Opinion
MORENO, J.
This appeal requires us to determine what an insured must prove in order to establish its rights under a lost or destroyed insurance policy. The trial court found that the insured had introduced sufficient secondary evidence to prove the substance of the policy’s material provisions, and rendered judgment for the insured. The Court of Appeal held that the insured was required instead to prove the actual words of those provisions, and reversed the judgment. As will appear, we conclude that the holding was erroneous, and therefore reverse the Court of Appeal’s judgment, with some issues outside the scope of review to be determined on remand.
Facts
From the 1940’s until the late 1960’s, Rexall Drug Company (Rexall), predecessor in interest of plaintiff Dart Industries, Inc. (Dart), was one of a number of pharmaceutical companies that manufactured and marketed the prescription drug diethylstilbestrol (DES), a synthetic estrogen widely used at the time to prevent miscarriages.
Throughout this period, Rexall carried comprehensive general liability (CGL) insurance under policies issued in sequence by its three primary [1065]*1065carriers: Employers Liability Assurance Corporation, Ltd. (Employers), predecessor in interest of defendant Commercial Union Insurance Corporation (Commercial Union), for the years 1946-1951; Liberty Mutual Insurance Company (Liberty Mutual), for the years 1951-1966; and Continental Insurance Company (Continental), for the years 1967-1981.
From the mid-1970’s onward, a large number of claimants alleging personal injuries caused by exposure to DES began filing actions for damages against DES manufacturers nationwide. The majority of claimants were adult women whose mothers had ingested DBS while the claimants were in útero and who, when they reached childbearing age themselves, manifested precancerous and cancerous vaginal and cervical lesions, as well as various deformities of their reproductive organs resulting in infertility or miscarriages. (See generally Brown v. Superior Court (1988) 44 Cal.3d 1049 [245 Cal.Rptr. 412, 751 P.2d 470]; Sindell v. Abbott Laboratories (1980) 26 Cal.3d 588 [163 Cal.Rptr. 132, 607 P.2d 924, 2 A.L.R.4th 1061].) Many of the complaints were filed against Dart, both in California and in other jurisdictions, alleging exposure to DES in each of the foregoing time periods. Dart tendered the defense of these actions to Commercial Union, Liberty Mutual, and Continental. All three carriers denied coverage and refused to provide a defense.
In October 1984, Dart filed the present action for declaratory relief and damages against Commercial Union, Liberty Mutual, and Continental. The primary relief sought was a declaratory judgment (Code Civ. Proc., § 1060) establishing that the policies defendants sold to Dart and its predecessor obligate them to provide a defense to the DES actions and to indemnify Dart for any liability assessed against it in those actions.
In October 1986, Dart settled with Liberty Mutual and Continental. Commercial Union declined to participate in the settlement. The remaining issue in the case, accordingly, was whether Dart was entitled to a defense and indemnity from Commercial Union under the policy in effect for the period of September 1, 1946, to September 1, 1951, i.e., Employers Policy No. CL92302.1 The question was complicated by the undisputed fact that the policy is lost and neither party has been able to find it or a copy of it. This fact, as we shall see, has led to the principal issue of law we address in this case.
This case has a long and convoluted procedural history, and has been the subject of no fewer than four appeals and four decisions of the Court of Appeal. We briefly review that history.
[1066]*1066After various pretrial proceedings, Dart moved for preference in trial setting. The motion was denied, and shortly thereafter the court dismissed the action for failure to bring it to trial within five years. (Code Civ. Proc., §§ 583.310, 583.360.) Dart appealed. In the first appeal, the Court of Appeal held the trial court abused its discretion in denying the motion for preference, and reversed and remanded for trial. (Dart Industries, Inc. v. Commercial Union Ins. Co. (Feb. 28, 1992, B047651) [nonpub. opn.] (Dart Industries I).) The court also concluded that the trial court had properly denied Dart’s motion for summary judgment, and held that at trial Dart must prove the contents of the insurance policy by a preponderance of evidence, rejecting Commercial Union’s contention that proof must be by clear and convincing evidence.
The case was initially tried in 1993 before Judge Newell Barrett, sitting without a jury. Dart undertook to prove the existence and the substance of the material terms of the lost policy by means of the testimony of Charles Pyne, an employee of Obrion, Russell & Co. (Obrion), which served as the insurance agent of Employers, Commercial Union’s predecessor and a broker for the Dart account during the relevant period, as well as by various items of documentary evidence. Judge Barrett entered a minute order announcing a tentative decision in favor of Commercial Union on the ground that Dart had not proved the lost policy’s limits of liability. Dart requested a statement of decision (Code Civ. Proc., § 632), but Judge Barrett died before issuing such a statement. A second judge thereupon signed a judgment for Commercial Union based on Judge Barrett’s minute order. Dart appealed and the Court of Appeal reversed the judgment on the ground that a second judge is not authorized to sign a judgment based only on the first judge’s tentative decision. (Dart Industries, Inc. v. Commercial Union Ins. Co. (May 26, 1995, B083165) [nonpub. opn.].)
