MATHESON, Circuit Judge.
Philadelphia Indemnity Insurance Company (“Philadelphia”) and Lexington Insurance Company (“Lexington”) insured the same school building that suffered fire damage. In this declaratory judgment action, they dispute their relative responsibilities to pay for the loss.
Charter school Tulsa School of Arts and Sciences (“TSAS”) leased the Barnard Elementary School building from the Independent School District No. 1 of Tulsa County, Oklahoma (“District”). As required under the lease, TSAS acquired an insurance policy for the Barnard building. The policy TSAS purchased through Philadelphia named the District as the loss payee. The District had a separate insurance policy with Lexington that also covered the Barnard building, among other District bufldings.
The district court ordered Philadelphia to pay 54 percent and Lexington to pay 46 percent of the approximately $6 million loss. Lexington appeals, arguing it should have no obligation to pay. Philadelphia cross-appeals, arguing Lexington should have to pay more.
Exercising jurisdiction under 28 U.S.C. § 1291, we affirm.
I. BACKGROUND
A. Factual History
In 2012, an Oklahoma charter school, TSAS, leased a building for its operations from the District. The building — the Barnard Elementary School — was one of more than 100 facilities owned by the District and covered for fire damage under its insurance policy from Lexington.
The lease agreement required TSAS to acquire its own insurance policy for the building. TSAS secured a policy from Philadelphia, under which TSAS was the named insured and the District was the loss payee.
The Lexington and Philadelphia policies were similar. They had the same effective dates: July 1, 2012, to July 1, 2013. They both protected against fire damage to the Barnard building. And the policies had identically worded “Other Insurance” provisions, which stated:
[1333]*13331. You may have other insurance subject to the same plan, terms, conditions and provisions as the insurance under this Coverage Part. If you do, we will pay our share of the covered loss or damage. Our share is the proportion that the applicable Limit of Insurance under this Coverage Part bears to the Limits of Insurance of all insurance covering on the same basis.
2. If there is other insurance covering the same loss or damage, other than that described in 1 above, we will pay only for the amount of covered loss or damage in excess of the amount due from that other insurance, whether you can collect on it or not. But we will not pay more than the applicable Limit of Insurance.
App., Vol. 1 at 125 (Philadelphia policy); App., Vol. 2 at 286 (Lexington policy).
An important difference between the policies was their coverage limits. The parties stipulate that Philadelphia’s policy limit was $7 million. The Lexington policy, which covered many District buildings, had a total coverage limit of $100 million per occurrence, but, as we discuss below, the parties dispute whether that is the relevant limit here.
Fire damaged the Barnard building on September 5, 2012. The insurers agreed the total adjusted loss was $6,014,359.06. It is unclear from the record whether the District or TSAS, or both, made claims under either the Lexington or Philadelphia policies, or how long it took the insurer(s) to pay the claim(s), but counsel for Philadelphia said at oral argument that “the insureds have been paid, and as far as they’re concerned it’s over.” Oral Arg. at 20:18-25. As between the insurers, however, litigation ensued.
B. Procedural History
In March 2013, Philadelphia filed a complaint in the U.S. District Court for the Northern District of Oklahoma seeking a declaratory judgment.1
The parties cross-moved for summary judgment. In a December 2015 order, the district court granted Philadelphia’s motion and denied Lexington’s motion. Phila. Indem. Ins. Co. v. Lexington Ins. Co., No. 13-CV-165-JED-FHM, 2015 WL 8485249, at *3 (N.D. Okla. Dec. 9, 2015).
The district court concluded that, under Oklahoma law, the policies’ “other insurance” provisions canceled each other out because the “two insurers have provided insurance policies that cover the same loss.” Id. at *2, The court rejected Lexington’s arguments that (1) Philadelphia lacked standing to sue, (2) the different named insureds on the two policies and the alleged different interests insured precluded sharing, (3) the parties to the lease— TSAS and the District — had agreed that TSAS would acquire primary insurance and that Philadelphia’s policy was therefore the policy of first resort, and (4) the Philadelphia policy was more “specific” such that its coverage was primary with Lexington providing only excess coverage. Id. at *1-3.
Having concluded that the “other insurance” clauses were mutually defeating, the district court ruled that “Philadelphia and Lexington shall share coverage of the loss ‘on a pro rata basis according to the ratio each respective policy limit bears to the [1334]*1334cumulative limit of all concurrent policies.’ ” Id. at *3 (quoting Equity Mut. Ins. Co. v. Spring Valley Wholesale Nursery, Inc., 747 P.2d 947, 954 (Okla. 1987)). But because the relevant limit on the Lexington policy was unclear, the court ordered briefing on that narrow issue. Id.
After briefing, the district court ruled that Lexington’s relevant policy limit for purposes of the pro rata calculation equaled the total amount of damage to the building, $6,014,359.06. App., Vol. 2 at 459. The court selected this number based on an “Occurrence Limit of Liability Endorsement” (“Endorsement”) in the Lexington policy. The court concluded that applying the $100 million overall limit per occurrence, which Philadelphia argued should apply, would require it to ignore the unambiguous terms of the Endorsement. Id.
The district court arrived at its pro rata apportionment as follows:
• $6,014,359.06 was the total amount of the loss.
• $7 million was the Philadelphia policy limit.
• $6,014,359.06 was the relevant Lexington policy limit per the Endorsement.
