SEC May Approve IEX Options’ Anti–Latency-Arbitrage Model Without CAT Modeling; ORP Quotes Are “Protected” Under the Options Plan

Case: Citadel Securities LLC v. U.S. Securities and Exchange Commission (11th Cir. May 29, 2026)
Posture: Petition for review of SEC approval order for IEX Options (SR-IEX-2025-02), 90 Fed. Reg. 45861 (Sept. 23, 2025)

1. Introduction

This case sits at the intersection of modern market microstructure and administrative law. Petitioner Citadel Securities LLC (“Citadel”), a major options market maker and high-frequency trader, challenged the U.S. Securities and Exchange Commission’s (“SEC” or “Commission”) approval of Investors Exchange, LLC’s (“IEX”) proposal to launch IEX Options—an options exchange designed to blunt “latency arbitrage” by combining (i) a 350-microsecond “speedbump” access delay with (ii) software that cancels or reprices certain market-maker quotations when the system detects fleeting price dislocations.

Citadel presented five issues: (1) whether the SEC had substantial evidence that latency arbitrage exists in options markets; (2) whether the SEC had substantial evidence that IEX’s Options Risk Parameter (“ORP”) accurately targets latency arbitrage; (3) whether ORP-affected quotations qualify as “protected” quotations; and (4)–(5) whether the SEC acted arbitrarily and capriciously in finding the proposal neither unfairly discriminatory nor an undue burden on competition.

2. Summary of the Opinion

The Eleventh Circuit (Rosenbaum, J.) denied the petition. Applying the APA’s arbitrary-and-capricious standard and Exchange Act substantial-evidence review, the court held:

  • Latency arbitrage in options: Substantial evidence supported the SEC’s finding that latency arbitrage exists and harms options-market liquidity and spreads, even without “gold-standard” quantitative proof.
  • ORP targeting: Substantial evidence supported the SEC’s conclusion that ORP activates during microsecond-scale dislocations and therefore primarily affects latency-arbitrage strategies rather than ordinary investors; the SEC was not required to build a bespoke model from Consolidated Audit Trail (“CAT”) data.
  • Protected quotations: ORP quotations qualify as “protected” under the Options Order Protection and Locked/Crossed Market Plan (“Options Plan”); Regulation NMS’s “immediate” execution concept does not govern options protected-quote status.
  • No unfair discrimination: The SEC reasonably found ORP is not “unfair discrimination” because it is a risk protection commensurate with market makers’ unique continuous-quoting risks, is rule-bound and transparent, and is expected to improve liquidity for all market participants.
  • No undue burden on competition: The SEC reasonably found IEX Options would not unduly burden competition, and could increase competition (including by lowering entry barriers for market making) while advancing Exchange Act goals like efficient execution and investor protection.

3. Analysis

3.1. Precedents Cited (and How They Shape the Decision)

The opinion is as much about administrative-law restraint and deference as it is about market design. Several cited decisions supply the scaffolding for the court’s review posture and its treatment of evidence.

A. Prior IEX / latency-arbitrage litigation and market-structure deference

  • Citadel Sec., LLC v. SEC (D.C. Cir. 2022): The Eleventh Circuit repeatedly leveraged this case as persuasive authority validating (i) the concept that latency arbitrage “actually harm[s] the market” and resembles a “surcharge,” and (ii) the SEC’s ability to credit market-structure explanations and proxies (like time-sliced indicators) when evaluating anti-latency tools. The court emphasized that the options proposal is an extension of the same basic IEX architecture previously upheld in equities.
  • Cboe Glob. Mkts., Inc. v. SEC (D.C. Cir. 2025): Cited for technical details (matching engines) and, more importantly, for the principle that agencies receive heightened deference when making predictive judgments grounded in technical expertise—an idea that recurs in the “targeting,” discrimination, and competition analyses.

B. APA review, “zone of reasonableness,” and evidentiary sufficiency

  • Am. Sec. Ass'n v. SEC (11th Cir. 2025): Supplies the Eleventh Circuit’s baseline framework for applying the APA arbitrary-and-capricious standard to SEC orders.
  • FCC v. Prometheus Radio Project (2021): Central to the court’s approach. It is cited for (i) the “zone of reasonableness” framing, (ii) the admonition that courts cannot substitute policy judgment for the agency’s, and (iii) the proposition that the APA imposes no general obligation to conduct or commission empirical or statistical studies—key to rejecting Citadel’s CAT-based demands.
  • City of North Miami v. Fed. Aviation Admin. (11th Cir. 2022) and Miccosukee Tribe of In- dians of Fla. v. United States (11th Cir. 2009): Reinforce that arbitrary-and-capricious review is “exceedingly deferential,” especially where an agency makes technical judgments.
  • Orkin v. SEC (11th Cir. 1994): Anchors the court’s treatment of SEC legal conclusions (de novo) versus factual findings (substantial evidence) and interacts with the Exchange Act’s command that supported SEC findings are conclusive.
  • Nat'l Parks Conservation Ass'n v. U.S. Dep't of the Interior (11th Cir. 2016) and Stone & Webster Constr., Inc. v. U.S. Dep't of Lab. (11th Cir. 2012): Provide the court’s definition of “substantial evidence.”