The case was assigned for retrial to Judge Loren Miller, Jr. By stipulation, the parties submitted the matter to him for decision on the basis of the record of the first trial, including the reporter’s and clerk’s transcripts and the exhibits, a videotape of the testimony given at the first trial, and new briefs. After reviewing this record, Judge Miller entered a minute order announcing a tentative decision in favor of Dart. He then issued a statement of decision at Commercial Union’s request, and entered judgment accordingly. Commercial Union appealed.
The Court of Appeal held that Judge Miller’s statement of decision was insufficient as a matter of law because it failed to adequately explain the factual or legal basis for his decision on a number of issues identified by Commercial Union and deemed by the Court of Appeal to constitute “principal controverted issues at trial” (Code Civ. Proc., § 632). On this ground, [1067]*1067the Court of Appeal reversed the judgment with directions to Judge Miller to issue a new statement of decision addressing certain additional issues and to enter a new judgment based thereon. (Dart Industries, Inc. v. Commercial Union Ins. Co. (June 27, 1997, B105886) [nonpub. opn.].)
As directed by the Court of Appeal and after further hearing, Judge Miller issued a substantially more detailed 29-page statement of decision and judgment. In the statement of decision, the court reviewed at length the evidence to support its findings that (1) Dart conducted a diligent and exhaustive search for the lost policy, but was unable to find it; (2) Dart proved the existence of the policy by secondary evidence; (3) the policy was a “manuscript policy, meaning that it was custom tailored to the insured,” rather than a standard form; (4) the policy covered Dart’s predecessor Rexall for the single five-year term from September 1, 1946, to September 1, 1951; (5) the policy provided coverage for product liability claims for bodily injuries occurring during the policy period, including injuries caused by DES exposure during that period but not manifesting themselves until after the period ended; (6) the policy provided occurrence-based coverage that would cover injuries initiated by the ingestion of the drug during the policy period but manifesting after that period; (7) the policy’s limits of liability were $100,000 per person and $300,000 per occurrence or accident, with a $300,000 annual product liability aggregate limit; (8) the policy did not contain a deductible; and (9) the policy covered defense costs separately from the limits of liability.
The court then reviewed the case law to support its conclusion that the “trigger of coverage” under the policy was exposure to DES rather than manifestation of the ensuing injury many years later. Because Dart’s several primary insurance carriers (Commercial Union, Liberty Mutual, and Continental) proposed different triggers of coverage, the court recognized the potential for either inadequate recovery or double recovery of benefits; to avoid that result, the court found equitable—and adopted—the method of allocating the insurers’ obligations devised by the parties, i.e., that Commercial Union should pay only 50 percent of the costs of defense and indemnity in actions against Dart arising from DES exposure during the 1946-1951 policy period.
In the ensuing judgment, entered on December 21, 1998, the court (1) declared the parties’ rights and liabilities under the policy in accord with the foregoing statement of decision; (2) awarded Dart damages of some $1.9 million for local defense costs, $550,000 for the allocated portion of the insured’s national coordinating defense counsel costs, and $260,000 for indemnity, plus prejudgment interest through 1992 of some $1.4 million; and [1068]*1068(3) declared that Dart is entitled to recover from Commercial Union 50 percent of the defense costs and settlements incurred by Dart after 1992 in actions for damages arising from DES exposure during the 1945-1951 policy period. Commercial Union appealed again.
The Court of Appeal again reversed the judgment, this time on the ground that the evidence was insufficient to support the findings of the trier of fact on the material terms of the lost policy, principally because Dart failed to produce the actual language of the policy. From Dart’s earlier stipulation to retry the case on the record of the first trial, the Court of Appeal inferred that Dart has conceded it has no additional evidence to offer, and on that ground remanded the cause to the trial court with directions to enter judgment in favor of Commercial Union.
Dart petitioned for review. We granted its petition to determine what an insured must prove in order to establish its rights under a lost insurance policy, and whether the Court of Appeal correctly held that the evidence in the case at bar was insufficient to support the trial court’s findings regarding the contents of the lost policy.
I
We begin with the statutory law. Evidence Code section 1521, subdivision (a), provides that “[t]he content of a writing may be proved by otherwise admissible secondary evidence,” excepting when “[a] genuine dispute exists concerning material terms of the writing and justice requires the exclusion” or when “[ajdmission of the secondary evidence would be unfair.” The admission of oral testimony regarding the contents of a writing is specifically governed by section 1523, which provides, in pertinent part, that such testimony is admissible “if the proponent does not have possession or control of the original or a copy of the writing and . . . [f ] . . . [n] either the writing nor a copy of the writing was reasonably procurable by the proponent by use of the court’s process or by other available means. . . .” (Cf. Fed. Rules Evid., rule 1004(1), 28 U.S.C.)