• $13,014,359.06 was the total amount of coverage ($7 million plus $6,014,-359.06 — the sum of the policy limits).
• 53.79 percent was Philadelphia’s percentage share of the loss. This was the proportion of Philadelphia’s policy limit to the total amount of coverage: ($7 million / $13,014,359.06) x 100 = 53.79 percent.
• $3,235,123.74 was Philadelphia’s share of the loss (53.79 percent x $6,014,359.06).
• 46.21 percent was Lexington’s percentage share of the loss. This was the proportion of Lexington’s policy limit to the total amount of coverage: ($6,014,359.06 / $13,014,359.06) x 100 = 46.21 percent.
• $2,779,235.32 was Lexington’s share of the loss (46.21 percent x $6,014,359.06).
The court ordered each insurer to pay its respective percentage of the loss and entered a final judgment to that effect. Id. at 459-60.
Lexington appeals the district court’s summary judgment order requiring that the loss be shared. Philadelphia cross-appeals and challenges the district court’s apportionment of the loss.
II. DISCUSSION
We (A) consider, and reject, Lexington’s argument that Philadelphia lacks standing. We then (B) reach the merits issues and (1) conclude the insurers here must share the loss under Oklahoma law and (2) reject Philadelphia’s argument against the district court’s pro rata apportionment. Accordingly, we affirm the district court’s judgment.
A. Philadelphia’s Standing
We must first address Lexington’s argument that Philadelphia lacks standing under Article III of the Constitution. See Cornhusker Cas. Co. v. Skaj, 786 F.3d 842, 851 (10th Cir. 2015) (“Standing is a threshold issue in every case before a federal court ... because the standing issue implicates a federal court’s subject-matter jurisdiction.” (quotations, internal citation, and emphasis omitted)). “[W]e review questions of standing de novo.” Colo. Outfitters Ass’n v. Hickenlooper, 823 F.3d 537, 544 (10th Cir. 2016). We agree with the district court that Philadelphia has standing.
1. Legal Background
The Constitution limits federal court jurisdiction to certain cases and controver[1335]*1335sies. See U.S. Const, art. Ill, § 2. In the declaratory judgment context, the Supreme Court has explained Article Ill’s case-or-controversy requirement as follows: “ ‘Basically, the question in each case is whether the facts alleged, under all the circumstances, show that there is a substantial controversy, between parties having adverse legal interests, of sufficient immediacy and reality to warrant the issuance of a declaratory judgment.’ ” MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118, 127, 127 S.Ct. 764, 166 L.Ed.2d 604 (2007) (quoting Md. Cas. Co. v. Pac. Coal & Oil Co., 312 U.S. 270, 273, 61 S.Ct. 510, 85 L.Ed. 826 (1941)).
The case-or-controversy requirement raised here is standing. As a general matter, “a plaintiff must demonstrate standing to sue by establishing ‘(1) an injury in fact, (2) a sufficient causal connection between the injury and the conduct complained of, and (3) a likelihood that the injury will be redressed by a favorable decision.’ ” Colo. Outfitters Ass’n, 823 F.3d at 543 (alterations and some internal quotation marks omitted) (quoting Susan B. Anthony List v. Driehaus, — U.S.-, 134 S.Ct. 2334, 2341, 189 L.Ed.2d 246 (2014)).
2. Analysis
Philadelphia has standing. Its alleged injury is financial, definite, and concrete. Philadelphia’s interests are adverse to Lexington’s: One insurer or the other will bear the loss or they will share it in some manner. Because Philadelphia’s injury is causally connected to how its policy interacts with Lexington’s policy, a judicial determination of the insurers’ respective responsibilities under the policies will redress and resolve this dispute.2
Lexington argues that because Philadelphia’s insured, TSAS, is neither a party to nor a third-party beneficiary of the Lexington-District insurance contract, and, because an insurer can have no greater rights in a policy than its insured, Philadelphia cannot “file suit on the Lexington Policy” and therefore lacks standing. Aplt. Br. at 12. Lexington cites May v. Mid-Century Insurance Co., 151 P.3d 132 (Okla. 2006), for the proposition that “only parties to the insurance contract may bring suit thereon.” Aplt. Br. at 12. We reject this argument because May is inap-posite and because it misapprehends Philadelphia’s declaratory judgment action.
In May, a fire damaged a condominium complex. 151 P.3d at 134. One of the unit owners sued the condominium association’s insurer for bad-faith refusal to pay benefits allegedly owing to her through the association’s policy. Id. The Oklahoma Supreme Court affirmed the trial court’s dismissal of the case because the unit owner was not an insured under the association’s insurance contract, nor was she a third-party beneficiary. Id. at 140-41.
Lexington’s reliance on May is misplaced. It misapprehends Philadelphia’s declaratory judgment claim. Philadelphia is not seeking to enforce rights as an insured under the Lexington policy; it is suing Lexington insurer-to-insurer. Philadelphia seeks a judicial determination of how its policy interacts with the Lexington policy. As we explained more than 40 years [1336]*1336ago in rejecting a similar standing argument, “this action is not one to enforce a contract but rather seeks a declaration of the relative rights and duties of [the insurers].” United Servs. Auto. Ass’n v. Royal-Globe Ins. Co., 511 F.2d 1094, 1096 (10th Cir. 1975).3
And as we recently pointed out, Lexington’s argument about who is covered under its policy does not concern Article III standing: “Although denominated a matter of standing, [the insurer’s] argument is actually of another genus entirely — one that does not implicate a court’s jurisdiction. Indeed, it presents no more than a garden-variety question regarding the proper interpretation of the [p]olicy.” Comhusker, 786 F.3d at 851. Lexington’s argument that it has a different insured does not concern jurisdiction. It concerns the merits, to which we turn next.
