McHUGH, Circuit Judge.
I. INTRODUCTION
The Pioneer Centres Holding Company Employee Stock Ownership Plan and [1327]*1327Trust (the “Plan” or “ESOP”) and its trustees sued Alerus Financial, N.A. (Alerus) for breach of fiduciary duty in connection with the failure of a proposed employee stock purchase. The district court granted summary judgment to Alerus after determining the evidence of causation did not rise above speculation. The Plan appeals, claiming the district court erred in placing the burden to prove causation on the Plan rather than shifting the burden to Alerus to disprove causation once the Plan made out its prima facie case. In the alternative, the Plan contends that even if the district court correctly assigned the burden of proof, the Plan established, or at the very least raised a genuine issue of material fact regarding, causation. We affirm.
II. BACKGROUND
A. Factual History
1. The Proposed Transaction
Pioneer Centres Holding Company (Pioneer) owned and operated (through its subsidiaries) several automobile dealerships in Colorado and California, including Land Rover, Audi, and Porsche. In 2001, Pioneer sponsored the Plan under the Employee Retirement Income Security Act of 1974 (ERISA). Matthew “Jack” Brewer (Pioneer’s founder), Robert Jensen (Pioneer’s President), and Susan Dukes (Pioneer’s Chief Financial Officer), served as the Plan’s trustees. Mr. Brewer initially owned 100% of Pioneer’s stock. Over the course of several years, Mr. Brewer sold 37.5% of his Pioneer stock to the Plan and retained 62.5% ownership.
In 2009, the Plan’s trustees “proposed a stock transaction whereby the ESOP would become the 100% owner of Pioneer” (the “Transaction”). The Transaction included a stock redemption agreement, whereby Pioneer would redeem most of Mr. Brewer’s shares, and a stock purchase agreement, whereby the Plan would purchase the remaining shares. Mr. Jensen and Ms. Dukes also held stock options that they would exercise and that Pioneer would then redeem as part of the Transaction. Because the trustees’ interests in the transaction were adverse to those of the Plan, and to avoid any conflict of interest issues, the Plan hired Alerus as an independent “transactional trustee.” Alerus’s job was to determine whether, and on what terms, the Plan should purchase Mr. Brewer’s shares.
2. Correspondence Between Pioneer and Land Rover
Pioneer’s dealership agreement with Land Rover required approval before any changes in ownership or management occurred, stating: “[T]here will be no change in the foregoing [dealership ownership and management] in any respect without [Land Rover’s] prior written approval.” The agreement also granted Land Rover a right of first refusal to purchase any of Pioneer’s stock offered for sale.
Pioneer sent a letter to Land Rover on August 17, 2009, in which it asked Land Rover- to consent to Mr. Brewer’s transfer of his remaining Pioneer stock to the Plan to make the Plan the 100% owner of Pioneer. The letter included the proposed terms of the Transaction and informed Land Rover that Pioneer’s management would not change.
On August 31, 2009, Land Rover1 responded that it had not received any previous notice, or request for approval, of the [1328]*1328prior transfers of 37.5% of Mr. Brewer’s stock to the Plan. Instead, Land Rover indicated that its records still showed Mr. Brewer as owning 100% of Pioneer’s stock. Land Rover accordingly requested documentation of Mr. Brewer’s previous transfers to the Plan, and “reminded” Pioneer that “changes in ownership may not be made without our prior, written approval.” In addition, Land Rover “reserve[d] all of [its] rights with respect to any prior, unauthorized changes in ownership and any misrepresentations made in connection with the Dealer Agreement.” Finally, Land Rover explained that because Pioneer had failed to send a complete buy/sell agreement (a formal proposal), Land Rover’s right of first refusal with respect to the Transaction had not been triggered. To illustrate, Land Rover included a checklist of documents and information it requires before it will consider whether to approve a proposed ownership transfer.
On September 15, 2009, Pioneer sent a second letter to Land Rover, explaining that Pioneer’s August 17 letter was not intended as a formal proposal, but rather as “an informal request for an opinion from [Land Rover] regarding any significant issues [it] may have regarding the proposed change of ownership to being 100% ESOP owned.” Pioneer did not dispute that it never applied for or received Land Rover’s authorization for the prior transfers, it explained, however, that it thought permission was unnecessary because it was the ownership of the holding company (Pioneer) that had changed, not the ownership of Pioneer’s dealerships.
On October 30, 2009, Land Rover responded that it had “a substantial objection regarding the previous changes of ownership that have resulted in [Pioneer] being 37.5% [Plan] owned.” Because Pioneer had not previously disclosed or sought approval for these transfers, Land Rover maintained that each “was a material violation of the terms of the Land Rover Dealer Agreements,” which require prior written approval before any change in ownership.
Land Rover further complained that in addition to not informing it of these transfers, Pioneer affirmatively misrepresented its ownership interest when Mr. Brewer signed a new dealership agreement in February 2005. In that agreement, Mr. Brewer listed himself as the only beneficial owner of Pioneer and as the 100% owner of Pioneer’s stock. Land Rover considered this a “material misrepresentation” because “Mr. Brewer had already transferred 14.5%” of Pioneer’s stock to the Plan at that time. Consequently, Land Rover “demand[ed] that all prior, unauthorized transfers of beneficial ownership be reversed and that ownership be restored to comply with the representations made in the [dealership agreements].”
