HORIZON HEALTH CORPORATION, Petitioner/Cross-Respondent v. ACADIA HEALTHCARE COMPANY, INC.; Psychiatric Resource Partners, Inc.; Michael A. Saul; Timothy J. Palus; Peter D. Ulasewicz; Barbara H. Bayma; and John M. Piechocki, Respondents/Cross-Petitioners, 520 S.W.3d 848

Justice Green delivered the opinion of the Court. In this case, we must determine whether the court of appeals erred in concluding that the evidence of the petitioner’s future lost profits was legally insufficient to support the jury’s award and whether the exemplary damages award is unconstitutionally excessive despite the court of appeals’ suggested remittitur. We agree that the evidence is legally insufficient to support any award of future lost profits, but we conclude that the court of appeals’ remitted exemplary damages award is unconstitutionally excessive.

We also agree with the court of appeals that the entity defendants cannot be held jointly and severally liable for the exemplary damages awarded against the individual defendants, that it is proper to remand the issue of attorney’s fees given the insufficient evidence supporting any award of future lost profits, and that the trial court did.not err in imposing discovery sanctions against one of the individual defendants.

Accordingly, we reverse the judgment of the court of appeals in part and remand the case to that court so that it may reconsider its suggested remittitur of exemplary damages in light of our decision. I. Background Horizon Health Corporation provides contract management services to hospitals .and healthcare providers to manage their psychiatric and behavioral health programs. Originally formed in 1981, Horizon was acquired by Psychiatric Solutions, Inc. (PSI) in 2007.

In 2010, members of Horizon’s upper-management team, who called themselves “Project Shamrock,” attempted to purchase Horizon from PSI after learning that PSI was considering converting from a publicly traded company into a private entity. However, PSI was ultimately acquired by Universal Health Services (UHS), a large publicly traded company. The members of Project Shamrock attempted to purchase Horizon from UHS, but UHS rejected the proposal in' late 2010.

Subsequently, PSI’s chief executive officer, Joey Jacobs, left PSI and became chief executive officer of Acadia Healthcare Company, which owns “freestanding psychiatric, child and adolescent, residential, chemical dependency treatment” facilities. In May 2011, Horizon’s president, Michael Saul, approached Acadia about joining Acadia’s team and presented a business plan to Acadia’s president, Brent Turner, proposing that Acadia establish a subsidiary to manage mental-health programs for hospitals and other mental-health providers.

In his presentation, Saul identified several companies that would be “competition” for the proposed subsidiary, including Horizon, which Saul indicated was “lost in UHS bureaucracy” and would •lose customers “due to relationships.” Acadia agreed to the proposal, and Saul sent Turner his résumé and the résumés of other members of Horizon’s management who had been part of Project Shamrock (Peter Ulasewicz, Tim Palus, and Barbara Bayma) as a “proposed management team.” Saul also told Turner that they “would go hard” after John Piechocki, a member of Ulasewicz’s sales team, based on his successful sales record at Horizon, and Saul and Ulasewicz began to recruit Piechocki shortly thereafter.

A. Management Lift-Out and Direct Competition with Horizon On two separate occasions in June 2011, Saul, Ulasewicz, Palus, and Bayma met to discuss their anticipated move to Acadia and their plans for the Acadia subsidiary. In August and September 2011, Saul, Ula-sewicz, Bayma, Palus, and Piechocki resigned from Horizon and joined Acadia’s recently formed subsidiary based on Saul’s proposal, Psychiatric Resource Partners (PRP). After losing multiple members of its upper-management team in a two-month period, Horizon conducted a forensic investigation- of its computer system.

Horizon discovered that Saul, Ulasewicz, Bayma, and Palus had coordinated their departures from Horizon to join the newly formed Acadia subsidiary. Horizon also learned that some of the Horizon defectors had made copies of several Horizon documents preceding their move to PRP.

For example, Saul purchased an external hard drive in late 2010 and, after having Horizon’s internal encryption system disabled, copied what Jack DeVaney, Horizon’s president, described as “a massive, massive amount” of Horizon documents on it, such as policies and procedures, “non-standard” contract language, financial models, monthly account listings, sales presentations, orientation materials, and legal files.

Similarly, Piechocki copied Horizon contracts, financial models, and lists of Horizon’s sales leads, marking some of the leads “DEAD” before resigning from his position at Horizon and adding those leads to PRP’s “master contact list” after joining PRP. In September 2011, Horizon notified Saul, Ulasewicz, Bayma, Palus, and Pie-chocki that their resignations and subsequent employment with Acadia were in violation of their employment agreements, the restrictive covenants entered into “at the inception of [their] employment,” and of their common-law duties of good faith and loyalty.

Horizon demanded that they end their employment with Acadia and return all documents to Horizon. Upon their resignations from Horizon, and after receiving Horizon’s notice, the new members of PRP began competing with Horizon, soliciting business from Horizon’s prospective and existing client base—though it is undisputed that Horizon did not lose any existing customers to PRP. For example, Piechocki secured a consulting contract for PRP with Southwest Regional Medical Center, which was an active Horizon sales lead that Piechocki had emailed to his personal email address before his departure.

In January 2012, Piechocki signed Westlake Regional Hospital (Westlake) to a contract with PRP in “direet competition” with Horizon, using Horizon’s financial models to “crunch[] numbers” to win the contract. Similarly, Ulasewicz scheduled a meeting with Cottage Hospital, a potential client with which he had met while employed by Horizon. Ulasewicz had learned while working at Horizon that Cottage Hospital’s barrier to using contract-management services might be removed; however, Ulasewicz did not share this information with anyone at Horizon. B.

Procedural History In October 2011, Horizon filed suit against Saul, Ulasewicz, Bayma, Palus, and Piechocki (the individual defendants) for breach of fiduciary duty, misappropriation of trade secrets, conversion, liability under the Harmful Access by Computer Act, liability under the Theft Liability Act, tor-tious interference with existing contracts, tortious interference with prospective business relationships, and conspiracy. Horizon also brought claims against Saul, Palus, Ulasewicz, and Bayma for breach of the restrictive covenants not to compete, fraud, and breach of contract.

Horizon alleged that Acadia and PRP (the entity defendants) were liable for essentially all of these acts and omissions either because they were directly involved or under the doctrines of ratification and vicarious liability and because they aided and abetted the individual defendants in committing the underlying misconduct. At the ensuing trial, the jury entered a unanimous verdict in Horizon’s favor on many of its claims.

The jury found that Saul, Ulasewicz, Bayma, and Palus breached the terms of their non-compete agreements (NGAs); ' Saul and Ulasewicz breached the terms of their covenants not to solicit with respect to their efforts to ensure Piechocki’s move to PRP; and all of the individual defendants failed to comply with their fiduciary duties to Horizon. Furthermore, the jury found that Acadia and PRP ratified this conduct and will earn future profits as a result.

Additionally, the jury found that the individual defendants, while acting in the scope of their employment with Acadia and PRP, intentionally interfered with the NCAs; misappropriated Horizon’s trade secrets; converted Horizon’s proprietary information; intentionally committed theft of Horizon’s property and trade secrets; and knowingly accessed Horizon’s computers, computer network, or computer system without consent and with intent to harm Horizon. The jury also found that Acadia and PRP ratified this conduct.

The jury further found that Saul, Ulasewicz, Bayma, and Palus committed fraud and fraud by nondisclosure by submitting expense reports for trips taken in June 2011.

Moreover, the jury found that all of the defendants participated in a conspiracy that damaged Horizon and that Acadia and PRP intentionally aided and abetted the individual defendants in breaching some of their fiduciary duties, intentionally interfering with the NCAs, misappropriating trade secrets, and converting Horizon’s proprietary information, but only PRP aided and abetted the theft of Horizon’s property or trade secrets and the harmful computer access.

Finally, the jury found that the damage sustained by Horizon as a result of the individual defendants’ breach of fiduciary duties, intentional interference with the NCAs, misappropriation of trade secrets, conversion of Horizon’s proprietary information, and theft was attributable to the malice of the individual defendants, as well as the entity defendants, Acadia and PRP.

The jury awarded Horizon $898,000 in future lost profits from the Westlake contract based on Saul’s, Palus’s, Ulasewicz’s, and Bayma’s failure to comply with their covenants not to compete and $3,300,000 in future lost profits based on Saul’s and Ulasewicz’s failures to comply with their covenants not to solicit—constituting Horizon’s lost sales from Piechocki’s future sales production.

The jury also awarded Horizon $50,000 as the fair market value of the stolen property or trade secrets that were the subject of Horizon’s claim for theft of property or trade secrets; $5,049.24 for the travel expenses charged to Horizon by Ulasewicz, Palus, and Bay-ma that were not associated with Horizon’s business; $1,750,000 in exemplary damages; and $900,000 in attorney’s fees for representation costs incurred through the conclusion of trial. In total, the jury awarded Horizon $6,903,049.24.