B. Merits Issues
1. Sharing the Loss
On the merits, the first issue is whether both insurers each bear some responsibility for the loss. Philadelphia contends Oklahoma law mandates that the insurers share the loss. Lexington argues it has no obligation to pay for three reasons. First, Lexington argues the two policies list different named insureds with different interests. Second, it argues the lease agreement between the two insureds — TSAS and the District — made Philadelphia primarily responsible for the loss. And third, Lexington argues the Philadelphia policy is more specific to the loss.
The district court granted summary judgment to Philadelphia. It properly applied Oklahoma insurance law in concluding the loss must be shared. Lexington’s arguments to the contrary are unpersuasive.
a. Standard of review
We review summary judgment de novo and apply the same legal standard as the district court. Comhusker, 786 F.3d at 849. A court “shall grant summary judgment if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a); see also Cornhusker, 786 F.3d at 850; Am. Cas. Co. v. Health Care Indem., Inc., 520 F.3d 1131, 1135 (10th Cir. 2008) (articulating same standard in Oklahoma insurance case).4 We also review legal questions de novo, including the district court’s inter[1337]*1337pretations of Oklahoma law, which the parties agree governs here. See Bird v. West Valley City, 832 F.3d 1188, 1199 (10th Cir. 2016). “Where the state’s highest court has not addressed the issue presented, the federal court must determine what decision the state court would make if faced with the same facts and issue.” Id. (quotations omitted).
b. Legal background
Under Oklahoma law, “[a]n insurance policy is to be treated as a contract” and is “enforced according to its terms.” Equity Mut., 747 P.2d at 953. “The whole of a contract is to be taken together, so as to give effect to every part, if reasonably practicable, each clause helping to interpret the others.” Okla. Stat. tit. 15, § 157 (2016). “The terms of the parties’ contract, if unambiguous, clear, and consistent, are accepted in their plain and ordinary sense, and the contract will be enforced to carry out the intention of the parties....” Dodson v. St. Paul Ins. Co., 812 P.2d 372, 376 (Okla. 1991). “The interpretation of an insurance contract and whether it is ambiguous is a matter of law for the Court to determine and resolve accordingly.” Id.
In Equity Mutual, the Oklahoma Supreme Court established the framework we apply here. That case featured a complex coverage situation where the loss stemmed from a tractor-trailer accident. 747 P.2d at 950 & nn.2-4. The owner of the tractor had one insurer that covered the tractor, and the permissive user of the tractor — that is, the driver — carried his own insurance for the trailer through a different insurer. See id. The case came to the Oklahoma Supreme Court as a series of certified legal questions from the federal district court and was decided on a set of stipulated facts. Id. The state supreme court answered the certified questions of law and left to the federal court the task of applying those answers to the case at hand.5
In answering the certified questions, the Oklahoma Supreme Court explained several important terms and features of Oklahoma insurance law.
First, the court referred to policies covering the same loss as “concurrent policies.” Id. at 949-50.
Second, the court discussed different coverage levels. An insurance policy provides primary coverage — as opposed to excess or secondary coverage, which kicks in only after exhaustion of the primary policy — “when, under the terms of the policy, the insurer is liable without regard to any other insurance coverage available.” Id. at 954; see also id. (“By definition ... primary coverage up to the limits of the policy will be applied to a loss before resort is had to any excess coverage.”). Coverage is excess or secondary “when, under the terms of the policy, the insurer is liable for a loss only after any primary coverage — other insurance — has been exhausted.” Id.
Third, the court defined particular clauses that can be found within some insurance policies. As relevant here, “A pro rata clause ... limits coverage to a proportionate share in relation to all coverage available for the same risk,” Id. And an excess-coverage clause provides that, if there is other primary coverage, the insurer will provide excess coverage only. Id.6
[1338]*1338In Equity Mutual, the court recognized that “[questions of conflict and apportionment of liability between or among insurers arise when more than one policy covers the same loss.” Id. Equity Mutual established the principle that, absent an alternative agreement, “when one or more policies provide primary coverage for the same loss, that loss shall be shared by the insurers.” Id. at 949.
Equity Mutual further addressed when concurrent policies each contain excess-coverage clauses, potentially leaving policyholders with “only excess coverage [and] no primary coverage.” Id. at 954. To avoid this result, Equity Mutual decided the excess-coverage clauses cancel. Id. In other words, when “concurrent policies” have excess-coverage clauses, Oklahoma holds the clauses “are mutually repugnant and are to be disregarded, with the loss shared by the insurers on a pro rata basis.” Id. The policies thus provide primary coverage by operation of law, and each insurer pays its pro rata share.