Significant for our purposes, Land Rover also advised Pioneer that it would not approve a change in ownership to 100% Plan owned, because its “requirements for oymership/operation of its dealerships would foreclose such an arrangement.” Land Rover explained that the “identity, reputation, financial resources, personal and business qualifications and experience, and the marketing philosophy of the designated owners and management of [Pioneer] are of vital significance” to Land Rover. Land Rover further stated that
if majority ownership of a Land Rover Dealer were held by an ESOP, then the Dealer would ultimately be controlled by an ever-changing group of employees who have not been vetted for ownership and management by [Land Rover], and who may not have the requisite financial and personal capabilities, qualifications, experience and commitment. Further, control and management of the dealer[1329]*1329ship would be subject to internal politics and factions. This is an unacceptable ownership structure for a Land Rover Dealer.
Moreover, in this particular case, we are being asked to approve a transfer of full ownership following a series of undisclosed and unauthorized transfers in which both current ownership and the [Plan] participated.
On December 3, 2009, Pioneer (assisted by Aleras) responded, interpreting Land Rover’s October 30 letter as announcing a prohibition against ESOP-owned dealerships, and asserting that this position violated California and Colorado law, as well as federal public policy favoring ESOP ownership. Pioneer also challenged some of Land Rover’s reasoning for finding Plan ownership unacceptable. First, Pioneer explained that the same management team would stay in place after the transfer, because 100% of the stock in the dealerships would continue to be owned by the holding company (Pioneer), which in turn would be owned by the Plan. And the Plan would remain under the management of its trustees, Mr. Brewer, Mr. Jensen, and Ms. Dukes. Second, Pioneer claimed that “numerous studies” have shown ESOP-owned companies outperform competitors because the employees have a beneficial interest in the company and are motivated to help the company succeed. It is undisputed that Pioneer did not specifically identify or cite to a single study to support this claim.
On December 14, 2009, Land Rover wrote to Pioneer and clarified that it did not have a policy against all ESOP ownership. But Land Rover explained that its hesitation with respect to this Transaction was because “Pioneer never applied for nor received Land Rover’s authorization for any of these prior transfers, in direct contravention of the Dealer Agreements,” and because Pioneer misrepresented its ownership in 2005. Land Rover also took issue with Pioneer’s arguments in favor of ESOP ownership. To begin, Land Rover indicated that even if management continued after the Transaction, it could change at any time and there was no guarantee of “who will ultimately control the dealership and be able to make changes in management and business direction.” And Land Rover noted Pioneer’s assertion that ESOPs outperform competitors was “not supported by the performance of your own dealerships.” Whereas Land Rover sales were down 16% nationally that year, Pioneer’s three Land Rover dealerships’ sales were well below average, being down 45%, 35%, and 34%. Land Rover also observed that all three dealerships were below average on the Customer Satisfaction Index (CSI), and both Denver dealerships were unprofitable.2
Finally, Land Rover indicated that it had not yet received a formal ownership transfer proposal from Pioneer, but that Pioneer was
free to submit any ownership transfer proposal that you wish to submit, and [Land Rover] will consider it in good faith and on the merits.... Any proposal to transfer majority ownership to the [Plan], however, will have to address and satisfy the concerns [Land Rover] identified about the identity, business ability, and financial capability of the person(s) who will have ultimate legal control of the dealership.
[1330]*1330Pioneer never responded to the December 14 letter.
Almost a year later, in a letter dated November 8, 2010, Land Rover approved the prior transfers of 37.5% of Pioneer’s stock to the Plan. But Land Rover warned that it “would not support a future ownership change giving majority ownership or control” to the Plan, because Land Rover’s “policy is that the person(s) who hold majority ownership and control of [a dealership] must be person(s) who are capable and committed to achieving and maintaining the level of retail representation” that Land Rover requires. This was the last communication between Pioneer and Land Rover on this issue.
3. Negotiations Between Alerus and Pioneer
In November 2009, Alerus sent Mr. Brewer draft stock redemption and stock purchase agreements that required Mr. Brewer to make certain representations and warranties. Mr. Brewer’s attorney, Richard Eason, revised the drafts by adding thirty-two “best of knowledge” qualifiers. In his transmittal letter, Mr. Eason told Alerus: “This is as far with the reps and warranties as [Mr. Brewer] will go.” Mr. Eason hoped to obtain Alerus’s signature on these revised Transaction documents, subject to later completing acceptable schedules. Because the schedules were likely to take “substantial time” to complete and Mr. Eason believed they were not of particular interest to the manufacturers, he intended to submit the signed revised Transaction documents to Land Rover without the schedules. He hoped to trigger Land Rover’s obligation to review and approve or reject the Transaction, or to exercise its right of first refusal.
But Alerus decided that the revisions to the representations and warranties were unacceptable and refused to sign the revised Transaction documents. As a result, Pioneer could not submit a signed copy of the revised Transaction documents to Land Rover.
Mr. Eason notified Mr. Brewer of Ale-rus’s decision by email. In addition to explaining that Alerus was unwilling to accept the qualified representations and warranties, Mr. Eason opined that the “likelihood of a transfer of control to the ESOP appears even more unlikely in view of the letter” received from Land Rover. Mr. Eason explained that Land Rover
continues to assert their position that the previous transfers to the ESOP were unauthorized and still subject to their approval or non-approval. The tone of [Land Rover’s] letter suggests that they would probably not approve a change in control to the ESOP and we would be faced with some form of litigation with them.... I believe that [Land Rover] will ultimately approve the minority ownership by the ESOP but that change of control to the ESOP will be problematical.
Alerus ultimately determined that because Mr. Brewer was unwilling to make the unqualified representations and “assume the attendant risk, ... the Plan should not purchase Brewer’s stock.” Formal Land Rover review was thus never triggered and the Transaction was abandoned. Alerus sent Pioneer a final invoice for its services in November 2010.