C. Reliance on comments and non-quantitative records

  • Azar v. Al- lina Health Servs. (2019): Used to underscore why notice-and-comment exists—to gather information and avoid errors—supporting the SEC’s reliance on comment letters as evidence.
  • Stilwell v. Off. of Thrift Su- pervision (D.C. Cir. 2009): Cited for the point that the APA does not generally require empirical evidence; the court used it to reject the notion that the SEC had to replicate the D.C. Circuit-era quantitative showing from the equities context.
  • Am. Iron & Steel Inst. v. Occupational Safety & Health Admin. (11th Cir. 1999): Supports the proposition that comments can provide sufficient factual support.
  • Nasdaq Stock Mkt. LLC v. SEC (D.C. Cir. 2022) and Chamber of Com. of U.S. v. SEC (D.C. Cir. 2005): Reinforce that the SEC cannot—and need not—base every action on empirical data, further weakening Citadel’s data-compulsion theory.
  • Domestic Sec., Inc. v. SEC (D.C. Cir. 2003) and Ga. Dep't of Educ. v. U.S. Dep't of Educ. (11th Cir. 2018): Used to remind that substantial evidence does not require resolving all conflicts or establishing a preponderance.

D. Discrimination, competition, forfeiture, and the role of alternatives

  • Timpinaro v. SEC (D.C. Cir. 1993): Critical for the statutory reading that the Exchange Act bars “unfair discrimination,” not all discrimination—enabling the SEC to approve differentiated treatment for market makers if justified.
  • American Equity Investment Life Insurance Co. v. SEC (D.C. Cir. 2010): Citadel invoked it to demand a competition “baseline.” The court distinguished it: it addressed a different statutory provision (Exchange Act § 2(b) rulemaking) and different procedural context, and in any event did not impose the sweeping baseline requirement Citadel asserted.
  • Bradford Nat'l Clearing Corp. v. SEC (D.C. Cir. 1978): Used to rebut any claim the SEC must pursue objectives in the “least anticompetitive manner possible.”
  • LaCourse v. PAE Worldwide Inc. (11th Cir. 2020) and Access Now, Inc. v. Sw. Airlines Co. (11th Cir. 2004): Enforce issue-forfeiture rules; the court used them to dispose of a “less intrusive alternatives” argument raised only on reply.
  • FTC v. Procter & Gamble Co. (1967), Jacobs v. Tempur-Pedic Int'l, Inc. (11th Cir. 2010), Nat'l Collegiate Athletic Ass'n v. Bd. of Regents of Univ. of Okla. (1984), FTC v. Univ. Health, Inc. (11th Cir. 1991), and United States v. Apple, Inc. (2d Cir. 2015): Serve as general competition-law references supporting the SEC’s procompetitive narrative (innovation, lower barriers to entry, improved price/quality/output).

E. “Backing away” and execution mechanics

  • In re Credit Suisse Sec. (USA) LLC (SEC 2016): Provides the definition of “backing away” as failing to execute after orders were matched. The court used it to show Citadel’s critique mischaracterized ORP mechanics: ORP operates before matching/execution, not after.

3.2. Legal Reasoning

A. Framing: two review tracks (substantial evidence vs. arbitrary-and-capricious)

The opinion cleanly separates (i) Exchange Act “substantial evidence” review of SEC factual findings (treated as conclusive if supported) from (ii) APA arbitrary-and-capricious review of the SEC’s policy judgments and explanatory adequacy. This division is outcome-determinative because once the court deems the factual predicates (latency arbitrage existence; ORP targeting) supported, Citadel’s downstream “speculative problem” and “excessive benefit” theories lose traction.

B. Latency arbitrage in options: the SEC can prove it qualitatively

The court accepted that a record composed heavily of participant comment letters plus the SEC’s expert synthesis of market structure can amount to substantial evidence. The court emphasized:

  • Multiple market makers reported latency arbitrage as common and costly, driving wider spreads and reduced displayed liquidity.
  • The SEC tied those reports to structural features of options markets—massive option-series counts and the need to reprice many derivatives when an underlying moves; plus the lack of off-exchange alternatives.
  • Citadel’s demand for the same kind of quantitative, exchange-specific trading data used in the equities context was unrealistic because IEX Options was not yet operational.
  • As a matter of law, the APA/Exchange Act does not impose a general empirical-data requirement.

C. ORP targeting: time-based proxies and no mandatory CAT modeling

The court endorsed the SEC’s logic that microsecond-scale signals are a sensible proxy for identifying latency-arbitrage activity because only technologically advantaged firms can systematically trade in those windows. It also accepted the SEC’s explanation that volume-based statistics can be misleading: a more effective anti-latency tool may “affect” more attempted latency-arbitrage orders precisely because it prevents execution (changing fill rates).