The trial court rendered final judgment in July 2013, awarding Horizon the full amount of damages found by the jury, as well as $41,740 in sanctions against Saul based on the trial court’s pretrial discovery-abuse ruling. The trial court, however, reduced Horizon’s trial attorney’s fee award from $900,000 to $769,432, disregarded the jury’s zero award of appellate attorney’s fees, and awarded Horizon $97,500 for appellate attorney’s fees. Acadia appealed, presenting multiple issues for review, and Horizon cross-appealed on the limited issue of .the trial court’s reduction in the attorney’s fees award.

The court of appeals reversed the trial court’s judgment and rendered a take-nothing judgment in part and remanded in part. 472 S.W.3d 74, 105-06 (Tex. App.—Fort Worth 2015, pet. granted). The court of appeals held that Horizon was not entitled to any award of future lost , profits damages because its expert testimony was impermissibly speculative and legally insufficient. Id. at 89-91.

Thus, the court of appeals rendered a take-nothing judgment against Horizon on all of its contractual and tort claims, except for theft of property and trade secrets ($50,000 jointly and severally) and fraudulent expense repoi’ts ($5,049.24 jointly and severally). Id. at 92-93. Because only $55,049.24 in compensatory damages remained, the court of appeals held that the jury’s $1,750,000 exemplary damages award was unconstitutionally excessive and remanded for a new trial on Horizon’s attorney’s fees. Id. at 104.

In its initial decision, the court of appeals suggested a remittitur of the exemplary damages award to a total amount of $220,196.96, to be apportioned among the individual defendants in the same proportions awarded by the jury. Id. at 98. On rehearing, however, the court of appeals increased its suggested remittitur to $220,196.96 for each individual defendant, totaling $1,100,984.80. Id. at 99 & n.31. The court of appeals also reversed the trial court’s judgment holding Acadia and PRP jointly and severally liable for the exemplary damages assessed against the individual defendants. Id. at 101-02.

Finally, the court of appeals affirmed the trial court’s award of discovery sanctions against Saul. Id. at 105. Both sides filed petitions for review raising a number of issues.

The issues include whether (1) the court of appeals erred in holding that the evidence of future lost profits is. legally insufficient; (2) an exemplary damages award is proper in this case and whether the court of appeals’ suggested remittitur of exemplary damages violates the Due Process Clause of the Fourteenth Amendment; and (3) the court of appeals erred in reversing the joint-and-several exemplary damages awards against Acadia and PRP. II.

Sufficiency of the Evidence of Future Lost Profits Horizon argues that the court of appeals erred by holding that the evidence is legally insufficient to support the jury’s award of $4,198,000 for future lost profits. We disagree. A.

Standard of Review The court of appeals properly stated the law regarding legal-sufficiency review of a jury’s finding: “In a legal-sufficiency review, we determine whether more than a scintilla of evidence supports the jury’s finding by considering evidence favorable to the finding if a reasonable fact-finder could and disregarding evidence contrary to the finding unless a reasonable fact-finder could not.” Id. at 87 (citing Cent. Ready Mix Concrete Co. v. Islas, 228 S.W.3d 649, 651 (Tex. 2007); Cont’l Coffee Prods. Co. v. Cazarez, 937 S.W.2d 444, 450 (Tex. 1996)).

The evidence supporting a jury’s finding is legally insufficient if: (1) there is a complete absence of evidence of a vital fact; (2) the rules of law or of evidence bar a court from giving weight to the only evidence presented to prove a vital fact; (3) there is no more than a mere scintilla of evidence presented to prove a vital fact; or (4) the evidence offered conclusively establishes the opposite of a vital fact. Sw. Energy Prod. Co. v. Berry-Helfand, 491 S.W.3d 699, 713 (Tex. 2016) (citing City of Keller v. Wilson, 168 S.W.3d 802, 810 (Tex. 2005)).

In conducting a legal-sufficiency review, the court is “ ‘limited to reviewing only the evidence tending to support the jury’s verdict and must disregard all evidence to the contrary,’ except contrary evidence that is conclusive.” Id. (quoting Mancorp, Inc. v. Culpepper, 802 S.W.2d 226, 227 (Tex. 1990), and citing City of Keller, 168 S.W.3d at 817). B. Applicable Law As we have explained, the rules concerning the sufficiency of evidence of lost profits damages are well established: Recovery for lost profits does not require that the loss be susceptible of exact calculation.

However, the injured party must do more than show that [it] suffered some lost profits. The amount of the loss must be shown by competent evidence with reasonable certainty. What constitutes reasonably certain evidence of lost profits is a fact intensive determination. As a minimum, opinions or estimates of lost profits must be based on objective facts, figures, or data from which the amount of lost profits can be ascertained. Although supporting documentation-may affect the weight of the evidence, it is not necessary to produce in court the documents supporting the opinions or estimates.

ERI Consulting Eng’rs, Inc. v. Swinnea, 318 S.W.3d 867, 876 (Tex. 2010) (quoting Holt Atherton Indus., Inc. v. Heine, 835 S.W.2d 80, 84 (Tex. 1992)). Thus, lost profits damages can be recovered only when both the fact and amount of damages is proved with reasonable certainty. See, e.g., Heine, 835 S.W.2d at 85. A party’s bare assertion that a contract was lost does not establish lost profits with reasonable certainty. Id.

Rather, the general rule is that recovery of lost profits as damages is allowed “where it is shown that a loss of profits is the natural and probable consequence of the act or omission complained of, and their amount is shown with sufficient certainty.” Tex. Instruments, Inc. v. Teletron Energy Mgmt., Inc., 877 S.W.2d 276, 279 (Tex. 1994) (quoting Sw. Battery Corp. v. Owen, 131 Tex. 423, 115 S.W.2d 1097, 1098 (1938)).

However, “anticipated profits cannot be recovered where they are dependent upon uncertain and changing conditions, such as market fluctuations, or the chances of business, or where there is no evidence from which they may be intelligently estimated.” Id. (quoting Owen, 115 S.W.2d at 1098). Indeed, “[t]he law is wisely skeptical of claims of lost profits from untested ventures or in unpredictable circumstances, which in reality are little more than wishful thinking.” Phillips v. Carlton Energy Grp., LLC, 475 S.W.3d 265, 280 (Tex. 2015).

Although legally sufficient evidence might otherwise establish a breach-of contract or tort permitting an award of lost profits, “profits ‘not susceptible of being established by proof to that degree of certainty which the law demands’ cannot be recovered as damages.” Id. at 279 (quoting Owen, 115 S.W.2d at 1099). When the evidence supporting a claim for lost profits damages is largely speculative or a mere hope for success, lost profits have not been established with reasonable certainty. Tex. Instruments, 877 S.W.2d at 279.

With these rules in mind, we now consider whether legally sufficient evidence establishes the fact and amount of Horizon’s lost profits damages with reasonable certainty. C. Analysis As a preliminary matter, we need not distinguish between Horizon’s causes of action supporting the award of lost profits damages because these damages are recoverable for any one of those claims. Swinnea, 318 S.W.3d at 877 n.3 (citing Waite Hill Servs., Inc. v. World Class Metal Works, Inc., 959 S.W.2d 182, 184-85 (Tex. 1998) (per curiam)).

However, because Horizon presented two different bases supporting the award of lost profits damages—the loss of the Westlake contract' and the loss of Piechocki’s future contract acquisitions—we must consider whether Horizon established the fact and amount of its lost profits arising from each basis independently. 1.

Westlake Contract The jury awarded Horizon $898,000 in damages for lost profits based on Saul’s, Ulasewicz’s, Bayma’s, and Pa-lus’s breach of their NCAs, breaches of fiduciary duties, intentional interference with the employment agreements, misappropriation of trade secrets, conversion of proprietary information, theft of trade secrets, knowing access of Horizon’s computer system, and fraud.

In doing so, the jury apparently concluded that it was more likely than not that Horizon would have won the Westlake contract absent the misconduct of these defendants and that, in reasonable probability, Horizon would suffer $898,000 in lost profits as a result. For the reasons explained below, we hold that the evidence is legally insufficient to establish the fact of lost profits damages regarding the Westlake contract with reasonable certainty.

Texas courts require that a plaintiff seeking damages based on lost profits from future business opportunities adduce evidence establishing that prospective customers would have done business with the plaintiff absent the defendant’s misconduct. See, e.g., City of Dall. v. Vills. of Forest Hills, L.P., Phase I, 931 S.W.2d 601, 606-07 (Tex.

