A SUMMARY OF ON THE OPTIMALITY OF SHAREHOLDER CONTROL: EVIDENCE FROM THE DODD-FRANK FINANCIAL REFORM
by Robert Battalio, Professor, Mendoza College of Business, University of Notre Dame.
The options exchanges recently began producing voluntary execution quality statistics in an effort to forestall the extension of mandated execution quality statistics to options markets. Battalio, Shkilko and Van Ness (2011) find that the reports produced by several options exchanges overstate the cost of liquidity by a considerable amount. The apparent ambivalence shown by certain options exchanges toward producing accurate execution quality statistics suggests these statistics have little, if any, impact an exchange’s market share of trades.
How did execution quality statistics arise in equity markets? Why don’t we have them in options markets? Do we need them in options markets? I address each of these questions below.
In the beginning, execution quality statistics were used to attract retail order flow. In the late eighties and early nineties, brokers who either sold or internalized their retail orders argued they were obtaining best execution for customer orders since they were executed at the best prevailing quote. These claims frustrated officials at the NYSE, since retail orders executed by the NYSE often received prices that were better than the best quoted prices. To reverse the loss of retail orders to markets internalizing or purchasing orders, in November of 1995 the NYSE began a pilot program named “ML/NYSE PRIME.” In this program, the NYSE began informing Merrill Lynch’s retail investors as to the savings they received because of “price improvement”—the execution of market orders at a price better than the best current bid or offer for that stock.
Soon after the inception of this program, the NYSE’s competitors complained to the SEC that the NYSE should also inform Merrill Lynch’s customers when they receive prices that are worse than the best quote. The SEC agreed. Rather than comply with this mandate, the NYSE terminated “ML/NYSE PRIME.”
By this time, however, the horse had left the barn. Two private firms soon began generating private execution quality reports for brokers and for trading venues. In addition to price improvement, these firms produced a myriad of other statistics, such as effective spread, realized spread, execution speed and depth enhancement. Among other things, these statistics were used by brokers to defend selling or internalizing their retail orders.
The SEC mandates the production of execution quality statistics to foster competition and improve execution quality. In January 2001, the SEC introduced Rule 605 (originally named Rule 11Ac1-5), which requires market centers trading equities “to prepare and make available to the public monthly reports in electronic form that categorize their order executions and include statistical measures of execution quality” and provides specific instructions on how to compute the measures (see SEC Release No. 34-43590). Each month, these market centers produce more than 240 execution quality statistics per stock!
While the SEC would have liked to extend this rule to options markets, as of January 2001, the market data needed to compute many of the execution quality statistics (a consolidated quote) was not available. Rule 606 (originally named Rule 11Ac1- 6), which applies to both equity and option markets, requires brokers handling customer orders to prepare quarterly reports detailing where they sent their orders to be executed. By increasing the visibility of order execution and routing practices, the SEC hoped that competition would improve execution quality for retail investors by making it easier for brokers (and others) to evaluate their order-routing decisions.
Have these statistics affected order-routing decisions? Boehmer, Jennings and Wei (2007) examine Rule 605 data between 2001 and 2004 and find that markets reporting better execution quality statistics in month t receive more orders in month t+1, even after controlling for other publicly available measures of execution quality. These authors conclude that the “SEC’s emphasis on disclosure to effect public policy can produce beneficial effects.”
Should SEC Rule 605 be extended to options markets? Despite the apparent benefits of Rule 605 data, Stigler (1964) notes that “information costs money, and no society is rich enough to get all the available information.” In the adopting release, the SEC suggested individual market centers executing equities could hire outside firms to compute the monthly execution quality statistics for about $30,000 per year (current estimates by industry participants suggest this figure is now more than $48,000 per year). While seemingly trivial, in aggregate the SEC estimated that in 2001 it would cost equity markets approximately $21.8 million per year to comply with Rule 605. These costs will most certainly be higher in options markets since storage and analysis of data is a much more difficult problem.
Even if these data could be costlessly generated, it is not clear that they are valuable to market participants. First, these data are produced with a one month lag. This means that the data are useless in a world where brokers give customers the option of choosing their smart-order routers. In its first quarter order-routing report for 2011, Interactive Brokers recommends its customers use its smart-order routing system, “which continually scans competing market centers and automatically seeks to route orders to the best market, taking into account factors such as quote size, quote price, liquidity-taker costs, liquidity-provider rebates and the availability of automatic order execution.” Since execution quality varies both across and within trading days, order-routing decisions based on real-time factors dominate those made conditional on statistics that are at least one month old. Moreover, the average cost of liquidity is not relevant for brokers who are tied into a specific liquidity providers on a given exchange.
Finally, as noted in the introduction, recent experience with voluntarily produced execution quality statistics in the options market suggests these statistics are not important. On July 17, 2008, the Equity Options Trading Committee of the Securities Industry and Financial Markets Association (SIFMA) recommended that each of the options exchanges begin voluntarily publishing monthly standardized execution quality reports according to its guidelines. For a set of options traded on the major options exchanges, Battalio, Shkilko and Van Ness (2011) compare effective spreads computed using the exchange statistics with those computed using publicly available data.
In Panel B of Table 3, they find that effective spreads estimated from publicly available data match those produced by the CBOE in March, April, and June 2010. However, in May of 2010, the CBOE’s estimated cost of liquidity is seven times higher than the Battalio et al. estimate. While this discrepancy may be due to outliers caused by the Flash Crash, results for the ISE are more systemic. In March through June 2010, the ISE’s effective spreads are approximately three times higher than the Battalio et al. spread estimates. Giving validation to the statistics computed by Battalio et al. is the fact that their effective spreads are within $0.005 of the effective spreads produced by S3 (a company that, among other things, produces execution quality statistics) for the NYSE AMEX and NYSE Arca in each of the four months.
The apparent ambivalence toward the voluntarily produced execution quality statistics suggest options exchanges do not view these numbers as relevant or important. The existence of smart routers coupled with the importance of individual liquidity providers on exchanges help explain why some exchanges may not care about getting their monthly execution quality statistics correct. While Rule 605 may have had the desired impact in equity markets in the early part of the last decade, technology has progressed to a point that Rule 605 appears to have outlived its usefulness.
1. Battalio, R., A. Shkilko, and R. Van Ness, 2011. To pay or be paid? The impact of taker fees and order flow inducements on trading costs in US options markets. Mendoza College of Business working paper.
2. Boehmer, E., R. Jennings, and L. Wei, 2007. Public disclosure and private decisions: Equity market execution quality and order routing. Review of Financial Studies 20, 315-358.
3. PR Newswire, 1995. Merrill Lynch announces pilot program to inform investors how much money they save on listed equity trades. November 2.
4. Stigler, G., 1964b. [Professor Stigler revisited]: Comment. Journal of Business 37, 414-422.