after-acquired subsidiaries. Because this policy explicitly applies to “after-acquired” subsidiaries, it provided coverage for McGhan.

The parties dispute whether McGhan was covered by the 1978-81 First State policies. These policies provided for extension of coverage to subsidiaries upon notice to the insurer as follows:

The words “Named Insured” includes The Named Insured stated in The Declarations forming a part hereof and/or subsidiary, associated, affiliated companies or owned and controlled companies as now or hereafter constituted and of which prompt notice has been given to the Company.

The district court analyzed the notice provision of the policy and found the provision ambiguous. It interpreted the notice provision in favor of 3M, so that 3M as then constituted — including its then-existing subsidiaries — was the named insured and therefore no notice of the acquisition of McGhan was required.

First State contends that this ruling is incorrect for several reasons. First, it argues that where the provisions are ambiguous, they must be resolved by trial. It contends that the caselaw, though sparse, makes clear that under this type of named-insured provision, notice is required to include a subsidiary as a named insured, regardless of whether the subsidiary is acquired before or after the policy’s inception. It states that there is no evidence of notice in 1978 and only enough evidence of notice to create a factual question in 1979, 1980, and 1981.

First State cites Am. Motor Inns, Inc. v. Harbor Ins. Co., 590 F.Supp. 468, 470 (W-D.Va.1984), in which the court interpreted similar language and concluded that due to the failure to give notice of a subsidiary before an accident occurred, the subsidiary was not covered. But unlike the insurer in American Motor Inns, there was no evidence that First State kept a separate endorsement listing all subsidiaries for which an additional premium had been paid or that specific premiums were charged as additional subsidiaries were added. Id. Consequently, American Motor Inns is distinguishable.

Finally, First State challenges the court’s ruling that it waived its “named insured” defense. The court found waiver due to First State’s failure to raise it on the many occasions the parties were ordered to identify “with particularity and precision” the issues in the case that still needed to be resolved. Although we affirm on the merits of the defense, we do not find the waiver ruling erroneous.

VII.

Two affiliated insurers, Twin City Fire Insurance Co. and First State Insurance Co., challenge the district court’s award of prejudgment interest to 3M. It is undisputed that 3M is entitled to prejudgment interest. Further, the parties agree that the commencement date of prejudgment interest relates to the fact that these are reimbursement policies. At issue are the dates from which the interest should be calculated. An appellate court will review questions as to construction of the prejudgment interest statute de novo, but issues as to whether the claim is readily ascertainable or liquidated may raise questions of fact reviewable under a clearly erroneous standard. Trapp v. Hancuh, 587 N.W.2d 61, 63 (Minn.App.1998).

A court may order prejudgment interest on damages. Minn.Stat. § 549.09, subd. 1 (2000). Such an award serves two purposes: First, it compensates the plaintiff for loss of use of its money while depriving the defendant of any gain; and second, it promotes settlement. Bumiece v. Illinois Farmers Ins. Co., 398 N.W.2d 542, 544 (Minn.1987). The interest is calculated from either the start of the action, demand for arbitration, or written notice of claim, whichever occurs first. Minn.Stat. § 549.09, subd. 1(b).

The district court found that 3M “billed” all the occurrence carriers in 1998 under its “all sums” theory. It also gave the occurrence carriers “a stream of information chronicling 3M’s indemnity and defense costs over time.” The court found that the carriers were “unresponsive to 3M’s explicit and implicit payment ‘presentments.’ ” The district court concluded that 3M “made specific demands for payment and provided sufficient information to constitute a demand for payment.” The court also found that 3M demonstrated that any further effort on its part to make further demands for payment would have been futile. It concluded that

as to each of the Insurers, prejudgment interest on the amounts they owe under the judgment and decree in this action shall run from the date on which their individual policies were reached by reason of 3M’s payment of covered costs or its obligation to pay such costs, whichever first occurs.

