PER CURIAM:
Aspen Technology (Aspen) sued Tekin Kunt, a former employee, for violating a noncompete clause in his employment contract. It subsequently added Kunt’s new employer, M3 Technology (M3), as a defendant and, brought claims for tortious interference with contract, trade secret misappropriation, and copyright infringement. A jury found in favor of Aspen and
the district court issued a permanent injunction. M3 appeals the district court’s denial of its motion for judgment as a matter of law as to Aspen’s misappropriation and infringement claims and further appeals the damages award and the grant of a permanent injunction. We affirm the district court’s rulings on M3’s motion and the grant of a permanent injunction. However, we vacate the award of attorney’s fees and remand for the issuance of an adjusted damages award.
I
Aspen and M3 both develop, sell, and service specialized software for chemical and petrochemical companies. They compete directly in three areas:
(1) scheduling (Aspen’s Orion vs. M3’s SIMTO Scheduling);
(2) blending (Aspen’s MBO vs. M3’s SIMTO M-Blend); and
(3) distribution (Aspen’s DPO vs. M3’s SIMTO Distribution).
Aspen is the older of the two companies. M3 was founded in 2002 by Dong Dong, Lei Wu, and Lawrence Pan, all of whom are former Aspen employees. In 2005, David Jasper, also a former Aspen employee, became the fourth employee and shareholder of M3. In 2008 and 2009 respectively, Robert Hutchings and Craig Acuff— also former Aspen employees—joined M3.
Initially, M3 developed and sold terminal software, a product not sold by Aspen, because Dong’s employment agreement with Aspen included a two-year noncom-petition clause. Following this two-year period, M3 began selling its scheduling product, which competes with Aspen’s Orion. M3 developed its blending software, M-Blend, which competes with Aspen’s MBO, shortly thereafter. At least by 2011, all three of M3’s products that compete with Aspen’s products were on the market.
In January 2010, Kunt, Aspen’s director of technology and research, resigned and joined M3. Kunt had previously signed a noncompete agreement with Aspen, which required Kunt to refrain from competitive employment and from soliciting Aspen customers with whom he had worked for one year after leaving Aspen. In April 2010, Aspen initiated this suit against Kunt to enforce its contract. In June and July 2010, Aspen filed copyright registrations for versions of its software beginning in 2006 through 2010. In July 2010, Aspen amended its complaint to assert claims against M3 for trade-secret misappropriation, copyright infringement, and tortious interference with its contract with Kunt. Aspen and Kunt eventually settled their dispute, leaving only Aspen’s claims against M3.
During discovery, M3 employees engaged in an array of misconduct. Kunt admitted to having concealed Aspen property at a friend’s home, including a laptop with Aspen documents, an external hard drive containing Aspen’s source code, and technical notes Kunt had made while working on Aspen’s software. Acuff revealed that he had removed an external hard drive containing Aspen’s confidential pricing calculators from Aspen. Furthermore, Acuff admitted to initiating a defragmentation of his laptop—a process that makes the subsequent recovery of a deleted file “nearly impossible”—before it could be imaged during discovery. On advice from counsel, Acuff invoked the Fifth Amendment during his deposition and at trial. Hutchings also initiated a defragmentation of his laptop before it could be imaged.
At trial, the jury heard evidence of additional misconduct, namely that Pan, the chief developer of M3’s M-Blend software, was in possession of a confidential manual
for Aspen’s MBO blending software, and that Jasper had a customer list on his computer, the metadata of which suggested that it had been created while Jasper was still in Aspen’s employ.
At the close of Aspen’s case, M3 moved for judgment as a matter of law pursuant to Federal Rule of Civil Procedure 50(a) on the grounds that Aspen’s misappropriation and infringement claims were barred by statutes of limitations and, alternatively, that Aspen had failed to present sufficient evidence on its claims. This motion was denied. The case was tried before a jury, which found, in relevant part, that none of Aspen’s claims was time-barred; that M3 had misappropriated eight of Aspen’s trade secrets; that M3 had infringed all of Aspen’s asserted copyrights; and that M3 had interfered with Aspen’s contract with Kunt.