These statutes are codifications of the venerable common law rule that lost documents may be proved by secondary evidence. In Folsom’s Executors v. Scott (1856) 6 Cal. 460, 461, the court stated: “The rule ... for the admission of secondary evidence of a lost paper, requires ‘that a bona fide and diligent search has been unsuccessfully made for it in the place where it was most likely to be found;’ and further, ‘the party is expected to show that he has in good faith exhausted in a reasonable degree all the sources of information and means of discovery which the nature of the case would naturally suggest, and which were accessible to him.[’]”
[1069]*1069As was elaborated in Kenniff v. Caulfield (1903) 140 Cal. 34 [73 P. 803] (Kenniff): “ ‘If any suspicion hangs over the instrument, or that it is designedly withheld, a rigid inquiry should be made into the reasons for its non-production. But where there is no such suspicion, all that ought to be required is reasonable diligence to obtain the original—in fact, courts in such cases are extremely liberal.’ ” (Id. at p. 42.) Questions whether the search was sufficient in scope and was conducted in good faith are addressed to the discretion of the trial court, and will not be disturbed on appeal absent abuse of discretion. (Id. at pp. 41-42; see also Guardianship of Levy (1955) 137 Cal.App.2d 237, 249-250 [290 P.2d 320].)
A corollary of the rule that the contents of lost documents may be proved by secondary evidence is that the law does not require the contents of such documents be proved verbatim. In Posten v. Rassette (1855) 5 Cal. 467, for example, the owner of real property gave a written power of attorney to his agent, Parker. The document was accidentally destroyed by fire, but the owner took no steps to revoke the power. A year later, acting on the same power, Parker sold the property to a third person. This court affirmed a judgment recognizing the validity of the sale. At trial, the notary who witnessed the deed of sale testified he had seen the power of attorney in question, that it had authorized Parker to sell the property, and that the document had been destroyed by fire. This court held that the testimony was sufficient to establish the existence, loss, and contents of the power of attorney. “In the case of lost instruments where no copy has been preserved, it is not to be expected that witnesses can recite its contents, word for word.” (Id. at p. 469.)
This rule was reaffirmed in Kenniff, supra, 140 Cal. 34. There the only evidence of the contents of a lost and unrecorded deed to a parcel of real property was the testimony of the conveyancer who drafted it: he identified the blank form he had used, and said he had copied into it a description of the property taken from a prior deed of the property to the grantor and he was sure the deed was to the lot in issue. Holding this testimony to be sufficient proof of the contents of the lost deed, we stated: “It is not necessary, in order to admit evidence of the contents of a lost instrument, that the witnesses should be able to testify with verbal accuracy to its contents; it is sufficient if they are able to state it in substance.” (Id. at p. 43.) Subsequent cases are in accord. For example, in Seaboard National Bank v. Ackerman (1911) 16 Cal.App. 55, when all record of a judgment in a lawsuit had been destroyed by fire, the court reiterated: “‘In the case of a lost instrument where no copy has been preserved, it is not to be expected that witnesses can state the contents, word for word.’ [Citation.] The substance of a lost or destroyed document is all that is required.” [1070]*1070(Id. at p. 58; accord, Von Hasseln v. Von Hasseln (1953) 122 Cal.App.2d 7, 13 [264 P.2d 205] ,)2
The lost document cases illustrate a few of the many types of secondary evidence that courts have admitted to prove the contents of a missing instrument. For example, courts have often admitted oral testimony for this purpose. (Kenniff, supra, 140 Cal. at p. 43; Posten v. Rassette, supra, 5 Cal. at pp. 469-470; Robinson v. Thornton (1969) 271 Cal.App.2d 605, 611-612 [76 Cal.Rptr. 835] [lost contract to repurchase gas lease; testimony of contracting party]; Guardianship of Levy, supra, 137 Cal.App.2d at p. 249 [testimony of recipient of lost letters]; Deacon v. Bryans (1928) 88 Cal.App.322, 325 [testimony of witness who saw lost promissory notes]; Seaboard National Bank v. Ackerman, supra, 16 Cal.App. at p. 57 [testimony establishing record of judgment destroyed by fire]; Hedstrom v. Union Trust Co. (1908) 7 Cal.App. 278, 286 [94 P. 386] [testimony of drafter of lost lease].) Courts have also admitted a standard form of the lost document (Kenniff, supra, 140 Cal. at p. 43 [blank form used in drafting lost deed]), as well as evidence of a routine practice of a party (Amoco Production Co. v. United States (10th Cir. 1980) 619 F.2d 1383, 1389-1390 [lost deed; grantor’s routine practice of reserving a mineral interest in all property deeded]).
The use of secondary evidence was affirmed in the only two published California decisions involving lost insurance policies. In Clendenin v. Benson (1931) 117 Cal.App. 674, a driver sued for damages to his automobile caused by a collision with the defendant’s truck, and his insurer joined a subrogation cause of action. The court rendered judgment for the insurer even though the latter could not produce the policy at trial. Affirming the judgment, the Court of Appeal held that the evidence “shows that a diligent and bona fide but unsuccessful search for the document was made at the place where it was most likely to be found; further, the contents of the policy were sufficiently shown by the testimony of employees of the insurer and by its records.” (Id. at p. 678.)