4. Pioneer Sells Its Assets to Kuni
More than a year after the Transaction was abandoned, Pioneer sold most of its assets to Kuni Enterprises for more than $10 million above what the Plan would have paid for Pioneer’s stock. Perhaps because it was an asset purchase rather than a stock purchase, Kuni did not require the unconditional representations and warran[1331]*1331ties that Alerus had demanded.3 During the course of negotiations between Pioneer and Kuni, Mr. Jensen met with Kuni’s representative, Greg Goodwin. When Mr. Goodwin inquired whether Pioneer employees would be disappointed that the Transaction had failed, Mr. Jensen told him that Land Rover had “indicated that the employee ownership in the company had maxed out ... and [Pioneer] was not going to be allowed to add any [Plan] ownership in the future.”
Mr. Goodwin was also present when Pioneer announced Kuni’s purchase of Pioneer’s assets. Mr. Goodwin recalled that Mr. Brewer “expressed regret that he was unable to complete his plan to sell Pioneer to the employees,” and identified “the resistance or disapproval of one of the manufacturers” as a cause of that failure. After the asset sale to Kuni, the Denver Business Journal reported that Mr. Jensen had identified the reason for the failure of the stock sale to the Plan as “the [fact that] manufacturers weren’t willing to deal with a company that was 100 percent employee-owned.”
B. Procedural History
After Pioneer sold its assets to Kuni, the Plan filed suit against Alerus4 for breach of fiduciary duty under 29 U.S.C. § 1109. Alerus moved for summary judgment,5 arguing (1) it did not breach any fiduciary duties, and (2) .even if there was a breach, Alerus did not cause any losses to the Plan because the Plan did not establish that Land Rover would have approved the Transaction.
1. The Plan’s Experts
The Plan hired two experts to opine on whether Land Rover would have approved a formal proposal for 100% ESOP ownership of Pioneer’s stock. The first expert, Carl Woodward, is a certified public accountant who believed Land Rover would have approved the transaction because: (1) multi-owner dealerships have become common; (2) the factors Land Rover considers in deciding whether to approve a change of ownership weigh in the Plan’s favor; (3) applicable state law requires manufacturers to be objectively reasonable in deciding whether to approve, and there is no objective reason for Land Rover to withhold approval; (4) Land Rover would not have exercised its right of first refusal because it would have had to own and operate other brands’ dealerships (Porsche and Audi); (5) Land Rover’s statement that it [1332]*1332“would not support” the Plan’s ownership change was just “code” that it “might allow but would not yet positively ‘approve’ and (6) Porsche and Audi had approved the Transaction.
The second proposed expert, Oren Tasi-ni, is an attorney “with decades’ experience in the purchase and sale of automobile dealerships,” who opined that “under California and Colorado law, Land Rover would have been required to consent to the sale to the [Plan].” He based this opinion on his view that California and Colorado laws favor approval of dealership transfers and prohibit manufacturers from “unreasonably” withholding consent. Mr. Tasini believed it would have been unreasonable for Land Rover not to approve the Transaction, particularly because it had agreed to the prior transfer of 37.5% of the stock.
Neither Mr. Woodward nor Mr. Tasini had any personal involvement with a proposed ESOP stock purchase in the automotive industry generally, or with Land Rover’s approval process specifically.
2. Land Rover’s 30(b)(6) Testimony
George Delaney, one of Land Rover’s Franchise Development Managers, testified as Land Rover’s 30(b)(6) witness. Mr. Delaney confirmed that Land Rover follows the law, acts reasonably, and has no policy against ESOPs. He estimated that Land Rover rejects less than one-third of change of control applications. Mr. Delaney also testified that Land Rover exercises its right of first refusal a “minority of the time.”
But Mr. Delaney could not “say ... without speculation” whether Land Rover would have approved the Transaction.6 When presented with the revised Transaction documents, Mr. Delaney testified, “If I had received this document, it still would not be a complete amount of information for me to make a recommendation.” He explained that the documents were missing several items contained on the checklist Land Rover sent Pioneer with its first letter. Mr. Delaney indicated that he “certainly” would have had an interest in seeing at least some of the schedules. Thus, even if Aleras had sent the signed Transaction documents without the schedules as hoped by Mr. Eason, Mr. Delaney stated the documents as submitted would not have triggered Land Rover’s formal approval obligation because the documents were “not complete.”
3. District Court Grants Summary Judgment
The district court bypassed the issue of whether Aleras had breached its fiduciary duty because it concluded the Plan had not established loss, an element of its prima facie case. The court explained that the Plan’s claimed damages were from the “proposed transaction and resulting benefits,” but there was insufficient “evidence that the proposed transaction would have been consummated” because it was only speculative that Land Rover’s “approval would have occurred.” The court reasoned that at “all material times, Land Rover indicated it would not approve and/or recommend the approval of the complete change of ownership from [Mr.] Brewer to the ESOP.” Additionally, “[Mr.] Delaney, the only Land Rover representative relied on by the parties, stated it would be speculative as to whether Land Rover would have approved.” The court explained that “[s]peculation ... is not sufficient to establish causation,” and “in the absence of [1333]*1333causation, the [Plan] is unable to show it has been damaged by the [alleged] breach of any duty.”
Although the court acknowledged a split, of authority on whether the burden shifts to the defendant to disprove causation once the plaintiff has established a prima facie case of breach of fiduciary duty under ERISA, it did not resolve that issue. Instead, the district court found the Plan had failed to meet its initial burden of “showing ... a causal connection between the breach of fiduciary duty and claimed loss as part of its prima facie case of loss.” The court therefore determined that “even assuming Alerus, as the fiduciary, must disprove causation, [the Plan] has not established a prima facie case of a loss in the first instance.”