Twin City and First State challenge these findings, contending there is no support in the record. They contend that their reimbursement obligation did not arise until 3M submitted a specific claim to them. They assert that this occurred on October 18, 1999, when 3M for the first time submitted a report by its computational expert, assigning settlement and de fense costs to particular years, using the allocation and other coverage principles established by the court or agreed to by the parties. Consequently, the insurers contend they had no payment obligation until this time. Instead, they assert that once claim and cost information was received through discovery, they made an interim payment. Had they received a properly prepared reimbursement request applying the court’s rulings, they claim, they would have likewise made a payment. They contend there is no evidence that further demands for payment would have been futile.

The question of when the insurers had notice of the claim is a question of fact. While the insurers contend that this could be measured only after the computational expert submitted his report, the obligation to reimburse does not necessarily arise at only that time. See, e.g., Solid Gold Realty, Inc. v. Mondry, 399 N.W.2d 681, 684 (Minn.App.1987) (noting that determination as to date demand for payment made was factual question). The district court was the trier of fact here and had extensive knowledge of the issues. We decline to reverse the findings as to when notice of the claim occurred and whether any further demands would have been futile.

VIII.

We next address whether the policies provide coverage for supplemental payments 3M made to plaintiffs on mixed claims for disease and implant rupture. Although 3M had initially waived coverage for damages from rupture, the district court ultimately ruled that 3M would be covered for supplemental payments made to plaintiffs who suffered from disease and whose implants also had ruptured. The insurers challenge this decision.

To obtain coverage under the policies at issue, the injury to the claimant must be unexpected and unintended from the insured’s point of view. In their declaratory judgment complaint, the insurers sought a determination that they had no duty to indemnify 3M for expected-and-intended claims. In 1995, 3M moved for summary judgment to dismiss this defense as to the claims for disease, asserting there was no evidence it expected or intended that the implants would cause the systemic diseases alleged by most plaintiffs. It explained that the only evidence of expected injuries concerned purely mechanical breast implant failures, such as ruptures, that a small number of recipients had experienced, and it explicitly waived coverage for these mechanical-failure claims. The court granted 3M’s motion for summary judgment and dismissed the insurers’ intended-and-expected defense as to the diseases.

An issue later arose as to coverage for “supplemental rupture payments” that 3M made pursuant to the global settlement. These payments — ranging in amount from $15,000 to $50,000 — were made to those claimants whose implants had ruptured, but who also asserted that their systemic diseases were caused by their implants. These payments were made in addition to the payments made for disease. In November 1997, the insurers and 3M moved for summary judgment as to whether these supplemental rupture payments were covered. The court, by the predecessor judge, denied 3M’s motion and granted the insurers’ motion, determining that 3M had renounced insurance coverage for those rupture payments when it “voluntarily narrowed the [expected or intended] issue to include disease claims and to exclude mechanical claims.”

Upon reconsideration of the 1997 decision, the court determined that 3M had waived coverage for purely mechanical failure claims, but concluded that the supple mental coverage for mixed -claims that “merely grant[ed] a systemic disease claimant a higher settlement amount because there has been a rupture” had not been withdrawn. After the court trial on this defense, the court concluded that these payments were covered and that 3M did not clearly waive coverage for them. It later applied the same reasoning to the claims of plaintiffs who had opted out of the global settlement. The insurers challenge this ruling.

Interpretation of language in an insurance policy is reviewed de novo. See-feld, 481 N.W.2d at 64. Questions of waiver that do not involve questions of fact, as is the case here, are also reviewed as a matter of law. Montgomery Ward & Co. v. County of Hennepin, 450 N.W.2d 299, 304 (Minn.1990). In 1995, 3M waived its right to seek payment for mechanical failure claims such as rupture that were expected or intended and these were not in any event covered by the policy. The district court in 1997 addressed whether the supplemental payments for rupture were included in that waiver and ruled that they were. While 3M seeks to avoid the waiver by characterizing the payments as a mere enticement to encourage these mixed-claim plaintiffs to settle, neither we nor the 1999 district court can characterize these supplemental payments as anything but payments for rupture. 3M waived coverage for expected and intended injuries — as it had to under the terms of the policy — and this waiver applies to the supplemental payments for mixed-claim plaintiffs. Consequently, we reverse the district court’s ruling to the contrary as an error of law.

IX.