The jury also awarded the following damages to Aspen in the following categories:
A. Tortious interference:
1. Damages (including attorney’s fees): $896,332
2. Exemplary damages: $100,000
B. Trade-secret misappropriation:
1. Aspen’s lost profits: $2,000,000
2. M3’s profits: $2,800,000
3. Exemplary damages: $1,000,000
C. Copyright infringement:
2. M3’s profits: $2,900,000
TOTAL: $11,696,332
Pursuant to Aspen’s request, the district court reduced the tortious interference damages to $546,329 to reflect only Aspen’s requested attorney’s fees. The district court awarded the adjusted total of $11,346,329, which reflects the remainder of the jury verdict in its entirety. The district court also issued a permanent injunction prohibiting M3 from selling any presently existing or future derivative versions of those products that use or contain any information substantially derived from Aspen’s protected material. M3 renewed its motion for judgment as a matter of law under Rule 50(b), which the court also denied. This appeal followed.
II
“This court reviews a district court’s denial of a motion for judgment as a matter of law de novo.”
Such motions “in an action tried by jury [are] a challenge to the legal sufficiency of the evidence supporting the jury’s verdict.”
In reviewing a jury verdict, this court considers “all of the evidence—not just that evidence which supports the non-mover’s case—but in the light and with all reasonable inferences most favorable to the party opposed to the motion,” that is, Aspen.
“If the facts and inferences point so strongly and overwhelmingly in favor of one party that the Court believes that reasonable men could not arrive at a contrary verdict, granting [judgment as a matter of law] is proper.”
III
We first address M3’s contention that the statutes of limitations barred Aspen’s
misappropriation and infringement claims.
Causes of action for trade-secret misappropriation and copyright infringement both have a three-year limitations period.
While causes of action generally accrue “when a wrongful act causes some legal injury, even if the fact of the injury is not discovered until later, and even if all resulting damages have not yet occurred,”
several equitable tolling doctrines may defer the accrual of a claim. Specifically, the discovery rule and the doctrine of fraudulent concealment apply to both misappropriation and infringement claims.
Here, the jury had sufficient evidence to find that Aspen did not discover and in the exercise of reasonable diligence should not have discovered the asserted misappropriation until after Aspen filed suit against Kunt. First, M3’s marketing of competing products alone did not constitute “discovery of the injury.” Aspen must have known or have been reasonably able to discover that M3 had used Aspen’s proprietary information to create competing products, and M3 points to no such conclusive evidence.
Furthermore, the jury could have found that M3 had fraudulently concealed its misconduct. In a letter to Aspen, M3 denied all allegations of misappropriation and infringement. The denial of wrongdoing “may constitute concealment ... [if] the circumstances indicate that it was reasonable for [Aspen] to rely on [M3’s] denial.”
Given that the jury found that M3 had misappropriated and infringed Aspen’s protected informa
tion, it could have also reasonably concluded that MB’s denial was intended to conceal its wrongdoing. Therefore, M3 was not entitled to judgment as a matter of law on its limitations defense.
IV
M3 next argues that the jury’s verdict, which found that M3 had misappropriated eight of Aspen’s trade secrets, was not supported by legally sufficient evidence. These eight trade secrets are as follows: (1) DPO source code, (2) MBO source code, (3) Orion and MBO V7 Roadmap, (4) Scheduling
&
Blending V8
&
Beyond, (5) MBO 1.3.1 User’s Manual, (6) SDPIMS code, (7) Customer list, and (8) Pricing calculators. “A trade secret is any formula, pattern, device or compilation of information which is used in one’s business and presents an opportunity to obtain an advantage over competitors who do not know or use it.”
Trade-secret misappropriation in Texas requires (1) the existence of a trade secret, (2) improper acquisition of that secret, and (3) use of the trade secret without authorization.
“The word ‘use’ ... has been equated with ‘commercial use,’ and has been liberally defined to include both the sale of the trade secrets themselves, as well as the ‘commercial operation’ of those trade secrets within profit-seeking endeavors.”