App.—Dallas 1996, no writ); see also Heine, 835 S.W.2d at 85 (holding that a business owner’s testimony was insufficient to establish lost profits where he “was not able to specify which contracts they lost, how many they lost, how much profit they would have had from the contracts, or who would have awarded them contracts” and explaining that the plaintiffs “could have supported their lost profits with testimony that they had lost out on specific contracts” but failed to do so).

Thus, to recover lost profits from a contract with Westlake as damages, Horizon needed to present evidence showing that Westlake would have entered into a contract with it. We conclude that there is no evidence that Westlake would have entered into a contract with Horizon, as opposed to some other company, had it not signed a contract with PRP. Rather, the evidence shows only that PRP would not have won the Westlake contract absent the misconduct of the individual defendants; no evidence supports the conclusion that Horizon would.have won Westlake’s business.

Horizon’s damages expert, Jeff Balcombe, simply assumed that Horizon would have won the Westlake contract if not for the defendants’ wrongful conduct, and he specifically stated that he had no opinion as to whether Horizon would have retained Westlake. Further, Balcombe testified that PRP’s contract with Westlake included a term providing for a $150,000 loan or advance of construction costs. Although Bal-combe apparently deducted this amount from his calculations, he also acknowledged that he could not recall seeing any Horizon contract containing a similar provision.

Piechocki’s unrefuted testimony confirms that the Westlake contract included the $150,000 loan provision and that Horizon never included such terms in its contracts. Additionally, we cannot identify any evidence supporting a conclusion that Westlake would have accepted Horizon’s bid had it not accepted PRP’s. Indeed, it is equally plausible that Westlake would have rejected all of the bids had it not accepted PRP’s proposal. Thus, Balcombe’s testimony is insufficient because he admittedly assumed that Horizon would have won the Westlake contract.

Horizon argues that the evidence establishes that Westlake was on its lead list before the individual defendants’ departures and that the defendants utilized Horizon’s trade secrets in preparing PRP’s more attractive and winning bid for West-lake’s business, supporting an inference that Horizon would have won Westlake’s business absent the defendants’ misconduct.

However, even if we assume that the jury could have reasonably found that the defendants’ misconduct allowed PRP to bid on and ultimately win the Westlake contract, it is pure speculation to conclude that Horizon would have won the bid had PRP not made a more attractive offer. This is especially true given that there is no evidence that Horizon would have included a- $150,000 loan or advance of construction costs in its own bid.

Put simply, there is no evidence in the record establishing to a reasonable degree of certainty that Horizon would have won the Westlake contract absent the defendants’ misconduct, and Balcombe specifically acknowledged that he based his calculations on the assumption that Horizon would have won the Westlake contract had the defendants not committed the underlying torts. Thus, Horizon failed to establish the fact of damages relating to the Westlake contract with reasonable certainty.

See Heine, 835 S.W.2d at 85 (“[T]he bare assertion that contracts were lost does not demonstrate a reasonably certain objective determination of lost profits.”); see also Coastal Transp. Co. v. Crown Cent. Petroleum Corp., 136 S.W.3d 227, 232 (Tex. 2004) (“Opinion testimony that is conclusory or speculative is not relevant evidence, because it does not tend to make the existence of a material fact ‘more probable or less probable.’” (quoting Tex. R. Evid. 401)).

In sum, we hold that the evidence is legally insufficient to support the jury’s finding that Horizon sustained lost profits with respect to the Westlake contract because the evidence does not establish the fact of damages with reasonable certainty. 2. Piechocki’s Future Sales We also conclude that the evidence is legally insufficient with respect to Horizon’s claim for lost profits relating to the defendants’ wrongful solicitation of Piechocki.

As explained below, Balcombe’s estimates regarding the amount of time Piechocki would have remained at Horizon and the number of contracts he would have sold during that time are too speculative to constitute competent evidence. Furthermore, no evidence speaks to the profitability of the contracts Piechocki would have sold had he remained at Horizon. Put simply, the record supports the court of appeals’ conclusion that Balcombe’s testimony regarding the length of Piechocki’s future employment at Horizon and the success he would have had during that time was based on improper assumptions and thus conclusory.

The evidence shows that Piechocki was Horizon’s best salesperson and that he was in the top 90% of salespeople in his field. In calculating the damages that resulted from the wrongful solicitation of Piechocki, Balcombe assumed, among other things, that

(1) Piechocki, an at-will employee, would have stayed at Horizon for two years if not promoted, or four years if promoted, and

(2) Piechocki would have outperformed his average replacement by selling six contracts per year instead of four and that the profitability and initial contract term of each of those contracts would have comported with Horizon’s historical averages. Balcombe testified that he calculated Piechocki’s anticipated length of employment at Horizon based on a statistical analysis, estimating how long Piechocki would have stayed with Horizon absent the defendants’ wrongful conduct.

Balcombe testified that he “prepared an analysis, looking at employee historical tenure of people in the business development role at the VP level, as well as the senior VP level.” This analysis was based on ten years of Horizon’s retention data for employees in Piechocki’s position, including twenty-four sales vice presidents. Balcom-be presented two alternatives.

For Alternative 1, Balcombe assumed that Piechocki would have left Horizon after two years had the wrongful conduct not occurred, based on his analysis that the probability of an individual with Piechocki’s tenure dropped “below 50 percent after year two if they have been there as long as Mr. Piechocki had.” Thus, Balcombe estimated that even if Piechocki had not been promoted, a sales professional with Pie-chocki’s tenure would have remained at Horizon for at least two more years.

For Alternative 2, Balcombe assumed that Pie-chocki would have been promoted to the position of senior vice president of business development and would have left Horizon after four years, based on his calculation of the statistical probability of his continued employment dropping to less than 50% in year four. Balcombe testified that the four-year length of employment assumed in Alternative 2 was “based upon the fact that senior vice presidents stayed longer.” Balcombe confirmed that, in his Alternative 2 analysis, if Piechocki stayed an additional four years, it was because he was promoted.

But Balcombe’s assumption that Piechocki would have been promoted was based on a statement by DeVaney and did not take into account whether Piechocki himself intended to stay with Horizon. Indeed, Piechocki’s testimony supports that his decision to leave Horizon was motivated in part by DeVanejfs decision not to promote him when DeVaney had the chance, upon Ulasewicz’s departure, despite Piechocki’s specific request that De-Vaney consider him for the position.

Additionally, nothing in the record suggests that Balcombe considered whether the employees who were the subject of his analysis had signed employment agreements or NCAs and whether that affected the length of their employment. This is particularly significant because Piechocki was not bound by an NCA or any other type of employment agreement—a fact Balcombe conceded but did not factor into his analysis. However, Balcombe’s assumptions about the prospective length of Piechocki’s future employment at Horizon appear to have some ground in fact. See Burroughs Wellcome Co. v.

Crye, 907 S.W.2d 497, 499 (Tex. 1995) (explaining that an expert may offer an opinion based on assumed facts that are established by legally sufficient evidence). Piechocki admitted that: (1) he wanted the promotion; (2) he expressed his interest to DeVaney after the other defendants left; and

(3) Horizon’s sales manager told Piechocki that she wanted him to take the position. DeVaney essentially scheduled an interview with Piechocki, and De-Vaney testified that Piechocki was the “only real option” Horizon had to fill the position from within the company. Pie-chocki also testified that he had no plans to leave Horizon before he was solicited to join PRP. Rather, Piechocki’s decision was motivated in part by DeVaney’s decision not to promote him and in part by the opportunity to join PRP.

Thus, while scant evidence supports Balcombe’s Alternative 2 analysis, at least some evidence supports that Piechocki would not have left Horizon when he did absent the defendants’ solicitation in violation of their employment agreements. As to the number of sales Piechocki would have won during his continued employment at Horizon, in addition to Horizon’s historical sales data showing that Piechocki outpaced his peers two-to-one, an email from Ulasewicz to Saul supports the conclusion that Horizon lost sales as a result of Piechocki’s departure.

Ulasewicz wrote: I cannot think of a bigger body blow relative to impacting future new sales for Horizon than to get Piechocki out of there. If he is working for us, you have a very, very formidable sales team.... You take him out of Horizon and that will set their sales back at least one year—I mean it will stop—less than five deals. Recruiting effective sales people is very difficult and training takes along [sic] time. DeVaney similarly testified to a one-year period before any replacement salesperson could be expected to close a contract.

In fact, the loss of Piechocki was particularly harmful because Piechocki had experience in both psychiatric operations and sales, making him uniquely qualified to sell services to hospital operations executives. Furthermore, an email from Saul to Turner stated that, in addition to Ulasewicz, Bayma, and Palus, “we would go hard after John Piechocki, VP Sales—as recruiting the 2 best salesmen in the country would be a significant advantage to us and put a real hurt on the competition.” In fact, DeVaney specifically testified about how Horizon’s sales had been impacted by the loss of Piechocki.

