Before Smith, Clement, and Wilson, Circuit Judges. Edith B. Clement, Circuit Judge: Clay and Lindsay Buchholz own a large ten-thousand-square-foot house in Austin, Texas. They insured their home with Crestbrook Insurance Company. Their policy included “Biological Deterioration or Damage Clean Up and Removal” coverage (“mold coverage”). In April 2019, the Buchholz family discovered a widespread mold infestation in their home. Although Crestbrook covered many of their losses, it denied a generalized claim for mold growing in the Buchholzes’ walls and heating, ventilation, and air conditioning system. On cross-motions for summary judgment, a magistrate judge issued a report and recommendation in favor of Crestbrook, and the district court adopted the magistrate judge’s conclusions. We AFFIRM. I A The crux of this appeal is whether the optional mold coverage the Buchholzes purchased in their Crestbrook policy, which provided $1.6 million in mold damage insurance in exchange for $4,554.53 in additional premiums, covers a generalized mold loss. In April 2019, the Buchholzes discovered mold growing inside the wall of their indoor basketball court when a ball accidentally smashed through it. They hired a contractor to inspect the home, and his inspection stumbled upon a mold infestation that affected the entire house. The Buchholzes submitted six claims to Crestbrook. It paid $745,778 in covered losses on five of the claims. However, Crestbrook sent a reservation of rights letter regarding the sixth claim for general mold growth and mold in the HVAC system. In response to that letter, the Buchholz family retained MLAW Forensics, Inc., to investigate the cause of their mold infestation. Crestbrook agreed to pay for the costs of MLAW’s investigation. Dean R. Read, P.E.,[1] wrote a causation report on what he concluded led to the mold growth at the Buchholzes’ house. His report found that “discrete leaks and a ‘global’ issue due to interruption or restriction of the moisture vapor drive drying process” caused the mold. Specifically, he concluded that the house’s HVAC system was “[i]mproperly designed or configured and non-functional,” which resulted in “elevated moisture content[]” and subsequent mold growth. Based on MLAW’s causation report and a supplemental investigation by Tom Green, P.E., on the malfunctioning HVAC system, Crestbrook denied Appellants’ mold claim. The denial letter stated: As noted in the MLAW Forensics report, workmanship and construction issues were discovered with the HVAC system, wall paint, and paneling and trim. Your policy contains exclusions for biological deterioration or damage, a defect or inadequacy in design, workmanship, construction and materials. In addition, the policy contains exclusions for weather conditions or dampness, and gradual or sudden loss due to a mechanical breakdown. Given the aforementioned exclusions, the biological deterioration or damage additional limited coverage would not apply to this claim. B On March 13, 2020, the Buchholzes sued Crestbrook in Travis County District Court, alleging their insurer wrongfully denied their sixth claim. Crestbrook timely removed the case to federal court. In their final complaint, the Buchholz family alleges that Crestbrook breached their insurance contract in bad faith and violated the Texas Insurance Code. They sought a declaratory judgment that Crestbrook’s policy covered their mold claim, an award of actual and exemplary damages, attorney’s fees, and costs. At the district court, the parties filed cross-motions for summary judgment. The Buchholzes moved for a partial judgment on their bad faith breach of contract claim and asked the court to construe the various exclusions found in the policy in their favor. Crestbrook counter-moved for entry of summary judgment, arguing that the Buchholz family failed to show they suffered from a covered cause of loss and that provisions in the policy excluded their claims. The magistrate judge handling the case recommended the district court rule in Crestbrook’s favor, concluding that the Buchholz family had failed to demonstrate a “covered cause of loss” as required by their mold coverage. The district court summarily adopted the magistrate judge’s report and recommendation and entered judgment. The Buchholz family timely appealed. II “This court reviews a district court’s grant of summary judgment de novo, applying the same legal standards as the district court.” Tradewinds Env’t Restoration, Inc. v. St. Tammany Park, LLC, 578 F.3d 255, 258 (5th Cir. 2009) (quoting Condrey v. SunTrust Bank of Ga., 429 F.3d 556, 562 (5th Cir. 2005)). “Summary judgment is appropriate when ‘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.’” United States v. Nature’s Way Marine, L.L.C., 904 F.3d 416, 419 (5th Cir. 2018) (quoting Fed. R. Civ. P. 56(a)). The parties agree that this case is in federal court based on diversity jurisdiction and Texas law controls. III A The Buchholzes argue that the magistrate judge improperly required them to demonstrate a