OPINION
CYNTHIA HOLCOMB HALL, Circuit Judge:The Resolution Trust Corporation and the former directors and officers of the insolvent Pacific Savings Bank appeal the district court’s summary judgment that no insurance coverage is available under policies issued by American Casualty Company and its predecessors. American Casualty appeals the district court’s order requiring it to advance defense costs incurred by the directors and officers in a lawsuit with the RTC until a final determination of coverage under the policies. We affirm the district court’s conclusion that no coverage exists under the policies and dismiss as moot the insurer’s appeal of the advancement order.
I.
In February 1989, federal regulators declared Pacific Savings Bank insolvent. The Federal Deposit Insurance Corporation, seeking to recover investment losses of more than $70 million by the bank, filed suit in district court against Pacific’s directors and officers (the “Directors/Officers”), alleging negligence, breach of fiduciary duty, and breach of contract.1 American Casualty Company of Reading, Pennsylvania, the issuer of Pacific’s several directors’ and officers’ insurance policies (“D & 0 policies”), subsequently sought a declaratory judgment that the D & 0 policies did not cover regulatory claims against the Directors/Officers. The Resolution Trust Corporation, which had replaced the FDIC as plaintiff in the underlying litigation, intervened and, with the Directors/Officers, counterclaimed against American Casualty and related entities Continental Casualty Company, CNA Financial Corporation, and CNA Insurance Companies (together “CNA”) for a declaration that the policies did in fact encompass the RTC claims.
A. Background
The coverage litigation implicated five D & 0 policies, each of which contained two identical provisions. First, the “Discovery Clause” gave Pacific the right to purchase “discovery coverage” in the event that CNA chose not to renew the policy. This coverage provided for additional time to report losses based on acts occurring prior to the policy’s expiration date. Second, the “Notice Clause” required CNA to inform Pacific at least thirty days before expiration of a policy that it would not renew coverage. Relevant provisions of the other policies, and circumstances surrounding their adoption, are as follows:
1. The 1980-83 Policy
In September 1980, MGIC Indemnity Corporation sold Pacific a three-year, $10 million D & 0 policy. After the policy expired, CNA purchased MGIC’s D & 0 business and, through an “Assumption Agreement,” assumed liability for “those expired and can-celled [MGIC] policies ... where a claim can be reported after the policy expiration date.” CNA and WMBIC, MGIC’s successor in interest, later executed a “Rescission Agreement” purporting to eliminate CNA’s liability for “claims in existence” under the assumed MGIC policies.
2. The 1983-85 Policy
Prior to expiration of the 1980-83 Policy, MGIC sent Pacific’s Broker, Pacifica Insurance Services (“PIS”), a Commitment for Insurance (“1983 Commitment”) setting forth proposed terms of a successor policy. The 1983 Commitment stated that “[t]he proposed terms and conditions under this Commitment for Insurance differ from the expiring Policy’s terms and conditions.” Specifically, MGIC had added a “Regulatory Exclusion,” which, as explained below, exempted from coverage the claims of certain government agencies. Although MGIC had also increased the premium and reduced the coverage limits, Pacific purchased the policy.
CNA later executed the Assumption Agreement and took over the policy. After *885researching the market for policies with more favorable terms and finding none, Pacific and PIS subsequently contacted CNA and renegotiated the policy terms, agreeing to a higher premium in exchange for removal of the Regulatory Exclusion.
3. The 1985-86 Policy
Before expiration of the 1983-85 Policy, CNA sent Pacific, through PIS, a formal quotation (“the 1985 Quotation”) with attached endorsement copies of new policy terms, including the Regulatory Exclusion. The 1985 Quotation prominently stated that “[a]cceptance of the terms and conditions as stated on this Quotation of Insurance ... waives any and all rights to” discovery coverage under the 1983-85 Policy. Kenneth Bo-vard, Pacific’s Vice President and General Counsel, subsequently met with the Directors/Officers to describe the proposed policy. Although Bovard knew at that time that Pacific was unable to purchase a D & 0 policy from any carrier without a Regulatory Exclusion, he neglected to mention the exclusion to the Directors/Officers. Pacific then purchased the policy and Bovard circulated a copy to the Directors/Officers, attaching a cover memorandum explaining changes from the 1983-85 Policy, describing the Regulatory Exclusion, and noting that such an exclusion “is uniformly issued in all similar policies throughout the country.” The Directors/Officers retained the policy without protest.
4. The 1986-87 and 1987-88 Policies
In late 1986, CNA again sent Pacific a quotation (the “1986 Quotation”) listing provisions of a proposed policy and attaching copies of proposed endorsements. Despite further increases in premium and reductions in coverage, Pacific purchased the policy, which contained a similar waiver of discovery coverage. CNA and Pacific repeated this process in late 1987 and, in late 1988, CNA refused to issue another policy.
B. District Court Proceedings
In the course of the declaratory judgment proceedings, the district court issued several relevant orders.
1. The Regulatory Exclusion Order
In 1991, on consideration of summaxy judgment motions, the district court held that the Regulatory Exclusion did not violate public policy and facially applied to RTC claims against the Directors/Officers. See 758 F.Supp. 1340 (C.D.Cal.1991) (Baker I). At that time, however, the court declined to consider the RTC’s2 assertions of estoppel (to prevent CNA from relying on the exclusion) and reformation (to delete the exclusion from the policies because of inadequate notice) and therefore it did not decide whether the exclusion was actually enforceable in the D & 0 policies. Id. at 1342.
2. The Advancement Order
From the inception of the declaratory judgment proceedings, CNA maintained that the D & 0 policies did not create a duty to pay the Directors/Officers’ defense costs. CNA noted that the “Option Clause” in the policies supported this position: “The Insurer may at its option and upon request, advance on behalf of the Directors or Officers ... expenses which they have incurred in connection with claims made against them, prior to disposition of such claims.”
