Written by Chris Walton, JDA verdict reversed for an evidentiary failure in the damages model is the worst kind of trial outcome. The plaintiff has won the substantive case — the breach is established, the misconduct is on the record, the jury credited the story — but the damages number does not survive appellate review because the expert built the model on the wrong company. The result is years of litigation, a jury finding of $138m in entity-level damages, a judgment of nearly $54m — and an appellate opinion holding that no evidence supports any part of it.
That is Kay v. Yosowitz, No. 14-23-00710-CV (Tex. App.—Houston [14th Dist.] Oct. 16, 2025) (op. on reh’g) (Christopher, C.J.), in which Houston’s Fourteenth Court of Appeals, withdrawing its original July 10, 2025 opinion on rehearing, affirmed a $378,470.08 direct-claim award, reversed the $53,900,980.99 awarded on the derivative fiduciary-duty claim, rejected the plaintiff’s offer of voluntary remittitur, reversed the associated fees and investigation costs — and, in the rehearing’s significant procedural turn, remanded so the plaintiff may make a new election of remedies from favorable jury findings that went unchallenged on appeal. The opinion is a clean teaching case for commercial litigators and damages experts about how to build, and how to attack, a damages model where the facts involve successor entities, diverted technology, and a defendant who never honored the structure he agreed to. It is also, as discussed below, a case in which the damages model failed for a deeper reason than thin counterfactual evidence: the counterfactual it assumed was foreclosed by the very contracts the plaintiff was enforcing.
Before their divorce, Martin Kay and Laura Yosowitz owned, as community property, a 78% membership interest in Greenlet, LLC, a Nevada limited liability company; Robert Salmons Jr. owned the other 22%. Greenlet ran two lines: a traditional real estate “Brokerage Business” and a “Software Business.” The divorce decree incorporated an Agreement Incident to Divorce (the AID) — the Texas instrument family lawyers will recognize — which in turn incorporated a Memorandum of Agreement (MOA, effective October 18, 2016) signed by Kay, Yosowitz, and Salmons. The MOA’s architecture: Kay and Yosowitz would form “Holdco” (Kay 76.65% and all voting rights; Yosowitz 23.35%, non-voting) to hold their Greenlet interest, and all three would form “IP Newco” (95% Holdco, 5% Salmons), to which Greenlet and its members would contribute all intellectual property and assets of the Software Business — a transfer, not a sale or license. Holdco could redeem Yosowitz’s shares for $3.75m within five years; she stood to receive a one-time bonus of up to $500,000 (plus $750,000 in trust for the children) if the companies’ aggregate value reached $40m; and Kay committed to twelve months of commercially reasonable efforts to raise outside funding and commercialize both businesses.
None of it happened. Kay told Yosowitz he had formed Holdco as “Greenlet Brokerage, LLC” and IP Newco as “Greenlet Technology, LLC.” Neither company ever existed. He formed Kiva Technologies, LLC, ostensibly to play the IP Newco role, but never transferred the Software Business to it. What Kay did instead: immediately after the divorce, he formed a Wyoming company, Realtech, LLC (later Entera Technology, LLC, a wholly owned subsidiary of Entera Holdings, Inc., which also owns the brokerage Entera Realty, LLC), and caused Greenlet to license its software to Realtech. Using some of the same developers, Realtech updated and expanded the software; Entera Holdings raised roughly $38m from institutional investors largely on that technology’s success; and Greenlet — its software frozen as of the MOA — was wound down, closing at the end of 2022. Yosowitz received a single check from Greenlet for less than $39,000.
She sued Kay (along with Salmons and several Kay entities), individually for breach of the MOA, breach of AID § 2.5, breach of an informal fiduciary duty, and fraud — and derivatively on Greenlet’s behalf for breach of the fiduciary duties Kay owed as Greenlet’s manager. The jury found for her across the board, finding Greenlet’s damages at $138,207,643.56. The trial court then did something structurally notable: citing “equitable considerations and the interests of justice,” it awarded Yosowitz directly her “allocable share” of the derivative damages — 39%, or $53,900,980.99, apparently reflecting half of the couple’s former 78% community interest — together with the $378,470.08 the jury assessed on her AID § 2.5 claim, plus investigation costs, fees, and interest.
The damage cap failed four independent ways. Kay argued the MOA capped Yosowitz’s recovery at $4.25m (the $3.75m redemption price plus the $500,000 maximum bonus). The court disagreed on four freestanding grounds: Holdco was never formed, so there was neither an interest to redeem nor an entity capable of redeeming it — nothing for the cap to attach to; the MOA itself requires payments beyond the capped items, including mandatory annual distributions to Yosowitz of at least 25% of Holdco taxable income allocated to her, expressly not credited against the redemption price; her elected individual recovery arose under AID § 2.5, outside the MOA entirely; and the MOA addresses what Holdco must pay, not what is recoverable from Kay. The drafting lesson runs both directions: a cap tied to a structure is only as good as the structure’s existence, and a defendant who never builds the structure cannot claim its ceilings.