In considering the evidence of causation, the district court excluded the testimony of Mr. Woodward and Mr. Tasini insofar as it related to whether Land Rover would have approved the Transaction. The court explained that “both opinions would require the experts to ‘read the mind’ of Land Rover, predict how Land Rover would have weighed factors it deemed relevant, and find that Land Rover would not only reach the conclusion that it must consent but also do so.” The court observed that not even Land Rover would speculate as to whether it would have approved the Transaction, and further concluded that such a prediction “is beyond the scope of any expert in this instance.” As an alternative basis for its ruling, the court found that both experts offered impermissible legal conclusions, when “neither expert may opine as to what the law requires.” The court also found Mr. Woodward unqualified to testify as to what the law requires because he is a CPA, not an attorney.
The district court granted summary judgment to Alerus, and the Plan now appeals. We have jurisdiction under 28 U.S.C. § 1291.
III. DISCUSSION
The Plan claims Alerus breached its fiduciary duties under ERISA by failing to execute the revised Transaction documents so that Alerus could send them to Land Rover for approval, and is thus liable under 29 U.S.C. § 1109(a), which provides:
Any person who is a fiduciary with respect to a plan who breaches any of the responsibilities, obligations, or duties imposed upon fiduciaries by this subchap-ter shall be personally liable to make good to such plan any losses to the plan resulting from each such breach....
29 U.S.C. § 1109(a) (emphasis added).
The district court concluded the Plan could not demonstrate a resulting loss because the evidence that Land Rover would have approved the Transaction was too speculative. The Plan contends this was error because the district court improperly required the Plan to prove causation, rather than shifting the burden to Alerus to disprove causation. And even if the district court did not erroneously assign the burden of proof, the Plan contends the evidence proved, or at least created a genuine dispute of material fact, that Land Rover would have approved the Transaction because: (1) record evidence of Land Rover and Pioneer’s relationship showed Land Rover would have approved; (2) California and Colorado state law would have required Land Rover to approve; and (3) the Plan’s experts would have testified that Land Rover would have approved, had the district court not abused its discretion in excluding their testimony.
We review the district court’s grant of summary judgment de novo. Birch v. Polaris Indus., Inc., 812 F.3d 1238, 1251 (10th Cir. 2015). Although cau[1334]*1334sation is generally a question of fact for a jury, where “the facts are undisputed and reasonable minds can draw only one conclusion from them,” causation is a question of law for the court. Berg v. United States, 806 F.2d 978, 981 (10th Cir. 1986). After the moving party has met its initial burden of showing an absence of a genuine issue of material fact, “the burden then shifts to the nonmoving party, who must offer evidence of specific facts that is sufficient to raise a ‘genuine issue of material fact.’ ” BancOklahoma Mortg. Corp. v. Capital Title Co., 194 F.3d 1089, 1097 (10th Cir. 1999) (citation omitted). “To defeat a motion for summary judgment, evidence, including testimony, must be based on more than mere speculation, conjecture, or surmise.” Bones v. Honeywell Int'l, Inc., 366 F.3d 869, 875 (10th Cir. 2004). The district court must draw all reasonable inferences in favor of the nonmoving party. Hornady Mfg. Co. v. Doubletap, Inc., 746 F.3d 995, 1004 (10th Cir. 2014). But an inference is unreasonable if it requires “a degree of speculation and conjecture that renders [the factfinder’s] findings a guess or mere possibility.” United States v. Bowen, 527 F.3d 1065, 1076 (10th Cir. 2008) (internal quotation marks omitted). The nonmoving party “must set forth evidence sufficient for a reasonable jury to return a verdict in [its] favor.” Rice v. United States, 166 F.3d 1088, 1092 (10th Cir. 1999); Schutz v. Thorne, 415 F.3d 1128, 1132 (10th Cir. 2005).
In reviewing the Plan’s challenges to the district court’s decision, we first address the proper allocation of the burden of proof with respect to the element of causation in an ERISA breach of fiduciary duty claim. We conclude that the plaintiff bears the burden on each element of its claim because Congress has given no indication that it intended to depart from that general rule. We also reject the Plan’s argument that a burden-shifting framework should be incorporated into ERISA from the common law of trusts.
Next, we consider whether the Plan met its burden at summary judgment to come forward with evidence from which a reasonable jury could find in its favor on each element of its breach of fiduciary duty claim. We agree with the district court that a reasonable factfinder could not conclude that Alerus caused the Transaction to fail because even if Land Rover had received the revised Transaction documents, the evidence does not rise beyond speculation that Land Rover would have approved the Transaction. In fact, all of the record evidence demonstrates that Land Rover would not have approved. Finally, we conclude the district court did not abuse its discretion in excluding the Plan’s expert testimony on causation.
A. The Burden of Proving Causation Falls on the Plaintiff in an ERISA Breach of Fiduciary Duty Claim
29 U.S.C. § 1109(a) provides that a fiduciary who breaches its duties under ERISA shall be personally liable for “any losses to the plan resulting from each such breach.” The plain language of § 1109(a) establishes liability for losses “resulting from” the breach, which we have recognized indicates that “there must be a showing of some causal link between the alleged breach and the loss plaintiff seeks to recover.” Allison v. Bank One-Denver, 289 F.3d 1223, 1239 (10th Cir. 2002) (internal quotation marks omitted); see also Plasterers’ Local Union No. 96 Pension Plan v. Pepper, 663 F.3d 210, 217 (4th Cir. 2011) (holding a breach of fiduciary duty “does not automatically equate to causation of loss and therefore liability” and consequently a “fiduciary can only be held liable upon a finding that the breach actually caused a loss to the plan”); Willett v. Blue Cross & Blue Shield, 953 F.2d 1335, [1335]*13351343 (11th Cir. 1992) (“Section 409 of ERISA establishes that an action exists to recover losses that ‘resulted’ from the breach of a fiduciary duty; thus, the statute does require that the breach of the fiduciary duty be the proximate cause of the losses claimed.... ”).