Under the pro rata method, “coverage of the loss is ... shared ... according to the ratio each respective policy limit bears to the cumulative limit of all concurrent policies.” Id.7 This default rule for apportioning the loss applies unless “the insurers have designated in their policies the same method of apportionment,” in which case “the contracts will control.” Id.8
Here, Philadelphia and Lexington have provided in their policies for the same method of apportionment. But, as it happens, they have both selected the default rule. We thus give effect to the parties’ pro rata clauses, a result that would be the same under Equity Mutual’s default method.
c. Analysis
The following analysis of the concurrent Philadelphia and Lexington policies concludes that their respective excess-coverage clauses cancel each other out and that their identical pro rata clauses require the two insurers share the loss to the Barnard building.
The “other insurance” provisions in the Philadelphia and Lexington policies are identical and each contains two clauses. The first clause of the “other insurance” provision in each policy provides that if the insured has other insurance, the insurer will pay only “the proportion that the applicable Limit of Insurance under this Coverage Part bears to the Limits of Insurance of all insurance covering on the same basis.” App., Vol. 1 at 125 (Philadelphia policy); App., Vol. 2 at 286 (Lexington policy). This is a pro rata clause. See Equity Mut.,14RV2d at 9M.
The second clause provides each insurer “will pay only for the amount of covered loss or damage in excess of the amount due from that other insurance.” App., Vol. [1339]*13391 at 125 (Philadelphia policy); App., Vol. 2 at 286 (Lexington policy) (emphasis added). This is an excess-coverage clause. Under Equity Mutual, because each policy provides excess coverage when there is a concurrent policy, the second clauses cancel. See Equity Mut,, 747 P.2d at 954.
Equity Mutual instructs that, after “disregard[ing]” the conflicting clauses, the insurers must share the loss “on a pro rata basis,” unless, that is, the policies contain “concurring provisions for apportionment” that provide otherwise, in which case, we follow the parties’ preferred method. Id. The policies here have concurring clauses for apportionment — the first clauses in each policy’s “other insurance” provision. We therefore give effect to the parties’ preference expressed in their identical first clauses for pro rata apportionment, which, as previously explained, tracks the default rule.
The district court conducted this same analysis and properly applied Oklahoma law when it declared the excess-coverage clauses mutually repugnant, disregarded them, and declared the insurers would share coverage of the loss on a pro rata basis. Phila. Indem., 2015 WL 8485249, at *3.
Our reading of Equity Mutual leads us to conclude the district court properly applied the principles and framework of that opinion here. Both policies insured the same building against fire damage, and each policy provides primary coverage by virtue of the conflicting excess-coverage clauses. See Equity Mut., 747 P.2d at 954 (“When concurrent policies have such ‘other insurance’ clauses which cancel each other, we hold that they are mutually repugnant and are to be disregarded, with the loss shared by the insurers_”); see also id. at 949 (“[W]hen one or more policies provide primary coverage for the same loss, that loss shall be shared by the insurers.”). Equity Mutual requires these insurers to share their common loss, in this instance in accordance with their pro rata clauses.
We find additional, persuasive support from Southern Insurance Co. v. Affiliated FM Insurance Co., 830 F.3d 337 (5th Cir. 2016), in which the Fifth Circuit, applying Mississippi law9 to facts similar to those here, held two “other insurance” clauses were mutually repugnant and ordered a pro rata distribution. Id. at 350-51.
A university’s insurance policy covered multiple campus buildings, including a house that the university rented to an alumni association. Id. at 340. The lease required the association to acquire an insurance policy for the house. Id. The association did so, but its policy did not list the university as an additional insured or as a loss payee, and the university’s policy did not name the association as an insured or a loss payee. Id. After a tornado damaged the house, the association’s insurer filed a declaratory judgment action. Id. at 341-42. The policies contained “other insurance” clauses that each “purported] to provide excess coverage.” Id. at 348. The court held the clauses canceled each other out and ordered a pro rata calculation. Id. at 349, 351. It reasoned:
To hold otherwise under these facts would be illogical.... [W]ere the clauses held not mutually repugnant, insurers [1340]*1340who covered the same risk for different insureds would be incentivized to take no action, out of fear that they would bear the entire loss. Practically speaking, such a result makes no sense.10
Id. at 350-51. The court found it critical that “the property, interest, and risk are identical: both policies insure the house for property damage, to the mutual benefit of the association and the university.” Id. at 350.
The rationale for sharing the loss is even stronger in this case than in Southern Insurance because here the District was not only the named insured under its policy with Lexington, but it was also the loss payee under the Philadelphia policy. Thus, not only did the policies insure the same building against the same risk, both policies redounded to the District’s benefit.
In sum, applying Equity Mutual to the excess-coverage clauses here, we conclude the district court did not err in granting Philadelphia’s motion for summary judgment to distribute the loss with Lexington pro rata.
d. Lexington’s arguments
Lexington argues the loss should not be shared for three reasons. None is persuasive.
i. Different insureds and interests
Lexington argues Philadelphia cannot “invoke” the “other insurance” provisions. Aplt. Br. at 20. It asserts the district court erred because Equity Mutual’s clause-can-celation rule “only applies if the policies provided by each insurer cover the same insured, same interest and same risk.” Id. Lexington contends the excess-coverage clauses cannot cancel here because its named insured, the District, is different from Philadelphia’s named insured, TSAS, and these different insureds have different interests.
