Curatia Analysis for Wed, March 20: What the SEC Nailed and Neglected in Enhanced Rule 605 Disclosures: An Exclusive Q&A with S3 CEO Mark Davies

When the SEC finalized its proposal to enhance Rule 605 disclosures earlier this month, many industry participants breathed a sigh of relief due to both the agency’s responsiveness to feedback and its willingness to embrace a piecemeal approach to overhauling US stock trading.

Following that reveal, though, experts have yet to publicly peel back the onion on what changes regulators implemented in the final rule and how they could impact trade reporting and analysis. We sat down with S3 CEO and Rule 605 savant Mark Davies to do just that.

1. • How would you grade the SEC on its implementation of enhanced Rule 605 disclosures? How much industry feedback did the agency take into account when finalizing the new rule? Overall, I’d give their edits a B+. They incorporated a decent amount of feedback from the industry and implemented a lot of meaningful changes, but of course, they didn’t include everything I feel would have improved the rule.

• • What feedback did the SEC include in the rule that made it better? There were a lot of adjustments to the initial proposal that the industry will see as improvements. Here are some of the most consequential: 
    • • Size grouping will now be based on notional rather than arbitrary lot sizes, significantly improving comparability.
    • • The effective over quoted (E/Q) calculation will now use “spread weighting,” which is the current industry standard for internal analysis. (Quoted spread is not included on the current 605.) S3 has been calculating it this way for the last two decades at the request of many of our customers.
    • • Stop orders will now distinguish between stop market, stop limit, and stop marketable limit.
    • • The SEC made some changes to how the opening should be handled. In the initial proposal, the agency had some statistics based off the first NBBO disseminated after the time of order receipt for orders that occurred outside of market hours. This led to a situation where trades that occurred before the primary market open could skew statistics. They also added a concept called “interim opening period,” which represents the time between the first NBBO being disseminated and the primary market open. These orders will now be treated more like orders that occur outside of market hours.
    • • Regulators eliminated statistics that don’t make sense for non-marketable order types (effective spread, E/Q, etc.).
    • • In the initial proposal, there was a concept called “beyond-the-midpoint-order,” which refers to orders that were on the near side of the spread. They updated it to “midpoint-or-better” to include the midpoint.
    • • They added immediate-or-cancels (IOCs) on other order types, not just marketable order types.
    • • They eliminated speed stats for median and 99th percentile (which were impossible to aggregate), and instead opted for several columns of execution time groupings, all the way down to “less than 100 microseconds.”
    • • They corrected size improvement to be relative to order time instead of execution time and added a metric to show how much size improvement was achieved.
    • • They added a fill rate relative to exchange-only in addition to the market as a whole.

• • What industry feedback did regulators ignore or explicitly reject? While a lot of the changes to the initial proposal are positive, there are some issues that were ignored: 
    • • The SEC didn’t parameterize the order types (e.g. marketability as one parameter, stops as another, IOC, etc.), opting instead for a subset of possible scenarios. (e.g. They say they don’t think Market IOC orders will have a different execution profile from market orders that aren’t specified as IOC, indicating that market orders have implicit IOC.)
    • • The size buckets parameters don’t consider fractions. Regulators rejected a suggestion to segregate orders over one share that include a fraction from those that don’t.
    • • Speed statistics for non-marketable limit orders were not restricted to only the first execution. This may result in distorted fill times for some orders.

1. • What challenges will firms face in trying to do this reporting by themselves? For limit orders, in the current 605 report, all you need to do is look at the quote at order time and execution time. With the new version, you need to track every quote from order time to execution time or cancellation. It will require a market replay for every order rather than simply a point in time. Additionally, it will require looking at all BBO quotes to determine the depth.These changes will be effective 60 days after federal register, with an implementation timeline of 18 months. The biggest challenge with this timeline is that while the SEC has defined the data points for the detailed report, they have not actually defined the format. That will need to be decided by the plan participants — the self-regulatory organizations (FINRA and the exchanges). This will need to happen fairly quickly for firms to be able to stay on track to meet the implementation deadline.As the provider of the industry’s leading software solution for automating Rule 605 reporting, S3 will of course be on top of the substantial reporting changes the SEC’s final rule necessitates. Firms looking to meet their reporting needs painlessly can contact us for assistance.

1. • Besides defining the report formatting, are there any other outstanding questions that remain? There are several points that will need to be clarified or defined. For example, rulemakers reference share-weighted average quoted spread. What shares is this weighted by? Quoted spread is not currently included, so the way we calculate now is using the order volume. Is this appropriate? Or should it be the NBBO volume? Is it the aggregate volume at the NBBO price point? For share-weighted average midpoint, what shares is this weighted by?For stop orders they reference “triggered,” but I couldn’t find a definition of “triggered.” In the initial proposal, they define “executable” stop orders based on a relationship between the stop price and the market, which is a different definition than what can be found in FINRA 5350 (the current standard). They have now removed that definition and defined stop orders as executable relative to when they are triggered (which is in line with the industry’s feedback). But they didn’t specifically say what “triggered” meant. I’m assuming the SEC is leaving this to the industry to decide because it had already been defined by FINRA, but they did not explicitly point to the FINRA definition.

• Any other observations? Here are a few more that may prove useful for reporting firms: 
  • • The SEC removed the requirement of reporting for auctions and said they are still considering it but did not indicate any intention to include it in the future. They did, however, give indications of other changes, such as odd-lots going live.
  • • Substantially more statistics are required in the Summary. We are waiting on customer and industry feedback to determine if this change accomplishes what the industry was hoping for.
  • • Broker-dealers (BDs) that route to clearing firms still need to create their own 605 in addition to pointing to the one created by the clearing firm.
  • • Non-US customers are explicitly included in the count of 100K clients.
  • • BDs who internalize fractional and other orders will need to report that internalization separately.
  • • Market centers / BDs should use the SIP until MDI is live, not their direct feeds.
  • • More realized spread buckets were added.
    • The SEC is not requiring a centralized repository.

  
  There is quite a lot of information in the release, and it will take some time for everyone to get their arms around it. We at S3 will be living and breathing this rule as its full reporting implications become apparent, so I’m happy to answer any questions people may have as they work through it.