But the statute is silent as to who bears the burden of proving a resulting loss. Where a statute is silent on burden allocation, the “ordinary default rule [is] that plaintiffs bear the risk of failing to prove their claims.”7 Schaffer ex rel. Schaffer v. Weast, 546 U.S. 49, 56, 126 S.Ct. 528, 163 L.Ed.2d 387 (2005). This is because the “burdens of pleading and proof with regard to most facts have been and should be assigned to the plaintiff who generally seeks to change the present state of affairs and who therefore naturally should be expected to bear the risk of failure of proof or persuasion.” 2 McCormick on Evid. § 337 (7th ed. 2013).
There are exceptions to the default rule, such as when “certain elements of a plaintiffs claim ... can fairly be characterized as affirmative defenses or exemptions.” Schaffer, 546 U.S. at 57, 126 S.Ct. 528; see also FTC v. Morton Salt Co., 334 U.S. 37, 44-5, 68 S.Ct. 822, 92 L.Ed. 1196 (1948) (“[T]he burden of proving justification or exemption under a special exception to the prohibitions of a statute generally rests on one who claims its benefits .... ”). The Supreme Court cautioned, however, that “while the normal default rule does not solve all cases, it certainly solves most of them.... Absent some reason to believe that Congress intended otherwise, ... the burden of persuasion lies where it usually falls, upon the party seeking relief.” Schaffer, 546 U.S. at 57-58, 126 S.Ct. 528.
Another exception to the default rule unique to the fiduciary duty question arises under the common law of trusts. Trust law advocates a burden-shifting paradigm whereby once “a beneficiary has succeeded in proving that the trustee has committed a breach of trust and that a related loss has occurred, the burden shifts to the trustee to prove that the loss would have occurred in the absence of the breach.” Restatement (Third) of Trusts § 100 cmt. f (2012); see also George G. Bogert & George T. Bogert, The Law of Trusts and Trustees § 871 (2d rev. ed. 1995 & Supp. 2013) (“If [a beneficiary] seeks damages, a part of his burden will be proof that the breach caused him a loss.... If the beneficiary makes a prima facie case, the burden of contradicting it ... will shift to the trustee.” (emphasis added)).
Here, the district court found it unnecessary to decide whether to adopt the burden-shifting approach because, even assuming Alerus carries the burden to disprove causation once the Plan establishes a prima facie case, it concluded the Plan had not established a prima facie case of a loss in the first instance. We [1336]*1336adopt a different analytical approach and reject outright the Plan’s argument that ERISA breach of fiduciary duty claims should be resolved under a burden-shifting framework. But we affirm the district court because we agree the Plan has failed to meet its burden.
To begin, there is nothing in the language of § 1109(a) or in its legislative history that indicates a Congressional intent to shift the burden to the fiduciary to disprove causation. Nor is there anything that suggests Congress intended to make the lack of causation an affirmative defense or an exemption to liability. Whether something constitutes an element, as opposed to an affirmative defense or exception, turns on whether “one can omit the exception from the statute without doing violence to the definition of the offense.”8 United States v. Prentiss, 256 F.3d 971, 979 (10th Cir. 2001) (en banc) (internal quotation marks omitted), overruled in part on other grounds as recognized by United States v. Langford, 641 F.3d 1195 (10th Cir. 2011). Section 1109(a) of ERISA imposes liability on a breaching fiduciary for “any losses to the plan resulting from each such breach.” 29 U.S.C. § 1109(a). The requirement that the losses to the plan have resulted from the breach cannot be omitted from the statute without substantially changing the definition of the claim, thereby doing violence to it. We thus hold that causation is an element of the claim and that the plaintiff bears the burden of proving it.
The majority of federal circuits that have considered the issue agree. These courts have refused to incorporate any burden shifting into ERISA breach of fiduciary duty claims because the language “resulting from” in 29 U.S.C. § 1109(a) makes “[cjausation of damages ... an element of the claim, and the plaintiff bears the burden of proving it.” Silverman v. Mut. Benefit Life Ins. Co., 138 F.3d 98, 105 (2d Cir. 1998) (Jacobs, J. and Meskill, J., concurring); see also Wright v. Ore. Metallurgical Corp., 360 F.3d 1090, 1099 (9th Cir. 2004); Kuper v. Iovenko, 66 F.3d 1447, 1459-60 (6th Cir. 1995) abrogated on other grounds by Fifth Third Bancorp v. Dudenhoeffer, — U.S. -, 134 S.Ct. 2459, 189 L.Ed.2d 457 (2014); Willett v. Blue Cross & Blue Shield of Ala., 953 F.2d 1335, 1343-44 (11th Cir. 1992).
In contrast, some circuits have incorporated the common law of trust’s burden shifting into ERISA breach of fiduciary duty claims. The Fourth, Fifth, and Eighth Circuits have held that once an ERISA plaintiff has proven a breach and a prima facie case of loss to the plan, “the burden of persuasion shifts to the fiduciary to prove that the loss was not caused by the breach of duty.”9 See Tatum v. RJR Pen[1337]*1337sion Inv. Comm., 761 F.3d 346, 362 (4th Cir. 2014) (internal quotation marks omitted), cert. denied, — U.S.-, 135 S.Ct. 2887, 192 L.Ed.2d 924 (2015); McDonald v. Provident Indem. Life Ins. Co., 60 F.3d 234, 237 (5th Cir. 1995); Martin v. Feilen, 965 F.2d 660, 671 (8th Cir. 1992).