We disagree. Both policies protected the Barnard building against the risk of fire damage. As to different named insureds, Lexington fails to cite any Oklahoma case holding that the policies must list the same insured for the insurers to share the loss. Indeed, the insurance policies in Equity Mutual covered different named insureds. See Equity Mut, 747 P.2d at 954 (discussing loss sharing where policies cover the “same loss”). And in this case, both policies protect the District’s interests. The District is the named insured under the Lexington policy, and it is the loss payee under the Philadelphia policy. The District receives the payout regardless of which policy is invoked.11 Lexington asks us to overlook this economic reality.
[1341]*1341Lexington’s brief discusses three principal cases in which courts refused to split a loss between insurers whose policies listed different insureds and covered the same loss. None of these cases, however, involved a named insured that was also the loss payee under another policy covering the same loss. Lexington emphasizes the courts’ discussion of how the policies in those cases covered different interests. It argues that the policies here protected different interests because the District was a landlord and TSAS was a tenant.
These cases are distinguishable and do not persuade us. In addition to the fact that none applied Oklahoma law, they are all distinguishable from this case because the District is both a named insured and a loss payee under the policies here. And the landlord-tenant distinction is unconvincing because, although landlords and tenants have different interests, it does not necessarily follow that their insurers cannot share coverage of a loss to the insureds’ common property — especially when the beneficiary of any payout from either insurer will be the same payee.
First, Lexington leads with Society Insurance v. Capitol Indemnity Corp., 260 Wis.2d 549, 659 N.W.2d 875 (App. 2003), in which a restaurant burned down and the tenant (the restaurateur) filed a claim with his insurer. Id. at 877-78. The tenant’s insurer paid the claim and then sought contribution from the landlord’s insurer, who covered the same building against fire damage. Id. at 878. The Wisconsin court held contribution was unavailable because there was no identity of insureds and because the policies covered “separate and distinct insurable interests.” Id. at 879-81. The court went on to explain, “If the property is destroyed, the [landlord] loses both the rental income and the value of the real estate, whereas the tenant loses income from the business, goodwill and other items related to the business.” Id. at 881.
The Society Insurance court did not explain why a difference in the interests — fee versus leasehold — should lead to one insurer bearing the whole loss. More important for our purposes, Society Insurance is distinguishable because each policy paid out to its sole insured. Here, the District receives the proceeds under both policies.
Second, Lexington cites St. Paul Fire & Marine Insurance Co. v. Protection Mutual Insurance Co., 607 F.Supp. 388 (S.D.N.Y. 1985) (applying New York law), adhered to on rehearing 644 F.Supp. 38 (S.D.N.Y. 1986). In St. Paul, a fire swept through a commercial building in Manhattan. See id. at 389-90. The owner carried fire insurance on the building, and the commercial tenant had a fire insurance policy of its own, “which covered stock, inventory and other personalty as well as ... the betterments and improvements.” Id. at 389. The insurer for the building owner paid for the building repairs, and the tenant’s insurer paid the personal property losses of its insured.' Id. at 390. The court rejected the landlord-insurer’s attempt to split the cost of the building repairs with the tenant’s insurer because “the two policies covered different insureds, moreover the insurable interests are clearly different. [The tenant’s] interest was a leasehold while [the owner’s] was in fee.” Id. at 391.
[1342]*1342Similarly to Society Insurance, the St. Paul opinion did not explain its denial of a split beyond identifying the distinct interests of landlords and tenants and noting the policies had different named insureds. Id. at 390-91. Also like Society Insurance, St. Paul is distinguishable: Here both policies cover the same building and both policies pay out to the District (the landlord). The District’s interest is protected under both the Philadelphia and Lexington policies notwithstanding the different names of the insureds listed on the' policies.
Finally, Lexington points to Reliance Insurance Co. v. Liberty Mutual Fire Insurance Co., 13 F.3d 982 (6th Cir. 1994). There, the two insureds were a property owner and the construction company the owner enlisted to build an apartment complex on the property. Id. at 983. The complex suffered fire damage while under construction, and each insured reported claims to its respective insurer. Id. The owner’s insurer paid the claim, and the contractor’s insurer paid no part of the loss. Id. The owner’s insurer sued for indemnification or, alternatively, contribution. Id. The Sixth Circuit, applying Michigan law, affirmed a grant of summary judgment against the owner’s insurer. Id. at 985. The court reasoned that, although the property and risk were the same, the insured interests were different. Id. at 983. The owner’s insurance policy covered its “ownership interest in the complex,” but the contractor’s insurance policy “protect[ed] only [the builder’s] contractual interests in completing and delivering the complex.” Id.
As in the other cases Lexington cites, Reliance did not have, as we have here, an insured that was a loss payee under one policy and a named insured under another. Under the arrangement here, the District’s interest is protected under both policies.12
Lexington would have one insurer pay a loss in full when multiple insurers are responsible for primary coverage under policies with different named insureds. That result is at odds with our reading of Equity Mutual. And to the extent Lexington contends different insurable interests prevent loss splitting here, that argument fails because both policies protect the District’s interest.
ii. The TSAS-District lease
Lexington also argues it should not have to share coverage of the loss “[b]ecause TSAS and the School District had an agreement that TSAS would procure the property insurance.” Aplt. Br. at 26. This agreement between the insureds, Lexington contends, made Philadelphia the primary insurer.13 We disagree because the insurance policies, not the lease, control.