The insurers argue that the district court erred in awarding 3M its coverage-action fees and costs after finding that the insurers breached the implied covenant of good faith and fair dealing. Among other things, the insurers specifically argue that attorney fees are not recoverable under Minnesota law absent statutory authorization or breach of a contractual duty to defend and that the trial court’s determination that 3M suffered no legally recognized damages as a result of any breach precludes a fee award. 4

When the facts are undisputed, whether an award of attorney fees is proper in a declaratory-judgment action to determine insurance coverage is a question of law subject to de novo review. Spicer, Watson & Carp v. Minn. Lawyers Mut. Ins. Co., 502 N.W.2d 400, 402 (MinmApp. 1993), review denied (Minn. Sept. 30, 1993).

Relying on tort cases stating that whether a common-law duty exists raises a factual question, 3M argues that the court’s “delineation” of the insurers’ implied duty of good faith, on which the court premised 3M’s fee award, raises a factual question subject to a clearly erroneous standard of review. See Louwagie v. State Farm Fire & Cas. Co., 397 N.W.2d 567, 570 (Minn.App.1986), review denied (Minn. Feb. 13, 1987); Johnson v. Urie, 394 N.W.2d 846, 849 (Minn.App.1986), affd, 405 N.W.2d 887 (Minn.1987). But unlike a common-law duty, a contractual duty, such as the implied duty of good faith requires contract interpretation. See Lake Superior Paper Indus, v. State, 624 N.W.2d 254, 258 (Minn.2001). And like all issues involving contract interpretation, whether an implied contractual duty exists raises a question of law reviewable de novo. See U.S. v. Basin Elec. Power Coop., 248 F.3d 781, 796 (8th Cir.2001) (“[t]he application of the implied covenant of good faith is * * * an issue of contract interpretation that we review de novo”). Accordingly, even if our decision turned on the scope of the insurers’ implied duty of good faith, the standard of review is de novo.

The Morrison Exception

As a general rule, a litigant in Minnesota is not entitled to an award of attorney fees and costs absent explicit statutory or contractual authorization. Ly v. Nystrom, 615 N.W.2d 302, 314 (Minn. 2000). Under a limited exception, however, an insured may recover attorney fees it incurs in a coverage action when its insurer breaches a contractual duty to defend. Morrison v. Swenson, 274 Minn. 127, 137-38,142 N.W.2d 640, 647 (1966). In carving out this narrow exception, the supreme court reasoned that fees arising directly out of a breach of the duty to defend are recoverable “as a direct loss incident to the breach.” Id. at 138, 142 N.W.2d at 647.

In the 36 years since its decision in Morrison, the supreme court has consistently declined to extend Morrison to permit a fee award absent a breach of a contractual duty to defend. See, e.g., Garrick v. Northland Ins. Co., 469 N.W.2d 709, 714 (Minn.1991) (overruling line of cases finding authority in Uniform Declaratory Judgments Act, Minn.Stat. § 555.08 (1990), to award fees in any coverage dispute); Olsen v. Preferred Risk Mut. Ins. Co., 284 Minn. 498, 507-08, 170 N.W.2d 581, 587 (1969) (denying fees incurred in declaratory-judgment action for collision coverage under automobile policy); Abbey v. Farmers Ins. Exch., 281 Minn. 113, 119, 160 N.W.2d 709, 712 (1968) (denying fees incurred in suit to establish right to disability benefits). Significantly, in the two cases in which this court awarded fees absent a breach of the duty to defend, the supreme court reversed. See Jenoff, Inc. v. N.H. Ins. Co., 545 N.W.2d 385 (Minn. App.1996), rev’d, 558 N.W.2d 260 (Minn. 1997); Am. Standard Ins. Co. v. Le, 539 N.W.2d 810 (Minn.App.1995), rev’d, 551 N.W.2d 923 (Minn.1996).