DeVaney testified that Horizon signed fourteen new contracts in 2011, the year Piechocki left, but signed only seven new contracts the next year. In total, Horizon sold nine new contracts from the time that Ulasewicz and Piechocki left in the fall of 2011 through the date of the December 2012 trial, but none of those contracts were sold by Pie-chocki’s replacement. Contrasting this with the fact that PRP signed five contracts over that same period—two won by Pie-chocki and three won by Ulasewicz— makes the evidence detailed above even more compelling.

Thus, at least some evidence other than Balcombe’s testimony supports that the loss of Piechocki caused Horizon to lose future sales. Nevertheless, the fact that Piechocki might have stayed at Horizon a little long-’ er and might have sold contracts generating revenue for Horizon is not enough to establish that Horizon lost profits—evidence establishing the profitability of contracts Piechocki would have sold is necessary. See Miga v. Jensen, 96 S.W.3d 207, 213 (Tex. 2002) (“Lost profits are damages for the loss of net income to a business measured by reasonable certainty.” (emphasis added)).

But the record does not include evidence regarding the profitability of contracts Piechocki sold or would have sold had he remained at Horizon. Balcom-be assumed when formulating his opinion regarding the lost profits damages relating to Piechocki that the value of contracts Piechocki would have sold in the future would have comported with Horizon’s historical averages. Put differently, Balcombe based his calculations on the average profit generated by a contract sold by Horizon generally, not the profit generated by the contracts Piechocki sold.

Balcombe did this by compiling data of Horizon’s profit per contract over an approximately ten-year period. This assumes, however, that the profitability of Piechocki’s contracts corresponded with the average profitability of Horizon’s contracts. The fact that Pie-chocki sold more contracts than his peers does not speak to the profitability of those contracts. See id. In fact, we cannot identify any evidence regarding the profitability of any contract Piechocki sold while a Horizon employee.

Nor was there any evidence as to whether the contracts sold by Piechocki were more or less profitable than those sold by other salespeople. In the absence of evidence showing the profit associated with Piechocki’s sales, the number of contracts he might have sold is meaningless, and there is insufficient evidence to determine whether Horizon lost any profit as a result of his departure.

In short, Horizon failed to prove the fact of lost profits relating to Piechocki’s departure because the evidence speaks only to the number of contracts Piechocki sold previously and might have sold in the future, not the profitability of those contracts. This is fatal to Horizon’s claim for lost profits relating to Piechocki.

For Horizon to establish lost profits relating to the wrongful interference with a particular employee—that is, to establish with reasonable certainty damages for the lost sales that an employee would have obtained for Horizon absent the wrongful interference—Horizon’s evidence estimating those lost profits must .be tied to the performance of that employee. Horizon failed to offer such evidence, and the evidence is thus legally insufficient to establish that Horizon lost profits because of Piechocki’s departure.

In sum, we hold that the evidence is legally insufficient to support the jury’s finding that Horizon sustained lost profits because the evidence does not establish the fact of damages. D. Horizon’s Remaining Arguments Regarding Lost Profits Horizon argues that the court of appeals’ opinion conflicts with our decision in Arkoma Basin Exploration Co. v. FMF Associates 1990-A, Ltd., 249 S.W.3d 380 (Tex. 2008), in which we held that experts need not introduce foundational data supporting their conclusions unless the opposing party or trial court insists. Id at 389-90 (citing Tex. R. Evid. 705(a)).

Horizon’s argument is based entirely on a comment by the court of appeals that “Balcombe’s calculations, estimates, ‘statistical analysis,’ and ‘work papers’ supporting his conclusions were not admitted into evidence and were merely demonstrative aids” and that “[bjecause this information was not admitted into evidence, some of Balcombe’s explanations for his conclusions are difficult to decipher on appeal.” 472 S.W.3d at 89. According to Horizon, the court of appeals erred by finding Balcombe’s testimony “not competent” to the extent that his underlying work papers were admitted as demonstratives only.

Reading the court’s analysis as a whole, it is clear that the absence of Balcombe’s work papers did not impact the court of appeals’ decision. Rather, as shown above, Balcombe’s testimony is insufficient in light of his own testimony that he simply assumed Horizon would have won the Westlake contract and based his estimates regarding the profitability of Piechocki’s future sales on the average profit generated by contracts sold by Horizon generally, rather than the profit generated by the contracts Piechocki sold. We conclude that this argument has no merit.

Horizon also argues that we should remand for a new trial on the issue of lost profits. The court of appeals declined to do so, concluding that “Balcombe’s testimony did not establish any amount of lost-profit damages with reasonable certainty.” Id. at 90 n.20. According to Horizon, this was an error because “when there is sufficient evidence that lost profits were suffered, the appropriate remedy is either remittitur or remand, not a take nothing judgment.” In support of its argument for remand, Horizon relies on our decision in ERI Consulting Engineers, Inc. v.

Swinnea, in which we held that courts should suggest remittitur or remand for a new trial on damages where “competent evidence exists to establish some reasonably certain amount of lost profits—just not the particular amount awarded.” 318 S.W.3d at 880. We conclude that Swinnea has no applicability to the facts of this case because, as we have explained, the evidence is insufficient to establish that Horizon lost any amount of profits. Rendition is proper on this issue. III.

Exemplary Damages Against the Individual Defendants Acadia argues that the evidence is legally insufficient to support any award of exemplary damages against any defendant because no evidence supports the jury’s finding that the individual defendants acted with malice. Acadia also challenges the exemplary damages award as grossly excessive in light of the actual damages sustained by Horizon and thus unconstitutional, even after the court of appeals suggested a remittitur of the exemplary damages award.

For the reasons explained below, we hold that legally sufficient evidence supports an exemplary damages award, but we agree with Acadia that the exemplary damages award remains unconstitutionally excessive despite the court of appeals’ suggested remittitur. A. Sufficiency of the Evidence Supporting an Exemplary Damages Award The court of appeals held that legally sufficient evidence supports the jury’s finding that all of the individual defendants acted with malice. 472 S.W.3d at 95. We agree.

Pursuant to the Texas Civil Practice and Remedies Code, “exemplary damages may be awarded only if the claimant proves by clear and convincing evidence that the harm with respect to which the claimant seeks recovery of exemplary damages results from: (1) fraud; (2) malice; or (3) gross negligence.” Tex. Civ. Prac. & Rem.

Code § 41.003(a). “‘Clear and convincing’ means the measure or degree of proof that will produce in the mind of the trier of fact a firm belief or conviction as to the truth of the allegations sought to be established.” Id. § 41.001(2). “ ‘Malice’ means á specific intent by the defendant to cause substantial injury or harm to the claimant.” Id. § 41.001(7). “[I]n reviewing the legal sufficiency of evidence to support a finding that must be proved by clear and convincing evidence, an appellate court must ‘look at all the evidence in the light most favorable to the finding to determine whether a reasonable trier of fact could have formed a firm belief or conviction that its finding was true.’ ” Sw.

Bell Tel. Co. v. Garza, 164 S.W.3d 607, 609 (Tex. 2004) (quoting In re J.F.C., 96 S.W.3d 256, 266 (Tex. 2002)). In doing so, the “reviewing court must assume that the factfinder resolved disputed facts in favor of its finding if a reasonable factfinder could do so.” Id. at 627 (quoting In re J.F.C., 96 S.W.3d at 266). In Question 21, the jury found by clear and convincing evidence “that the harm to Horizon inquired about in Question No. 14 resulted from malice” by each individual defendant.

Question 14 inquired about both the harm from the theft of Horizon’s trade secrets and the harm from Ulasewicz’s, Bayma’s, and Palus’s submission of fraudulent expense reports. In Question 22, the jury found by clear and convincing evidence “that the harm to Horizon resulted from the fraud found in Question No. 12 or Question No. 13”—the questions relating to Ulasewicz’s, Bayma’s, and Palus’s submission of fraudulent expense reports for the June 2011 meetings.

Thus, the jury found that malice supports an exemplary damages award for all of the harms Horizon suffered and that fraud supports an exemplary damages award only for the harm relating to the fraudulent expense reports. The defendants have not argued in this Court or in the court of appeals that insufficient evidence supports the jury’s finding that the harm to Horizon resulting from Ulasewicz’s, Bayma’s, and Palus’s submission of improper expense reports constitutes fraud—an independent basis to award exemplary damages for that harm. See Tex. R. App. P. 38.1(f), 53.2(f).

