Reply Brief — Trinseo Europe GmbH, Petitioner v. Kellogg Brown & Root, L.L.C., et al.

Supreme Court briefAug 25, 2026

Ask Donna

What actually matters in this document.

Text

No. 25-1373

IN THE

Supreme Court of the United States

TRINSEO EUROPE GMBH,

Petitioner,

KELLOGG BROWN & ROOT, L.L.C.;

STEPHEN HARPER, also known as STEVE HARPER;

STEVE HARPER CONSULTING, INCORPORATED;

POLYCARBONATE CONSULTING SERVICES, INCORPORATED,

Respondents.

On Petition for Writ of Certiorari

to the United States Court of Appeals

for the Fifth Circuit

REPLY BRIEF OF PETITIONER

STEWART HOFFER

JUSTIN R. BRAGA

HICKS THOMAS LLP

700 Louisiana Street,

Suite 2300

Houston, Texas 77002

CANDICE C. WONG

Counsel of Record

ELIZABETH J. KALANCHOE

NICHOLAS ARDITO

FRIED, FRANK, HARRIS,

SHRIVER & JACOBSON LLP

801 17th Street, N.W.

Washington, DC 20006

(202) 639-7000

Candice.Wong@friedfrank.com

Counsel for Petitioner

TABLE OF CONTENTS

TABLE OF AUTHORITIES .................................... ii

INTRODUCTION .................................................... 1

ARGUMENT ............................................................ 2

I. The Decision Below Conflicts With Decisions

of this Court and Other Circuits. ................... 2

II. The Decision Below Is Wrong. ........................ 7

III.This Case Is a Clean Vehicle. ....................... 11

CONCLUSION....................................................... 12

(i)

(ii)

TABLE OF AUTHORITIES

Cases

Page(s)

Bigelow v. RKO Radio Pictures,

327 U.S. 251 (1946) ............................................ 2, 3

Caudill Seed & Warehouse v. Jarrow

Formulas, Inc.,

53 F.4th 368 (6th Cir. 2022) .............................. 5, 6

Eastman Kodak Co. of NY v. Southern

Photo Materials Co.,

273 U.S. 359 (1927) ............................................ 2, 3

EchoSpan, Inc. v. Medallia, Inc.,

No. 24-4751, 2025 WL 3046753 (9th

Cir. Oct. 31, 2025) .......................................... 5, 6, 7

Insulet Corp. v. EOFlow,

176 F.4th 1347 (Fed. Cir. 2026) ............................. 5

O2 Micro Int’l Ltd. v. Monolithic Power

Sys., Inc.,

399 F.Supp.2d 1064 (N.D. Cal. 2005),

aff’d per curiam, 221 Fed. App’x 996

(Fed. Cir. 2007) ................................................... 5, 6

Sea-Land Servs. Inc. v. Gaudet,

414 U.S. 573 (1974) ............................................ 2, 3

Story Parchment v. Paterson Parchment

Paper Co.,

282 U.S. 555 (1931) ................................................ 2

(iii)

TABLE OF AUTHORITIES—Continued

Page(s)

Syntel Sterling Best Shores Mauritius

Ltd. v. TriZetto Grp., Inc.,

No. 15 Civ. 211, 2024 WL 1116090

(S.D.N.Y. Mar. 13, 2024) ........................................ 8

Texas Advanced Optoelectronic Sols.,

Inc. v. Renesas Electronics America,

Inc.,

895 F.3d 1304 (Fed. Cir. 2018) .......................... 5, 6

University Computing Co. v. LykesYoungstown Corp.,

504 F.2d 518 (5th Cir. 1974) ................................ 10

OTHER AUTHORITIES

U.S. Const. amend. VII ............................................... 9

INTRODUCTION

In the face of rudimentary principles of jury

deference, the decisions below zeroed out a $77 million

damages award that the jury deemed warranted by

respondents’ theft of multiple Trinseo trade secrets.

Respondents’ brief in opposition clings to their $77

million windfall, principally by racing away from the

question presented. As if respondents had not just

spent years litigating for “per-trade-secret” damages

apportionment, respondents now act as if it has

precious little to do with this case.

Papering over lower-court divisions, respondents

trumpet consensus over the high-level principle that

evidence should link damages to the misappropriated

technology, attributing different outcomes to different

applications of that principle. But the real issue is one

level down: whether that evidence requires, as a

matter of law, individually valuating each jury-found

secret (or jury-found group) or providing a

methodology to do so.

That was the ruling

respondents sought and won. And that was the

measure by which Trinseo’s proffered evidentiary

basis was deemed no basis at all.