In Rogers v. Prudential Ins. Co. (1990) 218 Cal.App.3d 1132 [267 Cal.Rptr. 499], an employee who became totally disabled while covered by a group major medical policy issued by the defendant insurer sued after his benefits were terminated. Pursuant to company practice, the insurer had destroyed its copies of the policy several years after terminating the plan; the plaintiffs copy had been destroyed by fire, and he could produce only a brief announcement of the plan summarizing its provisions and referring to [1071]*1071the master policy for details of coverage. The insurer moved for summary judgment on the ground that it was relieved of liability by the terms of the missing policy, asking the court to determine those terms by construing the announcement alone. The court granted the motion. Reversing the ensuing judgment, the Court of Appeal explained that the trial court was not limited to construing the announcement. “Contents of a lost or destroyed memorandum ‘may be shown by an unsigned copy or by oral evidence.’ [Citation.] Such evidence may include testimony of long-time . . . employees [of the insurer] who were familiar with the policy’s standard provisions, or copies of other policies sold at the same time which utilized similar provisions.” (Id. at p. 1137.)
Having established that the contents of a lost insurance policy may be proved by secondary evidence, the question remains—what precisely must be proved. In an action on an insurance policy that has not been lost or destroyed, it is well settled that “[t]he burden is on an insured to establish that the occurrence forming the basis of its claim is within the basic scope of insurance coverage. [Citations.] And, once an insured has made this showing, the burden is on the insurer to prove the claim is specifically excluded.” (Aydin Corp. v. First State Ins. Co. (1998) 18 Cal.4th 1183, 1188 [77 Cal.Rptr.2d 537, 959 P.2d 1213].) The rule applies irrespective of whether the dispositive policy language is found in clauses described as exclusions, exceptions, limitations, conditions, or endorsements: “[I]t is the function served by policy language, not the location of language in an insurance policy, that is determinative.” (Id. at p. 1191, italics added.) This rule follows from the general evidentiary rule found in Evidence Code section 500 that “[e]xcept as otherwise provided by law, a party has the burden of proof as to each fact the existence or nonexistence of which is essential to the claim for relief or defense that he is asserting.”
We see no reason not to apply this rule to a policy that has been lost or destroyed without fraudulent intent on the part of the insured. Thus, the claimant has the burden of proving (1) the fact that he or she was insured under the lost policy during the period in issue, and (2) the substance of each policy provision essential to the claim for relief, i.e., essential to the particular coverage that the insured claims. Which provisions those are will vary from case to case;3 the decisions often refer to them simply as the material terms of the lost policy. (See, e.g., Nat. American Ins. Co. of Cal. v. [1072]*1072Underwriters (9th Cir. 1996) 93 F.3d 529, 534; Servants of Paraclete, Inc. v. Great American Ins. (D.N.M. 1994) 857 F.Supp. 822, 827.) In turn, the insurer has the burden of proving the substance of any policy provision “essential to the . . . defense” (Evid. Code, § 500), i.e., any provision that functions to defeat the insured’s claim. (See, e.g., Burroughs Wellcome Co. v. Commercial Union Ins. Co. (S.D.N.Y. 1986) 632 F.Supp. 1213, 1223; Emons Industries v. Liberty Mut. Fire Ins. Co. (S.D.N.Y. 1982) 545 F.Supp. 185, 189.) Those provisions, too, will be case-specific.4
II
We apply these rules to the case at bar. As noted at the outset, the primary issue at trial was whether Employers Policy No. CL92302 (hereafter, the policy) entitled Dart to a defense and indemnity from Commercial Union in litigation arising from exposure to DES between September 1, 1946, and September 1, 1951.
Some matters on which Dart, as the insured, has the burden of proof are undisputed in this case. It is undisputed that neither Dart nor Commercial Union now has a copy of the policy, and that Dart made a reasonably diligent but unsuccessful search for the policy. The trial court was therefore entitled to find that the policy was lost or destroyed, and that it was lost or destroyed without fraudulent intent on the part of Dart. Accordingly, the court properly admitted the secondary evidence offered by Dart to prove the material terms of the policy. Commercial Union does not dispute the point.
It is also undisputed that the policy was a CGL policy and was in effect from September 1, 1946, to September 1, 1951, and that Commercial Union’s predecessor in interest (Employers) issued the policy to Dart’s predecessor in interest (Rexall). The trial court was therefore entitled to find that Dart was insured under the policy during the period in issue. Again Commercial Union does not dispute the point.