This court has recently acknowledged the even more liberal construction of the term “use” contained in the
Restatement (Third) of Unfair Competition,
which states that “[a]s a general matter, any exploitation of the trade secret that is likely to result in injury to the trade secret owner or enrichment to the defendant is a ‘use’ under this Section.”
Of these eight violations, we first address the four unrelated to Aspen’s source code and thén turn to those related to its source code.
A
The first four materials refer to the customer list in Jasper’s possession, the pricing calculators in Acuffs possession, and two product development road maps.
M3 argues that Aspen never presented any evidence that M3 actually used the confidential information found in its possession.
M3 is incorrect that no evidence of use was presented. Jasper admitted to having used the customer list after joining M3. Jasper’s explanations that support a finding of non-use are not so “overwhelming” that reasonable jurors could not have concluded that the customer list was in existence while Jasper was employed at Aspen
arid therefore, that its use by M3 constituted misappropriation.
There was also sufficient evidence of use as to the pricing calculators. These confidential pricing calculators were discovered on Acuff s external hard drive and when he was questioned as to their use, Acuff invoked his Fifth Amendment right and refused to answer. The district court permitted an adverse-inference jury instruction, directing that the jury could “infer by [ ] Acuff s refusal to answer that his answers to the questions posed to him would have been adverse to M3’s interests” so long as the jury did “not base [its] verdict solely on that adverse inference.” The record demonstrates that Aspen also presented circumstantial evidence of use by eliciting testimony that - pricing is a highly individualized, confidential process; that Acuff was one of the primary negotiators for M3; that M3 directly competed against Aspen in almost all its projects; and that between 2009 and at least through early 2010—i.e., the time period Acuff worked at M3—M3 lost only one of its bids against Aspen. Furthermore, because Acuff defragmented his laptop hard drive, the district court permitted the jury to “infer that the content of those deleted ... documents or information would have been unfavorable to M3.” Given the circumstantial evidence presented and the two adverse-inference instructions, the jury had a legally sufficient basis from which to determine that M3 had misappropriated Aspen’s pricing calculators.
Finally, Aspen presented sufficient evidence on the use of the two road maps. Though evidence of possession itself cannot support an inference of actual use,
Aspen presented testimony on three specifie ways M3 could have used these road maps to “tremendous advantage”: (1) permitting M3 to create a competing product, (2) showing M3 where it need not waste its time and resources, and (3) revealing the needs that would not be met by Aspen’s anticipated design.
Aspen also presented evidence to suggest that Dong and Hutchings had personally solicited information about Aspen’s road maps from other companies. As Aspen’s expert testified, road maps are sometimes shared with Aspen’s customer community to obtain feedback, and M3 had contacts with customers.
The jury could have reasonably inferred from this calculated attempt to acquire confidential information that M3 used the road maps in its possession. Moreover, because Hutchings also violated instructions regarding discovery by initiating a defragmentation of his hard drive, the jury was once again entitled to infer that the deleted or altered information would have been unfavorable to M3.
As before, M3 presented contrary testimony that it had completed its own product plans years before Aspen’s road maps were created. However, the jury was entitled to make credibility determinations, and it found Aspen’s version to be more compelling.
B
The remaining trade-secret violations—MBO User’s Manual, MBO source code, DPO source code, and SDPIMS code—all concern Aspen’s source code. M3’s argument consists of isolating each of the similarities presented at trial between its source code and that of Aspen’s to assert that none of the common terms are
trade secrets. However, Aspen never contended that these common terms in and of themselves were trade secrets, but rather that the jury could infer from these commonalities that M3 had misappropriated Aspen’s source code—the actual trade secrets at issue. We hold that there was sufficient evidence as to all four source code violations.
First, there was evidence as to the misappropriation of Aspen’s MBO User’s Manual and MBO source code. As discussed above, trade-secret misappropriation requires a showing that a trade secret existed, that it was improperly acquired, and that it was used without authorization.
It is undisputed that the MBO User’s Manual and MBO source code were trade secrets and that Pan improperly acquired them. As to their unauthorized use, Aspen’s expert testified that M3 could have used the manual “to create source code” and to generally gain a competitive advantage by determining how the program operated and creating competing source code. Furthermore, “[ejvidence of a similar product may give rise to an inference of actual use under certain circumstances,” such as when the defendant’s product was quickly developed by someone who had recently resigned from the plaintiff-company.