On CAT, the court treated Citadel’s demand as inconsistent with FCC v. Prometheus Radio Project and with Congress’s tight statutory deadlines for SRO rule approvals under 15 U.S.C. § 78s(b)(2). The message is not that CAT is irrelevant, but that the SEC is not arbitrary for declining to build a complex, bespoke analytic pipeline from massive raw CAT datasets within an accelerated review cycle.

D. “Protected” quotations: Options Plan governs; Regulation NMS does not

On the protected-quotation challenge, the court’s reasoning is notably formal and rule-text-driven:

  • The parties agreed the Options Plan governs protected quotes in options markets.
  • Citadel did not successfully connect ORP to the Options Plan’s “non-firm quotation” exception because the Plan defines non-firm status by reference to an exchange’s own rules, and Citadel did not show IEX’s enumerated non-firm scenarios applied.
  • Citadel’s attempt to import Regulation NMS’s “immediate” execution concept failed because Regulation NMS governs equities protected quotations, and the SEC expressly noted it does not apply here.

E. No unfair discrimination: “unfair” is the statutory limiter

The court accepted the SEC’s rationale that ORP’s market-maker-only availability is not “unfair” because it is proportionate to market makers’ unique, continuous-quoting risks and is expected to increase displayed liquidity at competitive prices for everyone. The opinion also dismantled the rhetorical “rigged game” framing by rejecting the “backing away” premise: ORP reprices/cancels before an order can match and execute.

F. No undue burden on competition: procompetitive narrative + Exchange Act purposes

The court upheld the SEC’s Section 6(b)(8) analysis on two independent grounds:

  • No undue burden shown: The SEC reasonably found IEX Options can compete alongside other exchanges and that the model may spur competitive innovation (with IEX’s modest equities market share serving as reality-check evidence against claims of market foreclosure).
  • Even if burden existed: The SEC reasonably tied the proposal to Exchange Act purposes—efficient execution, fair competition, and investor protection—particularly by limiting a “surcharge” imposed by latency arbitrage and encouraging tighter spreads/deeper liquidity.

3.3. Impact

The decision has three likely doctrinal and practical effects:

  • Lower “quant proof” expectations in SRO approvals (within reason): The opinion signals that, in fast-moving, technical market-structure settings, the SEC may support key findings with participant comments and expert synthesis—especially when the product is not yet live—and still satisfy substantial evidence and APA review.
  • CAT is not a mandatory analytic prerequisite: Parties challenging SEC approvals in the Eleventh Circuit will face an uphill climb arguing the SEC acted arbitrarily by not extracting and modeling CAT data, particularly given statutory deadlines and Prometheus Radio.
  • Protected-quote status turns on the Options Plan text and the exchange’s rules: Challenges will need to engage directly with the Options Plan’s definitions and the exchange’s specific “non-firm” conditions, rather than analogizing to Regulation NMS equities concepts.

Substantively for market design, the decision gives additional appellate validation to IEX-style “speedbump + repricing/cancellation” mechanisms in a new asset class (options), potentially encouraging further experimentation by exchanges and additional SEC approvals that rely on similar evidentiary mixtures.

4. Complex Concepts Simplified

  • Latency arbitrage: Profiting from tiny delays in information/trade-message travel between venues. When prices move, a fast trader can “pick off” an outdated (“stale”) quote before the liquidity provider updates it.
  • Speedbump (access delay): A deliberate delay (here 350 microseconds) applied to incoming orders so that the exchange’s systems can update/cancel stale quotes before a would-be latency-arbitrage order reaches the matching engine.
  • Indicator and ORP: The Indicator detects moments of quote instability/dislocation by watching other exchanges; the ORP (if a market maker opts in) cancels or reprices that market maker’s quote when dislocation suggests high latency-arbitrage risk.
  • Protected quotation (options): A best displayed bid/offer that triggers order-protection routing obligations under the Options Plan. Citadel’s argument tried (unsuccessfully) to redefine “protected” by importing equities-only Regulation NMS concepts.
  • Substantial evidence vs. arbitrary-and-capricious: “Substantial evidence” asks whether enough relevant evidence supports a factual finding; “arbitrary-and-capricious” asks whether the agency reasonably considered the issues and explained itself within a “zone of reasonableness.”

5. Conclusion

Citadel Securities LLC v. U.S. Securities and Exchange Commission entrenches a deferential, text- and record-grounded approach to reviewing SEC approvals of exchange microstructure innovations. The Eleventh Circuit held the SEC can find latency arbitrage in options markets and approve targeted countermeasures on the strength of participant comments, structural analysis, and expert judgment—without being compelled to produce comprehensive quantitative studies or CAT-based modeling. It further clarified that options “protected quotation” status is governed by the Options Plan (not Regulation NMS), and it upheld the SEC’s determinations that IEX Options’ ORP is neither unfairly discriminatory nor an undue burden on competition. The opinion’s broader significance lies in how it aligns administrative-law deference with the realities of high-speed markets: agencies may act on imperfect but reasonable evidence when time, complexity, and statutory design require practical decision-making.