In so ruling, the Fifth Circuit set a legal baseline

that sharpens divisions over a commonly-recurring

scenario—where a jury finds some but not all the

alleged secrets, yet the damages evidence at trial went

to all the alleged secrets. Those divisions have

whipsawed litigants, alternately saving or felling jury

awards. And they are reflected in a common cadre of

cases routinely trotted out by courts whenever

apportionment rears its head—with the Federal and

now Fifth Circuits on one side, and Sixth and Ninth

Circuits on the other. Respondents brush off the

disarray as fact-specific, but the irreconcilable

outcomes stem from the different legal lenses through

which award underpinnings are viewed.

The Fifth Circuit’s ruling was also wrong. Seeking

to defend the indefensible—that Trinseo should get

zero damages for proven thefts of its technology—

respondents essentially claim that Trinseo asked for

this. But the notion that Trinseo strategically opted

to risk recovering nothing is absurd; in fact, market

realities about integrated PC packages precluded pertrade-secret apportionment. There is no defending the

extraordinary conclusion that there was no basis in

evidence—none—to uphold the jury’s more-thanreasonable approximations. The Fifth Circuit so

concluded only by redefining the ordinarily-modest

evidentiary threshold into a highly specific, inflexible

bar, defying the customary respect to which the jury’s

judgment is owed.

ARGUMENT

The Decision Below Conflicts With Decisions

of this Court and Other Circuits.

Respondents’ brief makes no effort to square the

jury-award-nullification below with this Court’s

longstanding

precedents

establishing

that

uncertainties that go “to the extent of the damage,” not

“the fact of damage,” do not preclude an award. Story

Parchment v. Paterson Parchment Paper, 282 U.S.

555, 562-63 (1931); cf. Pet.20, 29 (citing Eastman

Kodak v. Southern Photo Materials, 273 U.S. 359

(1927); Bigelow v. RKO Radio Pictures, 327 U.S. 251

(1946); and Sea-Land Servs. v. Gaudet, 414 U.S. 573

(1974)).

The fact of damage here was undisputable; the

liability evidence “establishing Trinseo’s right to …

damages” was “substantial.” Pet.App.89. Yet with

only the extent of Trinseo’s damages at issue,

respondents fail to acknowledge those precedents, let

alone reconcile the Fifth Circuit’s decision with the

jury’s

“fair

latitude

make

reasonable

approximations,” without “exactness,” on “probable

and inferential” bases. Sea-Land Servs., 414 U.S. at

590; Eastman Kodak, 273 U.S. at 379; Bigelow, 327

U.S. at 264.

Respondents fare little better trying to reconcile the

decisions of the Fifth and other Circuits. The Fifth

Circuit’s decision was not “strict,” they say, because

the Fifth Circuit did not call it that. Opp.Br.19. But

respondents cannot deny that “strict apportionment”

is lifted from the district court opinion the Fifth

Circuit affirmed.

Pet.App.52-53.

And while

respondents tie themselves into knots to avoid the

term “per-trade-secret apportionment,” that is what

they sought—at every stage of the proceedings1—and

what the courts delivered. The district court, in

granting JMOL, demanded “(1) evidence that

apportions value per trade secret, or (2) evidence that

provides some methodology … for how the jury may do

so itself.” Pet.App.64. Following suit, the Fifth Circuit

cited, as the means of apportionment that “come to

mind,” (1) “individually valu[ating] each alleged trade

E.g., ROA.13392 (instruction failed to require jury to

“apportion[] damages … trade secret by trade secret”);

D.C.Dkt.327 at 12, 25 (Trinseo was required to “analyze on a

trade secret-by-trade secret basis”); D.C.Dkt.334 at 12 (“plaintiffs

[must] try their cases … on a trade secret by trade secret basis”);

D.C.Dkt.385 at 12 (criticizing Trinseo’s “position that it was not

required to apportion … on a trade secret-by-trade secret basis”);

C.A.Dkt.134 at 43 (Trinseo failed to “apportion[] damages by

trade secret or otherwise provid[e] a methodology”).

secret” or the (jury-found) “group of trade secrets” or

(2) “provid[ing] a methodology for the jury to calculate

the value of a particular trade secret or group[.]”

Pet.App.18 n.10.

There is no meaningful daylight between those

formulations. While the Fifth Circuit hinted that

those means might not be “exhaustive,” it squarely

deemed their absence the deficiency below. See

Pet.App.8-9 (“Pastore did not individually valuate

each of the alleged trade secrets or any specific

combination of trade secrets, nor did he provide a

method for the jury to do so.”). And it deemed any

mismatch between the trade secrets valuated (at trial)

and those found by the jury (after trial) to “lend[] itself

to … speculation” and leave the award with “no basis.”

