Equity Market Structure Regulation: Time to Start Over

Paul G. Mahoney is the David and Mary Harrison Distinguished Professor at the University of Virginia School of Law. This post is based on his paper, forthcoming in the Michigan Business & Entrepreneurial Law Review.

Over the past 15 years, stock trading has switched from a mostly manual to a mostly automated process. The SEC’s Regulation NMS provides a regulatory framework for the new world of electronic exchanges. My recent paper, Equity Market Structure Regulation: Time to Start Over, forthcoming in the Michigan Business & Entrepreneurial Law Review, argues that Regulation NMS has not achieved its goals and should be replaced with a simpler, less prescriptive system.

In the early 2000s, the NYSE and Nasdaq dominated trading in large-company stocks. Today, trading is fragmented among 13 different exchanges, with three more on the way. The NYSE and Nasdaq previously used different trading systems. The NYSE operated a physical trading floor. On Nasdaq, competing dealers posted buying and selling prices electronically and brokers transacted with them by telephone.

Today, all exchanges are primarily electronic and automated. Brokers enter orders directly into the exchange’s hardware and software system, which continuously and automatically executes all trades that can be made. An exchange today is a computer server and a set of rules incorporated into code.

There are two basic types of orders a retail investor, through its broker, may enter into an exchange, a limit order and a market order. A limit order is an offer to buy (a “bid”) or to sell (an “ask” or “offer”) at a specified price. The trader is willing to transact, but only if someone else accepts its price. A market order is an instruction to buy or sell at the best available price. In contractual terms, it is an acceptance of the (best priced) offer to buy or sell. When a market order enters the system, it is matched against a limit order or dealer quotation on the other side of the market. The system is designed to find the best price for each incoming market order.

Limit order traders, therefore, sell liquidity (the ability to transact quickly at close to the market price) and market order traders purchase liquidity. The bid-ask spread, or gap between the highest bid and the lowest ask, sets the price of liquidity. Today, the primary liquidity providers are high-frequency traders, or HFTs, who use dedicated communications networks and computer algorithms to enter and cancel orders across all markets to earn a spread while minimizing risk. HFTs also search for “stale” quotes, or limit orders that slower traders have not yet updated in response to new information.

Why have equity markets changed so dramatically and quickly? Technology is the main reason. Electronic markets bring buyers and sellers together faster and more cheaply than the old NYSE or Nasdaq systems. Electronic exchanges are also less expensive to create than floor- or dealer-based systems.

Regulation is a subsidiary reason. In 1975, Congress instructed the SEC to create a “national market system” but left that concept largely undefined. Over the next 30 years, the SEC adopted increasingly prescriptive rules telling exchanges how to trade stocks. That process culminated in Regulation NMS, adopted in 2005.

The SEC’s preferred market structure consists of separate but linked markets. Exchanges publicly display their best bids and offers in a consolidated quotation stream that identifies a “National Best Bid and Offer” (NBBO) at any moment in time. Exchanges may not execute trades at prices inferior to the NBBO; if necessary, they must re-route orders that come to them to another exchange.

The theory behind the separate-but-linked design is that it simultaneously provides competition among orders to supply the best price and earn the spread, and competition among exchanges to attract trading interest and earn trading and data fees. In principle, this should produce a narrow bid-ask spread and low fees for trading and data access.

The reality has fallen short of the goal. Prices for market data are not set competitively because Regulation NMS collectivizes best-quote data rather than permitting exchanges to sell it individually. Regulation NMS and related rules encourage exchanges to operate essentially identical trading systems, reducing innovation. Exchanges compete for trading volume through complex fee structures that compensate brokers for trades. They also compete by offering increasingly complex variations on the basic limit and market orders that are particularly appealing to HFTs.

The model does not maximize competition among orders. In a consolidated market, a trader can improve its chances of earning a spread only by improving its price. The separate but linked market, by contrast, offers a trader many ways to improve its chances of trading without improving its price. The trader can move to an exchange that offers a rebate to the counterparty or one with a shorter trading queue. Traders thereby free-ride on prices quoted by others, reducing their risk but making prices less informative. HFTs are particularly adept at moving orders rapidly from one venue to another. The likely result is excessive investment in speed and insufficient investment in information.

The SEC has recently recognized some of these drawbacks and taken small steps to alleviate them. Earlier this year, it proposed a rule allowing competing information processors to replace the single consolidated quotation stream. Last October, it issued a policy statement requesting proposals for improvements to market structure for small-company stocks. It also adopted a rule, since vacated by the DC Circuit, instituting a pilot program to determine the effect of the rebates that exchanges offer brokers.

Rather than tinker at the margins with a system that produces so many unintended consequences, the SEC should replace Regulation NMS with a simpler set of rules. These should be based on the following principles:

  • Exchange autonomy—exchanges should be free to design their trading systems as they see fit, with no mandate to connect or route trades to any other exchange. Brokers, not exchanges, should search for and obtain the best price for their customers.
  • Regulatory consistency—brokers operate proprietary trading platforms known as Alternative Trading Systems (ATS) that execute trades but are not regulated as exchanges. The differing regulatory treatment should be eliminated.
  • Issuer choice—under current rules, every exchange can, and typically does, trade every listed company’s stock. The SEC should permit companies to select the venue(s) that trade their stock. Some companies might choose to concentrate all trading on a single exchange, while others might opt to spread trading among many venues.

There are two positive things to be said for the current system. It has coincided with a period of declining commissions and bid-ask spreads. It also forces exchanges to backstop the broker’s obligation to find the best price for its customer. The first, however, is a trend that long predates Regulation NMS and is not limited to the United States, so Regulation NMS is not the cause. The second is hard to justify on cost-benefit grounds. It would be simpler for the SEC to take a strict enforcement stance on violations of the broker’s duty of best execution. At the margin, Regulation NMS likely encourages order routing practices that maximize broker rebates and payments for order flow at the expense of customers.

Adopting simpler market structure regulations based on these principles would facilitate innovation in trading system design and give exchanges, brokers, and issuers more control over the use of technology to improve liquidity and price discovery.

The complete paper is available for download here.

Trackbacks are closed, but you can post a comment.

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>