Transport Indemnity Company (Transport) appeals from a declaratory judgment holding it liable to defend and indemnify certain parties against claims arising out of an automobile accident. Finding that the district court correctly interpreted the relevant insurance contracts in the light of controlling state law, we affirm.

Transport brought this action for declaratory relief pursuant to 28 U.S.C. §§ 2201, 2202 to determine the duties of Transport and Liberty Mutual Insurance Company (Liberty) to defend and indemnify certain parties against claims arising out of a June 5, 1974 automobile accident. Jurisdiction is based on diversity of citizenship.

A vehicle operated by Simons or Rice was involved in a collision with a vehicle operated by Hill. Hill and Rice were killed. Simons was a high-level employee of both SWF Plywood (SWF) and Carolina Pacific Plywood, Inc. (Carolina Pacific), and Rice was the Executive Vice President and Chief Operations Officer of both companies. Carolina Pacific was the registered owner of the vehicle involved, and the vehicle was assigned to Rice. As part of an interim arrangement between SWF and Carolina Pacific, in anticipation of a planned merger, the cost of operation, maintenance and depreciation of the vehicle was paid by Carolina Pacific which in turn charged SWF for one half of the expenses. As of the date of the accident, SWF and Carolina Pacific had also consolidated the management of the two corporations and shared equally in paying the salaries of Rice and Simons.

Hill's estate claims damages in excess of $250,000 and claims that Simons, the estate of Rice, SWF and Carolina Pacific are each liable. Transport, as the issuer of a liability and damage policy which listed Southwest Forest Industries, Inc. and "all divisions and subsidiaries" as the named insured (which includes both SWF and Carolina Pacific), called on Liberty, the issuer of a similar policy which named SWF as the insured, to participate in settling Hill's claim. Liberty declined, contending the policy it had issued to SWF did not cover any of the named defendants for liability arising from this accident.

Transport contends that the district court erred in concluding that (1) Liberty had no potential liability in this case because the facts fell within the range of the joint venture exclusion clause of SWF's policy with Liberty and (2) Transport's policy extended coverage to Rice, Simons, and SWF. We will address each issue in turn.

The district judge held that Rice, Simons, Carolina Pacific and SWF were not covered by Liberty's primary policy because of the following exclusion provision:

This insurance does not apply to bodily injury or property damage arising out of (1) a non-owned automobile used in the conduct of any partnership or joint venture of which the insured is a partner or member and which is not designated in this policy as a named insured.

(Emphasis added.) Transport claims that the district judge erred in finding that (1) Rice and Simons used a "non-owned" rather than a "hired" automobile, and (2) the automobile was being used in the conduct of a "joint venture." It is clear that this exclusion would not apply if Carolina Pacific and SWF were not engaged in a joint venture or if the automobile was a "hired" rather than a "non-owned" automobile.

Before we resolve the substantive issues, however, we must set forth our standard of review and the law we are to apply. First, the dispute between the parties concerns the correct interpretation of the exclusionary contract provision rather than the correct application of the facts of this case to an accepted definition. We are thus primarily faced with a question of law subject to our de novo review. See United States v. Haas and Haynie Corp., 577 F.2d 568, 572 (9th Cir. 1978); Fratis v. Fireman's American Ins. Co., 56 Cal. App. 3d 339, 341-42, 128 Cal. Rptr. 391, 393 (1976).

Second, in this diversity case we apply the case law of Oregon. Owens v. White, 380 F.2d 310 (9th Cir. 1967). Because the law of Oregon is not crystal clear on some issues presented to us, however, we attach great weight to the district judge's determination as to the law of the particular state in which he sits and we will reverse that determination only if it is clearly wrong. American Timber & Trading Co. v. First Nat'l Bank, 511 F.2d 980, 983 (9th Cir. 1973), cert. denied, 421 U.S. 921, 95 S. Ct. 1588, 43 L. Ed. 2d 789 (1975); Insurance Co. of North America v. Thompson, 381 F.2d 677, 681 (9th Cir. 1967).

Transport, in support of its claim that no joint venture existed, contends that an essential element in the legal definition of a joint venture is a joint, shared, or mutual profit. E. g., Elliott v. Murphy Timber Co., 117 Or. 387, 394, 244 P. 91 (1926); Adams Mfg. & Engineering Co. v. Coast Centerless Grinding Co., 184 Cal. App. 2d 649, 7 Cal. Rptr. 761 (1960). Transport argues that here, however, SWF and Carolina Pacific shared expenses and consolidated management to maximize efficiencies, but continued to sustain individual expenses and to calculate profits individually.