Pet.App.18, 23. That gloss on the minimum baseline

to sustain an award is “strict” in every sense.

Nor do respondents succeed in smoothing over interCircuit differences. Respondents fixate on the highlevel agreement—enshrined in the agreed-upon jury

instructions—that some evidence should support

attributing damages to proven wrongs. The issue,

however, is not whether a damages award needs

evidence (it does), nor even whether the Fifth Circuit

ignored Trinseo’s evidence (it did), but rather, what

that evidence need or need not encompass. The Fifth

Circuit held that a legally sufficient evidentiary basis

must encompass damages valuations for each juryfound secret (or group) or a methodology for the same.

On this point of law, there is emphatically no

agreement.

To be sure, evidence will vary by case. But what is

striking about the cases the petition details—which

are predictably invoked in every case where

apportionment arises—is their near-equivalence on

key facts: In each case, there was liability on somebut-not-all alleged secrets. Pet.App.9 (4-of-10); Texas

Advanced Optoelectronic Sols.. v. Renesas Elec. Am.,

895 F.3d 1304, (Fed. Cir. 2018) (1-of-3); O2 Micro Int’l

v. Monolithic Power Sys., 399 F.Supp.2d 1064 (N.D.

Cal. 2005), aff’d, 221 F. App’x 996 (Fed. Cir. 2007) (5of-11); Caudill Seed & Warehouse v. Jarrow Formulas,

53 F.4th 368 (6th Cir. 2022) (4-of-6); EchoSpan v.

Medallia, 2025 WL 3046753 (9th Cir. Oct. 31, 2025) (1of-9). In each case, there was an unapportioned

valuation of all the alleged secrets. Pet.App.18; TAOS,

895 F.3d at 1317; O2 Micro, 399 F.Supp.2d at 1076;

Caudill, 53 F.4th at 388; EchoSpan, 2025 WL

3046753, at *1. In none of the cases did the tradesecret owner valuate each jury-found secret (or group),

nor prescribe “a methodology” to do so. Under the

Fifth Circuit’s reasoning, those considerations alone

would render all the awards “speculation.”

Pet.App.18.

Respondents then try to cast the cases that

nonetheless upheld awards as fact-specific, but their

distinctions are not distinctions at all.2 The Sixth

Circuit in Caudill, they note, was unbothered by how

the “damages model assumed misappropriation of all

six” secrets because there was also testimony that

broccoli-seed research drove Caudill’s R&D expenses.

Respondents note another distinction without a difference—

that some cases involve Uniform Trade Secrets Act-based state

statutes. Opp.Br.17. “Given that the DTSA was enacted in light

of the UTSA and … uses virtually identical language, it is

appropriate to consider cases interpreting the UTSA . . . in

interpreting the DTSA.” Insulet v. EOFlow, 176 F.4th 1347,

1354–55 (Fed. Cir. 2026); accord Caudill, 53 F.4th at 380-81 &

n.2 (invoking DTSA cases because “Kentucky’s trade secret

protection statute is nearly identical”).

53 F.4th at 389. That testimony about one secret’s

outsized importance sufficed to eliminate any

“mismatch” between the expert’s testimony and the

award because it gave the jury “options” within the

unapportioned sum, even after the jury found no

liability on two alleged secrets. Id. at 389, 393. So too

with EchoSpan, where the Ninth Circuit held that

testimony about one secret’s importance gave the jury

a “reasonable basis” to approximate value within the

unapportioned sum, with no obligation to “completely

discount or completely credit” that testimony. 2025

WL 3046753, at *2. Here too, testimony that two juryfound secrets particularly drove value yielded leeway

for the jury to maneuver within the unapportioned

sum.

Nor is it significant that, in dicta, EchoSpan left

open the possibility that apportionment could be

required in a hypothetical case where a jury had no

opportunity “to understand how the system and each

component operated.” Id. This jury, like EchoSpan’s,

received detailed testimony about how the PC package

and each component operated and interrelated.