The Direetors/Officers disagreed and sought a declaration requiring CNA to advance defense costs. The district court granted the motion, holding that CNA “has a present duty to simultaneously reimburse the insured defendants’ defense fees and costs incurred in the underlying action.” The court agreed that the D & 0 policies seemed to make CNA’s duty to pay defense expenses “optional and therefore not [derivative of] any duty to defend enumerated in the policies.” It held, however, that Okada v. MGIC Indemnity Corp., 823 F.2d 276 (9th Cir.1986), compelled a different result. The court interpreted Okada, in which we required an insurer to advance defense costs under a D & 0 policy with identical language, as holding “that the duty to advance defense costs for an underlying action attaches so long as the underlying claims remain potentially covered under the policies.” The district court believed that the Directors/Offi*886cers’ estoppel and reformation claims, of which it had postponed consideration, created a possibility of coverage under the policies and, as a result, required CNA to pay the defense expenses. CNA appealed.
3. The Coverage Order
In early 1993, CNA moved for summary judgment on the coverage issues. The RTC opposed on three grounds: First, the RTC argued that coverage was available under the 1980-83, 1983-85, and 1985-86 Policies because, by offering successor policies with materially different terms, CNA had constructively failed to renew and therefore had breached the policies by not giving notice of nonrenewal. The RTC contended that, as a result, the Directors/Offieers had a current right to invoke discovery coverage. Second, the RTC contended that coverage was available under the 1985-86 Policy because the Directors/Officers could, as a remedy for CNA’s failure to give notice of the Regulatory Exclusion, reform the policy to excise the exclusion. And, finally, the RTC argued that the Regulatory Exclusions in the 1985-86, 1986-87, and 1987-88 Policies were inapplicable because the RTC is not a “regulatory agency.”
The district court granted CNA’s motion, holding that none of the D & 0 policies covered the RTC claims against the Directors/Offieers. The court based its decision on different grounds for the various policies:
*For the 1980-83 Policy, the Rescission Agreement nullified any liability CNA had under the Assumption Agreement.
* For the 1983-85 Policy, even if CNA had violated the Notice Clause, coverage was not available because the RTC’s only remedy would be to purchase discovery coverage for the year following the alleged nonrenewal. Because the RTC claims were not made until after the discovery period would have expired, any possible remedy would be ineffective.
* For the 1985-86 Policy, even if CNA initially had not renewed the policy, the RTC was not entitled to purchase discovery coverage because Pacific had accepted a successor policy with full notice of its terms.
* For the 1986-87 and 1987-88 Policies, the Regulatory Exclusion was enforceable and nullified coverage of RTC claims.
The RTC filed a timely appeal3 which has been consolidated with CNA’s appeal of the Advancement Order. We conduct de novo review, e.g., O’Shea v. City of San Francisco, 966 F.2d 503, 505 (9th Cir.1992), and affirm the district court’s excellent Coverage Order in all regards.4 As a result, we also dismiss as moot CNA’s appeal of the Advancement Order and vacate the district court’s decision on that issue.
*887II.
The RTC asserts that each of the five D & 0 policies provide coverage for the claims asserted in RTC v. Baker. We address each in turn.
A. The 1980-83 Policy
The district court held that “[t]he CNA parties have established that, as a matter of law, any assumption of liability by Continental Casualty Company for the 1980-83 policy was rescinded by the so-called ‘Rescission Agreement.’ ” In so holding, the court relied in part on Continental Casualty Co. v. Allen, 710 F.Supp. 1088, 1102 (N.D.Tex.1989), which found an identical rescission agreement to terminate CNA’s liability for previously-assumed policies. The RTC contends that Allen is inapposite because, in this case, a “Claims Service Agreement” between CNA and WMBIC, executed contemporaneously with the Rescission Agreement, illustrates that CNA intended to be liable on the 1980-83 Policy even after executing the Rescission Agreement. The RTC’s argument is as follows: The Rescission and Claims Service Agreements refer to “Claims in Existence.” The Claims Service Agreement contains an exhibit listing all such claims but not including the RTC claims. Some evidence indicates that the Claims Service Agreement represents “the extent of the universe of claims” covered by the Rescission Agreement. That evidence creates a material issue of fact regarding CNA’s liability.
The district court considered the extrinsic evidence and held that (1) the Rescission Agreement clearly encompassed CNA’s potential liability under the 1980-83 Policy and (2) “the relevant terms of the Rescission and Assumption Agreements [are] sufficiently unambiguous to merit disregarding the deposition testimony ... relied on by the RTC.” We agree.
Under California law,5 “[s]everal contracts relating to the same matters, between the same parties, and made as parts of substantially one transaction, are to be taken together.” Cal.Civ.Code § 1642. E.g., Nish Noroian Farms v. Agricultural Labor Relations Bd., 35 Cal.3d 726, 201 Cal.Rptr. 1, 5-6, 677 P.2d 1170, 1174-76 (1984). Even interpreted with the Claims Service Agreement, however, the plain meaning of the Rescission Agreement is clear. That agreement defines “Claims in Existence” as those covered by the Assumption Agreement. The Claims Service Agreement, on the other hand, defines the term to be “those claims which were in existence and reflected on the books and records of WMBIC as of November 1, 1983, and as set forth” in the relevant exhibit. As CNA notes, these definitions served the different purposes of the two agreements:
“Claims In Existence” is defined differently in the Rescission Agreement and the Claims Service Agreement, and it makes no sense to suppose that the terms have identical meanings. Under the Claims Service Agreement, Continental was agreeing to service claims for MGIC’s successor. Obviously, for Continental to service the claims, they had to be identified precisely. In the Rescission Agreement, on the other hand, the parties were rescinding Continental’s prior obligations under the Assumption Reinsurance Agreement. Accordingly, they adopted the same broad definition of “Claims In Existence” used in the Assumption Reinsurance Agreement in order to accomplish their manifest purpose, to rescind Continental’s responsibility for all Claims-in-Existence.
Representatives of CNA and WMBIC testified that they intended the Rescission Agreement to terminate all of CNA’s liability on the policy. “Where the parties have attached the same meaning to a promise or agreement or term thereof, it is interpreted in accordance with that meaning.” Restatement (Second) of Contracts § 201(1). “The construction given a contract by the parties before any controversy has arisen as to meaning will, when reasonable, be adopted by the courts.” Western Med. Enters, v. Albers, 166 Cal.App.3d 383, 212 Cal.Rptr. 434, 439 (1985).
CNA’s interpretation certainly is reasonable. At the time of the agreement, the RTC had not filed claims against the Directors/Officers and the parties could not have known *888of those claims. We affirm the district court’s holding that CNA is not responsible for the 1980-83 Policy.