discrete cause of covered loss within the framework of an all-risks insurance policy. The Buchholz family says that the only way they could have met this burden would be by inverting the Texas law standard and requiring them to disprove every exclusion found in the underlying policy. Because of these errors, they request that we reverse the district court’s entry of summary judgment and remand the case for trial. Under Texas law, when deciding a dispute regarding insurance coverage, we first look to “the language of the policy because we presume parties intend what the words of their contract say.” Gilbert Tex. Constr., L.P. v. Underwriters at Lloyd’s, London, 327 S.W.3d 118, 126 (Tex. 2010). We give the policy’s words “their ordinary and generally-accepted meaning unless the policy shows the words were meant in a technical or different sense.” Id. Insurance policies are contracts construed “according to general rules of contract construction to ascertain the parties’ intent.” Id. We also “examine the entire agreement and seek to harmonize and give effect to all provisions so that none will be meaningless.” Id. “Our analysis of the policy is confined within the four corners of the policy itself.” State Farm Lloyds v. Page, 315 S.W.3d 525, 527 (Tex. 2010). Whether an insurance contract is ambiguous is a question of law. Id. A disagreement between the parties regarding the meaning of policy terms or interaction between terms does not create ambiguity. Id. We may consider the policy ambiguous only if it is subject to two or more reasonable interpretations. Id. Where we determine policy language to be ambiguous, “we must resolve the uncertainty by adopting the construction that most favors the insured.” Nat’l Union Fire Ins. Co. of Pittsburgh, Pa. v. Hudson Energy Co., 811 S.W.2d 552, 555 (Tex. 1991). In a coverage dispute, the insured has the burden first to prove that their loss falls within the terms of the contract. Gilbert, 327 S.W.3d at 124 (citation omitted). Once the insured demonstrates this, the burden shifts to the insurer, who, to avoid liability, must show that the loss falls into an exclusion to the policy’s coverage. Id.; see also Tex. Ins. Code§ 554.002. Finally, “[i]f the insurer proves that an exclusion applies, the burden shifts back to the insured to show that an exception to the exclusion brings the claim back within coverage.” Gilbert, 327 S.W.3d at 124. The magistrate judge correctly laid out the Texas insurance dispute burden-shifting framework in her report and recommendation. 4392. When analyzing the Buchholz family’s insurance contract, she concluded: [The Buchholzes] fail to identify the cause of the mold damage. Instead, [the Buchholzes] submit that the Policy is an inclusive, all risk policy that covers all-risk of accidental direct physical loss to the property unless an exception applies . . . . [The Buchholzes] skip the essential step of showing that the mold damage was caused [by] a covered cause of loss. The burden does not shift to Crestbrook to show an exclusion applies until [the Buchholzes] first show the mold damage resulted from a covered loss. [The Buchholz family] ha[s] neither identified a cause of the mold growth nor shown that a cause was a covered loss. Accordingly, [the Buchholzes] fail to meet their burden to show that the Mold Claim is covered under the [mold coverage] provision The Court need not address whether an exclusion applies because [the Buchholzes] fail on their first burden. In support of the magistrate judge’s conclusions, Crestbrook asks us to see the mold coverage as an independent coverage that stands apart from the all-risks coverage in the primary insurance contract. Therefore, says Crestbrook, the Buchholzes must show that their mold loss fell within the mold coverage’s terms. That provision provides: Biological Deterioration or Damage Clean Up and Removal. In the event that a covered cause of loss results in Biological Deterioration or Damage to property covered under Coverages A., B., and C., we will pay, up to the amount shown on the Declarations . . . . In Crestbrook’s telling, the Buchholz family needed to establish a “covered cause of loss” to trigger their mold coverage. But according to the insurer they have failed to do so, and the district court correctly entered summary judgment. This seems logical as far as it goes. But Crestbrook glosses over an important step (as does the report and recommendation). If we accept the insurer’s argument that the mold coverage provision is separate from the rest of this all-risks policy, what is a “covered cause of loss”?