Standing survived on the charge as submitted — and on K-1s. Derivative claims for a foreign LLC are governed by the law of the formation state, Tex. Bus. Orgs. Code § 101.462(a), here Nevada, which requires the plaintiff to have been a member or noneconomic member at the time of the challenged transaction. Nev. Rev. Stat. § 86.485. Kay argued Yosowitz never satisfied the operating agreement’s membership prerequisites — but he had not objected to the charge’s membership instructions, so sufficiency was measured against the charge as submitted. Osterberg v. Peca, 12 S.W.3d 31, 55 (Tex. 2000). Greenlet’s own 2014 K-1s identified Yosowitz as a 1% member, which sufficed. (The record also contained two conflicting operating agreements, one undated — a familiar closely-held-company pathology.) Practice notes on both sides: tax filings are potent membership evidence, and charge objections are where standing fights are won or lost.
The derivative damages fell, for reasons deeper than a thin record. The jury found Kay breached fiduciary duties to Greenlet in four ways; for misappropriation the charge asked for out-of-pocket amounts, and for exploitation of business opportunities it asked for lost profits — both permissible measures. For the general duties (utmost good faith, scrupulous honesty, placing Greenlet’s interests first), the charge asked for “the value Greenlet would have received had Kay complied with his fiduciary duties, minus the value it did receive.” The $138.2m landed there — and the parties agreed it could only have come from valuing Entera Holdings’ shares. Yosowitz’s closing asked the jury to multiply Kay’s eight million Entera shares by $7.25 (about $58m) and find $60m; the jury appears instead to have valued all of Entera. The court of appeals held there was no evidence that, but for the breach, Greenlet would have received value measurable by Entera’s shares: “The two companies are not at all comparable.” Entera — which Yosowitz’s own damages expert, chartered financial analyst Tobin Reiff, described as “a start-up company based on unique technology” — owned both Realtech and a brokerage; Greenlet, by the express command of the MOA, the AID, and the divorce decree, was required to transfer its entire Software Business away, free of charge, leaving only a brokerage.
That last point is the opinion’s analytical core, delivered in footnote 14: the agreements that narrowed Greenlet’s business also narrowed Kay’s fiduciary duties. Kay could not have owed Greenlet a duty to develop the Software Business, “given that Greenlet was required — by the express will of all of its members, the MOA, the AID, and the divorce decree — to transfer all of its Software Business to another company. Kay still had a fiduciary duty to place Greenlet’s interests before his own, but Greenlet’s interests no longer included the Software Business.” In other words, the damages counterfactual — Greenlet keeping the software and flourishing as Entera did — was not merely unproven. It was foreclosed by the plaintiff’s own contracts. No evidence suggested institutional investors would fund an established company contractually bound to give its technology away. The breach was real (the software was supposed to go to IP Newco, 95% owned by a Holdco in which Yosowitz held 23.35% — instead it was licensed to Kay’s new venture); but the gap between the breach and Entera’s enterprise value was a gap the model never bridged and, on this record, could not.
Footnote 13 supplies the expert-practice landmine. Reiff testified that he was engaged to determine the fair market value of Entera Holdings shares “to measure damages sustained by Ms. Yosowitz based on allegations of business opportunities of Greenlet that were transferred to Entera.” But the jury was separately asked to assess damages for Kay’s exploitation of Greenlet’s business opportunities — and answered $0. The valuation was built for a theory the jury rejected, and the number migrated to a general-duties measure it did not fit. When the expert’s engagement theory and the charge’s damages theory diverge, the award lives in the gap — and dies there on review.
The remittitur rejection is its own lesson. On rehearing, Yosowitz offered to treat Greenlet’s damages as $60,004,425.25 and accept 39% — $23,401,725.85. The court refused for the same reason it reversed: “no evidence supports any part of the derivative damages.” Remittitur cures excessive awards supported by some evidence; it cannot rescue an award built on an impermissible measure. When the measure is wrong, there is no defensible fraction.