[1343]*1343Lexington mistakenly relies on a passage from Equity Mutual, The Oklahoma Supreme Court “h[e]ld that absent an agreement between the owner of a commercial vehicle and a permissive user, when one or more policies provide primary coverage for the same loss, that loss shall be shared by the insurers.” 747 P.2d at 949 (emphasis added). Based on this passage’s reference to the absence of an agreement, Lexington insists we should enforce the lease agreement between TSAS and the District, which, according to Lexington, stated that TSAS’s insurer — Philadelphia — would supply primary coverage.
Lexington’s argument selectively quotes and thereby oversimplifies Equity Mutual. Lexington ignores the Oklahoma Supreme Court’s statement that “[a] private agreement cannot expand the terms of an insurance policy” and that any consideration of an agreement between insureds “must be applied conformably to the general principle that contractual obligations cannot be expanded by agreements with strangers to the contract.” Id. at 955. The court further explained that an outside agreement “may not be invoked when the policy so secured provides only pro rata coverage or has a conflicting ‘other insurance’ clause.” Id. In that circumstance, “the loss will be shared on a pro rata basis.” Id.
The Philadelphia-TSAS insurance contract contains an “other insurance” provision that includes a pro rata clause and an excess-coverage clause. Lexington in essence asks us to allow the lease to amend the Philadelphia policy to delete the excess-coverage clause. We reject that invitation. See id.; see also Travelers Ins. Cos. v. Dickey, 799 P.2d 625, 628 (Okla. 1990) (“An insurer’s undertaking cannot be altered or modified by an insured’s agreement with a third party in the absence of the insurer’s consent.”).
The Fifth Circuit rejected a similar argument in Southern Insurance, stating “there [was] no reason to consider the lease” between the university and the alumni association because “nothing about the other-insurance clauses in either policy [was] ambiguous.” 830 F.3d at 351.
Lexington’s argument fails because we have no reason to look to the lease. Philadelphia’s policy unambiguously disclaims primary coverage in the presence of “other insurance.” So does Lexington’s. And under Equity Mutual, these conflicting excess-coverage clauses cancel. See 747 P.2d at 954. The district court did not err by ordering pro rata apportionment.14
iii. “More specific” coverage
Lexington posits that Philadelphia’s policy “is more specific to the risk and[ ] thus primary.” Aplt. Br. at 19 (capitalization altered). By “more specific,” Lexington observes that the Philadelphia policy covered [1344]*1344only the Barnard building, whereas Lexington’s “blanket” policy covered over 100 sites owned by the District, including the Barnard school. Id. This fact does not supply a rationale for Lexington to avoid its responsibility to pay.
First, “[w]hether a particular policy is ‘blanket’ has nothing to do with the number of buildings covered.” See Lexington Ins. Co. v. Peerless Ins. Co., No. 09-CV-500-GKF-TLW, 2010 WL 2079706, at *2 (N.D. Okla. May 21, 2010) (rejecting the same argument from Lexington).
Second, Lexington does not ground this argument in Oklahoma law. Its only authority is Holden v. Connex-Metalna Management Consulting GmbH, 302 F.3d 358 (5th Cir. 2002) (applying Louisiana law). But Holden acknowledged that its outcome was contrary to the rule applied in most jurisdictions — including Oklahoma — under which “policies that cover the same risks with respect to the damaged property must be treated as ‘concurrent insurance’ — even if the policies do not cover an identical set of property.” Id. at 366 n.11 (contrasting Indus. Indem. Co. v. Cont’l Cas. Co., 375 F.2d 183, 185-86 (10th Cir. 1967)).
In Industrial Indemnity, which Holden cited as an example of the majority rule, we applied Oklahoma law and affirmed a pro rata apportionment of a loss arising from a vehicle accident at an oil well. 375 F.2d at 184-86. Industrial Indemnity implicated two insurers. One covered all of the contractor’s activities at the well site; the other insured the oil company for losses stemming from automobiles. Id. at 184. Over the objection of the former, we affirmed the trial court’s proration of the loss and concluded Oklahoma law did not support resorting first to the more “specific” automobile policy when both policies were applicable. Id. at 185-86.
We adhere to Industrial Indemnity and note that the district court in another case recently rejected this same argument from Lexington. See Peerless Ins., 2010 WL 2079706, at *2. The Philadelphia and Lexington policies each provide primary coverage because, as explained in Equity Mutual, the policies’ excess-coverage clauses cancel. Each policy specifically protected the Barnard building against fire loss. The Philadelphia policy is no more “specific” than the Lexington policy.
[[Image here]]
We reject Lexington’s arguments and affirm the district court’s grant of summary judgment to Philadelphia for sharing the loss.
2. Pro Rata Apportionment
Having concluded the insurers must share the loss, the second merits issue concerns how the loss must be shared. “When the relevant facts are undisputed, we review the district court’s interpretation of an insurance contract de novo.” Houston Gen. Ins. Co. v. Am. Fence Co., Inc., 115 F.3d 805, 806 (10th Cir. 1997) (applying Oklahoma law). “The interpretation of an insurance contract is governed by state law....” Id.
a. Legal background
Because the policies have mutually repugnant excess-coverage provisions, we give effect to the policies’ identical pro rata clauses. See Equity Mut., 747 P.2d at 954 (“Where the insurers have designated in their policies the same method of apportionment, the contracts will control.”).15
[1345]*1345Pro rata distribution requires apportioning the loss “according to the ratio each respective policy limit bears to the cumulative limit of all concurrent policies.” Id. Equity Mutual provided an example of how this works involving three insurers: “[I]f one insurer has a policy limit of $100,000, another $200,000 and a third $300,000, the first would pay ⅛ of the loss up to $100,000, the second would pay ⅝ of the loss up to $200,000 and the third would pay ⅛ the loss up to $300,000.” Id. Earlier in this opinion we provided the mathematical steps for the district court’s pro rata apportionment in this case. See supra 1333-34.