An award of attorney fees in this case is neither authorized by statute nor justified by a breach of a contractual duty to defend. Nonetheless, noting that current Minnesota law is “baffling” and subject to “serious criticism on economic grounds,” the district court awarded 3M attorney fees on a determination that fees are recoverable when an insurer does not “deal fairly” with its insured. In departing from precedent, the court reasoned that the failure to award fees works as a disincentive to the prompt performance of insurance obligations and causes the insured to incur “real economic damage” — as distinct from contract damages. — because it prevents the insured from deploying its resources in a timely and efficient manner. Notwithstanding that this is not a case involving commercial general-liability insurance, the court further reasoned that fee awards are “required as a practical matter if commercial general liability insurance is to work in a coherent manner in [a] litigious modern industrial society and economy.”

Relying on American Standard, 3M argues that the fee award is proper because any breach, not just a breach of the duty to defend, will support a fee award. The American Standard court stated in dictum that an insured may not recover attorney fees incurred in maintaining or defending a coverage action “unless the insurer has breached the insurance contract in seme respect — usually by wrongfully refusing to defend the insured.” 551 N.W.2d at 927 (emphasis added). Read in context, however, the court’s reference to “breach[] * * ⅞ in some respect” may be interpreted only as a reference to breaches of obligations associated with the duty to defend, not other forms of breach.

Throughout the opinion, the American Standard court repeatedly stated that the Morrison exception is limited to damages resulting from the insurer’s breach of its duty to defend. It also noted that Minnesota courts have “consistently resisted efforts to expand the Morrison holding to allow collection of attorney fees in actions [that] do not involve the insurer’s breach of contract by failure to assume the duty to defend.” Id. at 926. And the court concluded by overruling two cases to the extent that they “are or can be interpreted as inconsistent with the [ ] rule adopted in Morrison * * * as a limited exception to the general rule that legal fees are not recoverable absent statutory authority.” Id. at 927-28. The American Standard court’s reference to a breach of the insurance contract “in some respect” cannot thus fairly be read as signaling a willingness to expand Morrison and abandon long-standing precedent. A contrary , interpretation would render American Standard meaningless.

In addition, absent a breach of the duty to defend, a fee award impermissibly expands the Morrison exception beyond its rationale. In cases involving a duty to defend, the insured has contracted to avoid the burdens of litigation and the insurer has contracted to shield the insured from those burdens. When the insurer breaches its duty to defend, the insured must do what it contracted to avoid-namely, assume the burden of litigation to enforce its contractual right to a defense. The cost of forcing the insurer to defend is thus a “direct loss incident to the breach.” Morrison, 274 Minn, at 138,142 N.W.2d at 647.

By contrast, the attorney fees spent in a declaratory-judgment action over reimbursement are not a “direct loss incident to the breach” because the promise to reimburse the insured’s liabilities does not involve a promise to relieve the insured from the burdens of litigation. An agreement to reimburse the insured’s liabilities is simply an agreement for the payment of money, an obligation typical of many contract disputes. A fee award absent a breach of the duty to defend, therefore, defeats the rationale on which the Morrison exception is premised.

Because neither the legislature nor the supreme court has clearly expressed an intent to permit a fee award absent statutory authorization or breach of a contractual duty to defend and such an award is inconsistent with the rationale behind the Morrison exception, the district court erred in awarding 3M its coverage-action fees and costs. See Garrick, 469 N.W.2d at 714 (“If the change in Minnesota’s historical doctrine is to be made, it seems to us that this argument ought to be directed to the legislature”).

Absence of Legally Recognized Damages

Even if attorney fees were recoverable absent a beach of a contractual duty to defend, the fee award in this case is barred by the district court’s determination that 3M suffered no legally recognized damages as a result of any breach. Under Minnesota law, a party may not proceed with a breach claim once the court has determined that the party has no recoverable damages. See Despatch Oven Co. v. Rauenhorst, 229 Minn. 436, 447, 40 N.W.2d 73, 80 (1949) (affirming judgment against party who could demonstrate only nominal damages arising from alleged breach of contract); Sloggy v. Crescent Creamery Co., 72 Minn. 316, 317-18, 75 N.W. 225, 226 (1898) (affirming dismissal of breach-of-contract claim where party could not demonstrate that damages arose from breach). The damages requirement assures that breach claims redress real, tangible harm.

