The Two-Tiered Market
During High School, I recited the pledge of allegiance over the loudspeaker to the school most mornings.
While there are many today that view the pledge as an anachronism (or worse), the words "with liberty and justice for all" have always meant something to me. One of my most deeply held beliefs is that we should strive to make equality of opportunity a reality in America. While I recognize that achieving this is very complex for society in general, it should not be the case in the financial markets. As the previous SEC chair has commented, we should strive to eliminate "two tiered markets".
Considering the market structure that existed when I started working on Wall Street in the mid-1980s, we have made a lot of progress.
The equity markets in 1985 were far from an egalitarian model. On the NYSE, firms with their own floor brokers generally received better executions than those without them and it even mattered where their booths were located (location of "booths" close to key specialist posts were highly valued). Specialists had a significant amount of discretion to treat orders "differently", so there was a lot of value in building relationships. Notably, when receiving orders, via the electronic DOT system, the specialists had discretion and could see the name of the firm sending the order. This point was underscored by a conversation I had with an ex-Morgan Stanley quantitative trader after we had both left the firm. He confided in me that his DOT executions, at his new firm, were far worse than he was used to receiving, based on a series of quantitative metrics. He was told that his new firm, which did not have a significant floor presence, had "issues" with several specialists.
This truly was a "two tiered market", but the NASDAQ market was arguably worse. First, there was outright collusion of NASDAQ market-makers, which resulted in wider quotes that disadvantaged clients. In addition, there was both a Public market, which displayed quotes from the NASDAQ system and a Private market that included the Instinet terminal. At that time, large brokers often used Instinet to trade with each other, inside of the posted bid and offer. After establishing a position at a superior price using Instinet, they sometimes filled customer orders from their inventory at the (wider) NASDAQ price. The good news is that both of these examples were put to an end by an enforcement action brought by the SEC and the passage of the "Order Handling Rules" that included "Limit Order Display" and "Manning" protection.
Given that history, it is interesting that many argue that technology was the driving force behind breaking the oligopolistic features of the equity markets. I believe that without the Order Handling Rules, NASDAQ would not have changed and without Regulation NMS essentially forcing the NYSE to become electronic, it is unlikely that the monopoly power of the specialists would have been broken. At a minimum, it is clear that regulation has been instrumental in driving our markets to be more accessible.
While today's market structure is clearly more egalitarian, there are still issues, particularly with regard to transparency. As we have stated, the vast majority of all order routing behavior is undisclosed and there is no public standard for best execution reporting, except for retail orders under 10,000 shares. Despite this, the SEC seems to be focused on "Industry Solutions" to routing transparency and does not seem to be interested in changing the best execution reporting rule. I hope that I am wrong about both, as I believe that changes to both rules are necessary.
The notion that institutional routing disclosures should be kept, exclusively, between investors and their broker dealers, does have some intuitive appeal. The institutional trading business is nuanced, and very prescriptive disclosures of order routing could become overly complex. It potentially could reveal sensitive trading information or inadvertently reveal intellectual property related to smart order routing. That said, having a standard set of disclosures at an aggregate level (not stock by stock), which is applied to all Smart Order Routing activity, would be extremely helpful without those risks. At a minimum, such a regime would allow investors to understand how the linkages in our markets really work and would force all brokers and exchanges to participate. It would also become the basis for better private disclosures, as the exact same infrastructure would support more detailed, private disclosures with enhanced analytics. If, on the other hand, disclosures were only required "upon request," without a public component, it would essentially create a two tiered market for information. Smaller broker dealers would be less likely to participate, making industry comparisons even more difficult. Smaller investors would have no basis of comparison between their orders and other investors and would have a harder time getting a standard format to analyze. Further, the lack of public disclosure would allow the critics of our markets, with no access to the data, to continue to claim that routing practices are opaque. The result would be the continued impairment of investor confidence.
The rhetoric surrounding best execution reporting is similar. Despite the consensus that we have need of more and better data on best execution, very few people support specific improvements. It seems to be an almost perfect example of a problem that economists often refer to as the "public good" problem. Everyone wants better streetlights, but no one wants to pay for them. In this case, everyone wants to see better data on execution quality, but no one wants to disclose their own statistics, unless forced. Once again, the argument that institutional orders are too nuanced to measure is made, but in this case, it is, as the saying goes, "a load of crap" to conclude that all executions are hard to measure.
Many commenters explain that the key to analyzing best execution is to measure the slippage of institutional orders at the "parent" level and cite a variety of difficulties in doing so. They argue how problematic it is to conduct institutional best execution analysis due to the complexity and variability of the process. The flaw in this rhetoric is that, despite the fact that there are many steps in the institutional trading process, the last step is always the same. In every case, there are immediately executable orders sent to brokers, ATSs, or exchanges. Those orders could all be measured if the SEC made some straightforward changes to rule 605. Today, most immediately executable orders sent to exchanges or ATSs can be excluded from the rule for being part of a "not held" or oversized order, but eliminating those exclusions does not seem to have the support of the consensus. If those exclusions were to be eliminated and those orders were categorized and reported on in statistically equivalent groupings, investors of all types would gain. It is important to remember that, despite major imperfections in Rule 605, it does provide data for much of the order flow submitted by retail investors.
The history of how those retail orders have been treated is particularly instructive. Retail brokers have determined how to separate out the relevant from the irrelevant statistics for their own orders, and use that data in their negotiations with their market makers. The result has been an incredible increase in execution quality over the past decade. Despite not being part of the rule, the metric of effective divided by quoted spread (E/Q) has become a de-facto industry standard. By that metric, the execution cost for market orders has shown more than a 31% decrease in costs. (Q1 2007 for all sizes showed an E/Q of 130.69 and the average E/Q in Q1 2015 had declined to 89.55) This improvement, while significant, actually understates the improvement to retail investors, whose private statistics show far better execution quality. This history has proven that a rule which forces aggregate disclosure will also spawn more detailed private analysis with a result of a much higher level of transparency and improvements in execution quality.
Ultimately, the question is, what should be the government's role in the availability of data for analyzing the markets. It's interesting in this political atmosphere, with debates over the government's use of very broad domestic surveillance techniques, that requiring more information disclosure for equity trading is even an issue. This is a perfect example of needed surveillance. While many people (including yours truly) are deeply concerned about the NSA and DEA's, seemingly, overarching use of surveillance techniques, without requiring probable cause, securities regulators are being overly careful. This is unfortunate, since public disclosure of information is the best method of exposing and countering bad actors, whether they are private citizens, corporations, or in government itself. That conclusion was articulated well by James Madison
"A popular government, without popular information, or the means of acquiring it, is but a prologue to a farce or a tragedy; or perhaps both. Knowledge will forever govern ignorance; and a people who mean to be their own governors must arm themselves with the power which knowledge gives."
We would do well to heed those words.
David Weisberger, Managing Director, Trading Services at Markit
Posted 6 June 2015
S&P Global provides industry-leading data, software and technology platforms and managed services to tackle some of the most difficult challenges in financial markets. We help our customers better understand complicated markets, reduce risk, operate more efficiently and comply with financial regulation.
This article was published by S&P Global Market Intelligence and not by S&P Global Ratings, which is a separately managed division of S&P Global.