b. Additional factual background
Although the parties agree Philadelphia’s policy limit is $7 million, they dispute the relevant limit for the Lexington policy. The dispute centers on the “Occurrence Limit of Liability Endorsement” found in the Lexington policy. The Endorsement consists of two paragraphs, the first of which provides in relevant part:
1. The limit of liability or Amount of Insurance shown on the face of this policy, or endorsed on this policy, is the total limit of the Company’s liability applicable to each occurrence. ... Notwithstanding any other terms and conditions of this policy to the contrary, in no event shall the liability of the Company exceed this limit or amount irrespective of the number of locations involved.
App., Vol. 2 at 299. As stated on the first page of the policy, Lexington’s overall
“[l]imit of [i]nsurance” is “$100,000,000,, PER ANY ONE OCCURRENCE.” Id. at 282.
The Endorsement’s second paragraph contemplates that the relevant limit could be lower. It states:
2. In the event of loss under the policy, the liability of the Insurer(s) shall be limited to the least of the following:
a.) The actual adjusted amount of loss, less applicable deductible(s);
b.) [NOTE: the policy contains language deleting paragraph 2(b), which the parties agree is inapplicable here.16] [; or]
c.) Any other Limit of Liability or Sublimit of Insurance or Amount of Insurance specifically stated in this policy to apply to any particular insured loss or coverage or location.
Id. at 299 (emphasis added). The district court relied on this second paragraph of the Endorsement to hold Lexington’s relevant limit was the actual adjusted amount of the loss, or $6,014,359.06, because using the $100 million limit would require it to ignore the Endorsement. Id. at 459.
c.Analysis
In its cross-appeal, Philadelphia argues the district court erred in setting Lexington’s policy limit at $6,014,359.06 (the amount of the loss) instead of $100 million (Lexington’s overall policy limit). Using Philadelphia’s approach — which relies on the same steps as the district court’s method except that it substitutes $100 million for Lexington’s policy limit — Philadelphia [1346]*1346would bear 6.54 percent of the loss,17 and Lexington would pay the remaining 93.46 percent,18 or $5,621,019.98. Lexington argues that, if it must share pro rata, the district court properly selected $6,014,359.06 as its relevant policy limit. This issue thus comes down to the selection of one number — the relevant Lexington policy limit — and this turns on how to interpret the Endorsement to the Lexington policy. We agree with Lexington and the district court.
We must identify the relevant Lexington policy limit because the pro rata calculation requires determining the relevant limit for each policy. The pro rata clause in the Lexington policy (like the Philadelphia policy) states that its share of the loss “is the proportion that the applicable Limit of Insurance” represents in terms of the total available insurance. Id. at 286 (emphasis added); see also Equity Mut., 747 P.2d at 954 (providing for pro rata apportionment “according to the ratio each respective policy limit bears to the cumulative limit of all concurrent policies.”). To identify Lexington’s “applicable [l]imit of [i]nsurance,” we look to the policy as a whole, including the Endorsement. See Okla. Stat. tit. 15, § 157 (2016) (“The whole of a contract is to be taken together, so as to give effect to every part, if reasonably practicable, each clause helping to interpret the others.”).
The parties agree Lexington’s policy is a blanket policy.19 That is, the $100 million amount is available for every covered loss; the policy is not scheduled20 such that different covered items have specific sub-limits that together total $100 million. In addition, the parties do not dispute that (1) the first paragraph of the Endorsement refers to the $100 million limit, (2) the “adjusted amount of loss” language in paragraph 2(a) equates to the $6,014,359.06 loss here, and (3) paragraph 2(b) is inapplicable. The parties further agree that under paragraph 2(c) there is no relevant “Sub-limit of Insurance.”
The Lexington policy (1) declares that its “Limit of Insurance” is “$100,000,000 PER ANY ONE OCCURRENCE,” App., Vol. 2 at 282, and (2) states that, when Lexington shares coverage of a loss with another insurer, its pro rata share depends on its “applicable Limit of Insurance,” id. at 286. Looking only at these two provisions, it appears $100 million is the proper number to use in calculating Lexington’s proportional payment for the loss. But the Endorsement in Lexington’s policy must also be considered, and it more specifically restricts the limit in a given case.
The first paragraph of the Endorsement states that “[t]he limit of liability or Amount of Insurance shown on the face of this policy, or endorsed on this policy, is the total limit of the Company’s liability applicable to each occurrence, as hereafter defined.” Id. at 299. This sentence equates “limit of liability” with “Amount of Insurance shown on the face of this policy,” which is the $100 million “Limit of Insurance” that appears on the first page of the [1347]*1347policy.21 Philadelphia recognizes that “[t]he $100,000,000 policy limit is the subject of paragraph 1 of the [Endorsement].” Aplee. Reply Br. at 3.