Here, the court granted the insurers a directed verdict on all of 3M’s breach claims, including breach of the implied covenant of good faith and fair dealing, reasoning that 3M had “failed to create a jury question for legally recognized consequential contract damages.” The court noted that the damages 3M sought were “inherently speculative,” and it rejected 3M’s efforts to use the efficient use of capital as a measure of damages in lieu of lost profits.

Nonetheless, the court subsequently awarded 3M attorney fees and costs to remedy and deter “economic damage.” The only “economic damage” the district court identified as resulting from the insurers’ breach was 3M’s inability to “put [its] money to work immediately,” which resulted in damage to 3M’s standing in the financial markets. But the insurers correctly argue that even if economic damage were an adequate measure of damages and the record in fact supported the court’s finding of economic damage, the damage did not result from the alleged breach. 3M’s litigation costs do not measure its inability to put its money to work immediately because the alleged breach did not arise from delayed receipt of insurance proceeds but from inadequate responses to 3M’s prelitigation demands. We agree that if this case is to be likened to duty-to-defend cases, there must be damage directly caused by the breach.

Thus, even if this court, as an error-correcting court, had the authority to expand Morrison to permit a fee award based on a breach of the implied covenant of good faith and fair dealing, the fee award in this case would be improper because 3M failed to establish damage directly flowing from the alleged breach.

X.

3M argues that the district court erred in denying its motion to amend its counterclaim to add a claim for punitive damages. 3M recognizes that under Minnesota law punitive damages may not be recovered for breach of contract except in exceptional cases where the breach is accompanied by an independent tort. See Wild v. Rarig, 302 Minn. 419, 440, 234 N.W.2d 775, 789 (1975). It argues, however, that the findings on which the court premised its determination that the insurers breached the implied covenant of good faith make this the “exceptional case” in which a claim for punitive damages should be allowed. Whether punitive damages are allowed is a question of law subject to de novo review. Jensen v. Walsh, 623 N.W.2d 247, 249 (Minn.2001).

Minnesota law allows punitive damages when a party acts with “deliberate disregard for the rights or safety of others.” Minn.Stat. § 549.20 (2000). Punitive damages are not permitted in breach-of-contract cases absent an independent tort, however, even when the breach is malicious. See, e.g., Lickteig v. Alderson, Ondov, Leonard & Sween, P.A., 556 N.W.2d 557, 561 (Minn.1996); WUd, 302 Minn, at 440, 234 N.W.2d at 789; Pillsbury Co. v. Nat’l Union Fire Ins. Co., 425 N.W.2d 244, 249-50 (Minn.App.1988), appeal dismissed (Minn. Mar. 13, 1989); Saltou v. Dependable Ins. Co., 394 N.W.2d 629, 633 (Minn.App.1986). Because breach of the implied covenant of good faith and fair dealing is not a tort, the district court did not err in denying 3M’s motion to amend its counterclaim to add a punitive-damages claim. Cheme Contracting Corp. v. Wausau Ins. Cos., 572 N.W.2d 339, 343-44 (Minn.App.1997) (stating that insured’s bad-faith claim arises from contract and, therefore, sounds in contract, not tort), review denied (Minn. Feb. 19,1998); Pillsbury, 425 N.W.2d at 250 (refusing to recognize tort of bad-faith denial of insurance claim).

Relying on Jensen, where the court held that a plaintiff could seek punitive damages in an action involving vandalism, 3M argues that the word “rights” in Minn. Stat. § 549.20 refers to all rights, including contract rights, and reflects a legislative intent to allow punitive damages .for the “deliberate disregard” of contract rights. The Jensen court specifically noted, however, that Minn.Stat. § 549.20 was intended to codify the common law on punitive damages. Jensen, 623 N.W.2d at 249-50; see also Minn-Iowa Television v. Watonwan T.V. Improvement Ass’n, 294 N.W.2d 297, 311 (Minn.1980) (stating that statute “leaves intact the caselaw prohibition against punitive damages in contract actions”). Jensen may not, therefore, be read as signaling a change. Indeed, if the legislature had intended to overrule the line of cases prohibiting punitive damages in contract cases, it would have specifically provided for such awards in the statute. See Barr/Nelson, Inc. v. Tonto’s, Inc., 336 N.W.2d 46, 52 (Minn.1983). Instead, when the legislature amended Minn.Stat. § 549.20 in 1986 and 1990, it chose not to extend punitive damages to breach-of-contract claims.