Nevertheless, we must determine whether the jury could have reasonably concluded that the individual defendants acted with malice in order to support an exemplary damages award for the harm Horizon sustained as a result of the defendants’ theft of trade secrets. In concluding that legally sufficient evidence supported the jury’s finding that the individual defendants specifically intended to cause substantial injury or harm to Horizon, the court of appeals failed to analyze the evidence supporting the malice of each defendant and instead grouped them together. See 472 S.W.3d at 95-96.

Further, the court of appeals relied on evidence of the tort itself, with little more, to support the jury’s finding. Id. This was improper under the circumstances of this case. Under this Court’s precedent, to recover exemplary damages based on malice, Horizon needed to prove actual damages and submit clear and convincing evidence of outrageous, malicious, or otherwise reprehensible conduct. Safeshred, Inc. v. Martinez, 365 S.W.3d 655, 660 (Tex. 2012).

It was also required to prove that the defendants specifically intended for Horizon to suffer substantial injury that was “independent and qualitatively different” from the compensable harms associated with the underlying causes of action. Id. at 662 (quoting Transp. Ins. Co. v. Moriel, 879 S.W.2d 10, 19 (Tex. 1994)). Thus, we conclude that the court of appeals failed to properly analyze whether legally sufficient evidence supports the malice finding on which the trial court based its exemplary damages award.

Nevertheless, as shown below, we hold that the evidence is sufficient to support the jury’s finding of malice as to each individual defendant. 1. Saul We agree with the court of appeals that clear and convincing evidence supports malice on Saul’s part. First, the evidence supports that Saul orchestrated the management team “lift-out” and worked with Acadia to form PRP. Further, part of the business plan that Saul presented to Acadia in advocating for the formation of an Acadia subsidiary recognized that part of the subsidiary’s strategy would include targeting Horizon’s customers.

Indeed, this strategy was reflécted in materials Saul presented to Acadia before his departure from Horizon, One of the five “revenue sources” Saul highlighted in his presentation was “Horizon Business.” The presentation also projected that the revenue from “HOR Business” would be $2.28 million a year for 2012 and 2013, and that future operating management contracts, derived in part from Horizon’s business, would produce over $25 million by 2016.

Moreover, Saul sent Acadia’s president an email recognizing that his planned “management team ‘lift-out’ ” while the other individual defendants were employed by Horizon carried “riskfs],” such as a “claim [of] tortious interference.” Saul also sent Acadia’s president an email explaining that “we would go hard after John Pie-chocki, VP Sales—as recruiting the 2 best salesmen in the country would be a significant advantage to us and put a real hurt on the competition.” Once Acadia decided to proceed with Saul’s plan, Saul requested an external hard drive for his Horizon computer, paid for by Horizon, and, after instructing his secretary to disable the encryption on his computer, downloaded “everything that was non-financial on [Horizon’s] server.” Saul also emailed many of Horizon’s confidential documents to himself before resigning.

This evidence is sufficient to have produced a firm belief in the minds of the jury that Saul acted with specific intent to cause substantial injury or harm to Horizon. 2. Ulasewicz The court of appeals also correctly held that the evidence is sufficient to support the jury’s finding that Ulasewicz acted with malice. Specifically, we. agree that one email from Ulasewicz to Saul, which the court of appeals considered “particularly damning,” supports the jury’s conclusion. 472 S.W.3d at 95-96.

The email, which was sent while Ulasewicz and Saul were both Horizon employees and before Saul’s presentation to Acadia, stated: Here are my thoughts on a 12-24 month stratetgy [sic] relative to positioning. This time frame is critical to us in terms of success.

Based on our preliminary sales plan as presented, we are in fact saying that we are going to take [a] certain number of agreements out of Horizon’s hide, both new deals but also terming contracts_I would also recommend you begin to group the contracts we know ai-e coming up over the next two years and place them in maybe three categories from In Play to Unlikely to Switch....

We also need to know not only the termination dates but much more importantly any rollover dates, this is critical.... .., The more members of our senior management we bring over the greater our ability to shape and hone a message to potential clients that is based implicitly and explicitly on our knowledge that Horizon exists in name only.... I do advocate we get either Palus or [Pie-chocki] and ... we should bring in Bay-ma.

Hurting Horizon early and often is a business stratetgy [sic] and a good one.... .,, I cannot think of a bigger body blow relative to impacting future new sales for Horizon than to get Piechocki out of there. ... The message to potential clients is Pedigree—we need to convey this is not a startup, this is a logical continuation of the undeniably established Leadership, Experience and Expertise that maintained Horizon in its number one position for the last ten years.... ... [Transition timing is very important I believe.

We need to gut punch them as we leave, to me that means having all of our ducks in a row so we can move quickly into the market. Let’s make sure we talk around timelines before you commit, I know you are anxious to leave but if you wait for the right time, it will be all the sweeter. Business first—success is the best rvenge [sic]—trust me on this.

We conclude that this email supports the jury’s finding that Ulasewicz acted with malice because it shows that he intended to “[h]urt[ ] Horizon early and often” and “gut punch” Horizon in executing the scheme to form a business that would compete directly with Horizon. 3. Piechocki We also hold that the evidence supports the jury’s conclusion that Pie-chocki specifically intended to cause substantial harm to Horizon. First, Piechocki admitted that he secured the Westlake contract for PRP and that he used Horizon’s confidential documents in doing so.

This shows that Piechocki used Horizon’s trade secrets and proprietary information after moving to PRP to acquire the business of a prospective Horizon client. Pie-chocki also surreptitiously forwarded emails containing proprietary and confidential information, including lists of active Horizon sales leads, to his personal email account on the eve of his departure (and in violation of the employee handbook) for the purpose of providing key client information to a Horizon competitor.

Pie-chocki also marked certain sales prospects as “DEAD” before he resigned, and those same leads were added to Acadia’s sales list when he joined PRP. Piechocki even admitted that he used Horizon’s stolen documents to create PRP’s bid for West-lake and that he met with Horizon’s existing customers after he took copies of those customers’ contracts from Horizon. PRP used Horizon’s most recent Master Contact List to create a new sales lead list for PRP—and Ulasewicz described disclosure of sales leads as

an act of “treason.” Finally, after joining PRP, Piechocki simply replaced Horizon’s name with PRP’s on a confidential document containing specific language for customer contracts. This evidence, on balance, supports the jury’s malice finding with respect to Piechocki without focusing solely on the underlying torts because it shows that Piechocki subsequently used all of the materials he stole to benefit Acadia and PRP at Horizon’s expense. That is, this circumstantial evidence is legally sufficient to support a finding of malice. See, e.g., Bentley v.

Bunton, 94 S.W.3d 561, 596 (Tex. 2002) (explaining in the context of defamation that “[t]he defendant’s state of mind can—-indeed, must usually—be proved by circumstantial evidence”). 4. Bayma We also hold that legally sufficient evidence supports the jury’s finding that Bayma acted with malice. The evidence shows that Bayma first began communicating with Saul about his plans in May 2011—the month Saul pitched the idea to Acadia of a new subsidiary that would compete with Horizon.

In his presentation, Saul identified several companies that would be “competition” for the proposed subsidiary, including Horizon, which Saul indicated was “lost in UHS bureaucracy” and would lose customers “due to relationships.” On June 8, 2011, all of the individual defendants except Piechocki met at Saul’s home to discuss Saul’s new plans with Acadia. The next day, Bayma sent Saul an email outlining her questions and recommendations regarding the new Acadia subsidiary. Further, on June 24, 2011, Bayma forwarded Horizon training documents to Saul relating to freestanding psychiatric hospitals.

That same day, Saul sent Palus and Bayma information regarding Acadia’s benefits package, the proposed dates for submitting resignations, and invited Bayma and Palus to his home for a second “planning session,” cautioning them to “[k]eep [their] plans discrete.” Five days later, all of the individual defendants except Piechocki met at Saul’s home, and Bayma asked Hqrizon to pay for those preparations without disclosing her true purpose.

Further, the evidence confirms that the individual defendants—a large portion of Horizon’s upper-management team—coordinated their departures from the outset so that they would each announce their resignation in a rapid succession, leaving Horizon without much of its leadership and in disarray. As planned, Ulasewicz, Bayma, and Palus all announced their resignations within seven days of each other. Unsurprisingly, DeVaney testified that the defendants’ successive departures caused “a lot of unrest” within Horizon.

Weeks after their departures, the newly formed Acadia subsidiary began soliciting Horizon’s current and prospective ■ clients using sales materials and contracts stolen from Horizon. In fact, the evidence shows that Bayma and Palus supported the preparation of PRP’s sales proposals and contracts—based on documents stolen from Horizon—shortly after joining PRP.