B. The 1983-85 Policy
The district court held that no coverage was available under the 1983-85 Policy because, even if the RTC proved at trial that CNA had violated the Notice Clause, the only remedy would be to allow Pacific to purchase discovery coverage for the year following the alleged nonrenewal. The court decided that the Discovery Clause permitted coverage only for actual claims filed during the discovery period (and not for notice of occurrences that might ripen into claims) and held that, because the RTC claims did not arise until the discovery period would have expired, no remedy was available for CNA’s alleged breach. The RTC takes issue with several of the district court’s conclusions. First, the RTC asserts that the Discovery Clause encompasses notice of occurrences as well as claims. Second, the RTC argues that the proper remedy is to permit the Directors/Officers to purchase discovery coverage now (and not relate coverage back to the time of the alleged breach). And, finally, the RTC claims that CNA should be estopped from relying on the one-year time limitation in the Discovery Clause.
1. Notice of Occurrences
The Discovery Clause provides that, if CNA decides not to renew the policy, Pacific may purchase coverage for “any claim or claims ... made against the Directors or Officers during the period of twelve calendar months after the date of such ... refusal to renew, but only with respect to [acts] committed before the date of such ... non-renewal.” CNA contends that this clause specifically limits coverage to actual claims made within the discovery period. The RTC, however, argues that notice of occurrences that could ripen into claims is adequate to trigger discovery coverage. Specifically, the RTC points to the “Notice of Claims” clause in the policy:
If during the policy period [Pacific] ... becomefs] aware of any occurrence which may subsequently give rise to a claim being made against the Directors and Officers ... and shall during such period, give written notice thereof to the Insurer as soon as practicable and prior to the date of termination of the policy, then any claim which may subsequently be made against the Directors or Officers arising out of the [occurrence] shall, for the purposes of this policy, be treated as a claim made during the policy year in which such notice was given.
The RTC contends that this clause, which clearly applies to notice of occurrences given during the policy period, also applies to notice of occurrences during the discovery period. In support, the RTC presented evidence allegedly illustrating that CNA previously had interpreted discovery coverage to encompass notice of occurrences.
CNA argues that the Notice of Claims clause applies only to notice of occurrences given during the policy period itself. CNA points to the “Coverage Clause” of the policy, which specifically distinguishes between the “policy period” and the “discovery period:”
This policy shall cover Loss ... arising from any claim made (i) within the policy period or (ii) within the discovery period if [such coverage is purchased].... [A]ny claim made subsequent to the policy period as to which notice was given to the Insurer within the policy period as provided in [the Notice of Claims clause] shall be treated as a claim made during the policy period.
CNA contends that the RTC’s extrinsic evidence is irrelevant to interpretation of this “unambiguous” policy language.
The district court agreed with CNA. The court noted that the Tenth Circuit, in American Casualty Co. v. FDIC, 958 F.2d 324, 326-28 (10th Cir.1992) (Oklahoma Federal ),6 interpreted identical policy language as limiting discovery coverage to actual claims made within the discovery period. The court recognized that the Eighth Circuit, in McCuen v. American Casualty Co., 946 F.2d 1401, 1404-06 (8th Cir.1991), reached the opposite conclusion, but concluded that Oklahoma Federal was “better reasoned.” We agree.
*889a. Ambiguity
The RTC contends that the policy is ambiguous and that, as a result, the court must construe policy terms against the insurer. In so arguing, the RTC seeks to elevate an abstract general rule over concrete policy terms. As the California Supreme Court recently clarified, we may not ignore straightforward contract terms in order to create “ambiguities.”
While insurance contracts have special features, they are still contracts to which the ordinary rules of contract interpretation apply.... If contractual language is clear and explicit, it governs. On the other hand, if the terms of a promise are in any respect ambiguous or uncertain, it must be interpreted in the sense in which the promisor believed, at the time of making it, that the promisee understood it. This rule, as applied to a promise of coverage in an insurance policy, protects not the subjective beliefs of the insurer but, rather, the objectively reasonable expectations of the insured. Only if this rule does not resolve the ambiguity do we then resolve it against the insurer.
... [T]he court must interpret the language in context, with regard to its intended function in the policy. This is because language in a contract must be construed in the context of that instrument as a whole, and in the circumstances of that case, and cannot be found to be ambiguous in the abstract.
Bank of the West v. Superior Court, 2 Cal. 4th 1254, 10 Cal.Rptr.2d 538, 544-45, 833 P.2d 545, 551-52 (1992) (emphasis added) (internal quotations and citations omitted). See also AIU Ins. Co. v. Superior Court, 51 Cal.3d 807, 274 Cal.Rptr. 820, 831-32, 799 P.2d 1253, 1264-65 (1990).
Three courts have held that CNA’s policy language unambiguously limits notice of occurrences to the policy period. Each used an analysis similar to that of the Tenth Circuit in Oklahoma Federal:
Given the clear dichotomy drawn in [the Coverage Clause] between the “policy period” and the “discovery period,” coupled with reference in the same clause to the nature of notice that can be given during the policy period under [the Notice of Claims Clause], we cannot conclude that the term “policy period” — as used in both [the Notice of Claims Clause] and [the Coverage Clause] — may reasonably be interpreted to encompass the “discovery period.”
Oklahoma Federal, 958 F.2d at 327 (applying Oklahoma law). See American Casualty Co. v. FDIC, 999 F.2d 480, 482 (10th Cir.1993) (Casper) (applying Wyoming law); Baltimore Federal, 845 F.Supp. at 325-26 (applying Maryland law) (“The plain language of [the Discovery Clause] — ‘any claim or claims which shall be made’ — precludes coverage based on mere notice of potential claims. This is confirmed by a careful reading of the unambiguous language in [the Notice of Claims Clause], requiring that notice of potential claims be given ‘during the policy period,’ and [the Coverage Clause], distinguishing between ‘the policy period’ and ‘the discovery period.’”).