In response, Liberty relies on cases that speak of a "community of interest" or "mutual benefit or profit," as the requisite element. West v. Soto, 85 Ariz. 255, 336 P.2d 153, 157 (1959) (community of interest). Giddings Convalescent Home, Inc. v. Wilson, 473 S.W.2d 246, 248 (Tex.Civ.App.1971) (mutual benefit or profit). Liberty concludes that the mutual benefit flowing from an agreement to share expenses for the joint employment of men and machinery provides an alternative to the traditional element of shared profits. Transport's rejoinder is that, even if Liberty is correct that the term joint venture may be read more broadly than the traditional definition, the term as used in the policy is at least ambiguous and should thus be construed, consistent with traditional canons of construction, in favor of coverage.

Apart from standard legal definitions, the real question is what the parties intended when they contracted to exclude coverage of non-owned automobiles used in the conduct of a "joint venture." None of the cases cited by the parties deal with the meaning and scope of "joint venture" as it is used in this type of insurance clause. The district judge, applying the law of Oregon, found that a joint venture existed. We cannot say he was clearly wrong. We agree with Liberty that the purpose of the exclusion was to prevent the insured from engaging in a cooperative venture with another party, thereby incurring additional liability for injury caused by an automobile belonging to the uninsured party. We believe that such provisions contemplate the broader concept of joint venture. See Sprow v. Hartford Ins. Co., 594 F.2d 418 (5th Cir. 1979) (joint venture exclusion held to preclude coverage where the owners of two seafood businesses had agreed to share expenses in the joint operation of a refrigerated delivery truck). See also Levine v. Goldberg, 2 A.D.2d 409, 156 N.Y.S.2d 587 (1956) (per curiam) (finding joint venture in agreement to jointly lease office space and share certain office expenses).

SWF and Carolina Pacific were engaged in a joint operation, including the sharing of management personnel and certain expenses; it is such a joint enterprise that was contemplated by the exclusion. We thus hold that there is no ambiguity to be construed and that SWF and Carolina Pacific were engaged in a joint venture.

Under the policy, a "non-owned" automobile is defined as "an automobile which is neither an owned nor a hired automobile." A "hired automobile" is

(1) an automobile not owned by the named insured (2) which is used under contract in behalf of, or loaned to, the named insured, (3) provided such automobile is not owned by or registered in the name of (a) a partner or executive officer of the named insured or (b) an employee or agent of the named insured who is granted an operating allowance of any sort for the use of such automobile.

While Transport argues that the contract between SWF and Carolina Pacific to share the cost of operating the automobile obviously fits the above definition, it is clear that not all contracts contemplating the use of an automobile imply that the automobile is used "under contract in behalf of" the insured.

Courts have, for example, attempted to draw a line between mere service contracts, involving independent contractors, and "truck and driver" situations in which the insured is viewed as having contracted for the use of the automobile. It has thus been stated that "for a vehicle to constitute a hired automobile, there must be a separate contract by which the vehicle is hired or leased to the named insured for his exclusive use or control." Sprow v. Hartford Ins. Co., supra, 594 F.2d at 422. Although Sprow stated the "separate contract" and "control" tests in the conjunctive, the tendency to resolve doubt in favor of coverage has led some courts to find that the automobile is hired where there is a separate lease of the vehicle, even if there is no actual exercise of control by the insured, Russom v. Insurance Co. of North America, 421 F.2d 985, 993 (6th Cir. 1970), or where there is evidence of actual control, as in several loading cases, even if there is no separate lease of the vehicle. E. g., Wolverine Ins. Co. v. State Auto. Mut. Ins. Co., 415 F.2d 1182, 1184 (6th Cir. 1969) (insured's employee "was in control of the (vehicle) at the time of the accident"); Bituminous Cas. Corp. v. Travelers Ins. Co., 122 F. Supp. 197, 202-03 (D.Minn.1954) (independent contractor's truck in loading accident at defendant's lime quarry was a hired automobile). Thus, under these cases, either the existence of a separate lease or the exercise of actual control sufficiently implicates the insured in the use of the vehicle for a court to conclude that the automobile is being used "under contract in behalf of" the insured.

These cases reflect the efforts of courts to extend coverage to situations in which the insured is fairly viewed as having augmented his automobile fleet to meet his business needs, without expanding coverage to every contract that might literally fit the definition of a hired automobile. They teach that our concern is not merely with a literal application of the language but with a sense of the underlying purpose of the parties. We acknowledge, however, that the "separate contract" and "control" tests are not completely decisive on the facts before us. There is no separate lease of the automobile here. Thus, it might be inferred that SWF bargained for the benefits of the joint venture rather than for the use of an automobile. See Sprow v. Hartford Ins. Co., supra, 594 F.2d at 422. This was apparently the view taken by the district judge, who found that the contractual arrangement between Carolina Pacific and SWF did not amount to SWF renting the automobile under contract; rather, SWF simply "shared expenses with" Carolina Pacific in anticipation of merger. Liberty also argues that SWF lacked exclusive control of the automobile. But Transport contends that SWF exercised control over the use of the automobile through its agents, Rice and Simons, and we are hard-pressed to say that the degree of control exercised by SWF here is distinguishable from the loading cases described above.