In sum, we think the jury could have concluded that Bayma’s participation in the scheme to form a business that would compete directly with Horizon, using Horizon’s trade • secrets and proprietary materials, while still employed by Horizon, evidences her intent to injure Horizon—a conclusion the jury obviously reached in light of its finding that each of the defendants participated in a conspiracy to damage Horizon.

Indeed, the defendants recognized that keeping their plans secret, coordinating their departures, and their immediate use of Horizon’s documents were essential to PRP’s success and to their business strategy of “[h]urting Horizon early and often” and their goal of “gut punch[ing]” Horizon on their way out the door. 5. Palus Much of the evidence supporting the jury’s malice finding as to Bayma equally supports a finding that Palus acted with malice. Therefore, and as explained further below, we hold that legally sufficient evidence supports the jury’s finding of malice with respect to Palus.

In addition to the evidence supporting Bayma’s malice, the evidence shows that Palus first began communicating with Saul about his plans in April 2011—one month before Saul presented to Acadia his idea for a new subsidiary that would compete directly with Horizon. Palus also joined the other individual defendants at Saul’s home in June 2011, and Palus asked Horizon to reimburse him for the expenses incurred on those trips while concealing the true purpose of those meetings.

Further, Palus waited to announce his resignation, staying for a few days longer than Bayma and Ulasewicz, feeding information to Saul, Ulasewicz,. and Bayma regarding the fallout happening at Horizon. In one of Palus’s final emails as a Horizon employee, he told his cohorts that “[f]inal transitions are in full swing and things seem to be setting up well,” and he noted that he had contacted one of Horizon’s clients to

schedule a meeting. Additionally, the evidence shows that, following his resignation, Palus supported the preparation of PRP’s sales proposals and contracts based on documents stolen from Horizon. Palus also helped prepare PRP’s “new proforma model” and other documents using “former templates” (i.e., Horizon’s property) that he had forwarded to his personal email account on the eve of his departure.

In sum, we hold that sufficient evidence supports the jury’s conclusion that Palus acted with the specific intent to harm Horizon in light of his participation in the scheme to form a business that would compete directly with Horizon, using Horizon’s trade secrets and proprietary materials. Thus, the record supports the jury’s assessment of exemplary damages against each of the individual defendants. B.

Excessiveness of Exemplary Damages Award In light of our decision that insufficient evidence supports the jury’s award of lost profits damages, $55,049.24 of actual damages remain: $50,000 for the fair market value of the trade-secret items that the individual defendants misappropriated and $5,049.24 for the fraudulent reimbursements Ulasewicz, Bayma, and Palus requested from Horizon for their pre-resig-nation trip to discuss their plans for PRP.

Therefore, like the court of appeals, we must determine whether the exemplary damages award is excessive in light of the remaining actual damages award and, thus, unconstitutional. 472 S.W.3d at 96-97. For the reasons explained below, we hold that the exemplary damages award is excessive despite the court of appeals’ suggested remittitur and thus remand the case to the court of appeals so it may consider a remittitur in a constitutionally permissible amount.

The jury found that each of the individual defendants caused the $50,000 damages relating to the theft of property and theft of trade secrets in varying degrees. The jury found that Saul was 40% responsible for the harm and that Ulasewicz, Bayma, Palus, and Piechocki were each 15% responsible for the theft of trade secret damages. The jury also found that Ulasewicz, Bayma, and Palus submitted fraudulent expense reports in the following amounts:

(1) Ulasewicz: $1,398.45;

(2) Bayma: $1,049.38; and

(3) Palus: $2,601.4L Ultimately, the jury found that the individual defendants caused harm to Horizon as follows: Table 1 Defendant Saul Ulasewicz Bayma Palus Piechocki Total Theft of Property or Trade Secrets $20,000.00 .$7,500.00 $7,500.00 $7,500.00 $7,500.00 $50,000.00 Fraudulent Expense Reports $1,398.45 $1,049.38 $2,601.41 $5,049.24 Total Damages Caused $20,000.00 $8,898.45 $8,549.38 $10,101.41 $7,5000.00 $55,049.24 Because the jury found that the individual defendants committed felonious theft of trade secrets as defined in the Penal Code, the trial court assessed joint and several liability for the full amount' of compensatory damages against each of the individual defendants.

See Tex. Civ. Peac. & Rem, Code § 33.013(b)(2); Tex Penal Code § 31.05. The jury awarded a total of $1,750,000 in exemplary damages against the individual defendants: $500,000 each against Saul and Ulasewicz and $250,000 each against Bay-ma, Palus, and Piechocki.

The table below summarizes the jury’s assessment of exemplary damages: Table 2 Defendant Saul Ulasewicz Bayma Palus Piechocki Total Liability for Exemplary Damages $500,000 $500,000 $250,000 $250,000 $250,000 $1,750,000 Proportion 29% 29% 14% 14% 14% 100% The court of appeals’ initial decision suggested a remittitur of total exemplary damages to $220,196.96, allocated among the individuals in the same proportion as the jury assessed exemplary damages (a “per-judgment” approach). 472 S.W.3d at 98-99.

The table below depicts the ratio of compensatory to exemplary damages liability (joint and several) for each individual defendant in the court of appeals’ initial opinion, and the cumulative totals: Table 3 Defendant Saul Ulasewicz Bayma Palus Piechocki Total Remitted Exemplary Damages $63,857.13 (29%) $63,857.13 (29%) $30,827.57 (15%) $30,827.57 (15%) $30,827.57 (15%) $220,196.96 Joint & Several Liability $55,049.24 $55,049.24 $55,049.24 $55,049.24 $55,049.24 $55,049.24 Ratio 1.16:1 1.16:1 0.56:1 0.56:1 0.56:1 4:1 On rehearing, however, the court of appeals held that each individual defendant should be assessed $220,196.96 in exemplary damages, using a “per-defendant” rather than “per-judgment” approach.

Id. Thus, the court of appeals’ suggested re-mittitur assesses exemplary damages liability as follows: Table 4 Defendant Saul Ulasewicz Bayma Palus Piechocki Total Remitted Exemplary Damages $220,196.96 $220,196.96 $220,196.96 $220,196.96 $220,196.96 $1,100,984.80 Joint & Several Liability $55,049.24 $55,049.24 $55,049.24 $55,049.24 $55,049.24 $55,049.24 Ratio 4:1 4:1 ' 4:1 4:1 4:1 20:1 As explained below, we conclude that the court of appeals’ suggested remittitur on rehearing results in an award that violates due process because a 4:1 ratio of compensatory to exemplary damages is not appropriate in this case.

The more interesting and complicated question, however, is how the constitutional excessiveness of an exemplary damages award should be assessed when multiple defendants face joint and several liability for compensatory damages but are individually liable for exemplary damages, as required by Texas law. See Tex. Civ. Prac. & Rem. Code § 41.006. A few key principles emerge in reviewing the Supreme Court’s decisions exploring the constitutional limits of exemplary damages.

First, compensatory and exemplary damages serve different purposes; compensatory damages redress concrete losses caused by the defendant’s wrongful conduct, while exemplary damages are aimed at deterrence and retribution. See State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408, 416, 123 S.Ct. 1513, 155 L.Ed.2d 585 (2003); Cooper Indus., Inc. v. Leatherman Tool Grp., Inc., 532 U.S. 424, 432, 121 S.Ct. 1678, 149 L.Ed.2d 674 (2001); see also Bennett v.

Grant (Bennett II), — S.W.3d —, —, 2017 WL 1553157, *5 (Tex. 2017) (“As an overarching premise, exemplary damages further the state’s interest in punishing and deterring unlawful conduct.”). Second, “[elementary notions of fairness enshrined in [the Supreme Court’s] constitutional jurisprudence dictate that a person receive fair notice not only of the conduct that will subject him to punishment, but also of the severity of the penalty” that may be imposed. BMW of N. Am., Inc. v. Gore, 517 U.S. 559, 574, 116 S.Ct. 1589, 134 L.Ed.2d 809 (1996).

Accordingly, “[t]he Due Process Clause of the Fourteenth Amendment prohibits the imposition of grossly excessive or arbitrary punishments on a tortfea-sor.” Campbell, 538 U.S. at 416, 123 S.Ct. 1513.

Whether an award comports with due process is measured by three guideposts: “(1) the degree of reprehensibility of the defendant’s misconduct; (2) the disparity between the actual or potential harm suffered by the plaintiff and the punitive damages award; and (3) the difference between the punitive damages awarded by the jury and the civil penalties authorized or imposed in comparable cases.” Id. at 418, 123 S.Ct. 1513 (citing Gore, 517 U.S. at 575, 116 S.Ct. 1589). We review the constitutionality of an exemplary damages award de novo. Bunion v. Bentley, 153 S.W.3d 50, 54 (Tex. 2004) (per curiam).