Although McCuen reached the opposite conclusion, that decision is not persuasive because the court apparently failed to consider the distinction between the policy and discovery periods. See Oklahoma Federal, 958 F.2d at 328 (“The analysis of the term ‘claims’ in McCuen ignores the issue of whether [the Notice of Claims] clause applies to notice of occurrences given dining the discovery period.”). The RTC’s argument that the very fact that courts disagree illustrates an ambiguity is unpersuasive: “The mere fact that judges of diverse jurisdictions disagree does not establish ambiguity under the particular principles which govern the interpretation of insurance contracts in California.” ACL Technologies, Inc. v. Northbrook Property & Casualty Ins. Co., 17 Cal.App. 4th 1773, 22 Cal.Rptr.2d 206, 215 n. 39 (1993). Accordingly, we use California’s rule of following clear and unambiguous contract language and find for CNA on this issue.
b. Extrinsic Evidence
The RTC asserts that the evidence illustrates that MGIC and CNA interpreted the Discovery Clause to encompass notice of occurrences. Specifically, the RTC notes that a 1974 memorandum from MGIC to a cus*890tomer stated that discovery coverage under MGIC policies included notice of occurrences. The RTC also points out that several recent CNA memoranda refer to changes in Discovery Clause language, made after the dates relevant here, indicating an attempt to “tighten[] up” the discovery period “by allowing only claims to be reported.”
The district court correctly found the RTC’s evidence unpersuasive. First, the court noted that each document referred to policies with Discovery Clause language different from the 1983-85 Policy. Second, the court held that the language of the 1983-85 Policy could not support inferences reasonably susceptible to those which the RTC sought to draw from the evidence. And, finally, the court found the evidence irrelevant because it did not bear on the parties mutual intent in forming the contract.
In California, “[i]t is immaterial that one of the parties to a contract had an undisclosed intention or belief as to what it meant. [DJrafting history ... [is irrelevant when] it contradicts the objective theory [of interpreting the contract].” ACL Technologies, 17 Cal.App.4th 1773, 22 Cal.Rptr.2d at 217 n. 45 (internal quotation omitted). See Zubia v. Farmers Ins. Exch., 14 Cal.App.4th 790, 18 Cal.Rptr.2d 65, 71 n. 3 (1993) (“Since the [policy] provision at issue in this case is not reasonably susceptible to an interpretation which excludes medical expense coverage, the internal memoranda written by [the insurer’s] personnel are not relevant to our analysis.”).
In identical circumstances, three other courts have rejected the RTC’s contention that this evidence compels a different interpretation of the Discovery Clause. See Baltimore Federal, 845 F.Supp. at 325-26; American Casualty Co. v. FDIC, 821 F.Supp. 655, 666 (W.D.Okla.1993) (Blanchard) (“[The FDIC’s evidence] is completely irrelevant to the issues, as the written contract is not ambiguous.”); American Casualty Co. v. FDIC, No. 91-CV-0058-J, slip op. at 4 (D.Wyo. June 24,1993) (American National) (“[T]he newly discovered evidence offered by the [FDIC] is immaterial to the interpretation of the unambiguous policy language.”). We agree.
2. Remedy
The RTC argues that, even if discovery coverage does not include notice of occurrences, the proper remedy for breach of the Notice Clause is to permit the Directors/Officers to purchase discovery coverage now (and not to relate the coverage back to the time of the breach). The district court disagreed, concluding that “[a]llowing the RTC’s preferred solution would give the insureds much more than they bargained for.”
Courts considering the propriety of the RTC’s requested remedy have split on the issue. The Baltimore Federal court used the approach taken by the district court:
Assuming, only for the purposes of argument, that American Casualty failed to provide the required notice of nonrenewal, the RTC’s motion still must fail because it is not entitled to any meaningful remedy - To allow greater coverage [by extending the discovery period] would be to allow the insured, in effect, to write the new policy, adding or increasing coverage, eliminating conditions and exclusions.... If enforcement of the discovery period is the appropriate remedy, the RTC is left empty-handed because ... the bank gave no notice of a claim, nor was any claim filed against it, during that period.
Baltimore Federal, 845 F.Supp. at 324-25 (internal quotation omitted). The court in Benafield v. Continental Casualty Co., No. LR-C-92-389, 1993 WL 723510 (E.DArk. Oct. 1, 1993), used the same reasoning:
It is one thing to say that plaintiffs may exercise the discovery clause at this late date, and yet another to say that the discovery period should only now be in effect thus giving plaintiffs an additional seven or eight years of coverage. This construction of the policy was not bargained for by the parties, and this Court believes that the plain meaning of the clause is that the discovery period would be that twelve months following the expiration of the policy.
Id., slip op. at 11.
The Allen court, however, held that where the insurer fails to give notice of nonrenewal, *891the “defendants’ right to exercise the discovery option still exists due to a lack of notice of non-renewal” and concluded that the “right to exercise the discovery option runs ten days from the date of this judgment.” Allen, 710 F.Supp. at 1100, 1103. The Allen opinion is unclear, however, about whether actual claims had been filed against the insured within the discovery period or whether the court intended to extend that period until the date of its order. The opinion in American Casualty Co. v. FDIC, No. 86-4018, 1990 WL 66505 (N.D.Iowa Feb. 26, 1990) (Christensen ),7 in which the court followed Allen, is similarly ambiguous.
There are no circuit court decisions on point.8 Unfortunately, California law also provides little guidance on the issue. The RTC cites several cases which hold that insurance policies remain in effect until proper notice of nonrenewal is given. See, e.g., National Auto. & Casualty Ins. Co. v. California Casualty Ins. Co., 139 Cal.App.3d 336, 188 Cal.Rptr. 670, 672 (1983); Borders v. Great Falls Yosemite Ins. Co., 72 Cal.App.3d 86, 140 Cal.Rptr. 33, 35 (1977). These cases, however, involve specific statutes that apply only to automobile insurance. See Cabins. Code § 663(c) (“In the event that an insurer fails to give the named insured ... notice of nonrenewal ..., the existing policy, with no change in its terms and conditions, shall remain in effect for 30 days from the date ... the notice of nonrenewal is delivered.”).9
We think Baltimore Federal, Benafield, and the district court were correct. Permitting the RTC to purchase discovery coverage for claims not filed within the one-year period following expiration of the policy would give the Directors/Officers a windfall: indemnity for claims not otherwise covered under the policy. The remedy accorded by the district court, on the other hand, gives the Direetors/Officers the benefit of their bargain and no more. The RTC’s fears that this remedy gives insurers an incentive to “flout[ ] its notice provisions” are misplaced. Intentional disregard of policy provisions would, undoubtedly, expose an insurer to liability and punitive damages for bad faith. See, e.g., Tibbs v. Great Am. Ins. Co., 755 F.2d 1370, 1373-75 (9th Cir.1985) (applying California law). Insurers are not likely to undertake such a risk in the speculative hope of depriving an insured of discovery coverage. We affirm the district court on this point.