Moreover, both the Sixth and Ninth Circuits

couched their rationales in legal terms. The Sixth

Circuit differentiated O2 Micro and TAOS—Federal

Circuit cases the Fifth Circuit drew from—by

emphasizing its more “flexib[le]” and less “rigid” view

of the “evidence that can support a trade-secrets

damages award” in the “malleable context of trade

secrets.” Caudill, 53 F.4th at 389-90. The Ninth

Circuit underscored that the district court’s JMOL

ruling that EchoSpan had failed to “apportion … relief

on a trade-secret-by-trade-secret basis” erred by

presuming the jury’s lack of apportionment, rather

than “draw[ing] all reasonable inferences in favor of

preserving the jury’s verdict.” 2025 WL 3046753, at

*1-2. And both courts, unlike the courts below,

emphasized that damages figures need not correspond

with any specific testimony or obvious justification.

See id. at *2 (affirming award of “50%,” not “100%,” for

the value-driving secret despite lack of “coherent

justification” because “drawing … legitimate

inferences” is a “jury function[]”).

All these fissures—among the circuits and from this

Court’s precedents—warrant review, given the costly

stakes of jury-award-nullification and importance of

uniformity in trade-secrets litigation.

II. The Decision Below Is Wrong.

On the merits, respondents offer no tenable defense

of the Fifth Circuit’s harsh and arbitrary ruling.

1. Respondents advance the outlandish theory that

Trinseo “gamble[d]” on eschewing per-trade-secret

valuations “to maximize its damages.” Opp.Br.25-27.

But no rational actor would elect to risk recovering

nothing in the event of anything less than a cleansweep verdict.

Market realities simply dictated

Trinseo’s path.

As even the district court

acknowledged, evidence including respondents’ own

expert testimony showed that “in the industry, the

entire PC package … was the product” and “Trinseo

would never have sold its PC package piecemeal.”

Pet.App.53–54, 86-87; ROA.13723.

Respondents nonetheless assert that Pastore “could

have come up with an apportionment of damages for a

subset of Trinseo’s alleged trade secrets, but he simply

did not do so.” Opp.Br.25. Yet this wrests one

sentence—“I could do that work, if I was ordered to do

it”—badly out of context. ROA.11924. Pastore was

asked why his report did not offer opinions valuating

any number of hypothetical split verdicts (“if [jurors]

don’t find that each and every trade secret … [was]

misappropriated”). Id. Pastore was merely explaining

that he would not opine on that universe of

hypotheticals unless ordered to; as his next sentence

clarified, “there isn’t” “a verdict yet.” Id. This was a

comment on procedure, not substance. Nowhere did

Pastore indicate that he could apportion Trinseo’s PC

package into 10 individual values, or that any subset

would be valued at something other than the package.

To the contrary, Pastore testified repeatedly that the

royalty paid by a licensee would not vary with the

number of secrets. See ROA.11922-923 (“I’ve not done

a separation of value. Trinseo doesn’t sell a la carte its

trade secrets.”); ROA.11899 (agreeing that “if Trinseo

were to actually license its PC technology … it would

provide a full … package”); ROA.11902 (agreeing

“Trinseo would only license an entire technology

package” because “[t]hey don’t do a la carte”);

ROA.11928 (whether for 10 or 11 alleged secrets,

package would still be “$40 million per train”).

Weaker still is respondents’ claim that Trinseo

“gamble[d]” by taking the legal position that pertrade-secret apportionment was not required.

Opp.Br.25-27. Trinseo cannot be penalized for hewing

to a position well-supported by other Circuits and

uncontradicted by binding precedent. Cf. Syntel

Sterling Best Shores Mauritius v. TriZetto Grp., 2024

WL 1116090, at *8 (S.D.N.Y. Mar. 13, 2024).

Respondents’ other complaints—that Trinseo

should have but failed to pursue “compilation,” “builtin apportionment” or “entire market value” theories—

are more of the same. Opp.Br.26. With Trinseo’s

rejection of strict-apportionment, it would make no

sense (and would have been a bona fide gamble) for

Trinseo to send the jury one compilation secret in lieu

of 10 secrets based on impracticality concerns about

implementing strict-apportionment.

So too with

“built-in apportionment” and “entire market value”;

both theories could only have flowed from the premise

that strict-apportionment was required.

2. Respondents fare no better in painting this as the

rare case where no basis existed to support the award.

The petition proffered multiple alternative bases,

including those unrelated to Pastore, on which the

jury may have reached its damages figure. Pet.31 &

n.3. Respondents, like the Fifth Circuit, simply ignore

them.

