Back to News

Rethinking the Economic Analysis in the SEC’s Best Execution Proposal

Introduction

In this follow-up to my April 17th guest blog post, I evaluate the U.S. Securities and Exchange Commission’s proposed Regulation Best Execution rule (the “Best Ex Proposal”), which is part of a package of four equity market structure rules proposed by the Commission on December 14, 2022.

My comments focus on the economic analysis within the Proposal. I believe the Commission has not adequately justified the need for new regulations. Instead, it relies on internal analyses and unsupported conjecture without a solid economic foundation. I argue that the Best Ex Proposal fails to provide clear evidence of a market failure that would necessitate new regulations. Moreover, the SEC’s attempts to quantify the economic impacts are fundamentally flawed, making their conclusions unreliable.

The SEC implicitly suggests that the current system for equity trading lacks sufficient oversight, implying that the current best execution evaluation regimes used by FINRA and the MSRB are inadequate and fail to serve retail investors adequately. However, instead of providing concrete evidence of conflicts of interest between investors and broker-dealers that lead to inferior executions, the SEC relies on weak justifications. These include acknowledging that any benefits “may be small” and using flawed analyses often accompanied by qualifying language like “might” and “may.”

The Current Role of Retail Brokers and Wholesalers

In today’s equity markets, retail brokers offer their customers a comprehensive suite of services that includes more than trade execution. These services include managing customer accounts, providing custody services, offering margin credit, facilitating access to competitive banking products, and connecting customers with investment advisors. Additionally, brokers enhance the trading experience with customer service, educational materials, and robust, user-friendly trading interfaces and applications.

In today’s competitive market for retail order flow, retail brokers largely act as “introducing” brokers who outsource order execution to wholesalers, rather than executing trades themselves. Thus, when a retail investor places an order, the retail broker may route it to a wholesaler when the broker deems it likely to obtain the best execution quality. Wholesalers play a critical role as both executing brokers and market makers. The wholesaler can execute the order by either trading from its own inventory or by facilitating its execution by routing it to an exchange or alternative trading system (ATS).

This system allows retail brokers to focus on providing high-quality customer service while leveraging the expertise and infrastructure of wholesalers to achieve optimal trade execution by forcing wholesalers to compete for the order flow based on the execution quality they provide. Rule 605 reports and the SEC’s own proposals demonstrate that wholesalers consistently provide better prices than those publicly displayed, often executing orders at the midpoint price or better.

Both retail brokers and wholesalers have a best execution obligation, meaning they must strive to provide the best possible terms for their customers’ trades. Retail brokers meet this obligation by adjusting their routing decisions based on the relative performance among wholesalers and marketplaces. This competitive dynamic encourages wholesalers and marketplaces to deliver superior execution quality, benefiting retail investors through better prices and increased liquidity.

Two Critiques of the SEC’s Economic Analyses

The SEC’s rationale for a new best execution standard primarily rests on two key analyses: the Midpoint Liquidity Analysis and the Execution Quality Analysis. The former analysis points to untapped liquidity at certain price points that could enhance execution quality, while the latter identifies potential price improvement that is not being captured under the current system. I conclude that both analyses are problematic, fundamentally flawed, and therefore unreliable.

Midpoint Liquidity Analysis

The Midpoint Liquidity Analysis examines the potential availability of midpoint liquidity. Midpoint liquidity is used to identify potential trades that could be executed at the midpoint of the national best bid and offer (NBBO) prices. Midpoint pricing is an aggressive benchmark for evaluating best execution because a broker that executes every trade at the midpoint would earn zero profits.

The SEC’s analysis of midpoint pricing uses Consolidated Audit Trail (“CAT”) data to identify hidden liquidity at the midpoint on at least one venue at the same time wholesalers internalize retail orders at prices worse than the midpoint.1 The SEC concludes that there is substantial untapped liquidity at the midpoint of trades that is not being accessed under the current system.

Although it is not completely clear how the SEC performed this analysis, it does appear overly simplistic. For example, it does not adequately account for factors such as the minimum quantity restrictions imposed by many trading venues.  If an institutional investor deliberately places hidden orders with a minimum quantity restriction, those orders would not trade against other smaller-sized (such as retail) orders even if they were at the same venue. Institutional investors often trade large blocks of stock for various reasons, including minimizing information leakage. In contrast, retail investors usually trade in smaller quantities. As a result, much of the hidden institutional liquidity identified by the Commission is likely inaccessible to most retail investors. This would overstate the amount of accessible midpoint liquidity, thus discounting a significant fraction of the Commission’s estimated benefits.

An interesting and underappreciated finding is that, according to the SEC’s Midpoint Liquidity Analysis, 45% of all retail orders are executed at the midpoint or better.2 The fact that almost half of all orders receive midpoint pricing or better suggests that markets for retail order flow are highly competitive.

Execution Quality Analysis

The SEC found that wholesalers “compare favorably to exchanges in the execution quality of orders routed to them, suggesting that execution quality could be another key factor in the decision of retail brokers to route to wholesalers. In particular, marketable orders routed to wholesalers appear to have higher fill rates, lower effective spreads, and lower E/Q ratios.”3 Despite these favorable findings for the current market structure, the SEC believes that retail investors should be able to obtain even better executions.

After noting that the “realized spread measure is an imprecise proxy for the profits market makers earn supplying liquidity,” the SEC produces an analysis of realized spreads in Table 5 of the Best Execution Proposal and conjectures that:

“The higher realized spreads associated with orders handled by wholesalers observed in Table 5 suggest that wholesalers have an opportunity to earn higher economic profits than liquidity suppliers on exchanges after accounting for adverse selection costs (i.e., after adjusting for price impact).”4

There are two significant methodological shortcomings in this analysis. First, realized spreads are a poor proxy for market maker profitability. Second, reliance on mean realized spreads for wholesalers and exchanges fails to control for the distribution of order types.