3. Estoppel
Finally, the RTC asserts that the district court should have estopped CNA from relying on the Discovery Clause time limitation because the insurer engaged in wrongful conduct by disguising its nonrenewal of the policy. We disagree.
California courts apply equitable estoppel to prevent injustice where one party has, through action or inaction, caused another to act to his detriment. The party invok*892ing the doctrine must prove four elements: “(1) that the party to be estopped was apprised of the facts; (2) that he intended his conduct to be acted on, or that he acted in such a way that the party asserting estoppel reasonably could believe that he intended his conduct to be acted upon; (3) that the party asserting estoppel was ignorant of the actual facts; and (4) that the party asserting estop-pel relied upon the conduct to his injury.” Adams v. Johns-Manville Corp., 876 F.2d 702, 707 (9th Cir.1989) (applying California law). See, e.g., Hair v. State, 2 Cal.App.4th 321, 2 Cal.Rptr.2d 871, 875 (1991). ‘"Where any one of the elements of equitable estoppel is absent, the claim must fail.” Hair, 2 Cal.Rptr.2d at 875.
The RTC presented some evidence allegedly illustrating CNA’s intent to “disguise the fact that the policies were being changed rather than renewed on the same terms, and that the successor policies would contain restrictive endorsements which would bar coverage for claims that would have been covered under the older policies.” The district court found this evidence untenable, a conclusion supported by Farmers, in which the Eighth Circuit rejected contentions that CNA had acted improperly: “It is true that American Casualty increased its premium and decreased its coverage for officers and directors when it discovered the state of the Bank’s loan portfolio. That strikes us, however, as a reasonable business decision rather than bad faith.” Farmers, 944 F.2d at 459.
Even if the evidence did indicate an attempt by CNA to “hoodwink” Pacific into foregoing discovery coverage, estoppel would not be appropriate in this case because the Directors/Officers could not possibly have relied to their detriment on CNA’s conduct. The RTC’s arguments about reliance boil down to assertions that Pacific might have purchased discovery coverage and reported occurrences in the twelve-month discovery period. As explained above, however, notice of occurrences does not suffice in the discovery period. Because the RTC did not file claims against the Directors/Offieers until expiration of the discovery period, the policy would not in any event have covered those claims.
Thus, the district court correctly concluded .that no coverage was available under the 1983-85 Policy because, even if the RTC proved at trial that CNA had not renewed the policy and had not provided notice of such nonrenewal, the only remedy would be to allow the RTC to purchase discovery coverage for the year following the alleged non-renewal. Because the Discovery Clause permitted coverage only for actual claims filed during the discovery period (and not for notices of occurrences that might ripen into claims), no remedy was available for CNA’s alleged policy breach.
C. The 1985-86 Policy
The district court held that the RTC could not assert a claim for nonrenewal of the 1985-86 Policy because Pacific had accepted a successor policy (the 1986-87 Policy) with full notice of its terms: “[T]he 1986-87 policy could not have constituted a non-renewal of the previous policy. Defendants admit they had clear and conspicuous notice of all the terms of the new policy and they accepted the policy on those terms. As a matter of law, then, the insureds in this case cannot accept a policy with clear notice of its terms, yet years later assert a right to discovery coverage under the [prior] policy.” The RTC asserts that, under California law, acceptance of a successor policy does not transform a nonrenewal into a renewal or cure CNA’s failure to give notice of nonrenewal. Thus, the RTC maintains that CNA breached the Notice Clause and that it should now be able to purchase discovery coverage.10 We disagree.
*893The circuit courts that have considered the issue have held that the knowing acceptance of a successor policy with different terms constitutes a renewal of the prior policy. Adams v. Greenwood, 10 F.3d 568 (8th Cir.1993), a recent Eighth Circuit decision, is directly on point:
[The RTC’s] argument has been foreclosed by the district court’s determination ... that [the Bank] was adequately notified of the existence of the [Regulatory Exclusion] before entering into the [successor policy] contract.... [A]n insured who agrees to accept an altered ‘renewal’ policy forgoes his claim for extended discovery rights under the old policy. In this ease, [the Bank]’s acceptance of a series of policies after notice of the regulatory exclusion in 1982 and afterwards, is inconsistent with its assertion of constructive non-renewal. The RTC, as successor to [the Bank], has no right to any extended discovery period under the [earlier] policy.
Id. at 572 (applying Minnesota law). See also American Casualty Co. v. Continisio, 17 F.3d 62 (3d Cir.1994) (“Even in a state where ‘renewal’ is strictly defined as ‘continuation of coverage on the same, or nearly the same, terms as the policy being renewed,’ acceptance of new terms constitutes a renewal.” (applying New Jersey law); Farmers, 944 F.2d at 460 (“[A]greement to new terms creates a renewal. [The bank]’s agreement to the [successor] policy ... prevents the Board from securing after-the-fact coverage under the [earlier] policy”) (applying Iowa law).11 Cf. Mundy v. Lumberman’s Mutual Casualty Co., 783 F.2d 21, 22-23 (1st Cir.1986) (same result under Massachusetts law in context other than D & 0 insurance); GEICO v. United States, 400 F.2d 172, 174-75 (10th Cir.1968) (same result under federal law in context other than D & O insurance).