The direct claim survived because its measure was contractual. Kay attacked the $378,470.08 award on the theory that Yosowitz offered no fair-market-value evidence for the eight rental homes held by Greenlet Property, LLC (a distinct entity). Wrong premise, the court held: AID § 2.5 entitles Yosowitz to 46.8% of defined “net proceeds” — rent and sale price minus defined Permitted and Recurring Expenses — and her complaint was that Kay diluted the proceeds with impermissible offsets: construction and repair costs she never consented to in writing, a “rebate” paid to a buyer two weeks after closing, and principal-and-interest payments on a revolving line of credit from Sterrett Capital, LLC — a lender she later learned was affiliated with Kay. No FMV evidence was needed to decide whether those were “Permitted Expenses.” The contract defined the entitlement; the evidence proved the deviation. Fees on that contract claim are mandatory, Tex. Civ. Prac. & Rem. Code § 38.001(8), but the fee award blended contract-claim and derivative-claim work, so it was reversed and remanded for segregation — while the derivative fees and investigation costs fell outright with their statutory basis, the substantial-benefit predicate of Tex. Bus. Orgs. Code §§ 101.461(b)(1) and 101.451(a).
And the case is not over. The rehearing’s procedural turn matters as much as its damages holding: because Yosowitz obtained favorable jury findings on multiple claims and elected the derivative claim now reversed, the court remanded for her to make a new election of remedies among the unchallenged alternatives — with ancillary issues, including fees and investigation costs, to follow the election. Preserved alternative findings are appellate insurance; Kay shows the policy paying out.
Texas law gave the court its catalogue of permissible fiduciary-breach measures before the court held this model fit none of them: out-of-pocket losses and benefit-of-the-bargain damages, Combs v. Crepeau, No. 05-23-00088-CV, 2024 WL 4432324 (Tex. App.—Dallas Oct. 7, 2024, pet. denied) (mem. op.); lost profits, Salas v. Total Air Servs., LLC, 550 S.W.3d 683 (Tex. App.—El Paso 2018, no pet.); lost business value, measured by the market value of the business destroyed or the before-and-after difference, Sawyer v. Fitts, 630 S.W.2d 872 (Tex. App.—Fort Worth 1982, no writ); and, without proof of causation, equitable remedies such as constructive trust, KCM Fin. LLC v. Bradshaw, 457 S.W.3d 70 (Tex. 2015), and disgorgement, Longview Energy Co. v. Huff Energy Fund LP, 533 S.W.3d 866 (Tex. 2017). The successor-entity model failed not because successor evidence is categorically inadmissible, but because three distinct defects stacked.
The first was entity non-comparability. The successor was a differently constituted enterprise — a venture-backed “start-up company based on unique technology” owning both a technology company and a brokerage — whose realized value reflected new capital (roughly $38m of institutional investment), continued development, and a different asset mix. Equating its share value with the harmed entity’s lost value assumed away every difference. The second was duty-scope foreclosure. The governing contracts required the harmed entity to divest the very business line driving the successor’s value; the counterfactual in which Greenlet keeps and develops the software was not available as a matter of the parties’ own agreements, so a model premised on it had no evidentiary path, however rigorous its mechanics. The third was the theory–charge mismatch. The valuation was commissioned for a business-opportunity-diversion theory; the jury awarded $0 on that theory’s damages question, and the number was poured into a general-duties measure instead. The model must be built for the measure the charge will actually submit — and the charge must submit the measure the model supports.
For litigators and their experts, the working discipline in a successor-entity damages case runs as follows. Value the harmed entity directly — its own pre-breach financials, projections, assets, and trajectory — with the successor’s realized performance as an input or cross-check, never as the answer. Before building anything, map the counterfactual against the governing contracts: identify what the harmed entity was contractually permitted, required, and forbidden to do, because a counterfactual the operative agreements foreclose (as the MOA, AID, and decree foreclosed Greenlet’s software future) is not a damages theory; it is a legal-sufficiency reversal waiting to happen. Then build the explicit bridge: which assets and capabilities transferred, what value they had in the harmed entity’s hands at the time of breach, and which components of the successor’s value are attributable to the diversion versus post-diversion capital, development, management, and market conditions. Align the model with the charge — confirm the damages question the jury will actually answer, and that the expert’s engagement theory, report, and testimony support that measure; if alternative measures will be submitted, know which the model supports and say so. Structure direct claims with independently defensible, preferably contract-defined measures: Kay’s $378,470.08 net-proceeds award survived precisely because the contract defined the entitlement and no enterprise valuation was required. And preserve alternative jury findings and segregate fees from day one — unchallenged alternative findings funded Yosowitz’s second chance on remand, while unsegregated fees cost her the immediate recovery of even the mandatory contract-claim fees.
Forensic work was central to the underlying case: tracing the software license to Realtech, the assumed-name games (Entera Realty briefly did business as “Greenlet LLC”), the Sterrett Capital affiliation, and the offsets that shrank the § 2.5 net proceeds. That tracing established liability and the factual predicate for both claim types. But Kay is a sharp reminder that forensic tracing and damages valuation are different exercises carrying different evidentiary loads. Tracing answers what moved, with documentary certainty. The damages model must answer what was lost — the value of the diverted assets in the harmed entity’s hands and what that entity would have realized absent the breach, within the bounds its contracts allowed. Where one engagement covers both, the report should keep the exercises visibly distinct; treating the successor’s realized value as the answer to both questions is the failure mode this opinion now documents at $53.9m of reversed judgment.