Dart contends the Court of Appeal erred in requiring it to prove the material provisions of the policy by introducing evidence of the specific [1073]*1073language used in those provisions. The point is well taken. As discussed above, there is no such evidentiary requirement to prove the contents of lost documents, be they insurance policies or otherwise. Although the Court of Appeal opinion acknowledges the general rule that lost documents can be proved by secondary evidence, the court repeatedly faulted Dart for presenting no evidence of the policy language: “There is no evidence at all concerning the words used in the policy. There is no evidence to show what sort of language was used by Commercial (in standard or manuscript policies or at all) at the time Dart’s policy was issued (or at any time). There is no evidence to show what sort of language was used in the industry at the time Dart’s policy was issued (or at any time). There is no evidence to show what standard provisions, if any, were required by statute at the time Dart’s policy was issued, or what standardized provisions, if any, were usually incorporated into Commercial’s CGL policies (in the 1940’s or at any time).” The court concluded: “[W]e believe that a policy’s language is critical whenever an insured demands a defense or indemnity under a CGL policy,” and “we hold in this case that reference to the language of the policy is indispensable” because of the substantial time between exposure and injury in DES cases.
Thus, we understand the Court of Appeal’s holding as fashioning a special evidentiary rule with respect to insurance policies, an “actual language” rule, providing that the contents of such policies may only be proved by reproductions of the policy language, and that anything less does not, as a matter of law, constitute substantial evidence of these contents. The Court of Appeal reasoned that the meaning of an insurance policy, like that of any contract, is derived from examining its language and concluded that without such language, the contents of a policy cannot be determined.
Although it is a truism that we look to the language of a contract to ascertain its meaning, it is equally true that when a contract has been lost in good faith, and the actual language is unavailable, the law does not require proof of such language. Rather, as discussed above, the proponent of the lost document need only prove the relevant substance of the document. (Kennijf, supra, 140 Cal. at pp. 43-44; Posten v. Rassette, supra, 5 Cal. at p. 469; Von Hasseln v. Von Hasseln, supra, 122 Cal.App.2d at p. 13; Seaboard National Bank v. Ackerman, supra, 16 Cal.App. at p. 58.)
Commercial Union contends that in most lost insurance policy cases from other jurisdictions, there is at least an approximation of the actual language through specimen policies, standardized policies, policies from preceding or subsequent years, and the like. (See, e.g., Remington Arms Co. v. Liberty Mut. Ins. Co. (D.Del. 1992) 810 F.Supp. 1420, 1427; Burroughs Wellcome Co. v. [1074]*1074Commercial Union Ins. Co., supra, 632 F.Supp. at pp. 1222-1223; Employers’ Liability Assurance Corp. v. Hoechst Celanese Corp. (1997) 43 Mass.App. 465 [684 N.E.2d 600, 612].) Yet it is undisputed that in the present case there was a manuscript policy5 specifically written for Dart for the five-year period of 1946-1951; it is therefore less likely that specimen policies, standardized policies or successor policies would be available or useful in establishing the contents of the policy. We see no reason why the lack of such “actual language” evidence should preclude Dart from obtaining the benefits of its policy. When, as here, it is undisputed that there was an insurance policy covering the relevant time period and that the policy was lost in good faith and not recovered after diligent search, there is no reason either in the law of contract or of evidence why secondary evidence that attests to the substance but not.the precise language of an insurance policy should be insufficient as a matter of law to establish the insurer’s contractual obligations.
Commercial Union seeks to distinguish this case from other, noninsurance lost document cases on the ground that the latter involved fewer and simpler terms. The distinction is not legally significant. However few or many terms there may be, it remains necessary for the proponent of a lost deed, power of attorney or a money judgment—no less than a proponent of a lost insurance policy—to prove each of its provisions “essential to the claim for relief.” (Evid. Code, § 500.)
The trial court was thus correct in declaring that “Dart must prove with secondary evidence the existence and terms of the coverage material to the risk at issue. This does not mean that Dart is required to prove the policy’s contents verbatim. Rather, it must prove the substance of the Employers’ Policy with sufficient evidence to show coverage for the DES claims.” As noted above, the trial court found in detail that Dart had proved the substance of the material terms of the policy by sufficient secondary evidence, and entered judgment accordingly. We turn next to specific challenges to the trial court’s findings on the grounds that they were not supported by substantial evidence.
[1075]*1075III
The Court of Appeal opinion identified only one finding assertedly unsupported by the evidence: that the lost policy was an occurrence-based policy in that it provided coverage for exposure to DES during the policy period even if the injuries caused by that exposure were not discovered until after the policy period ended. The Court of Appeal’s opinion was not entirely clear as to the basis of its objection to this finding, but it appears it had two principal concerns. First, Charles Pyne’s testimony, which identified the policy as occurrence-based, essentially lacked probative value because it was “conclusory” and not corroborated by documentary evidence. Second, even if Pyne’s testimony established that the policy was in fact occurrence-based, such testimony was not sufficiently detailed to define that term and related terms so as to establish Commercial Union’s obligation to cover claims of individuals whose injuries ultimately arose from the ingestion of DES during the policy period. “Without any evidence of the language used in Dart’s missing policy, judicial interpretations of common words and phrases actually used in other CGL policies (e.g., ‘accident,’ ‘occurrence,’ ‘bodily injury’) are irrelevant.” We disagree with both points.