Nevertheless, we conclude that the district judge was not clearly wrong in concluding that under Oregon law an agreement to share automobile expenses as part of a joint venture should not be treated as a contract of hire so as to bring the automobile within the hired automobile coverage of the insured's policy. The same conclusion was reached in Sprow. The hired automobile provision and the joint venture exclusion should be read together, if possible, so as to make sense of the purposes of the parties. The joint venture exclusion was designed to prevent the insured from gaining the systematic bonus coverage that would come from utilizing vehicles owned by a partner or joint venturer. If every contractual arrangement between joint venturers for sharing automobile expenses were viewed as a contract of hire, so that even an automobile used in the conduct of the joint venture was considered a hired automobile, the purpose of the joint venture exclusion would be largely undermined.

The fact that the automobile here was being used in the conduct of the joint venture, rather than uniquely on behalf of SWF, distinguishes this case from Miller v. National Farmers Union Prop. & Cas. Co., 470 F.2d 700 (8th Cir. 1972), and Fratis v. Fireman's Fund American Ins. Co., supra, 56 Cal. App. 3d 339, 128 Cal. Rptr. 391.1 Viewing the holding in Sprow v. Hartford Ins. Co., supra, 594 F.2d 422, as the most consonant with the intent of the parties, we hold that this was a non-owned automobile used in the conduct of a joint venture.

We now address Transport's contention that the district judge incorrectly found that its policy covered Rice, Simons, and SWF.

Transport asserts that the district court's decision should have been limited to determining the purely legal question whether Rice was covered only if he was in the course and scope of his duties as Executive Vice President at the time of the accident. Instead, the district judge not only decided this legal question, but also found that Rice was in fact acting within the scope of his duties as Executive Vice President and that Transport was therefore liable. We conclude that the district judge was within his bounds to rule on Transport's own contention that it had "no duty to defend David Rice or . . . to pay any resulting judgment . . . ." The district judge was not limited to determination of the purely legal question as Transport contends.

In addition, Transport argues that the district judge inappropriately relied on the factual findings of Simons v. SWF Plywood Co., 26 Or.App. 137, 552 P.2d 268 (1976), a case arising out of the same accident. Although the parties have provided extensive argument as to whether Transport can be collaterally estopped from denying the findings of fact developed in Simons, we are not convinced that the district judge ruled that Transport was so estopped. Nowhere in the district court opinion does it mention the doctrine of collateral estoppel or assert that Transport is bound by the fact-finding in Simons. It is equally plausible that the district judge cited Simons because he agreed with the conclusion of law based on similar facts.

This conclusion is strengthened by the fact that the parties are not in dispute as to the pertinent facts. Transport does not deny that Rice was Executive Vice President of SWF and Carolina Pacific. Evidence in the record indicates that he was riding in a vehicle owned by his corporate employer and assigned to him. It is also agreed that Rice and Simons had attended a top-level business meeting in Albany the day of the accident, and had returned to Medford together. Indeed, Transport acknowledges that Rice was driving Simons to Grants Pass at the end of this business day.

Transport argues before us not that the district judge incorrectly found these facts, but that he drew an unwarranted legal conclusion from them. Indeed, Transport appears to acknowledge that Rice was acting in the course of his employment with the company, but denies that Rice was acting "as an executive officer" at the time of the accident. In support of its position, Transport cites cases holding that directors and officers may also act in separate capacities as employees of the corporation. Wharton v. Fidelity-Baltimore Nat'l Bank, 222 Md. 177, 158 A.2d 887 (1960) (director); Solheim v. Hastings Housing Co., 151 Neb. 264, 37 N.W.2d 212 (1949) (officer). In turn, Transport argues that Rice was necessarily acting only as an employee in driving Simons to Grants Pass because no business decisions were being discussed during the trip. In Solheim, however, the insured actually held two distinct positions, as President of the corporation and general manager of construction, and was clearly functioning in the latter capacity at the time of the accident. 37 N.W.2d at 218-19. Transport does not allege that Rice held two distinct positions in the company, and provides no legal authority for the view that a corporate officer can be seen as somehow acting in another capacity merely because his work-related activity does not involve the exercise of his executive expertise. If Rice was acting in an employment capacity, we think it is clear that he was acting in his sole capacity as an executive officer. The lack of proof of business discussion during the automobile trip hardly forecloses the finding that Simons and Rice were still in the process of completing their day of business on behalf of the companies. While neither Simons nor the district court decision addressed Transport's "dual capacity" argument, we conclude that it is clearly wrong in this case as a matter of law, and that the district judge was within his bounds to reach the conclusion he did.