Drawing all reasonable inferences in the verdict’s

favor, there is plainly evidentiary basis for the jury’s

reasonable approximations. The courts held otherwise

only by redefining the usually-modest evidentiary

threshold as one requiring damages evidence to

correspond neatly to the number and permutation of

jury-found secrets. See Pet.App.16, 18 (“Allowing

damages to be awarded for the misappropriation of

four trade secrets based on an estimation that

presumed misappropriation of ten trade secrets”

meant “the jury did not have a reasonable basis to

award damages based on the misappropriation of only

four”). The district court went so far as to parse

percentage deltas in assessing that correspondence,

Pet. 16—further underscoring the subversion of the

customary respect for jury inferences.

Nor can

respondents feign surprise that this deference sounds

in the Seventh Amendment. Opp.Br.29. Trinseo has

consistently touted the “constitutionally-required …

deference to a jury’s verdict” in this litigation.

C.A.Dkt.233 at 18.3 And respondents have no answer

to the bottom line: that strict-apportionment renders

trade-secret jury awards an aberration singled out for

judicial dissections of their underlying bases.

Meanwhile, respondents say nothing about why the

admittedly “imperfect” fit of patent law, Pet.App.86, is

properly overlaid on trade secrets—where what is a

trade secret (or not) is a jury question, not a formallydefined, known quantity. Those distinctions only

compound the harshness of the all-or-nothing regime,

as the difference between recovery and nothing turns

on any mismatch between the damages evidence at

trial and the trade secrets found after trial. This

arbitrary approach makes no remedial sense, simply

incentivizing litigants to conjure per-trade-secret

proof that may not exist.

Nor do respondents explain how the inflexible

prerequisites of per-trade-secret apportionment

square with the “flexible and imaginative approach”

that has guided trade-secret damages for 50 years.

Univ. Computing v. Lykes-Youngstown, 504 F.2d 518,

538 (5th Cir. 1974). By endorsing the take that the

Fifth Circuit was duty-bound to award Trinseo

nothing—despite the impossibility of conducting pertrade-secret valuations here and the “substantial”

evidence of Trinseo’s “right to … damages,”

Pet.App.89—respondents muster only a wan reference

to the prospective injunction. Opp.Br.16. That

E.g., D.C.Dkt.355 at 8 (noting “‘highly deferential’” review

“afforded to jury verdicts” and “high hurdles that KBR faces to

usurp the province of the jury”); D.C.Dkt.400 at 5 (noting “strong

presumption in favor of affirming a jury award”); C.A.Dkt.160 at

15 (“A court is ‘required to accept the verdict of a properly

instructed jury unless … there is a complete absence of evidence

to support it.”).

toothless “relief” does nothing to remedy the tens of

millions of dollars of harm already inflicted on Trinseo

by respondents’ years of enriching themselves off of

Trinseo’s technology.

At minimum, even absent certiorari, summary

reversal for a new damages trial is readily justified.

As both courts conceded, the apportionment

requirements were not the governing law at trial. See

Pet.App.13 (the Fifth Circuit has “never explicitly

adopted patent law’s apportionment principles in the

trade secret[s] context”); Pet.App.58 (“the Fifth Circuit

does not appear to have encountered a case involving

apportionment in trade secrets”). That the district

court did not adopt strict-apportionment as its legal

framework until its post-verdict JMOL ruling further

counsels in favor of this step. Whatever the merits of

the Fifth Circuit’s newly-minted legal baseline,

Trinseo should be afforded an opportunity to meet it.

III. This Case Is a Clean Vehicle.

Respondents’ last-ditch attempts to conjure vehicle

problems are meritless. Respondents submit that

“KBR presented multiple independent grounds for

affirmance that the Fifth Circuit did not reach.”

Opp.Br.28. Unaddressed alternative arguments—

encompassing sufficiency-of-evidence challenges, no

less—need give this Court no pause.

The district court’s “contingent alternative finding”

that the Fifth Circuit also did not reach, pertaining to

just one component of the damages award (not

reasonable royalty damages) and one defendant (not

the Harper Respondents), is similarly irrelevant. The

district court held that, in the event its apportionment

rationale fell, it would downward-adjust the jury’s

unjust enrichment damages against KBR to $10.5

million. Nothing about that limited reduction in the

alternative dims the stakes, and prospects of

meaningful relief, for Trinseo. And even $10.5 million

is substantially more than $0.

At bottom, far from “factbound,” Opp.Br.3, this case

arises from a Judgment-as-a-Matter-Of-Law ruling

reversing the factfinder’s conclusions after an

evidentiary trial, in which both courts below conceded

they were breaking new legal ground. There could

hardly be a more ideal posture for teeing up a

dispositive legal question.

CONCLUSION

The petition for a writ of certiorari should be

granted.

In the alternative, this Court should

summarily reverse the judgment below.

Respectfully submitted,