First, Pyne’s testimony was not conclusory. Pyne, a licensed insurance broker, was the manager of the Obrion office in Los Angeles during the years the policy was in effect. He became familiar with the product liability coverage of the policy by participating in a number of meetings at which that coverage was discussed; on another occasion, the insurance manager of Dart showed Pyne the policy and questioned him about a coverage matter. The trial court specifically asked Pyne, “Do you recall whether or not you saw an occurrence provision in this contract or do you just assume that there was an occurrence provision in the contract?” Pyne replied, “I saw it in the insurance contract,” and went on to make it clear he was referring to an occurrence-based coverage provision of the policy. This is direct testimony of a fact in issue.
Nor was it necessary that Pyne’s testimony be corroborated by documentary evidence or otherwise. It is well established that “ ‘ “[t]he testimony of a [single] witness . . . may be sufficient” [to support a judgment].’ ” (In re Marriage of Slivka (1986) 183 Cal.App.3d 159, 163 [228 Cal.Rptr. 76]; see also Evid. Code, § 411 [“[e]xcept where additional evidence is required by statute, the direct evidence of one witness who is entitled to full credit is sufficient for proof of any fact”].) Moreover, Pyne’s testimony was in fact corroborated. Dart introduced evidence of two other product liability claims against Dart—known as the Hinkle and Boone claims—that Commercial Union paid under the policy here in issue. In each, the plaintiff successfully [1076]*1076claimed damages for bodily injuries caused by ingestion of an allegedly harmful drug during the policy period but not discovered until after the period ended. Referring to the parties by the names of their predecessors, the trial court expressly found that “the Hinkle and Boone lawsuits against Rexall in the 1950’s are further specific evidence of occurrence coverage. Both cases involved ingestion of a drug during the Policy period with filing of claims after the Employers Policy period ended. Both claims were also paid for by Employers under the Employers Policy.”
The Court of Appeal dismissed this evidence on the ground that it “proves nothing.” The court reasoned that “[i]t is likely as not” that the Hinkle and Boone claims “were defended or paid as an accommodation to Dart” rather than as an obligation of coverage. But the trial court had expressly rejected this assertion as speculative. “There was no evidence produced that Employers ever paid claims that were not covered under the Employers Policy or that there was any practice of Employers to pay uncovered claims as an accommodation to its insureds.” Neither Commercial Union nor the Court of Appeal cites any evidence to the contrary. On this record, the trial court’s conclusion that it was probable that Commercial Union paid the Hinkle and Boone claims to comply with its contractual obligation is a reasonable inference, and as such it must be upheld on appeal. (Overton v. Vita-Food Corp. (1949) 94 Cal.App.2d 367, 370 [210 P.2d 757]; see also Crawford v. Southern Pac. Co. (1935) 3 Cal.2d 427, 429 [45 P.2d 183].) Nor may an appellate court substitute its determination of Pyne’s credibility for that of the trial court. Commercial Union’s contention that his testimony was so untrustworthy, self-contradictory and lacking in foundation as to deserve no credence is nothing more than a request that we redetermine his credibility. This we may not do.
The Court of Appeal is also incorrect that Pyne’s testimony was insufficiently detailed to permit determination of whether the policy covered injuries stemming from ingestion of DES during the policy period. He testified that he knew the meaning of the term “occurrence policy,” and defined it as one in which coverage is provided for “a happening . . . [that] would . . . take place during the policy term, but which might not be discovered until the expiration of the policy term.” He also testified that the policy covered personal injuries from defective products. Moreover, the trial court also found that DES-related injuries were factually similar to the injuries covered in the Hinkle and Boone claims and reasonably inferred that Dart’s policy covered DES injury as well. Such evidence, although it did not reproduce the policy language, was sufficient to support the trial court’s finding that Dart’s policy covered DES injuries initiated during the policy period but manifesting outside it.
[1077]*1077In sum, we conclude that the Court of Appeal erred in holding there was no substantial evidence to support the finding that the lost policy provided occurrence-based coverage and that such coverage included claims for injuries caused by DES exposure during the policy term but not manifesting until after the term ended.
IV
Commercial Union contends that Dart failed to prove a number of other material terms of the lost policy not discussed in the Court of Appeal opinion. The trial court, however, expressly found that Dart had sufficiently proved each such term. Commercial Union must therefore show the record is devoid of substantial evidence to support those findings. We conclude that it has failed to do so.
Commercial Union first complains of findings that the limits of coverage under the policy were $100,000 per person and $300,000 per occurrence, with a $300,000 product liability aggregate limit per year. For the first two of these limits, the trial court relied on both documentary evidence and testimony. The documentary evidence was a letter written in 1970 to Pyne by Herb Bennett, an agent in the Boston office of Obrion, replying to a query by Pyne as to whether the policy covered a certain claim by Rexall (now Dart). In his letter Bennett explained that Commercial Union had discarded the policy and could not produce it, but stated that according to Bob Forrest the primary policy limits were $100,000 and $300,000. The court found that Forrest was the Obrion partner who had negotiated the terms of the policy and was personally responsible for the Rexall account for many years. The court further found that Obrion had agency authority to give out information about the terms of a Commercial Union policy. On this basis, the court found the letter admissible as an authorized admission of a party. (Evid. Code, § 1222.) Commercial Union contends the trial court erred in admitting such testimony.