The Plan asks us to follow these decisions and shift the burden to the fiduciary once the plaintiff establishes a prima facie showing of a loss related to the breach. We decline that invitation. The “law of trusts often will inform, but will not necessarily determine the outcome of, an effort to interpret ERISA’s fiduciary duties.” Varity Corp. v. Howe, 516 U.S. 489, 497, 116 S.Ct. 1065, 134 L.Ed.2d 130 (1996). Where the plain language of the statute limits the fiduciary’s liability to losses resulting from a breach of fiduciary duty, there seems little reason to read the statute as requiring the plaintiff to show only that the loss is related to the breach.' And, as the Second Circuit observed in Silverman, the burden-shifting framework could result in removing an important check on the otherwise sweeping liability of fiduciaries under ERISA. See Silver-man, 138 F.3d at 106 (Jacobs, J. and Meskill, J., concurring) (“The causation requirement of § 1109(a) acts as a check on this broadly sweeping liability, to ensure that solvent companies remain willing to undertake fiduciary responsibilities with respect to ERISA plans.”).
In sum, we see no reason to depart from the “ordinary default rule that plaintiffs bear the risk of failing to prove their claims.” Schaffer, 546 U.S. at 56, 126 S.Ct. 528. Viewing the plain language, causation cannot “fairly be characterized as [an] affirmative defense[ ] or exemption[ ],” id. at 57, 126 S.Ct. 528, but is an express element of a claim for breach of fiduciary duty under 29 U.S.O. § 1109(a). We therefore hold that the burden falls squarely on the plaintiff asserting a breach of fiduciary duty claim under § 1109(a) of ERISA to prove losses to the plan “resulting from” the alleged breach of fiduciary duty.
We next turn to the question of whether the Plan met that burden. For the reasons we now explain, we agree with the district court that it did not.
B. The District Court Correctly Concliid-ed the Plan Failed to Come Forward with Sufficient Evidence of Causation to Survive Summary Judgment
As explained above, by suing Ale-ras for breach of fiduciary duty, the Plan assumed the burden of proof on each element of its claim. In order to prove the causation element, the Plan must demonstrate that Alerus’s alleged breach (refusal to sign the revised Transaction documents) caused the Plan to suffer damages (failure of the Transaction). The Plan therefore bears the burden of establishing by a preponderance of the evidence—more likely than not—that Land Rover would have approved the sale had Aleras signed the revised Transaction documents, which would have allowed Pioneer to submit them to Land Rover for review. For the Plan to defeat Aleras’s motion for summary judgment, it “must set forth evidence sufficient for a reasonable jury to return a verdict in [its] favor.” Rice, 166 F.3d at 1092. And this evidence “must be based on more than mere speculation, conjecture, or surmise.” Bones, 366 F.3d at 875.
[1338]*1338The district court concluded the evidence of causation could not rise above speculation because Land Rover gave every indication it would not approve the sale, and thus Alerus’s failure to sign the documents more likely than not did not result in the loss of the Transaction.
On appeal, the Plan ignores Land Rover’s repeated and consistent statements that it would not approve the Transaction and contends that “[h]ad the court appropriately considered and applied state law and presumed Land Rover would obey that law, it could only have drawn the conclusion that Land Rover approval ... was probable or, at a minimum, a genuine issue of fact in that regard had been established.”
In addressing this argument, we begin by reviewing the record evidence as a whole and determining that it overwhelmingly points to only one conclusion: Land Rover would not have approved the Transaction, even if Aleras had signed the revised Transaction documents. We next reject the Plan’s argument that Land Rover would have approved the sale if Aleras had signed the documents because Colorado and California law would have required Land Rover to do so. To the contrary, the record evidence indicates that even in the face of references to the state laws and Land Rover’s alleged legal obligations, Land Rover steadfastly refused to approve 100% Plan ownership. And the dissent’s suggestion that Pioneer would have successfully sued Land Rover if it had not approved is irrelevant and forfeited. Last, we conclude the district court did not abuse its discretion in excluding the Plan’s expert testimony regarding whether Land Rover would have approved the Transaction. Accordingly, we hold the district court correctly granted summary judgment in favor of Aleras because the Plan failed to come forward with any evidence from which the jury could find causation without engaging in speculation.
1. Record Evidence Regarding Approval
Throughout the communications with Pioneer, Land Rover repeatedly and consistently indicated it would not approve any further attempts to transfer stock from Mr. Brewer to the Plan:
• Land Rover “would not support a future ownership change giving majority ownership or control to an ESOP.”
• Land Rover’s “requirements for ownership/operation of its dealerships would foreclose such an arrangement.”
• “This is an unacceptable ownership structure for a Land Rover Dealer.”
• Pioneer was asking Land Rover “to approve a transfer of full ownership following a series of undisclosed and unauthorized transfers in which both current ownership and the ESOP participated.”
• Land Rover believed Pioneer committed a “material misrepresentation” when Mr. Brewer listed himself as Pioneer’s 100% owner in 2005 when he had already transferred 14.5% of Pioneer’s stock to the Plan at that time.
• Land Rover did not “condone the pri- or, unauthorized transfers of stock” and would “not tolerate any recurrence of the practice of making transfers without [Land Rover’s] prior written approval.”