The district courts have taken a similar position. “Where ... it can be shown that the parties agreed to a renewal on changed terms, it is not for this Court, in hindsight, to reorganize their affairs.” Baltimore Federal, 845 F.Supp. at 324 (applying Maryland law). See Blanchard, 821 F.Supp. at 662-663 (applying Oklahoma law); American Casualty Co. v. FDIC, No. 91-CV-692, slip op. at 8 (W.D.Mich. Jan. 8,1993) (White Cloud) (“the critical question is whether the changes in the [successor policy] were called to [the bank]’s attention at the time of renewal”) (applying Michigan law).12
Despite all this authority to the contrary, the RTC claims that California law mandates a different result. The RTC correctly notes that in California, as in most other jurisdictions, the renewal of an insurance policy means the continuation of coverage on substantially the same terms. See, e.g., Kemmerer Eng’g Co. v. Continental Casualty Co., 253 Cal.App.2d 188, 61 Cal.Rptr. 94, 96 (1967). That proposition, however, is only half the story. A renewal policy “perpetuates the terms and conditions of the original policy unless the parties provide otherwise.” Sandgren v. Fire Ins. Exch., 64 Cal.App.3d 634, 134 Cal.Rptr. 590, 591 (1976) (emphasis added). “An insurer when renewing a policy may not change the terms of the policy, without first notifying the insured.” Fields v. Blue Shield, 163 Cal.App.3d 570, 209 Cal.Rptr. 781, 785 (1985) (emphasis added) (internal quotation omitted). See also Cal.Ins.Code § 678.1(f)(6) (notice of nonrenewal not required where “[t]he insurer has made a written offer to the insured ... to *894renew the policy under changed terms or conditions or at a changed premium rate”).13
The RTC does not dispute the district court’s finding that Pacific accepted the 1986-87 Policy with notice of its new terms.14 Therefore, under California law, the 1986-87 Policy constituted a renewal of the 1986-86 Policy. As a result, the district court correctly held that the RTC cannot now elect discovery coverage under the earlier policy.
D. The 1986-87 and 1987-88 Policies
The district court held that the Regulatory Exclusion in the 1986-87 and 1987-88 Policies barred coverage of the RTC claims. On appeal, the RTC contends only that the exclusion is inapplicable because the RTC is not a “regulatory agency.” We reject this strained argument.
The Regulatory Exclusion provides that CNA will not be liable for claims against the Directors/Officers based upon or attributable to:
[A]ny action or proceeding brought by or on behalf of the [FDIC], the Federal Savings & Loan Insurance Corporation, any other depository insurance organization, the Comptroller of the Currency, the Federal Home Loan Bank Board, or any other national or state regulatory agency ... including any type of legal action which such agencies have the legal right to bring as receiver, conservator, liquidator, or otherwise.
Because every circuit that has considered the exclusion has found it valid and not violative of public policy, see FDIC v. American Casualty Co., 998 F.2d 404, 407-10 (7th Cir.1993) (Cuba); FDIC v. American Casualty Co., 996 F.2d 471, 472-74 (4th Cir.1993) (Fidelity); FDIC v. American Casualty Co., 975 F.2d 677, 680-83 (10th Cir.1992) (Security); FDIC v. Conner, 973 F.2d 1236, 1239-44 (5th Cir.1992); St. Paul Fire & Marine Ins., 968 F.2d at 702-03, the RTC now claims only that it is not a “regulatory agency.” The RTC’s reasoning is as follows: Because the exclusion does not list the RTC specifically, and because the RTC is not a “depository institution,” the exclusion only applies if the RTC fits within the general category “other national regulatory agency.” Because that general term follows a list of specific agencies, the maxim of ejusdem generis (“the general terms will be construed as applicable only to persons or things of the same general nature as the enumerated terms”) applies. The RTC is unlike the FDIC, FSLIC, or FHLBB and, accordingly, the exclusion is inapplicable.
As the district court realized, the RTC’s reasoning is fatally flawed:
The RTC replaced FSLIC as conservator or receiver for failed thrifts.... [W]hen the underlying action was instituted in 1989, it was done by the FDIC acting as the Managing Agent for the FSLIC.
The FDIC, also specifically named in the exclusion clause, was the FSLIC’s manager then, and it is now the exclusive manager of the RTC; it provides the employees to carry out the RTC’s functions; and, it will replace the RTC as receiver and conservator when the RTC is terminated ... It is illogical ... to conclude that the policy exclusion does not extend to the new agency [the RTC], created as part of a reorganization of federal regulation of the industry, that has assumed junctions previously performed by one agency named in the exclusion, FSLIC, and is essentially *895controlled by another agency named in the exclusion, FDIC. Lastly, ... the policy exclusion also expressly states that no coverage is available for “any type of legal action which such Agencies have the legal right to bring as receiver, conservator, liquidator, or otherwise.”
Baker I, 758 F.Supp. at 1346 (emphasis added) (internal citation omitted).
The RTC has the same powers of conser-vatorship and receivership as the FDIC when it acts in that capacity. 12 U.S.C. § 1441a(b)(4). Indeed, the RTC, “when it is acting as a conservator or receiver of an insured depository institution, shall be deemed to be an agency of the United States to the same extent as the [FDIC] when it is acting as a conservator or receiver of an insured depository institution.” Id. § 1441a(b)(l)(B). Congress created the RTC after CNA drafted the Regulatory Exclusion clauses at issue. It is unreasonable to protect the RTC from operation of the exclusion merely because CNA was unable to predict that Congress would restructure its regulatory framework of depository institutions and create a new federal agency.
Every other court to consider the issue has concluded that the RTC falls within the scope of the Regulatory Exclusion. See RTC v. Shall, No. 92-2654, 1993 WL 534022 (E.D.La. Dec. 14, 1993) (“RTC falls within the regulatory exclusion”); Baltimore Federal, 845 F.Supp. at 327 (“Created after the exclusion was drafted, the RTC was not mentioned in the provision. However, the fundamental reasons for the RTC’s creation reveal that it is essentially the alter ego, reincarnation, or affiliate of the FDIC or FSLIC, organizations specifically addressed by the exclusion.”); American Casualty Co. v. RTC, No. 92-10543-WF, slip op. at 27 (D.Mass. Oct. 19, 1993) (Comfed) (“this court thinks it clear that the RTC is a ‘regulatory agency’ within the meaning of the Policies”); Benafield, slip op. at 3; American Casualty Co. v. RTC, 839 F.Supp. 282, 292 (D.N.J.1993) (First Atlantic) (“RTC is a regulatory agency within the meaning of the regulatory exclusion”); Adams v. RTC, No. 4-89-CV-330, 1998 WL 181308 (D.Minn. May 19, 1993) (“The Court will not blind its eye, the RTC is certainly a regulatory agency within the contemplation of the contract”), aff'd 10 F.3d 568 (8th Cir.1993); Chandler v. American Casualty Co., 833 F.Supp. 735, 738 (E.D.Ark.1993) (“Although the RTC is not mentioned in the exclusion, it did not exist at the time the parties entered into the contract. It is inconceivable that the RTC would not have been mentioned in the exclusion had the parties known of its existence.”); RTC v. Walke, No. 92-0430, 1993 WL 455195 (W.D.La. Apr. 15, 1993); National Union Fire Ins. Co. v. RTC, No. H-92-1157, slip op. at 4-5, 1992 WL 611463 (S.D.Tex. Aug. 18, 1992) (“The phrase ‘or any other national or state Bank regulatory agency was added to include agencies similar to those named. This would include new agencies such as the RTC.”). aff'd in relevant part, No. 92-2835, slip op. at 5 (5th Cir. Feb. 21, 1994), (“the proposition that the RTC is not a regulatory agency is squarely contradicted by the agency’s express duties under FIRREA”). We follow this overwhelming authority and affirm the district court.