Relying on Jensen once again, 3M also argues that punitive damages are allowed in any case involving economic injury. Jensen involved a tort action for intentional damage to property, however, not a contract action. Thus, Jensen may not fairly be read as standing for the proposition that economic injury justifies a punitive-damages award in a breach-of-contract action.

Finally, 3M argues that punitive damages would provide an incentive to the prompt performance of insurance obligations and that this court should bring Minnesota in line with other states to prevent Minnesota from becoming a “haven for insurer misconduct.” 3M asserts that “most other states” allow punitive damages to be recovered from insurers who breach their contractual obligations in bad faith. But the Minnesota legislature has made a clear choice that punitive damages are not available for breach of contract and that prejudgment interest is an appropriate remedy for the delayed performance of contractual obligations. And “the task of extending existing law falls to the supreme court or the legislature,” not to this court. Tereault v. Palmer, 413 N.W.2d 283, 286 (Minn.App.1987), review denied (Minn. Dec. 18,1987); Flynn v. Am. Home Prods. Corp., 627 N.W.2d 342, 346 (MinnApp. 2001).

In any event, as the insurers correctly point out, only one jurisdiction — Delaware — allows punitive damages for breach of the implied covenant on a contract theory. See Tackett v. State Farm Fire & Cas. Ins. Co., 653 A.2d 254, 264-66 (Del.1995). Of the remaining jurisdictions 3M cites, some bar punitive damages in first-party insurance cases absent an independent tort, some have enacted a statutory action for insurer bad faith and allow punitive damages, and others allow punitive damages based on a bad-faith tort applicable to first-party insurance contracts. We agree with the insurers, therefore, that Minneso ta’s punitive-damages rule is “comfortably within the mainstream.”

Because breach of the implied covenant of good faith is not an independent tort, the district court properly denied 3M’s request to amend its counterclaim to add a punitive-damages claim.

XI.

3M argues that the district court abused its discretion in denying its request for extra-contractual equitable relief that would make it whole and hold the insurers accountable for their conduct. Specifically, 3M sought (1) restitution for the unjust enrichment the insurers achieved by withholding payment; (2) interest at a market rather than statutory rate; (3) compound rather than simple interest; (4) an order requiring the insurers to “drop down” or fill gaps in the coverage created by the insolvencies of other insurers in the program; and (5) access to its insurance on a continuous-trigger, all-sums basis. In denying 3M’s request for relief, the district court reasoned that equitable relief may not be granted where the parties’ rights are governed by a valid contract and that, given the equities, the requested relief was inappropriate.

Equity does not authorize courts to rewrite or abrogate contracts, impose obligations beyond contract terms, or change contractual obligations. See U.S. Fire Ins. Co. v. Minn. State Zoological BcL, 307 N.W.2d 490, 497 (Minn.1981); Cady v. Bush, 283 Minn. 105, 110, 166 N.W.2d 358, 362 (1969). Specifically in the insurance context, equitable doctrines “may not be used to create coverage where none is provided for in the contract.” Malakowsky v. Johannsen, 374 N.W.2d 816, 819 (Minn.App.1985) (stating it would be “wholly improper to impose coverage liability upon an insurer for a risk not specifically undertaken and for which no consideration has been paid”). A determination of damages may be disturbed only for clear abuse of discretion. Holiday Recreational Indus. v. Manheim Servs. Corp., 599 N.W.2d 179, 183 (Minn.App.1999).

3M concedes that the relief it is i-equest-ing is “not within the realm of typical contract remedies” and that there are “contract or common-law limitations that might otherwise preclude such remedies.” Nonetheless, it argues that extra-contractual equitable remedies are appropriate because (1) it lacks an adequate remedy at law; and (2) the insurers forfeited their right to rely on contract provisions to limit their liability. On this basis, 3M argues that the court may eliminate the contractual limitations on the insurers’ coverage obligations, including the policy limits, to award 3M more coverage than it purchased. But 3M offers neither precedent nor a rational basis for the relief it requests.