The rule is that “ 1 “whatever is said by an agent, either in the making of a contract for his principal, or at the time, and accompanying the performance of any act, within the scope of his authority, ... of the particular contract or transaction in which he is then engaged, is, in legal effect, said by his principal, and admissible as evidence .... But declarations or admissions by an agent, of his own authority, and not accompanying the making of a contract, or the doing of an act, in behalf of his principal, ... are not binding upon his principal . . . and are not admissible . . ’ (4 Wigmore on Evidence, § 1078, p. 119. . . .)” (Miller v. Anson-Smith Construction Co. (1960) 185 Cal.App.2d 161, 166 [8 Cal.Rptr. 131], italics added.) In the [1078]*1078present case, the undisputed evidence is that in 1970, when the letter was written, Obrion was still an agent for Commercial Union, and that Forrest was a partner in that firm, albeit inactive. The trial court made a reasonable inference that Obrion and its partners were authorized within the scope of their agency agreement to supply information in the course of business regarding Commercial Union’s insurance policies. “[A]n appellate court reviews any ruling by a trial court as to the admissibility of evidence for abuse of discretion.” (People v. Alvarez (1996) 14 Cal.4th 155, 203 [58 Cal.Rptr.2d 385, 926 P.2d 365].) We conclude the trial court did not abuse its discretion in admitting the Bennett letter as a party admission.
The court also relied on Pyne’s testimony on the point, who declared that to his knowledge, based on his recollection from his reading of the policy, the policy limits were $100,000 and $300,000. Commercial Union again questions the credibility of Pyne’s testimony, but the question was exclusively for the trial court to resolve. We conclude there was substantial evidence supporting the trial court’s finding regarding the policy limits.
For its finding that the policy had a products liability aggregate limit of $300,000 per year, the trial court again relied on Pyne’s testimony that the policy so provided. The court also correctly observed that if the annual product liability aggregate limit were less than the product liability per claim limit (i.e., $300,000), the coverage would be illusory.
Commercial Union contends there was no substantial evidence to support the finding that the lost policy provided coverage for a single five-year term (from Sept. 1, 1946, to Sept. 1, 1951). For this finding, the trial court relied on certain documentary exhibits from 1961 regarding retrospective premium adjustments that expressly referred to a single policy, designated as policy No. CL 92302, for that term, corroborated by Pyne’s testimony to the same effect. Commercial Union speculates that the lost policy could equally well have been for five successive one-year terms, with possible annual changes in its material provisions, but no evidence supports that speculation.
Commercial Union next contends its contractual obligations may have been relieved or reduced by an “other insurance” clause.6 It argues the trial court erroneously found that the insurer had the burden of proving the existence and terms of such a clause, if any, and that it failed to do so. It contends that an “other insurance” clause is not an [1079]*1079exclusion but a condition of coverage, that therefore it was Dart’s burden to prove the conditions of such a clause have been met, and that failure to prove the same is fatal to Dart’s case. This argument rests on several contentions.
First, Commercial Union points to testimony of Peter Fortuna, an underwriter at Commercial Union during the 1940’s and 1950’s, establishing that every policy issued by Commercial Union would have had one of three types of “other insurance” clauses: one that automatically rendered the policy null and void if there was any other insurance covering the same risk (a “null and void” clause); one that rendered it null and void up to the limit of coverage of the other policy, but paid any excess over that coverage within its own limits (a “null and void excess” clause); and one that prorated the. coverage between the two policies (a “pro rata” clause). Second, Commercial Union reasons that if Dart’s policy contained a “null and void” or “null and void with excess” clause, then it would not be liable for indemnification if there were other insurance with pro rata policies to cover these costs. At least some courts have held that when there are two primary insurance policies covering the same injury, one with a pro rata “other insurance” clause and another with a clause that agrees to indemnify only in excess of other insurance, the latter insurer does not become liable until the former’s policy limits have been reached. (See Donahue Constr. Co. v. Transport Indent. Co. (1970) 7 Cal.App.3d 291, 302 [86 Cal.Rptr. 632], and cases cited therein.) In fact, Dart’s successor insurers had policies with pro rata “other insurance” clauses. Third, Commercial Union argues that under a continuous injury trigger theory of coverage,- Dart’s successor insurers would be liable for occurrences initiated during that period but manifesting in a later period. (See Montrose Chemical Corp. v. Admiral Ins. Co. (1995) 10 Cal.4th 645, 689 [42 Cal.Rptr.2d 324, 913 P.2d 878].) Thus Commercial Union contends that Dart’s failure to prove the type of “other insurance” clause in the policy would leave a court unable to determine whether Commercial Union’s liability for the relevant period was reduced or extinguished. In other words, it argues Dart has failed to prove the substance of a policy provision essential to its claim for relief.