E. Conclusion
We affirm the district court as to each of the five relevant D & O policies. CNA is not responsible for the 1980-83 Policy. Pacific has no remedy for any breach of the 1983-85 Policy. Pacific nullified any rights it might have had under the 1985-86 Policy by accepting the 1986-87 Policy with notice of its terms. And, the Regulatory Exclusion bars coverage for RTC claims under the 1986-87 and 1987-88 Policies.
III.
As noted above, CNA appealed the district court’s order requiring it to advance defense costs to the Direetors/Offieers pending a final determination of coverage under the D & O policies. Because the Coverage Order was just such a final determination (and because we affirm that order), we dismiss as moot CNA’s appeal and vacate the Advancement Order.15
*896“A case becomes moot when interim relief or events have deprived the court of the ability to redress the party’s injuries.” United States v. Alder Creek Water Co., 823 F.2d 343, 345 (9th Cir.1987). “The basic question is whether there exists a present controversy as to which effective relief can be granted.” Village of Gambell v. Babbitt, 999 F.2d 403, 406 (9th Cir.1993) (internal quotation omitted). CNA seeks reversal of the Advancement Order, in which the district court held that it had “a duty to advance defense costs for [the] underlying action ... so long as the underlying claims remain potentially covered under the policies.” After issuing the Coverage Order, the district court clarified that CNA’s “duty to advance defense costs ceased when the Court” entered that order. As a result, the insurer is not now under any duty to advance funds.
We can give no further relief to CNA. “Were we to reverse the district court and hold in [CNA]’s favor, it would receive no relief beyond that already provided by [the Coverage Order]. Where events have occurred that prevent us from granting effective relief, we lack jurisdiction and must dismiss the appeal.” Enrico’s, Inc. v. Rice, 730 F.2d 1250, 1254 (9th Cir.1984) (internal citation omitted).
Although it now takes the opposite position, CNA recognized in its opening brief (when it was seeking to establish appellate jurisdiction) that final resolution of the coverage issues would in fact moot an appeal of the Advancement Order. CNA now asserts that the appeal is not moot because our decision on the merits will “have an impact on the parties.” Flagstaff Medical Ctr. v. Sullivan, 962 F.2d 879, 884 (9th Cir.1992). To support this assertion, the insurer claims that the Advancement Order might affect its efforts to recover defense costs it advanced to the Directors/Officers. This contention is specious. The D & 0 policies provide that, “in the event it is finally established the Insurer has no liability [under the policies], such Directors and Officers agree to repay to the insurer, upon demand, all monies advanced” for defense costs prior to resolution of coverage issues.
CNA therefore appears to have a contractual right to recover any fees it advanced to the Directors/Officers. See Okada, 823 F.2d at 282 (insurer can “reserve its rights under the agreement for the return of advances should the claims ultimately prove to be uncovered.”); Continisio, 819 F.Supp. at 404 (“disposition of the summary judgment motions [regarding insurance coverage] moots this claim” for advancement of defense costs).16 Reversal or affirmance of the Advancement Order will not in any way impact CNA’s ability to do so. See Allard v. DeLorean, 884 F.2d 464, 467 n. 1 (CNA’s “apprehension regarding [impact on repayment of advanced costs] is too remote a consequence to forestall the mootness of this appeal.”). As a result, we dismiss as moot the Advancement Order appeal.
We do, however, grant CNA’s request to vacate that order. “[W]hen an appeal is dismissed as moot, the established practice is for the appellate court to reverse or vacate the judgment below and dismiss the case.” Id. at 467. See United States v. Munsingwear, Inc., 340 U.S. 36, 39-40, 71 S.Ct. 104, 106-107, 95 L.Ed. 36 (1950). We see no reason not to do so in this case.
IV.
We affirm the district court’s Coverage Order. CNA is not responsible for the 1980-83 Policy. Pacific has no remedy for any breach of the 1983-85 Policy. Pacific nullified any rights it might have had under the 1985-86 Policy by accepting the 1986-87 Policy with notice of its terms. And, the Regulatory Exclusion bars coverage for the RTC claims under the 1986-87 and 1987-88 Policies. We dismiss as moot CNA’s appeal of the Advancement Order and vacate the district court’s decision on that issue.
*897AFFIRMED in part. APPEAL DISMISSED in part. District court order VACATED in part.
. The case remains pending before the district court. See RTC v. Baker, No. 89-386-AHS (C.D.Cal.); see also FDIC v. Baker, 739 F.Supp. 1401 (C.D.Cal.1990) (interlocutory order striking affirmative defenses of the Directors/Officers).
. Because the Directors/Officers have joined the RTC's briefs on the coverage issues, we refer to both as the "RTC.”
. The RTC also appealed an order excluding proposed expert testimony, with which it hoped to explain the nature of D & 0 policies to the jury. Because we affirm the Coverage Order, which the district court issued before trial, we need not consider the expert testimony ruling.
In any event, we think the district court was within its discretion by precluding the expert testimony as a sanction for the RTC's failure to seasonably supplement its interrogatory responses pursuant to Federal Rule of Civil Procedure 26(e)(1). See, e.g., Hancock v. Hobbs, 967 F.2d 462, 468 (11th Cir.1992); Thibeault v. Square D. Co., 960 F.2d 239, 244-48 (1st Cir.1992). See also Scott & Fetter Co. v. Dile, 643 F.2d 670, 673-75 (9th Cir.1981) (abuse of discretion for district court not to exclude testimony omitted from interrogatory responses).