3M’s claim that it lacks an adequate remedy at law is without merit. 3M seems to suggest that its inability to prove damages from the breach demonstrates the lack of an adequate remedy at law. As the insurers correctly point out, however, the inability to prove damages suggests only that 3M’s proof was deficient, not that the remedy was inadequate. 3M did not have a remedy at law because it could not prove legally recognized damages, an element of its breach claim.

3M’s claim that the breach of the duty of good faith in third-party cases gives rise to “a forfeiture [of the insurer’s] right to rely on contract provisions” and “yields an extra-contractual remedy” similarly has no merit. A bad-faith refusal to settle yields only a typical contractual remedy. See Short v. Dairyland Ins. Co., 334 N.W.2d 384 (Minn.1983) (insurer whose bad-faith refusal to settle resulted in judgment against the insured for amount in excess of policy limits held liable only for foreseeable damages arising from breach, i.e., damages caused by insurer’s failure to settle). Breach of the implied covenant similarly gives rise only to traditional contract remedies. Cherne, 572 N.W.2d at 343-44.

Because 3M has not established a legitimate basis for departing from the rule that equitable relief is not permitted when the parties’ rights are governed by a valid contract, the district court properly denied 3M’s request for equitable relief.

XII.

The insurers argue that the district court erred in instructing the jury on risk of loss and waiver in connection with their claim that 3M misrepresented material facts by failing to disclose its acquisition of McGhan Medical Corporation and its knowledge that minute amounts of silicone leaked from every breast implant 3M manufactured between 1977 and 1985. Trial courts are allowed “considerable latitude” in selecting the language in the jury charge. Alholm v. Wilt, 394 N.W.2d 488, 490 (Minn.1986) (quotations omitted). “The sole requirement is that the instructions convey to the jury a clear and correct understanding of the law.” Kirsebom v. Connelly, 486 N.W.2d 172, 174 (Minn.App.1992). Accordingly, errors in jury instructions warrant a new trial only if they destroy the substantial correctness of the charge, cause a miscarriage of justice, or result in substantial prejudice. Id. A new trial is also warranted if a reviewing court is unable to determine whether an erroneous instruction affected the jury. Apache Plaza, Ltd. v. Midwest Sav. Ass’n., 456 N.W.2d 729, 733 (Minn.App.1990), review denied (Minn. Aug. 23,1990).

a. Risk of Loss

An insurer may rescind an insurance policy when an insured’s misrepresentation increases the risk of loss. See Minn.Stat. § 60A.08, subd. 9 (2000). A misrepresentation increases the risk of loss when it impairs the insurer’s ability to make a reasonable decision initially to assume the risk of coverage. Howard v. Aid Ass’n for Lutherans, 272 N.W.2d 910, 913 (Minn.1978). The risk of loss is also increased when there is a greater likelihood that the insurer will be liable in the future than there would have been if the facts were as the insured stated them to be. Sec. Mut. Cas. Co. v. Affiliated FM Ins. Co., 471 F.2d 238, 242 (8th Cir.1972) (quoting approved jury instruction); Berthi-aume v. Minn. Mut. Life Ins. Co., 388 N.W.2d 15, 19 (Minn.App.1986) (insured’s failure to disclose, high-blood-pressure condition on life-insurance application entitled insurer to void policy because it influenced insurer’s decision to assume risk of coverage), review denied (Minn. July 31, 1986); cf Preferred Risk, 277 Minn, at 350, 152 N.W.2d at 483 (insured’s misrepresentation that he and his wife were to be principal drivers of insured vehicle did not increase insurer’s risk of loss).

The court instructed the jury that to find an increase in the risk of loss, it needed to find, among other things, that “at the time the policy was issued, the risk of loss associated with insuring breast implants was greater than the risk of loss associated with insuring other pharmaceutical products.” The insurers argue that the instruction “dramatically overstated insurers’ burden” and rendered meaningless the jury’s verdict on 3M’s misrepresentation claim. We agree that the instruction overstated the insurers’ burden.