[2] The Buchholzes reasonably suggest that in the absence of other contractual language, the underlying policy’s coverage of “any cause of loss resulting in accidental direct physical loss, except such causes as are excluded under the policy” should be considered a “covered cause of loss” under the mold coverage. Crestbrook does not provide its own definition for “covered cause of loss.” Instead, it argues that even if the report and recommendation skipped some analytical steps, the results are the same once the analysis is done correctly. However, we reject Crestbrook’s invitation to follow the report and recommendation’s lead and abridge the framework required by the Supreme Court of Texas. See Gilbert, 327 S.W.3d at 126. Simply put, the report and recommendation erred in its application of the Texas insurance dispute burden-shifting scheme. The mold coverage provision does not define “covered cause of loss,” rendering the term open to more than one reasonable interpretation. It is an ambiguous contractual term. See McDonnel Grp., L.L.C. v. Starr Surplus Lines Ins. Co., 15 F.4th 343, 346 (5th Cir. 2021) (“A policy provision is ambiguous only if it is susceptible to two or more reasonable interpretations.”) (quotation marks and citation omitted). So, the report recommendation should have deferred to the Buchholzes’ reasonable interpretation of the policy language. Nat’l Union Fire Ins., 811 S.W.2d at 555 (“[W]e must resolve the uncertainty by adopting the construction that most favors the insured.”). The Buchholz family’s proposed definition—that we should read “covered cause of loss” to mean “any cause of loss resulting in accidental direct physical loss, except such causes as are excluded under the policy”—is a reasonable adaptation of the underlying all-risk policy coverage.[3] As a result, we adopt this interpretation as the correct meaning of “covered cause of loss” within the context of the mold coverage, and we agree with the Buchholzes that the district court made an error by finding mold damage is not a direct physical loss. The report and recommendation then compounded its misapplication of “covered cause of loss” by concluding that the district court should enter summary judgment without first analyzing the exclusions and exceptions to the exclusions found in the policy. Crestbrook does not contend that mold damage is not a direct physical loss. Instead, Crestbrook argues it is an excluded loss. Consequently, the report and recommendation should have concluded that the Buchholzes suffered a direct physical loss covered by the terms of their insurance contract. The magistrate judge then needed to complete the Texas burden-shifting insurance coverage analysis by engaging with the parties’ arguments regarding exclusions and exceptions to the exclusions. In failing to do so, she incorrectly applied Texas insurance law. B Although the district court adopted an incorrect application of Texas insurance law, we agree with its ultimate judgment. We can affirm the district court on any grounds supported by the record on appeal. Sobranes Recovery Pool I, LLC v. Todd & Hughes Constr. Corp., 509 F.3d 216, 221 (5th Cir. 2007). In its motion for summary judgment, Crestbrook argued that a mold infestation is an excluded peril under the policy. So, contends Crestbrook, without another covered peril that caused the fungal growth, the mold coverage does not act as an exception to the “Biological Deterioration and Damage” exclusion (“mold exclusion”). Applying the Texas insurance burden-shifting framework, we agree with Crestbrook that the mold exclusion bars coverage for the Buchholz family’s claim. The MLAW causation report concluded that the most likely source of the mold was excess moisture in the walls due to a disruption of the “vapor drive drying process.” According to the report, the “vapor drive” or “moisture drive” is a natural process in which “water vapor migrates through the building . . . from a warmer[,] higher humidity” area to a “cooler[,] lower humidity” location “based on the rules of thermodynamics.” The report explained that polyester netting in the Buchholzes’ walls trapped particles that acted as a “food source” for the fungus. It described how the mold then relied on elevated moisture caused by a dysfunctional HVAC system to consume those food sources. In the MLAW engineer’s opinion, the HVAC system utilized two undersized and one significantly oversized air conditioning units, which failed to provide “the necessary dehumidification.” The interior paint, paneling, and trim further elevated humidity, preventing water vapor from moving freely through the home. Supplementing the MLAW report, Tom Green, P.E., concluded that the building was likely kept at too low a temperature, exasperating the house’s vapor drive issues. The Buchholzes challenged these conclusions with the help of another expert, Sean O’Brien, P.E., who wrote, in his sworn declaration, that the Buchholz family’s home was exposed to significant water intrusion that the vapor drive process and a faulty HVAC system could not explain. He found that for the MLAW report to be correct, the interior of the Buchholzes’ home would have had to be consistently at eighty degrees with 80% humidity— physically uncomfortable circumstances not attested to in the record. O’Brien also countered the claims made by Green, simulating the effects of keeping the temperature excessively low in the house. His model indicated that the Buchholz family