Allium Comments on FINRA's Proposal to Modernize Best Execution Guidance

Allium has submitted a comment letter to the Financial Industry Regulatory Authority on Regulatory Notice 26-15, which asks how FINRA should modernize its best execution guidance under Rule 5310.

The Notice poses a specific scenario. If the SEC rescinds Rule 611 of Regulation NMS, members would have greater discretion over which venues they access and how they handle customer orders. The duty under Rule 5310 would not change. What a firm needs in order to support that duty would.

Our recommendation is the one we made to the SEC in August, carried forward: standardize evidence, not outcomes. Establish a shared measurement vocabulary and reliable supporting evidence, without prescribing a single benchmark, execution method, or routing result.

What makes execution evidence useful

We propose four qualities, which describe supporting evidence rather than adding new execution-quality factors.

Traceable. A material factual claim connects back to its source records, with important transformations or assumptions identified.

Comparable. Prices, costs, instruments and timing are described consistently enough to understand what is being compared, without treating unlike instruments or trading conditions as equivalent.

Reconstructable. A reviewer can assess how the firm reached its result and, where appropriate, reproduce the material analysis.

Decision-relevant. The evidence lets a reviewer discern material differences in execution quality and identify what needs investigating.

Their practical purpose is to separate a record of what happened from support for why the firm's judgment was reasonable.

Shared definitions, not a shared benchmark

Recurring terms need enough content that unlike measurements are not presented as equivalent. "Available liquidity" should identify the access conditions and the quantity. "Price improvement" should identify the reference price and the measurement event. A "fill rate" should say whether its denominator is orders, attempts, or quantity. A measure of "completion" should state its endpoint.

Firms would keep their own data formats and map material definitions and assumptions to the shared concepts. This builds on the reporting and audit-trail terminology already in place rather than creating a parallel schema, and FINRA could develop the examples incrementally.

Decisions before results, and results that test decisions

A firm should set the criteria governing order objectives, venue access, benchmark selection and the treatment of costs and conflicts before outcomes are known. It should also establish who reviews, how often, and what triggers escalation.

The corresponding evidentiary principle is that a firm should not, without disclosure, substitute a post hoc standard for the one that originally governed its decision. Correcting data, improving a model or running a new diagnostic is legitimate, provided the original basis is preserved and the change is identified. Selecting the benchmark after seeing which one makes the result look better is retrospective benchmark shopping.

The reverse also holds. A well-documented decision is not automatically a reasonable one, and an unfavorable result is not automatically evidence of deficient diligence. We ask FINRA not to create a safe harbor resting on documentation, adherence to a methodology, or satisfaction of a self-selected benchmark.

Applying the existing factors to tokenized securities

The SEC granted temporary conditional exemptive relief on September 17, the Innovation Exemption, which gives this a concrete setting. It provides exchange-definition relief for qualifying Tokenized Securities Venues and dealer-definition relief for certain liquidity providers, in a context of permissioned automated market maker trading on public, permissionless ledgers. The relief runs to 2031 and is subject to modification. It does not establish that any particular venue is operating, that every member may access one, or that broker-dealer obligations are relieved in general. Our position is that FINRA should publish targeted examples applying the existing best-execution factors to this setting rather than write a separate standard for it.

The practical questions are familiar ones in unfamiliar surroundings. Evidence should establish what security the customer would actually acquire, rather than the issuer name or price reference a token displays. Where an AMM's price depends on quantity and pool state, the nominal pool price is not the executable price for the whole order. A reported trade, a blockchain inclusion, operational availability and legal settlement can be four different events, and a confirmation does not prove the customer received the required security or proceeds. Faster settlement does not by itself mean better execution.

Where Allium fits

Consistent blockchain data can organize observable activity across protocols and chains into comparable, traceable records that join to a firm's own decision evidence. That is the contribution we are describing, and it has limits worth stating plainly. Onchain records cannot establish customer instructions, eligibility, private quotations, or alternatives that were considered and never submitted. Review will still require authenticated offchain records.

What shared evidence offers is a common basis for comparing mechanisms. It does not make onchain execution superior, and no single dataset establishes compliance.

The aim of all of this is not to make every firm choose the same route. It is to make the basis for its choice understandable, its performance measurable, and material deficiencies actionable.

The letter is signed by Ethan Chan, CEO, and William Lai, Head of Institutions.

This letter continues the argument we made in our August comment to the SEC on Rules 611 and 610(e).

Read Allium's full comment letter to FINRA