We disagree that ignorance of the precise type of “other insurance” clause in a lost policy can be used to defeat the insurer’s obligations altogether. “Historically, ‘other insurance’ clauses were designed to prevent multiple recoveries when more than one policy provided coverage for a particular loss.” (Fireman’s Fund Ins. Co. v. Maryland Casualty Co. (1998) 65 Cal.App.4th 1279, 1304 [77 Cal.Rptr.2d 296].) On the other hand, “other [1080]*1080insurance” clauses that attempt to shift the burden away from one primary insurer wholly or largely to other insurers have been the objects of judicial distrust. “[Pjublic policy disfavors ‘escape’ clauses, whereby coverage purports to evaporate in the presence of other insurance. [Citations.] This disfavor should also apply, to a lesser extent, to excess-only clauses, by which carriers seek exculpation whenever the loss falls within another carrier’s policy limit.” (CSE Ins. Group v. Northbrook Property & Casualty Co. (1994) 23 Cal.App.4th 1839, 1845 [29 Cal.Rptr.2d 120]; see also Argonaut Ins. Co. v. Transport Indent. Co. (1972) 6 Cal.3d 496, 508 [99 Cal.Rptr. 617, 492 P.2d 673].) Partly for this reason, the modem trend is to require equitable contributions on a pro rata basis from all primary insurers regardless of the type of “other insurance” clause in their policies. (CSE Ins. Group, at p. 1845; see also Pacific Indemnity Co. v. Bellafonte Ins. Co. (2001) 80 Cal.App.4th 1226, 1236-1238 [95 Cal.Rptr.2d 911]; Fireman’s Fund Ins. Co. v. Maryland Casualty Co., supra, 65 Cal.App.4th at pp. 1305-1306.)
Whether or not the above rale is universally applicable, it is clear that the obligation of successive primary insurers to cover a continuously manifesting injury is a separate issue from the obligations of the insurers to each other. As one Court of Appeal explained: “[Apportionment among multiple insurers must be distinguished from apportionment between an insurer and its insured. When multiple policies are triggered on a single claim, the insurers’ liability is apportioned pursuant to the ‘other insurance’ clauses of the policies [citation] or under the equitable doctrine of contribution [citations]. That apportionment, however, has no bearing upon the insurers’ obligations to the policyholder. [Citation.] A pro rata allocation among insurers ‘does not reduce their respective obligations to their insured.’ [Citation.] The insurers’ contractual obligation to the policyholder is to cover the full extent of the policyholder’s liability (up to the policy limits).” (Armstrong World Industries, Inc. v. Aetna Casualty & Surety Co. (1996) 45 Cal.App.4th 1,105-106 [52 Cal.Rptr.2d 690]; see also FMC Corp. v. Plaisted & Companies (1998) 61 Cal.App.4th 1132, 1185 [72 Cal.Rptr.2d 467].) This principle is consistent with “the settled mle that an insurer on the risk when continuous or progressively deteriorating damage or injury first manifests itself remains obligated to indemnify the insured for the entirety of the ensuing damage or injury.” (Montrose Chemical Corp. v. Admiral Ins. Co., supra, 10 Cal.4th at p. 686.)
Here, the evidence is undisputed that Commercial Union was a primary insurer during the relevant time period. Therefore, even if Commercial Union had a “null and void with excess” “other insurance” clause, all that would be established is that it had a right to seek some kind of contribution from successive insurers also liable to Dart. It would not relieve Commercial [1081]*1081Union from either its obligation to indemnify or to defend Dart. Commercial Union is therefore incorrect that Dart’s inability to prove the type of “other insurance” clause found in its policy cancels its own obligations as a primary insurer.
Finally, Commercial Union contends there is insufficient evidence of the particular DES claims against Dart to justify the defense and indemnification obligations imposed on it by the trial court. We have already partly addressed this contention. As discussed, there is indeed substantial evidence to support the trial court’s finding that the policy in question covered injuries arising from DES ingestion during the policy period, and that therefore Commercial Union had a duty to defend and indemnify Dart. But there is a separate question of whether the evidence of DES claims presented by Dart justified the amount of damages. That question is not before us; it has nothing to do with proving the contents of a lost policy but rather concerns the validity of resolving insurance coverage claims in DES cases by relying on something less than an examination of each individual claim against the insured. (See Keene Corp. v. Ins. Co. of North America (D.C. Cir. 1981) 667 F.2d 1034, 1040; Burroughs Wellcome Co. v. Commercial Union Ins. Co., supra, 632 F.Supp. at p. 1218.) Our grant of review did not include this issue.
Therefore, on remand, the Court of Appeal is to apply the terms of the insurance policy, as delineated in the trial court’s findings, to determine whether there is sufficient evidence to support the amount of the trial court’s award of damages and defense costs.
V
The judgment of the Court of Appeal is reversed and the cause is remanded to the Court of Appeal for proceedings consistent with this opinion.
George, C. J., Kennard, J., Baxter, J., Werdegar, J., and Chin, J., concurred.