should not have experienced such a severe mold infestation even at sixty-five degrees and fifty percent humidity. Finally, O’Brien disagreed that the paint, trim, or paneling influenced the vapor drive process in a way that promoted widespread mold growth. The O’Brien declaration did not state any specific reason for the mold infestation, only noting that the walls had been exposed to “significant water intrusion.” Under the Texas insurance dispute framework, the Buchholzes must first show a direct physical loss as required under their all-risk policy. Then Crestbrook can identify any exclusions to coverage of that loss. Finally, the Buchholzes may attempt to establish an exception in favor of coverage that applies to the identified exclusion. The Buchholz family points to the mold infestation as a direct physical loss. Crestbrook does not dispute that this is a direct physical loss. The burden thus shifts to the insurer to show the policy excludes the claim. Crestbrook argues that the mold and other property exclusions bar coverage for a generalized mold claim. Because the mold exclusion is dispositive in the insurer’s favor, we focus on how it excludes the Buchholz family’s claim. The policy excluded coverage for “loss to any property resulting directly or indirectly from any of the following . . . Biological Deterioration or Damage, except as provided by [the mold coverage].”[4] By Crestbrook’s telling, although O’Brien attacks the MLAW and Green reports’ conclusions, he fails to identify a covered peril that resulted in fungal growth as required for the mold coverage to come into effect. Thus, says the insurer, the generalized mold claim must fail under mold exclusion. The Buchholzes respond, saying they have argued all along that excess “water within the walls of their home” caused the mold. Countering the Buchholz family’s “water within the walls” theory, Crestbrook cites Aetna Casualty & Surety Company v. Yates, 344 F.2d 939, 941 (5th Cir. 1965). In Yates, we found that an ensuing-loss provision in an all-risk policy with a mold exclusion did not cover mold damage resulting from unspecified excessive moisture inside a structure. We said, “[mold damage] may have ensued from water but not from water damage, and the damage ensuing from the rot was not the damage from the direct intrusion of water conveyed by the phrase ‘water damage.’” Id. Because mold can only result from excessive moisture in a building, we decided that generalized water intrusion could not be a covered peril when its primary effect was to cause a fungal infestation. Id. We reasoned to do otherwise would gut the mold exclusion. Id. The Supreme Court of Texas adopted our logic in Fiess v. State Farm Lloyds, 202 S.W.3d 744, 750–51 (Tex. 2006). The court wrote, “Surely [the Fifth Circuit] was correct. Mold does not grow without water; if every leak and drip is ‘water damage,’ then it is hard to imagine any mold, rust, or rot excluded by this policy, and the mold exclusion would be practically meaningless.” Id. at 276. Certainly, the case before us is slightly different in that Yates and Fiess evaluated an ensuing-loss provision rather than mold coverage. But their logic persuades us. The Supreme Court of Texas in Fiess noted that excess moisture in the walls is not “water damage” under an all-risks policy. Id. It follows that the ensuing-loss exception does not cover any resulting mold when faced with a mold exclusion. Id. We believe this reasoning is equally applicable to the Buchholzes’ mold coverage. To classify water intrusion as the covered peril underlying a generalized mold claim, the Buchholzes needed to identify “something more substantial than” excess water within their walls. Id. at 751. Ruling to the contrary would cause the mold coverage to completely nullify the mold exclusion, an outcome Texas law instructs us to avoid. Gilbert, 327 S.W.3d at 126 (We “examine the entire agreement and seek to harmonize and give effect to all provisions so that none will be meaningless.”); see also Fiess, 202 S.W.3d at 751 (“Mold does not grow without water; if every leak and drip is ‘water damage,’ then . . . the mold exclusion would be practically meaningless.”). The Buchholzes have shown they suffered a mold infestation, nothing more. Their theory is that water intrusion caused mold. But water intrusion as such is not a loss covered by the policy when its only manifested harm to covered property is fungal growth. Consequently, the Buchholzes have not shown that their mold coverage serves as an exception to the mold exclusion. So, their generalized mold claim is excluded by the terms of their policy. IV The district court incorrectly applied the Texas insurance coverage burden-shifting framework. Crestbrook is entitled to summary judgment regardless. It has demonstrated that a generalized mold claim is excluded under the policy. The Buchholzes have not met their burden of showing that an exception to the exclusion in their insurance contract brings their claim back within coverage. We AFFIRM.