The court of appeals’ evaluation of these factors was flawed because the court failed to analyze them as to each individual defendant and instead grouped them together. See 472 S.W.3d at 97-98 (analyzing the Gore/Campbell factors globally rather than on a defendant-by-defendant basis). Constitutional rights are personal in nature, and, therefore, inquiries into the constitutional excessiveness of an exemplary damages award must be defendant-specific.

See Gore, 517 U.S. at 574, 116 S.Ct. 1589 (explaining that the constitutional considerations of exemplary damages “dictate that a person receive fair notice ... of the conduct that will subject him to punishment”); see also Campbell, 538 U.S. at 417, 123 S.Ct. 1513 (explaining that the fair notice requirement guards against “arbitrary deprivation of [the defendant’s] property”). Indeed, the requirement that courts focus on the degree of culpability of each tortfeasor has long been established by this Court. For example, in St. Louis & Southwestern Railway Co. of Texas v.

Thompson, 102 Tex. 89, 113 S.W. 144 (1908), we explained: If the defendants or either of them were actuated by malice in making the charges against Thompson ... then the plaintiff may in the discretion of the jury recover exemplary damages against either or all of the said defendants in such sum as the jury may believe should be assessed against the said defendants or either of them.

It is not necessary, as in case of actual damages recovered, that all of the defendants should be subjected to the same verdict because some of the defendants may have acted without malice, but in combination with others, and as to such defendants there would be no right to recover exemplary damages. Id. at 147. Thompson clearly stands for the proposition that, unlike actual damages, all defendants should not be subject to the same verdict of exemplary damages in every case because they may have acted with varying degrees of malice, or with no malice at all. Id.

In fact, Horizon recognizes this principle in its briefing, arguing that “[a]s the Campbell guideposts reflect, the constitutionality of punitive damages is determined as to each defendant, not the resulting harm to plaintiff’ and that the “constitutionality of punitive damages must be evaluated defendant-by-defendant.” In compliance with this well-established principle, we next examine the constitutional propriety of the court of appeals’ suggested remittitur by analyzing each defendant separately. 1.

Reprehensibility The degree of reprehensibility of a defendant’s conduct is “the most important indicium of the reasonableness of a punitive damages award.” Gore, 517 U.S. at 575, 116 S.Ct. 1589.

Courts determine the reprehensibility of a defendant’s conduct by considering whether: (1) the harm caused was physical as opposed to economic; (2) the tortious conduct demonstrated ' an indifference to or a reckless disregard of the health or safety of others; (8) the target of the conduct had financial vulnerability; (4) the conduct involved repeated actions or was an isolated incident; and (5) the harm was the result of intentional malice, trickery, or deceit, or mere accident. Id. at 576-77, 116 S.Ct. 1589. Here, the harm caused by the individual defendants was entirely economic in nature.

Horizon incorrectly argues that this factor “do[es] not carry weight in this case.” To the contrary, this factor weighs against a finding of reprehensibility and, therefore, in favor of the individual defendants. Further, the parties, agree that the conduct of the individual defendants shows no indifference to the health or safety of others, and Horizon is not a financially vulnerable target. Thus, the first three reprehensibility factors run counter to the remitted exemplary damages award.

The fourth reprehensibility factor— whether the conduct involved repeated actions or was an isolated incident—also weighs against the court of appeals’ suggested remittitur. The court of appeals concluded that the individual defendants’ “repeated and intentional” conduct supported the increased reprehensibility of their actions. See 472 S.W.3d at 97. However, the Supreme Court and Texas courts have clearly stated that Campbell’s fourth reprehensibility factor is about recidivism, not the course of conduct giving rise to the plaintiffs ultimate harm or the wrongdoer’s attempts to conceal his wrongdoing.

See Gore, 517 U.S. at 577, 116 S.Ct. 1589 (“Our holdings that a recidivist may be punished more severely than a first offender recognize that repeated misconduct is more reprehensible than an individual instance of malfeasance.”); Bennett I, 315 S.W.3d at 874-78 & n.55 (explaining that Campbell’s repeated-actions factor “refers to recidivism” rather than the course of conduct giving rise to the plaintiffs cause of action (citing Tony Gullo Motors I, L.P. v. Chapa, 212 S.W.3d 299, 308-09 (Tex. 2006) (“The Supreme Court says ‘repeated conduct’ refers to recidivism.... ”))); see also Khorshid, Inc. v.

Christian, 257 S.W.3d 748, 768 (Tex. App.—Dallas 2008, no pet.) (“[Although appellants took possession of [the .plaintiffs] automobile on two occasions, there was no evidence appellants have been involved in other similar incidents.”). We conclude that there is no evidence of recidivism in this case because nothing in the record supports that the individual defendants are repeat offenders who have committed past transgressions. See, e.g., Gore, 517 U.S. at 577, 116 S.Ct. 1589.

Rather, the “repeated actions” by the individual defendants are the individual parts of their overall scheme to leave Horizon and form a competing business. Horizon is correct that all of the individual defendants were found to have committed multiple torts, but these findings do not. show recidivism. See id.; Bennett I, 315 S.W.3d at 878 n.55; Chapa, 212 S.W.3d at 308-09. We agree that the defendants’ misconduct was not isolated to theft of documents—they used the stolen property to solicit Horizon’s existing customers and secure future contracts for PRP.

It is also true that the individual defendants were found to have conspired to commit the underlying torts. Regardless, this is not the type of repeated misconduct contemplated by the Supreme Court or Texas courts when applying Campbell’s reprehensibility factors. This is not to say that we must look only to the precise contours of the defendants’ tortious actions. Rather, we have held that surrounding circumstances having a sufficient nexus to the underlying tort may be considered. Bennett I, 315 S.W.3d at 874-77.

For example, in Bennett I, the defendant rounded up and auctioned cattle that were on his land, despite questions from his own ranch hands about whether the cattle belonged to a neighbor with whom the defendant had long-running disputes over fencing and stray cattle. Id. at 869-71. When the neighbor sued for conversion, the defendant engaged in a remarkable variety of conduct intended to conceal his own culpability and instead implicate the plaintiff. Id. at 875-77.

The defendant’s efforts to cover up his actions included bribing and threatening witnesses, tampering with photographic evidence, suing and pm-suing the prosecution of a former ranch hand who cooperated with the neighbor’s investigation, and attempting to register the neighbor’s brand as his own. Id. We rejected the defendant’s argument that this conduct should not be considered in assessing the reprehensibility guidepost, holding that the malicious theft of cattle and the numerous efforts to conceal it showed that the plaintiffs harm “resulted from intentional malice, trickery, or deceit.” Id. at 877.

Similarly, when the former ranch hand later sued the defendant for malicious prosecution and sought exemplary damages, the Third Court of Appeals considered multiple extraneous acts of misconduct by the defendant and tied those misdeeds to the fifth reprehensibility factor but explicitly refused to say that the defendant’s misconduct was repeated. Bennett v. Grant, 460 S.W.3d 220, 250 n.18 (Tex. App.—Austin 2015), aff'd in part, rev’d in part on other grounds, — S.W.3d —, 2017 WL 1553157 (Tex. 2017).

Accordingly, none of the four reprehensibility factors examined to this point weigh in favor of the remitted exemplary damages award. Rather, as in Bennett I, much of the evidence on which Horizon relies to establish repeated misconduct speaks to the fifth factor—intentional malice, trickery, and deceit. Indeed, the final reprehensibility factor weighs in favor of an exemplary award because the evidence supports that the harm caused by each of the individual defendants was the result of intentional malice, trickery, or deceit rather than accident.

We discussed this point above, so we need not address it extensively here. As explained, there is evidence that each of the individual defendants acted with malice and that three of the individual defendants committed fraud, In sum, only one of the five reprehensibility factors is present in this case. We believe the fact that only Campbell’s fifth factor (harm resulting from intentional malice, trickery, or deceit) is present demands the conclusion that the exemplary damages award remains excessive despite the court of appeals’ suggested remittitur.

See Campbell, 538 U.S. at 425, 123 S.Ct. 1513 (refusing to adopt a bright-line rule but noting that “four times the amount of compensatory damages might be close to the line of constitutional impropriety”); Chapa, 212 S.W.3d at 308-10 (holding that a 4.33:1 ratio was unconstitutionally excessive where only the fifth reprehensibility factor was present and explaining that “[pjushing exemplary damages to the absolute constitutional limit in a case like this [e.g., purely economic harm] leaves no room for punishment in cases involving death, grievous physical injury, financial ruin, or actions that endanger a large segment of the public”). 2.