Most experienced commercial litigators handling fiduciary matters with successor-entity dynamics already know the legal-sufficiency standard — evidence viewed in the light favorable to the verdict, Am. Honda Motor Co. v. Milburn, 696 S.W.3d 612 (Tex. 2024); City of Keller v. Wilson, 168 S.W.3d 802 (Tex. 2005), with the proper measure of damages reviewed de novo, Signature Indus. Servs., LLC v. Int’l Paper Co., 638 S.W.3d 179 (Tex. 2022) — and structure damages models with the appellate posture in mind. If your expert values the harmed entity directly, checks the counterfactual against the operative contracts, builds the bridge to any successor evidence, and aligns the model with the charge, you are past the failure mode Kay documents. This article is a checklist for the cases where the diversion-to-successor pattern is unfamiliar, the number is large enough that appellate scrutiny is certain, or the litigator and expert have not worked together before.
Kay v. Yosowitz does not change Texas damages law; it catalogues the existing measures and holds a successor-share model to them. What it gives plaintiff-side teams is a recent, citable example of a $53.9m judgment — resting on a $138m jury finding — reversed because the expert valued the wrong company, for a counterfactual the plaintiff’s own contracts foreclosed, under a theory the jury separately zeroed out. The merits were not the problem; the jury believed everything. The breach was not the problem; the structure Kay promised never existed. The problem was the analytical span between “the assets were diverted to a successor that flourished” and “the harmed entity would have captured that value but for the breach” — a span this record could not support.
For defense teams, the same span is the cleanest line of attack, and Kay maps the preservation route: the instructed-verdict motion naming the improper measure (“the alleged equity value of Entera” is “not a proper measure of damages”) defeated the invited-error argument and carried the issue to reversal. For everyone: segregate fees in real time, and request alternative damages findings even on theories you do not expect to elect — in Kay, those unchallenged findings are the reason the plaintiff still has a case on remand.
Chris Walton, JD, is the President & CEO of Eton Venture Services. He can be reached at [email protected].
If you are litigating a fiduciary-duty matter with a successor-entity damages component — or defending against one — we’re glad to discuss how the damages model can be structured to survive, or expose, the kind of gap Kay reversed on appeal: the counterfactual against the contracts, the model against the charge, and the entity against its successor.
Kay v. Yosowitz, No. 14-23-00710-CV (Tex. App.—Houston [14th Dist.] Oct. 16, 2025) (op. on reh’g) (Christopher, C.J., joined by Jewell and McLaughlin, JJ.) (withdrawing op. of July 10, 2025).
Osterberg v. Peca, 12 S.W.3d 31 (Tex. 2000); City of Keller v. Wilson, 168 S.W.3d 802 (Tex. 2005); Am. Honda Motor Co. v. Milburn, 696 S.W.3d 612 (Tex. 2024); Signature Indus. Servs., LLC v. Int’l Paper Co., 638 S.W.3d 179 (Tex. 2022).
Combs v. Crepeau, No. 05-23-00088-CV, 2024 WL 4432324 (Tex. App.—Dallas Oct. 7, 2024, pet. denied) (mem. op.); Salas v. Total Air Servs., LLC, 550 S.W.3d 683 (Tex. App.—El Paso 2018, no pet.); Nelson v. Vernco Constr., Inc., 566 S.W.3d 716 (Tex. App.—El Paso 2018, pet. denied); Sawyer v. Fitts, 630 S.W.2d 872 (Tex. App.—Fort Worth 1982, no writ).
KCM Fin. LLC v. Bradshaw, 457 S.W.3d 70 (Tex. 2015); Longview Energy Co. v. Huff Energy Fund LP, 533 S.W.3d 866 (Tex. 2017); Bocquet v. Herring, 972 S.W.2d 19 (Tex. 1998).
Tex. Bus. Orgs. Code §§ 101.451(a), 101.461(b)(1), 101.462(a); Tex. Civ. Prac. & Rem. Code § 38.001(8); Nev. Rev. Stat. §§ 86.326(2)(b), 86.351(1), 86.485.
If you are litigating a fiduciary-duty matter with a successor-entity damages component — or defending against one — we’re glad to discuss how the damages model can be structured to survive, or expose, the kind of gap Kay reversed on appeal: the counterfactual against the contracts, the model against the charge, and the entity against its successor.
Schedule a free consultation meeting to discuss your valuation needs.