. We commend the district court for its thorough consideration of the barrage of arguments leveled by the parties. Although CNA and the RTC/ FDIC are well acquainted with each other (by our count they have litigated at least twenty different cases involving similar policies), neither presented a lucid case below or on appeal, preferring instead to make every conceivable (and occasionally inconceivable) argument. Unfortunately, these "shotgun" argument tactics are not unique to this case:
Because the RTC and American Casualty have litigated similar — but not identical — issues in many other cases, the matter was grossly overbriefed. ... [S]ome of the arguments presented appear to have been made because they are found on some checklist of all possible arguments, rather than through informed professional judgment regarding this particular case. ... [I]t would be best to pay more attention to the evidence and precedents as they exist rather than as [the RTC and CNA] wish they were.
American Casualty Co. v. RTC, 845 F.Supp. 318, 319 (D.Md.1993) (Baltimore Federal) (footnote omitted). After wading through twelve briefs in this case, we couldn't agree more.
. The parties agreed to apply California law to the agreements.
. Like everything else on appeal, the RTC and CNA disagree about how to refer to cases with identical lead parties. For this opinion, we adopt CNA’s style of referring to such cases by the name of the financial institution on whose behalf the RTC or FDIC was asserting claims.
. On appeal, the Eighth Circuit overruled Christensen in American Casualty Co. v. FDIC, 944 F.2d 455 (8th Cir.1991) (Farmers). However, because the Farmers court concluded that American Casualty had in fact renewed the policy at issue, it never considered the propriety of the district court’s proposed remedy.
. The RTC cites Casper to support its arguments. In that case, the Tenth Circuit did hold that, "if notice of nonrenewal is required, the policy runs either to the end of the period defined in the agreement or until the notice of nonrenewal becomes effective, whichever is later.” Casper, 999 F.2d at 482. Casper is inapposite, however, because it did not involve the right to buy discovery coverage retroactively. Rather, it merely held that, where the insurer sent notice of nonrenewal that set a specific date for termination of a policy, that date controlled over the earlier date of termination listed in the policy itself.
.Also unhelpful are the cases cited by the RTC that purportedly order specific performance of option contracts by extending the period in which the option is to run. In Bertero v. National General Corp., 254 Cal.App.2d 126, 62 Cal.Rptr. 714 (1967), the court gave the defendant the choice of paying money damages or allowing the plaintiff to exercise stock options despite expiration of the option time period. This remedy was merely an expeditious way to give the plaintiff “that which he had bargained for” without penalizing the defendant in case performance of the option would work undue hardship. Id. at 727. Similarly, the court in Warner Bros. Pictures v. Brodel, 31 Cal.2d 766, 192 P.2d 949, cert. denied, 335 U.S. 844, 69 S.Ct. 67, 93 L.Ed. 394 (1948), expressly declined to consider “the type of relief to which plaintiff may be entitled." Id. at 778, 192 P.2d 949.
. The RTC also argues that California law imposed on CNA an obligation to inform Pacific that the offer of a policy with different terms might constitute nonrenewal and therefore permit Pacific to elect discovery coverage. This contention is erroneous. California law does require an insurer to disclose facts upon which a claim may be based. See, e.g., Davis v. Blue Cross, 25 Cal.3d 418, 158 Cal.Rptr. 828, 834, 600 P.2d 1060, 1066 (1979). California law does not, however, require disclosure of a legal theory upon which coverage may exist. See, e.g., Love v. Fire Ins. Exch., 221 Cal.App.3d 1136, 271 Cal.Rptr. 246, 250-54 & nn. 5-8 (1990). Because *893the RTC’s “constructive nonrenewal” theory is a legal argument, CNA had no duty to notify Pacific that it might be able to reject the proposed policy and purchase discovery coverage.
. The Eighth Circuit's McCuen decision is not contrary to Adams and Farmers. In McCuen, the court held only that the insured had the right to purchase discovery coverage after it had rejected the insurer’s offer to renew on different terms. McCuen, 946 F.2d at 1404 (applying Iowa law). Read together, Farmers and McCuen stand "for the proposition that, under Iowa law, if an insurer offers to renew on substantially different terms, and the insured rejects those terms, the insurer is refusing to renew the policy. If on the other hand the insured is aware of the policy changes, and agrees to them, the policy has been renewed.” American Casualty Co. v. Continisio, 819 F.Supp. 385, 400 (D.N.J.1993), aff'd, 17 F.3d 62 (3d Cir.1994).
. The RTC cites Allen as an example of a decision contrary to these cases. In Allen, however, the court found that the insureds had not agreed to the new terms in the successor policy. Allen, 710 F.Supp. at 1094. The case is therefore inap-posite.
. The California legislature did not enact this provision until after the parties executed the policies at issue in this case. Nevertheless, the statute illustrates that no public policy is served by requiring insurers to send formal notices of "nonrenewal” along with offers to "renew” a policy on different terms.
. Two individual directors argue that a material issue of fact exists on this issue because CNA did not provide them with notice of the new terms directly. This argument has no merit. The D & 0 policies provided that, "[b]y acceptance of this Policy, the Directors and Officers ... agree that [Pacific] ... shall act on behalf of all Directors and Officers with respect to the giving and receiving of notice of claim, cancellation or any other notice required or permitted.” "[I]n the absence of a contrary contractual provision, the duty to give notice [to the individual directors] does not rest with the insurer.” Martz v. Union Labor Life Ins. Co., 757 F.2d 135, 141 (7th Cir.1985). See, e.g., St. Paul Fire & Marine Ins. Co. v. FDIC, 968 F.2d 695, 700-01 (8th Cir.1992) (notice sufficient where given to single designated officer); Farmers, 944 F.2d at 457-59 (same).
. The Directors/Officers initially contended that we had no jurisdiction over this interlocutory *896appeal. Now that the district court has issued a final judgment, however, nominal appellate jurisdiction clearly exists. See, e.g., Unioil, Inc. v. E.F. Hutton & Co., 809 F.2d 548, 554 (9th Cir.1986) ("we may treat an interlocutory order as a final order when that portion of the case that remained in the district court has subsequently been terminated"), cert. denied, 484 U.S. 822, 108 S.Ct. 83, 98 L.Ed.2d 45 (1987).
. We, of course, do not decide this issue.