Realized spreads, which reflect the difference between the buying and selling prices adjusted for the midpoint of the quote soon after a trade (the SEC 605 reports realized spreads using a five-minute delay), are not accurate indicators of a market maker’s profitability because they fail to account for critical factors affecting market maker profitability, such as inventory costs, fixed costs, and transaction fees.5 From the perspective of a market maker who seeks to earn the spread, the realized spread reflects the degree to which incoming order flow predicts future price movements, making it more of a proxy for the adverse selection faced by the liquidity provider and not an accurate reflection of the liquidity provider’s profitability. Ideally, to effectively measure wholesaler profitability using realized spreads, the time-period used to calculate realized spreads should approximate the market maker’s assumed holding period, i.e., the time it takes to flatten the position.

The second problem with the Execution Quality Analysis relates to its comparison of the levels of realized spreads for wholesalers and exchanges. Table 1, derived from Table 5 of the Best Execution Proposal, shows that realized spreads for all marketable orders are 0.61 basis points at wholesalers and -0.38 basis points at exchanges. According to the SEC, the difference in these spreads reflects potential price improvement that could be provided to retail investors if markets were more competitive.

Aside from the observation that realized spreads are poor proxies for market maker profitability, the SEC’s interpretations are not based on solid economic principles for two reasons. First, if realized spreads reflect market maker profitability, the negative realized spread for marketable orders executed at exchanges implies that, on average, on-exchange market makers operate at a loss, which seems unrealistic.6

A second more serious methodological concern is that this comparison is apples-to-oranges. The SEC’s calculation does not account for the different mix of market and marketable limit orders executed by wholesalers versus those on exchanges. I demonstrate below that a simple adjustment to reflect the actual distribution of order types executed by wholesalers would reverse the SEC’s conclusions about wholesaler profitability compared to on-exchange market makers. Consequently, the SEC’s assertion that wholesalers could offer additional price improvement because they earn higher realized spreads is invalid. After adjusting for the order type distribution, realized spreads on exchanges are actually higher.

For wholesalers, Table 1 shows that 20.81% of all marketable orders are marketable limit orders, while the remaining 79.19% are market orders. By contrast, 99.68% of all marketable orders executed on exchanges are marketable limit orders, while only 0.32% are market orders. This difference likely reflects the fact that wholesalers almost always use capital to internalize market orders, while they route some fraction of their marketable limit orders directly to exchanges.

Assuming, as the SEC does, that the quality of trade executions on exchanges remains unchanged, a recalculated realized spread for marketable orders that accounts for the distribution of order types on exchanges is 1.82 basis points (.2081 x -0.39 + 0.7919 x 2.40), instead of -0.38 basis points. This implies that after adjusting for the proportions of market and marketable limit orders but otherwise applying the Commission’s logic, it would be more expensive to route orders to exchanges than to have wholesalers handle their executions. This adjusted measure would necessarily lead to the conclusion that wholesalers earn fewer profits on a per-share basis than market makers on exchanges.

Conclusion

As a former SEC Chief Economist, I believe the economic analysis in the Best Ex Proposal is inadequate. It fails to identify a concrete market failure and relies on conjecture rather than a convincing, data-driven analysis. Furthermore, the SEC’s efforts to demonstrate a need for improved retail execution quality by focusing on midpoint liquidity and execution quality are fundamentally flawed and unreliable. Essentially, the Commission is pursuing regulation for its own sake. A more reasonable approach would be for the SEC to maintain the current competitive ecosystem while encouraging the industry to pursue data and market-driven incremental improvements that do not pose risks to the significant benefits that retail investors enjoy today.

Author

Craig LewisCraig M. Lewis is the Madison S. Wigginton Professor of Finance, Emeritus at the Owen Graduate School of Management at Vanderbilt University. He is a former Chief Economist and Director of the Division of Economic and Risk Analysis (DERA) at the U.S. Securities and Exchange Commission (SEC). 

 

 

Footnotes

  1. It should be noted that it is nearly impossible for anyone other than the Commission to perform this analysis, because it used CAT data and involved non-public, non-displayed orders. While the release contains a general description of what the analysis purports to do, it is difficult to understand what assumptions or decisions the staff had to make when implementing the analysis, or to verify that it was done correctly. I understand that SIFMA submitted a FOIA request for the public release of the CAT data used by the Commission in its analysis. See “Request to Extend Comment Period on Four Rule Proposals,” SIFMA, February 8, 2023). The release of this data would provide the public with an opportunity to assess the accuracy and reliability of the Commission’s analysis. [↩]
  2. According to Panel B of Table 7 in the SEC’s Order Competition Proposal, 44,57% of all wholesaler executions are executed at the midpoint of better. It also shows that 31.69% of all wholesaler executions are executed exactly at the midpoint. [↩]
  3. See Securities and Exchange Commission, 17 CFR Parts 240 and 242, Release No. 34-96496; File No. S7-32-22, RIN 3235-AN24, Regulation Best Execution, page 228. [↩]
  4. See Securities and Exchange Commission, 17 CFR Parts 240 and 242, Release No. 34-96496; File No. S7-32-22, RIN 3235-AN24, Regulation Best Execution, page 233. [↩]
  5. The SEC’s second analysis CAT data uses a one-minute time delay to measure realized spreads. Using a shorter time window does not affect any of my conclusions as it still represents a delay that is much longer than the amount of time it takes for a wholesaler to flatten its position. [↩]
  6. Given the delay in measuring realized spreads, a negative value suggests that the realized spread may be a better measure of adverse selection costs than profitability. [↩]

Related Resources

More News