Disparity Between Compensatory and Exemplary Damages The parties raise unique arguments regarding the disparity between the compensatory and exemplary damages awards. First, Acadia argues that the disparity ratio should be assessed for constitutional excessiveness on a per-judgment rather than per-defendant basis.

The court of appeals utilized this per-judgment approach in its original decision, calculating the constitutionally permissible award by allocating a four-times multiplier of the total compensatory damages award among the individuals in the same proportion as the jury assessed punitive damages. 472 S.W.3d at 98-99.

On rehearing, however, the court of appeals agreed with Horizon’s argument that the ratio should be assessed for constitutional excessiveness on a per-defendant basis, calculating the constitutionally permissible award by allocating a four-times multiplier to each defendant’s joint and several liability for compensatory damages. Id.

The result on rehearing is that each defendant is liable for more than $220,000 in exemplary damages, even though each defendant is jointly and severally liable for approximately $55,000 in compensatory damages and four of five of the individual defendants caused harm to Horizon in amounts less than $11,000. Before this Court, Horizon advocates for a “per-defendant, joint-and-several” approach, while Acadia argues that we must calculate the ratio based on a “per-judgment” or “per-defendant, causation” approach. We consider these various approaches in the sections that follow. a.

Per-Judgment or Per-Defendant Approach Assuming temporarily that the court of appeals’ universal 4:1 ratio was appropriate—which it was not—the first question we must answer is whether the court of appeals erred by assessing the disparity betwpen the compensatory and exemplary damages award on a per-defendant rather than a per-judgment basis. We hold that the court of appeals did not err on this point. Our prior decisions support that the proper basis for assessing the constitutional excessiveness of an exemplary damages award is per-defendant rather than per-judgment. See, e.g., Thompson, 113 S.W. at 147.

This approach is also consistent with the underlying purpose and focus of exemplary damages—to punish the wrongdoer rather than to compensate the claimant. Courts that have examined this question generally agree that the constitutional inquiry into the amount of exemplary damages turns on their effect on the individual defendant. For example, in Huynh v. Phung, No. 01-04-00267-CV, 2007 WL 495023 (Tex.

App.—Houston [1st Dist.] Feb. 16, 2007, no pet.) (mem. op.), the First Court of Appeals rejected the argument “that damages should be assessed based on a ratio of combined exemplary damages to compensatory damages, rather than ... based on a ratio of individual exemplary damages to compensatory damages.” Id. at *13.

As that court explained, “[bjecause exemplary damages are awarded on an individual basis and require separate jury findings as to each defendant, the proper method for reviewing the constitutionality of exemplary damages is to compare the exemplary damages awarded against each individual with the compensatory damages awarded against that individual.” Id. The First Court of Appeals reached the same conclusion in Carlton Energy Group, LLC v. Phillips, 369 S.W.3d 433, 460 (Tex. App.—Houston [1st Dist.] 2012), aff'd in part, rev’d in part on other grounds, 475 S.W.3d 265 (Tex. 2015).

In Phillips, the court upheld as constitutional exemplary damages awards of $8.5 million each against two defendants by “comparing the amount of exemplary damages awarded against [the defendants] separately to the amount of actual damages awarded against them jointly and severally.” Id. at 460-61. The outcome in those cases also corresponds with

section 41.006 of the Texas Civil Practice and Remedies Code, under which exemplary damages are assessed strictly on an individual basis. We have found no decision by a Texas court adopting the per-judgment rule that Acadia proposes. In fact, in other contexts this Court has evaluated exemplary damages by weighing their impact on the individual defendant. For example, when considering the statutory cap on wrongful-death damages, we held that the cap should be calculated on a per-defendant basis. See Rose v. Doctors Hosp., 801 S.W.2d 841, 846 (Tex. 1990).

In Rose, the jury awarded $1.3 million in actual damages jointly and severally against two defendants. Id. at 846-47. Because the statutory cap “clearly applies to the recovery against the individual defendant, not the award to the individual plaintiff,” the trial court’s judgment was proper because it was less than double the statutory cap, which is “multiplied by two since there are two culpable defendants.” Id. In shoi’t, our past decisions make clear that damages caps should be calculated on a “per defendant” basis.

Moreover, calculating the constitutionally permissible ratio by comparing the total compensatory damages recoverable to the total exemplary damages awarded fails to contemplate the scenarios in which defendants act with varying degrees of reprehensibility—as is the case here. It also runs counter to the court’s task of determining whether any of the defendants’ due-process rights were violated and shifts the court’s focus away from a particular defendant’s conduct to the defendants’ conduct collectively.

Because exemplary damages are assessed by the jury on a per-defendant basis according to each defendant’s culpability, the alleged excessiveness of those damages must also be evaluated on a per-defendant basis. See Tex. Civ. Prac. & Rem. Code § 41.006; Thompson, 113 S.W. at 147. Consequently, we hold that the court of appeals did not err by analyzing the constitutionality of the exemplary damages awards on a per-defendant basis on rehearing. b. Determining the Denominator for the Per-Defendant Approach Our determination that the ratio must be calculated on a per-defendant basis answers only half of the question.

We still must determine whether the joint-and-several compensatory award or the com-pensable harm caused by each defendant is the proper denominator for calculating the ratio of compensatory to exemplary damages. That is, we must determine whether courts must calculate the ratio based on the harm each individual defendant caused rather than the harm for which each individual defendant is jointly and severally liable.

Acadia argues that if ' exemplary damages are not to be assessed using the per-judgment approach, then they should be assessed by comparing the exemplary damages per individual defendant to the actual damages per individual defendant: Exemplary damages per individual defendant Actual damages per individual defendant In other words, Acadia argues that comparing the individual awards of exemplary damages to the aggregate amount of actual damages awarded against all defendants yields an unconstitutionally excessive exemplary damages award under the facts of this case. We agree.

Given that the purpose of exemplary damages is to punish defendants for civil wrongs and to deter future misconduct, the punishment should fit the misconduct rather than a state’s joint and several liability regime. Compensatory and exemplary damages undoubtedly serve different purposes, and the joint and several liability regime is intended to further the purpose of compensatory damages—redressing concrete losses caused by the defendant’s wrongful conduct.

Stated differently, we do not believe that a person should be punished by the assessment of exemplary damages simply because of his or her ability to pay, which is the case when a defendant is held jointly and severally liable for an entire compensatory damages award even though that defendant only caused a small portion of the plaintiffs injury.

We note that while Texas law allows defendants faced with joint-and-several compensatory awards to seek contribution from the other wrongdoers if required to pay the entire award, this is not the case with exemplary damages given that Texas law does not allow joint-and-several exemplary damages awards. See Tex. Civ. Prac. & Rem. Code § 41.006.

Thus, we hold that in determining the basis for a constitutionally permissible amount of exemplary damages, courts must consider the harm each defendant actually caused and assess the punishment based on that harm because this approach most closely matches the punishment to each defendant’s misconduct. As explained above, the jury found that of the $55,049.24 in actual damages for which Horizon is entitled to recovery, Saul caused $20,000, Ulasewicz caused $8,898.45, Palus caused $10,101.41, Bayma caused $8,549.39, and Piechocki caused $7,500.

Employing an approach that bases the ratio’s calculation on the harm caused by each defendant, as found by the jury, the court of appeals’ suggested remit-titur results in ratios of compensatory to exemplary damages against each individual defendant as follows: Table 5 Defendant Saul Ulasewicz Bayma Palus Piechocki Total Remitted Exemplary Damages $220,196.96 $220,196.96 $220,196.96 $220,196.96 $220,196.96 $1,100,984.80 Total Damages Caused $20,000.00 $8,898.45 $8,549.38 $10,101.41 $7,500.00 $55,049.24 Ratio 11:1 24.7:1 25.8:1 21.8:1 29.4:1 20:1 In light of these ratios, coupled with our conclusion that only one of the five reprehensibility factors is present in this case, we hold that the exemplary damages award remains unconstitutionally excessive despite the court of appeals’ suggested remittitur.

Given the purely economic nature of the harm, the fact that the individual defendants’ misconduct cannot properly be categorized as “repeated,” the lack of any reckless disregard for the health or safety of others, and Horizon’s financial status, we conclude that the imposition of ratios far exceeding 4:1 is improper as to the individual defendants. We recognize that the Supreme Court has stated that Gore’s second guidepost—the ratio—becomes less important as the amount of compensatory damages decreases.

That is, when compensatory damages are relatively modest but the other guideposts point strongly in favor of exemplary damages, even a relatively large ratio of compensatory to exemplary damages may not offend due process. See Gore, 517 U.S. at 582, 116 S.Ct. 1589 (“[L]ow awards of compensatory damages may properly support a hi

[…]