Trade Secrets

Brown v. TGS Management: When a Confidentiality Clause Becomes an Illegal Noncompete

A California appellate court voided an employer's sweeping confidentiality provisions as a de facto noncompete that barred a trader from his profession for life.

September 23, 2025
A financial trader studying multiple market data screens
The dispute pitted a statistical-arbitrage trader's right to work against an employer's expansive definition of 'confidential' information. Shutterstock
Educational content, not legal advice. This article explains general legal concepts. It does not create an attorney–client relationship. For your specific situation, consult a licensed attorney.

Brown v. TGS Management Co., LLC, 57 Cal.App.5th 303 (Cal. Ct. App. 2020), No. G058323 (4th Dist., Div. 3, Oct. 13, 2020), is the decision every drafter of employee confidentiality agreements in California now has to reckon with. The court of appeal vacated an arbitration award and held that an employer’s confidentiality provisions — drawn so broadly that they swept in essentially all knowledge “usable in” or “relating to” the securities industry — were void on their face under Business and Professions Code section 16600 because they operated as a de facto restraint on the employee’s right to work in his profession.

The case marks a doctrinal shift: a confidentiality clause, the supposedly benign cousin of the noncompete, can itself be struck down as an unlawful noncompete when its breadth leaves an employee nothing left to do.

At a glance

  • Case: Brown v. TGS Management Co., LLC, 57 Cal.App.5th 303
  • Court: California Court of Appeal, Fourth District, Division 3, No. G058323
  • Decision: Reversed and remanded, October 13, 2020
  • Posture: Petition to vacate an arbitration award; the arbitrator had upheld the confidentiality provisions and denied Brown’s claims
  • Governing law: Cal. Bus. & Prof. Code § 16600, voiding contracts that restrain anyone from engaging in a lawful profession, trade, or business
  • Holding: The confidentiality provisions were so broad they functioned as a de facto noncompete barring Brown from his field; they were void as a matter of law, and the arbitrator exceeded his powers by enforcing them rather than declaring them void

A confidentiality clause with no horizon

Richard Brown was a statistical-arbitrage trader at TGS. His employment agreement contained confidentiality provisions defining protected “Confidential Information” in sweeping terms — covering all information that was “usable in” or that “related to” the business, with “the Business” itself defined to encompass virtually every subject pertaining to securities trading. The agreement carved out only narrow exceptions, including information “generally known in the securities industry through legal means.”

When Brown sued TGS over unpaid bonuses, TGS counterclaimed that he had violated the confidentiality provisions and argued that no further compensation was owed — and that he should disgorge pay already received. The matter went to arbitration, where the arbitrator largely sided with TGS and enforced the confidentiality terms. Brown petitioned to vacate.

The court of appeal’s analysis began with the breadth of the definition. A statistical-arbitrage trader creates value precisely by using securities-related information that is not generally known. The “generally known” carve-out, the court found, was therefore essentially meaningless: it excluded only the information that had no value to Brown’s work in the first place. What remained inside the definition was everything Brown would actually need to practice his profession anywhere. The clause did not merely protect TGS’s secrets; it claimed dominion over the entire intellectual toolkit of the trade.

From confidentiality to restraint: the section 16600 reasoning

Section 16600 states a categorical California rule: every contract that restrains anyone from engaging in a lawful profession, trade, or business is, to that extent, void. California courts have long read it to bar most employee noncompetes outright, rejecting the “reasonableness” balancing that other states apply. Brown extends that logic to confidentiality clauses by asking a functional question: regardless of the label, does the provision operate to prevent the employee from working?

The court answered yes. Read together, the provisions “plainly bar Brown in perpetuity from doing any work in the securities field, much less in his chosen profession of statistical arbitrage.” A clause that forbids the use of all knowledge usable in an industry is, in operation, indistinguishable from a clause forbidding work in that industry. Because the restraint flowed from the contract’s own text, the court held the provisions void on their face — “patently” in violation of section 16600 — without needing to wait and see how TGS might enforce them in some future dispute.

That facial-invalidity holding answered TGS’s ripeness objection. TGS argued the challenge was premature because no one could yet say what conduct the clause would actually prohibit. The court disagreed: where a provision is void on its face, an employee may mount a facial attack on its legality without first risking violation, and a court or arbitrator must decide that legal question. An arbitrator who instead enforces a facially void restraint exceeds his powers — the narrow ground on which California courts will vacate an arbitration award — which is why the award could be set aside despite the usual deference to arbitral decisions.

The blurred line between protecting secrets and locking in employees

Brown is significant because it refuses to let drafting labels control. Employers wary of California’s hostility to noncompetes have long migrated the same protective impulse into confidentiality and non-solicitation clauses, which historically drew less scrutiny. Brown signals that California courts will look past the heading to the practical effect. A confidentiality provision is enforceable to protect genuine trade secrets and proprietary information; it is not a vehicle for accomplishing, through the back door, the career lock-in that section 16600 forbids at the front.

The decision also illustrates the danger of definitions that swallow their own exceptions. The “generally known” carve-out looked like a reasonable limitation, but because it excluded only valueless information, it left the operative scope unlimited. A confidentiality definition must leave the employee with a meaningful universe of general knowledge, skill, and experience to carry into the next job. When it does not, the clause stops protecting information and starts protecting the employer from competition by its former employee — the precise harm the statute targets.

Open questions

  • How broad is too broad? Brown involved a definition that captured essentially all industry knowledge. Where the line falls between a permissibly comprehensive secrecy clause and a de facto noncompete will be litigated case by case.
  • Does severance survive? The court treated the provisions as void, but whether a more carefully drafted agreement with a severability clause could preserve a narrowed confidentiality obligation after the overbroad portions fall is not fully resolved.
  • How far beyond securities does it reach? The reasoning is industry-agnostic. Its application to fields where “general knowledge” and “confidential information” are harder to separate — software, research, sales — will test the boundaries.
  • What is the interaction with statutory trade-secret law? A clause that merely restates trade-secret protection is on safer ground; the risk arises when the contract claims far more than trade-secret law itself would protect.

Implications

  • Draft confidentiality definitions narrowly and concretely. Tie protection to identifiable categories of proprietary information, not to everything “usable in” or “relating to” an industry.
  • Make carve-outs meaningful. Exceptions for “generally known” information are worthless if they exclude only information with no value; preserve the employee’s general knowledge, skill, and experience expressly.
  • Assume functional scrutiny in California. Courts will look past the “confidentiality” label to ask whether the clause actually prevents the employee from working.
  • Expect facial challenges. Employees need not breach an overbroad clause before attacking it; a facially void restraint can be challenged and struck without waiting for enforcement.
  • Recognize the arbitration exposure. An arbitrator who enforces a facially void restraint can be reversed for exceeding his powers, eroding the finality employers usually expect from arbitration.

Frequently asked questions

Can a confidentiality agreement really be an illegal noncompete? In California, yes. Brown v. TGS holds that if a confidentiality provision is so broad that it effectively prevents an employee from working in the field, it operates as a de facto noncompete and is void under Business and Professions Code section 16600 regardless of its label.

Why did the overbroad definition fail even though it had exceptions? The principal exception — for information “generally known in the securities industry through legal means” — excluded only information that had no value to the trader’s work. What remained inside the definition was everything he needed to practice his profession, so the carve-out did not meaningfully narrow the restraint.

Does this let an employee disregard a confidentiality clause? No. Genuine trade secrets and properly scoped proprietary information remain protectable. Brown targets clauses that reach far beyond that, claiming all industry knowledge and thereby restraining the employee’s right to work.

Authorities and sources