Here, Pan was the primary developer of M3’s competing M-Blend product, and he developed this software in six months. As a point of comparison, Aspen’s expert testified that Aspen’s MBO took multiple people years to develop. Therefore, the jury had a legally sufficient basis to find that M3 had misappropriated these trade secrets.
Next, the jury’s verdict as to Aspen’s DPO source code was also supported by sufficient evidence. Again, it is undisputed that this source code was a trade secret and that it was improperly acquired. As to M3’s unauthorized use, Aspen’s expert presented evidence that several unusually identical structures appeared in both Aspen’s DPO source code and in M3’s competing distribution product. This expert explained that there are “innumerable ways” to design source code since it involves a number of choices, left entirely to the programmer’s discretion, as to the structure, organization, and naming of lines of code. There were various instances where Aspen’s DPO source code appears verbatim in M3’s distribution source code, including not only sequences of numbers but also names of variables within the code. Additionally, both M3 and Aspen used the same large numbers as boundaries for infinity in their source code even though these numbers did not represent an industry convention. Furthermore, despite appearing in M3’s code, these boundary values did not actually serve a purpose.
Finally, the jury’s verdict that M3 had misappropriated Aspen’s SDPIMS source code—the predecessor of Aspen’s DPO software—also must be upheld. Aspen demonstrated that SDPIMS was undeniably source code and that it was improperly in M3’s possession. Its expert also noted that access to this code could harm Aspen in the same three ways it had specified in earlier testimony—by permitting M3 to create a competing product, save time and resources, and capitalize on gaps in the market. Although M3 elicited testimony from this same expert suggesting that SDPIMS never actually appeared in any Aspen product, this does not foreclose
the possibility that M3 used the older version of Aspen’s code to avoid making the same mistakes, assuming there were any, or as a baseline for developing its own competing software, either of which would constitute use under the
Restatement.
In sum, the jury had a legally sufficient basis for the eight misappropriation findings; accordingly, M3 was not entitled to judgment as a matter of law on any of Aspen’s trade-secret misappropriation claims.
V
Next, M3 challenges the jury’s verdict as to Aspen’s copyright infringement claims. To establish an infringement claim, a copyright owner must prove ownership of a valid copyright and copying of constituent elements of the work that are original.
M3 asserts that Aspen failed to provide sufficient evidence as to either element. We address both assertions below.
M3 first argues that Aspen did not present sufficient evidence of its ownership interest. “To establish ‘ownership,’ the plaintiff must prove that the material is original, that it can be copyrighted, and that he has complied with statutory formalities.”
M3 contests only the last element, arguing pursuant to Circular 61—a publication issued by the Copyright Office—that Aspen’s registered versions of its software do not protect its prior versions.
M3’s contention fails as a matter of law. As an initial matter, circulars may be instructive but are not authoritative.
More significantly, however, Circular 61 does not support M3’s position. Although M3 contends that under Circular 61, registration of a derivative work does not protect “preexisting material,” this Circular instructs only that derivative works must themselves be registered in order for newer material not contained in preexisting software to be protected.
Thus, although Circular 61 implies that copyrighting a derivative work affects a party’s rights as to the new material, it says nothing about the effects on preexisting work.
Furthermore, the case law weighs against M3’s position. In
In re Independent Service Organizations Antitrust Litigation.
the defendant made the same argument M3 makes here, urging that the plaintiff was “barred from bringing its copyright infringement [ ]claims because [the plaintiff] ha[d] not come forward with the registrations covering the preexisting materials from which the registered up
grades were derived.”
The court rejected this argument, holding that the plaintiffs “registration of the derivative works [wa]s sufficient to allow an infringement claim based on copying of the material, either newly added or contained in the underlying work.”
This position has been recognized by at least five circuits
and by district courts in two others.
Therefore, Aspen’s registration of its derivative materials permits Aspen to bring a claim that M3 had infringed preexisting versions of its software.