• Mr. Delaney testified that Land Rover’s relationship with Pioneer was not “very favorable at the time, because they didn’t tell us for half a dozen years that they had sold off a third of the company. That’s a substantive violation of the Land Rover Dealer Agreement, that’s important to us.”
[1339]*1339• Mr. Delaney testified it was a “red flag” that “Mr. Brewer hadn’t been completely honest with us for several years, [during] which he did not tell us that he had sold a portion of the company, and he had signed renewals of the Land Rover Agreement saying that he still owned 100 percent.”
On December 14, 2010, after Alerus and Pioneer warned Land Rover that its refusal to approve an E SOP-owned dealership might be unlawful, Land Rover reiterated its concerns about such an arrangement. And when, almost a year later, Land Rover approved the prior transfers of 37.5% of Pioneer’s stock to the Plan, it unequivocally stated that Land Rover “would not support a future ownership change giving majority ownership or control” to the Plan. This was Land Rover’s final word on the issue, and it came from Lee Maas, Land Rover’s Vice President of Franchise Operations, who had authority to reject the Transaction. Pioneer presented no admissible evidence that Land Rover would have changed course and approved the Transaction.10
There is also substantial evidence that Land Rover effectively communicated its position to Pioneer representatives and that they understood Land Rover would not approve the Transaction. First, Mr. Eason told Mr. Brewer in September 2010:
The likelihood of a transfer of control to the ESOP appears even more unlikely in view of the letter [Pioneer] received from Lee Maas ... which continues to assert their position that the previous transfers to the ESOP were unauthorized and still subject to their approval or non-approval. The tone of his letter suggests that they would probably not approve a change in control to the ESOP and we would be faced with some form of litigation with them.
And although Mr. Eason believed that Land Rover would “ultimately approve the [37.5%] ownership by the ESOP,” he told Mr. Brewer “that change of control to the ESOP will be problematical.” Second, Mr. Jensen told Kuni that Land Rover declined the Transaction. Third, Mr. Brewer attributed the failure of the Transaction to the “resistance or disapproval on the part of ... one manufacturer.” And because the evidence is undisputed that the other two manufacturers approved the Transaction, the only manufacturer who could have disapproved was Land Rover. Last, Mr. Jensen told the Denver Business Journal that “the manufacturers weren’t willing to deal with a company that was 100 percent employee-owned,” and therefore Pioneer “had no other options to look at but to find a buyer.” Mr. Jensen’s statement that Pioneer “had no other option[ ] ... but to find a buyer” would only be true if Land Rover was not going to approve the Transaction.
In light of this undisputed evidence, and lack of any evidence to the contrary, a reasonable jury could not conclude Land Rover would have approved the Transaction.11 Consequently, the district court did [1340]*1340not err in determining that the Plan failed to meet the threshold necessary to survive summary judgment on the issue of causation. See Schneider v. City of Grand Junction Police Dep’t, 717 F.3d 760, 780 (10th Cir. 2013) (“Mere speculation ... is not sufficient to establish causation.”).
2. State Law Arguments
Despite Land Rover’s clear indication it would not approve greater Plan ownership, the Plan claims summary judgment was improper because it would have been unreasonable—and thus illegal under California and Colorado law—for Land Rover not to approve the Transaction. The Plan argues “Land Rover was aware of [the] state laws and their application, and followed them as a matter of practice,” and because we “presume[ ] that a person obeys the law,” NLRB v. Shawnee Indus., Inc., 333 F.2d 221, 225 (10th Cir. 1964), it therefore follows that a jury could reasonably conclude that Land Rover would have followed the law and approved the Transaction. We disagree.
The record evidence on this topic supports but one conclusion: That Land Rover would not approve 100% Plan ownership. In its communications with Land Rover, the Plan raised the very state laws that it relies on here to argue that a failure to approve the Transaction would be illegal. Despite that thinly-veiled threat, Land Rover’s final letter stated that it would retroactively approve the prior transfer of 37.5% to the Plan, but that it “would not support a future ownership change giving majority ownership or control” to the Plan. Thus, despite the Plan’s position that approval was required by law, Land Rover clearly indicated that it would not approve the Transaction.12 Any contrary finding therefore would require the jury to engage in speculation that is unsupported by and, in fact, contradictory to the evidence. In the face of such speculation, the Plan cannot rely on Land Rover’s alleged obligations to Pioneer under state law to establish that Alerus caused the Transaction to fail.
The dissent takes the Plan’s argument one step further by assuming that if Land Rover had done exactly what it said it would do—reject any further transfers of stock from Mr. Brewer to the Plan—Pioneer could have successfully sued Land Rover for violations of Colorado and California law. But the Plan is not suing Land Rover; it is suing Alerus. And the issue here is whether Alerus’s alleged breach of its. duties to the Plan by not signing the revised Transaction documents so that Pioneer could submit the documents to Land Rover caused the Transaction to fail. Because the evidence supports only the finding that the Transaction failed because [1341]*1341Land Rover would not approve further transfers of stock to the Plan, summary judgment for Alerus is proper for lack of causation, irrespective of whether Pioneer has a valid claim against Land Rover under state law.13
And even if we were to assume that a hypothetical lawsuit between Pioneer and Land Rover is relevant to whether Alerus caused the Transaction to fail in the first instance, the Plan never raised this argument below and consequently, neither party argued or moved to admit evidence on whether a lawsuit between Pioneer and Land Rover was probable or likely to succeed. In the district court, the Plan argued Land Rover would have approved the Transaction in order to be in compliance with state law, but it never based causation on a hypothetical lawsuit between Pioneer and Land Rover. Thus, the Plan forfeited this argument by failing to raise it in the district court where it could be tested factually. See Finstuen v. Crutcher, 496 F.3d 1139, 1145 n.3 (10th Cir. 2007) (“The general rule is that an appellate court will not consider an issue raised for the first time on appeal.” (alterations and internal quotation marks omitted)).