Transport contends that the district court erred in concluding that Simons was an "executive officer" for the purposes of this contract. The district judge found that Simons was a general manager of production in charge of supervising plant managers, that he reported directly to Rice, and that on the date of the accident he attended a meeting on budgeting with "top management." He relied on Guillory v. Aetna Ins. Co., 415 F.2d 650, 652 (5th Cir. 1969), for the conclusion that executive officer "implies some . . . managerial responsibility for the affairs of the corporation . . . and a close connection with the (top) officers of the company," but is not restricted to corporate officers.

Transport distinguishes this case from Guillory on grounds that this is an arm's length manuscript contract to which the rule of construing insurance contracts against the carrier does not apply. But it fails to offer any reason for concluding that something more restrictive was intended by the use of "executive officer" here. Moreover, in contrast to Transport's reliance on a single case generally distinguishing "officers" and "agents," Vardeman v. Penn Mut. Life Ins. Co., 125 Ga. 117, 54 S.E. 66, 67 (1906), Liberty cites several other cases liberally construing "executive officer" as used in this type of insurance contract. Vega v. Southern Scrap Material Co., 517 F.2d 254, 257-58 (5th Cir. 1975); Strickland v. Transamerica Ins. Co., 481 F.2d 138, 148 (5th Cir. 1973); Galloway v. Employers Mut. of Wausau, 286 So. 2d 676, 679 (La.App.1973); Berry v. Aetna Cas. & Surety Co., 240 So. 2d 243, 246 (La.App.), writ refused, 256 La. 914, 240 So. 2d 374 (1970), cert. denied, 401 U.S. 1005, 91 S. Ct. 1255, 28 L. Ed. 2d 541 (1971). We cannot say the district judge was wrong in concluding that, under Oregon law, the contract is clear and unambiguous.

Finally, Transport contends that the district judge erred in refusing to consider an alleged oral agreement between Transport and Southwest Forest Industries providing that Transport would be liable as an excess insurer of SWF only. The district judge based his ruling in part on an Oregon statute which requires that "every contract of insurance shall be construed according to the terms and conditions of the policy." O.R.S. 743.045. Transport argues that the district judge should not have applied Oregon substantive law because Oregon courts would have applied the law of California or Arizona, as states with significant contacts which would uphold the validity of the oral agreement. In the alternative, Transport argues that the Oregon statute is merely a statutory embodiment of the parol evidence rule, the proper analysis of which would uphold the contract.

As to the choice-of-law issue, we cannot say that the district judge erred in applying Oregon law. Transport does not deny that Oregon has significant contacts with the transactions at issue here, but argues that the so-called rule of validation should be applied so as to uphold the alleged oral agreement between the parties. Although it is true that Oregon recognized the rule of validation in Lilienthal v. Kaufman, 239 Or. 1, 395 P.2d 543 (1964), it is also true that the Oregon Supreme Court applied the law of Oregon in that case because of Oregon's strong interest in applying its own law. We cannot say that the district judge, as one experienced in the law of Oregon, incorrectly concluded that the courts of Oregon would find a similar interest in the enforcement of that State's policy of having contracts of insurance interpreted according to the expectations arising from the written document.

Moreover, Transport failed to demonstrate that a different result would have been required under the law of Arizona or California. The full extent of its legal authority was the citation to one Arizona case holding that parol insurance contracts are valid under Arizona law. Beyond that, however, Transport simply argues that the alleged agreement should be viewed as collateral to the written policy or as an aid to its construction. We agree with Liberty, however, that the alleged oral agreement contradicts the face of the policy, which lists SWF as a named insured to receive its primary coverage. It is equally clear that the oral agreement goes far beyond aiding in construction of the policy's terms; rather than merely clarifying the identity of one of the named insureds, it varies the nature of the coverage provided. Finally, Transport's argument that the parol evidence rule does not apply where one of the parties to the proceeding is a stranger to the contract is also without merit. See 4 Williston on Contracts § 647, at 1161-67 (1961).

The district judge correctly applied Oregon law. We reject Transport's contention that the statutory provision in question is merely a codification of the parol evidence rule. This provision is part of the insurance code of the State. Under such statutes, "the policy . . . constitutes the sole contract, to the exclusion of all anterior or contemporaneous agreements not contained or expressed in the policy . . . ." Couch on Insurance 2d § 3:21, at 144 (1959). See, e. g., Spain v. Travelers Ins. Co., 332 So. 2d 827 (La.1976).

AFFIRMED.