To prove actionable copying, a plaintiff must show (1) “as a factual matter” that the defendant “actually used the eopyrighted material to create his own work” and (2) that the copying is “legally actionable.”
Copying is legally actionable if “the allegedly infringing work is substantially similar to protectable elements of the infringed work. [A] side-by-side comparison must be made between the original and the copy to determine whether a layman would view the two works as ‘substantially similar.’ ”
The question of whether two works are substantially similar is typically left to the ultimate factfinder.
To aid the factfinder in this comparison, this court has “generally endorsed the ‘abstraction-filtration-comparison’ test first outlined by the Second Circuit in
[Computer Associates International, Inc. v. Altai,
/tcc.
].”
This three-step procedure first
“dissect[s] the allegedly copied program’s structure and isolate[s] each level of abstraction contained within it.”
The court next filters out any unproteetable expression,
which, in this context, means removing from comparison any hardware and software standards, computer manufacturer design standards, target industry practices and demands, and computer industry programming practices.
Lastly, the court compares the “remaining protectable elements with the allegedly infringing program to determine whether the defendants have misappropriated substantial elements of the plaintiffs program.”
Here, Aspen’s expert explained that he had followed the
Altai
framework and discovered three nonliteral patterns that were shared between Aspen’s and M3’s codes: (1) both used the same three steps to process cells across different spreadsheets; (2) both used “parsing” to execute Excel formulas without using Excel; and (3) both used the power function device in a significant inflow/outflow calculation. None of the similarities were filtered out, the expert explained, since they were not dictated by technical limitations or required by petroleum industry standards, and these similarities were essential to the operation of Aspen’s product. Though M3 contends that this testimony was insufficient to demonstrate proper filtration or comparison, the jury heard M3’s account as to why the three similarities should have been removed from the comparison. The jury, as the ultimate factfinder, was entitled to determine whether the copied aspects of the program were entitled to copyright protection. We cannot say as a matter of law that the expert testimony was so lacking as would cause reasonable minds to reject the conclusion drawn by the jury in this case that Aspen’s claims were legally actionable. Accordingly, M3 was not entitled to judgment as a matter of law on Aspen’s copyright infringement claims.
VI
M3 next argues that the district court erred in denying its Rule 50(b) motion because Aspen failed to present sufficient evidence of damages. As an initial matter, M3’s argument that Aspen’s damages model does not account for the infringing fraction of M3’s source code fails. Although M3 is correct that, relatively speaking, few lines of its source code contained infringing material, it ignores the fact that those few lines could have been integral to the development of M3’s competing products. Aspen presented testimony that this was in fact the ease. Aspen was therefore permitted to make the same assumption in its damages model.
M3’s argument that Aspen did not present sufficient evidence to support its asserted amount of damages likewise fails. With respect to its damages on its copyright claims, Aspen was “permitted to recover [its] own actual damages, including
lost profits”
and
“any profits of [M3] that are attributable to the infringement.”
To demonstrate M3’s profits, Aspen was required to present proof of M3’s gross revenue, i.e., the “revenue
reasonably related
to the infringement,” after which M3 was required “to prove [its] deductible expenses and the elements of profit attributable to factors other than the copyrighted work.”
The record reflects that Aspen’s expert presented detailed testimony regarding both theories of damages based on Aspen’s and M3’s financial records. This testimony provided a legally sufficient basis for the award of damages on Aspen’s copyright claims.
Aspen also adequately supported its misappropriation damages award. Aspen was permitted to prove its damages on these claims in a number of ways, including by demonstrating its lost profits or the amount by which M3 was unjustly enriched based on its misappropriation.
As discussed above, Aspen’s expert testimony likewise provided sufficient evidence for this damages award.
M3’s next contention—that the district court impermissibly stacked the damages award for Aspen’s claims—also fails. Although “[a] plaintiff seeking compensation for the same injury under different legal theories is ... only entitled to one recovery,” a “jury’s award is not duplicafive simply because it allocates damages under two distinct causes of action.”