Minnesota courts have not previously grafted onto the statutory standard for risk of loss a requirement that the risk of loss be categorically greater than other risks for which the insurer provided coverage. For instance, in the absence of a finding that the risk of insuring an additional owner was greater than the risk of insuring the disclosed owner this court held that an insured’s failure to disclose multiple ownership of a vehicle increased the risk of loss and relieved the insurer from liability under the policy because “[mjultiple ownership logically increases authority for permissive use, which, in turn, increases the risk of loss.” Transam. Ins. Co. v. Austin Farm Ctr. Inc., 354 N.W.2d 503, 506 (Minn.App.1984), review denied (Minn. Feb. 6, 1985). And despite the absence of evidence that the hazard posed by asbestos was greater than the hazard posed by other products manufactured by the insured, this court also held that it was for the jury to decide whether the insured’s misrepresentation about its involvement in asbestos-related litigation increased the insurer’s risk of loss. Indep. Sch. Dist. No. 197 v. Accident & Cas. Ins. of Winterthur, 525 N.W.2d 600, 606 (Minn.App.1995), review denied (Minn. Apr. 27,1995). The court’s instruction was, therefore, erroneous.

But the error was harmless because the instruction did not result in substantial prejudice to the insurers. We recognize that, had the jury found that 3M misrepresented material facts in connection with its insurance applications, it could also have found, with an instruction that did not contain the challenged risk-of-loss language, that the addition of breast implants to 3M’s inventory of products increased the likelihood of future liability for the insurers and the potential size and number of payments. But the jury found that 3M had not misrepresented material facts in connection with its applications for insurance for the years 1977-1981. And the record supports the jury’s finding. 5 Absent a finding of misrepresentation, the erroneous risk-of-loss instruction had no effect on the jury’s verdict and is, therefore, harmless error.

b. Waiver

3M asserted waiver as a defense to the insurers’ misrepresentation claim. Specifically, 3M argued that the insurers waived any right to void the policies by continuing to insure 3M after they learned of the alleged misrepresentations. Over the insurers’ timely objection, the court instructed the jury that, to establish waiver, 3M had to show that the insurers knew the information on which they based their misrepresentation claim and either “continued to treat the policy as in full force anyway, or continued to insure 3M for additional periods.” Based on this instruction, the jury found that the insurers “knowingly waive[d] any prior misrepresentation that 3M may have made by continuing to insure 3M.”

Because the jury found that 3M made no material misrepresentations in its applications for insurance, we conclude that the risk-of-loss instruction was harmless error and we do not discuss the insurers’ claim that the waiver instruction was also erroneous.

DECISION

The decision by the district court is affirmed in part and reversed in part.

1 . After the last of the occurrence policies ended on December 31, 1985, the claims-made policies were in effect. This issue involves only the occurrence policies.

2 . 3M explains that material contained in its offer of proof was "virtually identical" with material it submitted in response to motions by insurers on other defense-cost issues earlier in 1999. 3M does not, however, assert that this material was before the court earlier, when it decided the summary judgment motions.'

3 . In 1983, the policy was replaced by a policy that contained almost identical language.

4 . The insurers also argue that the attorney-fees award is improper because (a) the district court’s finding that the insurers breached the implied covenant of good faith and fair dealing, on which the fee award was premised, is incorrect as a matter of law; (b) the court prejudicially erred in treating the breach issue, which is factual, as a declaratory-judgment issue to be resolved by the court; (c) the court abused its discretion in failing to make separate findings regarding the conduct of each insurer; and (d) the court erred in granting 3M prejudgment interest on the fee award in violation of Minn.Stat. § 549.09 (2000). Although we find merit in all of these arguments, we do not discuss them because the insurers’ arguments discussed above are dispositive.

5 . Phillip Purdy, the broker in charge of marketing 3M's casualty insurance, testified that 3M clearly indicated in its insurance applications that it was not disclosing all of its subsidiaries, products, or recent acquisitions, and that the insurers never asked for additional disclosures. He also testified that when he wrote the November 1981 memo on which the insurers premised part of their misrepresentation claim, he was not aware that minute amounts of silicone leaked from every silicone-gel breast implant 3M manufactured between 1977 and 1985. The jury apparently believed Purdy.