The second paragraph of the Endorsement states that Lexington’s “liability ... shall be limited to the least of’ the listed alternatives. App., Yol. 2 at 299. The alternatives include the adjusted amount of the loss (paragraph 2.a) and “[a]ny other Limit of Liability” (paragraph 2.c), which, based on the first paragraph, is $100 million. Applying these provisions, Lexington’s limit of liability for the pro rata calculation is $6,014,359.06.22 Lexington argues the district court correctly identified the adjusted amount of the loss as the relevant limit because applying a $100 million limit would render the Endorsement, and specifically paragraph 2(a), meaningless. It contends that paragraph 2(a) of the Endorsement concerns the actual amount of the loss and that, since under paragraph 2(c) there is no other applicable sublimit, the only other “Limit of Liability” paragraph 2(c) can refer to is the overall $100 million limit. Lexington concludes this sets up an easy choice between paragraph 2(a) and 2(c): Because the actual amount of the loss— $6,014,359.06 — is less than $100 million, the language in paragraph 2 stating the liability “shall be limited to the least of the following” mandates the application of paragraph 2(a) and the use of the $6,014,359.06 amount.
Philadelphia responds that the Endorsement “has no effect at all” on Lexington’s pro rata share. Aplee. Reply Br. at 2. Philadelphia agrees that no other sublimits are implicated under paragraph 2(c), but Philadelphia stops there. In Philadelphia’s view, 2(c) does not apply at all because there is no applicable sublimit. As for paragraph 2(a), Philadelphia argues it “simply protects Lexington from paying more than [the] actual adjusted amount of loss, less any applicable deductibles.” Id. at 3. That is, paragraph 2(a) is not a policy limit that should concern us when performing the pro rata calculation; it is only a limit on what Lexington can wind up paying in the end. Philadelphia goes on to explain that using the $100 million limit does not contravene paragraph 2(a) or require ignoring the Endorsement because, although Lexington would pay for nearly everything, 93 percent of the loss is still less than “the actual adjusted amount of the loss.” Philadelphia contends this is all that is required to satisfy paragraph 2(a).
We agree with Lexington that the $6,014,359.06 limit used by the district court is the applicable limit. First, Phila[1348]*1348delphia misreads paragraph 2(c). It ignores the paragraph 2(c) terms “other Limit of Liability” or “Amount of Insurance specifically stated in this policy to apply to any particular insured loss or coverage or location,” which recall the terms “[L]imit of [Liability” or “Amount of Insurance” in the Endorsement’s first paragraph concerning Lexington’s “liability applicable to each occurrence.” App., Vol. 2 at 299. Thus, on these facts, where there is no applicable sublimit, paragraph 2(c) ends up referring to the limit from the Endorsement’s first paragraph — $100 million. The choice is thus between the amount of the loss and $100 million. And, because paragraph two of the Endorsement specifies that “the least of’ the listed limits should apply, the $6,014,359.06 limit from paragraph 2(a), and not 2(c)’s $100 million limit, is the relevant limit.
But even if Philadelphia is right that paragraph 2(c) does not refer back to the Endorsement’s first paragraph and its $100 million limit — that is, even if paragraph 2(c) does not supply any limit and has no bearing here — the only limit left in play under paragraph 2 would be paragraph 2(a)’s limit equaling “the actual adjusted amount of loss.” The relevant policy limit would thus still be $6,014,359.06.
And we reject Philadelphia’s argument that paragraph 2(a) merely protects Lexington from a post-calculation outlay exceeding the total adjusted amount of the loss.23 This argument fails to treat paragraph 2(a) as the policy limit that it is. Philadelphia would look to paragraph 2(a) only after applying the $100 million limit in the calculation, but that approach ignores the Endorsement when it counts most— selecting the policy limit for the pro rata calculation. Paragraph 2(a) is one of several potential policy limits identified in the Endorsement. See App., Vol. 2 at 299 (stating “the liability of [Lexington] shall be limited to the least” of the amounts listed in the policy Endorsement (emphasis added)). Indeed, the Endorsement bears the name “Occurrence Limit of Liability Endorsement.” Here, paragraph 2(a) supplies the lowest limit, and therefore, following the Endorsement’s clear terms that the lowest limit applies, $6,014,359.06 supplies the limit needed for the pro rata calculation, which is the limit the district court used.
Philadelphia cites cases from other jurisdictions to argue for a contrary result, but the only case it cites applying Oklahoma law is our decision in American Casualty Co., 520 F.3d at 1135-36, where we approved a pro rata distribution that left one insurer with % of the loss and the other with the remaining ⅛. But American Casualty Co. does not advance Philadelphia’s argument for using the $100 million limit. The case is not instructive because the policy limits there were not in dispute. See id. at 1136. And none of Philadelphia’s other cases persuade us to disregard the language of Lexington’s Endorsement. See Okla. Stat. tit. 15, § 154 (2016) (“The language of a contract is to govern its interpretation, if the language is clear and explicit, and does not involve an absurdity.”).
As the district court observed, using the $100 million limit would require us to overlook the Endorsement. We apply the En[1349]*1349dorsement as written and give effect to paragraph 2(a). See Okla. Stat, tit. 15, § 157 (2016) (instructing courts to “give effect to every part” of a contract “if reasonably practicable”). Accordingly, we affirm the district court’s pro rata loss allocation ordering Lexington to pay 46.21 percent of the loss and Philadelphia to pay the other 53.79 percent.
III. CONCLUSION
For the foregoing reasons, we affirm the district court’s judgment.