Where there is evidence that the plaintiff suffered the total amount of the damage award on one of its theories, a court may rationally conclude that the jury found that the plaintiff suffered that amount of injuries and “merely allocated that amount between the two different causes of action.”
In this case, Aspen’s expert testified that M3 may have been unjustly enriched in an amount ranging from $5.5 million to $9.7 million; the total compensatory damages award fell within this range. Accordingly, the jury’s and, by extension, the district court’s damages award did not grant Aspen a double recovery.
Last, M3 argues that if judgment is upheld on the trade secret theory alone, the damages award would need to be modified since this theory permits recovery either of lost profits or unjust enrichment, but not both.
However, as discussed above, there is legally sufficient evidence to uphold the jury’s verdict on Aspen’s copyright claims, and copyright law permits the recovery of both lost profits
and
unjust enrichment.
VII
M3 next argues that the tortious interference award, which allotted only at
torney’s fees, fails as a matter of law. We agree.
“In Texas, attorney’s fees may not be recovered from an opposing party unless such recovery is provided for by statute or by contract between the parties.”
A number of equitable exceptions to this general rule have been recognized, one of which provides for attorney’s fees “where a plaintiff has been involved in litigation with a third party as a result of the tortious act of another.”
Certain prerequisites must be met, including that “the litigation must have involved a third party and must not have been brought against the defendant in the same action in which the fees are sought.”
Particularly on point is
Brown & Brown of Texas, Inc. v. Omni Metals,
Inc.,
in which a Texas court of appeals vacated the recovery of attorney’s fees incurred in an earlier stage of litigation, which was initially brought against other parties and to which the appellants were later added.
These fees could not be awarded because the earlier stage could “not properly be styled [as] ‘prior litigation.’ ”
It is undisputed that no statute or contract provided for the recovery of attor-
ney’s fees incurred by Aspen in its dispute involving Kunt.
Moreover, since both M3 and Kunt were involved in this litigation, Aspen cannot recover from M3 the attorney’s fees it expended to litigate against Kunt in the prior stage of litigation. We hold, therefore, that the district court erred in awarding Aspen its attorney’s fees under this equitable exception to the general rule. Because Aspen requested and was awarded $546,329 in attorney’s fees, we direct the district court to reduce the judgment-by this amount on remand.
VIII
M3’s final objection is to the district court’s grant of a permanent injunction. “According to well-established principles of equity, a plaintiff seeking a permanent injunction must ... demonstrate: (1) that it has suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balances of hardships between the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent
injunction.”
M3 argues that the district court erroneously presumed irreparable harm, ignored adequate alternatives, and failed to consider the public interest.
We review the grant of a permanent injunction for abuse of discretion.
“An abuse of discretion occurs if the district court [ ] relie[d] on clearly erroneous factual findings when deciding to grant or deny the permanent injunction, [] relie[d] on erroneous conclusions of law ..., or [ ] misapplied] the factual or legal conclusions when fashioning its injunctive relief.”
With these principles guiding us, we turn to M3’s challenges to the injunction. First, the district court did not abuse its discretion in finding that Aspen would suffer irreparable harm based on evidence that M3 had been “making use of As-pent ]’s information from the inception of its business,” and that there was a “substantial probability that Aspenf] [would] be unable to collect a judgment against M3” in light of M3’s initiation of bankruptcy proceedings.
Regarding adequate alternatives, requiring M3 to discontinue the use or sale of its products was not unreasonable because M3’s competing products were found to contain Aspen’s source code.
Finally, we agree that it was in the interest of public policy to prohibit the sale and use of M3 products containing infringing source code and that were derived from the improper misappropriation of trade secrets.
Nor was this permanent injunction “overbroad and non-specific.” The jury made specific findings as to which trade secrets and copyrights M3 had used, and the district court’s injunction accordingly enumerates the trade secrets and copyrights that M3 is enjoined from using. Accordingly, the district court did not abuse its discretion in granting the permanent injunction.
For the foregoing reasons, we AFFIRM the district court’s judgment in all respects except the grant of attorney’s fees for Aspen’s tortious interference claim. We direct that the district court reduce the judgment by the attorney’s fee amount on remand.