3. The Plan’s Experts
Finally, the Plan contends that the district court improperly excluded expert testimony from which the jury could have found that Land Rover would have approved the Transaction. We review the district court’s decision to exclude expert opinions for abuse of discretion. United States v. Avitia-Guillen, 680 F.3d 1253, 1256 (10th Cir. 2012). We will “reverse only if the district court’s conclusion is arbitrary, capricious, whimsical or manifestly unreasonable or when we are convinced that the district court made a clear error of judgment or exceeded the bounds of permissible choice in the circumstances.” Id. (internal quotation marks omitted). But “whether the district court applied the proper [legal] standard in admitting expert testimony” is reviewed de novo. Id.
Expert testimony must be relevant and reliable. Fed. R. Evid. 702; Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 588-89, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). To be reliable, expert testimony must be “based on actual knowledge, and not mere ‘subjective belief or [1342]*1342unsupported speculation.’ ” Mitchell v. Gencorp., Inc., 165 F.3d 778, 780 (10th Cir. 1999) (quoting Daubert, 509 U.S. at 590, 113 S.Ct. 2786). Further, a district court “is accorded great latitude in determining how to make Daubert reliability findings.” United States v. Velarde, 214 F.3d 1204, 1209 (10th Cir. 2000). When expert opinion “is not supported by sufficient facts to validate it in the eyes of the law, or when indisputable record facts contradict or otherwise render the opinion unreasonable, it cannot support a jury’s verdict” and will be excluded. Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 242, 113 S.Ct. 2578, 125 L.Ed.2d 168 (1993).
The district court excluded the testimony offered by the Plan’s experts regarding causation for several independent reasons. First, the court found the experts’ “proffered opinions amount to nothing more than speculation.” See Goebel v. Denver & Rio Grande W. R.R. Co., 215 F.3d 1083, 1088 (10th Cir. 2000) (“It is axiomatic that an expert, no matter how good his credentials, is not permitted to speculate.”). If Land Rover itself was unable to say, without speculation, whether the Transaction would have been approved, then the experts would similarly not be able to form an opinion without speculating. As the district court reasoned,
both opinions would require the experts to “read the mind” of Land Rover, predict how Land Rover would have weighed factors it deemed relevant, and find that Land Rover would not only reach the conclusion that it must consent but also do so. Such prediction, however, is beyond the scope of any expert in this instance.14
See Luciano v. E. Cent. Bd. of Coop. Educ. Servs., 885 F.Supp.2d 1063, 1072 (D. Colo. 2012) (excluding portion of expert’s opinion where he attempted to opine on what was motivating the behavior of a student’s parents, noting the expert could not “speculate on the parents’ state of mind”).
Second, the court excluded the testimony because it found the experts made impermissible legal conclusions.15 Mr. Tasini opined “that under California and Colorado law, Land Rover would have been required to consent to the sale to the [Plan].” But an “expert may not state his or her opinion as to legal standards nor may he or she state legal conclusions drawn by applying the law to the facts.” Okland Oil Co. v. Conoco Inc., 144 F.3d 1308, 1328 (10th Cir. 1998). This type of opinion does “not aid the jury in making a decision, but rather attempts to substitute [the expert’s] judgment for the jury’s.” Baumann v. Am. Family Mut. Ins. Co., 836 F.Supp.2d 1196, 1202 (D. Colo. 2011); see also United States v. Hill, 749 F.3d 1250, 1260 (10th Cir. 2014) (“[A]n expert may not go so far as to usurp the exclusive function of the jury to weigh the evidence and determine [1343]*1343credibility.” (internal quotation marks omitted)); Specht v. Jensen, 853 F.2d 805, 808 (10th Cir. 1988) (“[T]estimony on the ultimate factual questions aids the jury in reaching a verdict; testimony which articulates and applies the relevant law, however, circumvents the jury’s decision-making function by telling it how to decide the case.”); United States v. Jensen, 608 F.2d 1349, 1356 (10th Cir. 1979) (“[A]n expert witness cannot state legal conclusions by ... passing upon weight or credibility of the evidence.... ”).
Finally, the district court found Mr. Woodward “is a C.P.A. and is not qualified to testify on this subject.” The Plan argues the district court abused its discretion by finding Mr. Woodward unqualified because it only considered Mr. Woodward’s licensure, and “disregarded [Mr.] Woodward’s unparalleled experience and involvement in single and multiple franchise automobile dealership buy-sells requiring manufacturer approval.” But the district court did not abuse its discretion by identifying the fact that Mr. Woodward is not an attorney as an additional reason for not permitting him to testify as to whether refusal would have been contrary to law, where the decision is otherwise adequately supported.
Under these circumstances, we cannot conclude that the district court exceeded its considerable discretion in excluding the expert testimony. And in the absence of any evidence to contradict Land Rover’s unambiguous contemporaneous statements that it would not approve 100% ESOP ownership of the dealership, no reasonable jury could find causation.
IV. CONCLUSION
We decline to adopt a burden-shifting approach under 29 U.S.C. § 1109 and hold it is the plaintiffs burden to establish causation on a breach of fiduciary duty claim. We also affirm the district court’s conclusion that the Plan’s evidence of causation would have required the jury to engage in speculation. And we conclude the district court did not abuse its discretion in excluding the Plan’s expert testimony. We therefore affirm the district court’s grant of summary judgment to Aleras.