In early February, the options industry switched its minimum price increment from $0.05 ("nickels") to $0.01 ("pennies") in 13 key options classes (representing 13 percent of total volume and 3 percent of total quotations in January 2007) in a graduated pilot program mandated by the Securities and Exchange Commission (SEC). The move to penny pricing is seen as a way of reducing trading costs for investors by reducing the potential for market makers to earn a spread between options prices, but it also will increase market data volumes.
The doomsday scenario that some had predicted for the rollout of pennies -- bottlenecks, lost data and systems meltdowns -- did not materialize. However, the change presented a significant challenge, according to participants, and there still are systems capacity issues that will have to be reconciled in the future.
In the January 2006 issue of Advanced Trading, exchange technology leaders, vendors and analysts expressed concern that the entrance of new exchanges -- such as NYSE Arca, the International Securities Exchange (ISE) and the Boston Options Exchange (BOX) -- into the options market would compound the effect of changing from nickel to penny quotations. The change means that the 200,000 securities in equity options, all of which are updated as soon as the underlying stock price changes, now have the potential to update at five times as many price points. According to Jeff Wells, chair of the market data capacity committee at the Financial Information Forum (FIF), a trade group for data vendors, with six exchanges quoting in all classes, as many as 500,000 securities could be quoted simultaneously at the beginning of the trading day.
Regulators, exchanges and the Options Price Reporting Authority (OPRA) have been watching the phenomenon of options quotation growth since 1999, when exchanges stopped exclusive listing of options and began listing each other's options. In terms of data traffic, growth has happened at an exponential rate. According to the FIF, in July 2000 the peak message rate was 4,000 messages per second (mps), sustained for one minute. By September 2005 that rate was 83,339 mps.
According to a January memo issued by OPRA, all market data recipients are required to accept 456,000 mps as of March 13, 2007, and 716,000 mps by Jan. 8, 2008. OPRA expects the exchanges to churn out 4.3 billion messages per day in March and 6.2 billion by January 2008. On average, these rates are more than 60 times those seen in the equities market.
The SEC concluded that the only prudent solution was to institute a graduated pilot program and review the results before continuing to allow further securities to be penny-quoted. The pilot rollout ran from Jan. 26 to Feb. 9, starting with Whole Foods Markets and including the popular Nasdaq-100 (QQQ) series. The pilot will continue for six months, and another three months will be required to process the data collected from the pilot. All in all, participants say, the pilot rollout went smoothly, both from a technical standpoint and an organizational standpoint.
"The SEC was very cooperative in putting together a pilot," notes Phil Slocum, EVP of trading operations at the Chicago Board Options Exchange (CBOE). "It has been very collaborative between the Securities Industry and Financial Markets Association [SIFMA], the exchanges and the SEC. The real story is about how well the industry worked together." The participants have been holding biweekly meetings throughout the pilot period.
Although the stock market plunge of Feb. 27 changed the picture slightly, the industry picked an unusually smooth time to float a new trading methodology. Volatility was at an all-time low during the pilot rollout, so the preliminary recorded volume may be illusory.
"Quoting, though up, is not extravagant," says Slocum. "In 10 of the 13 classes, we generated more quotes per day than average, but half of the classes showed increased peak mps rates and half were down." Whole Foods showed the greatest jump, from 163 mps pre-pilot to 700 mps post-pilot. But recent news coming from the company may have contributed to the activity.
Significant Preparations Exchanges made significant systems and operations changes in advance of the pilot. The CBOE, for example, changed its mandated quoting requirements for market makers so that continuous quotation was required only for issues with nine-month-or-lower expiration dates; and it instituted a quote-mitigation fee to cut down on unnecessary quotes, according to Slocum.
The Philadelphia Stock Exchange (PHLX) increased processor power and disk capacity, and redesigned applications over the course of the past year in preparation for the pilot, according to Bill Morgan, PHLX's CIO. "It was an extraordinarily large operation," he says, though not as all-encompassing as the decimal-stock conversion or Y2K, he adds.
At NYSE Arca, which uses a reconfigured version of the ArcaEx equities software, a new fee structure was implemented. The new structure now offers a liquidity rebate to market makers and broker-dealers that provide liquidity on classes within the pilot group. But outside the pilot group, the market maker pays the same fee whether it is posting or taking liquidity, according to John Werts, EVP of derivatives, NYSE Arca. "It is similar to the way the equities market has worked for some time with the ECNs," Werts says. "We saw the benefits in the equities model and have effectively eliminated payment for order flow."
The practice of payment for order flow was one of the main reasons the SEC wanted to switch to penny quoting in the first place -- reasoning that smaller price increments would make each trade that much less lucrative, regulators were seeking to limit the potential for market makers to reward institutions for order flow. Werts argues that some of the "doomsday people" who had predicted an "industry meltdown" due to increased quote traffic had "economic incentives to retain the payment-for-order-flow model" and believes that the pilot group will be expanded to include more options securities.
In general, the move appears to be a boon for retail investors who are trading in relatively small lots. The benefit to institutional investors, however, is cloudier since narrower spreads and more-frequent trades make it more difficult to cross large trades, as has been the case in equities since decimalization, notes PHLX's Morgan. He adds that less-automated institutional players may find it that much more difficult to execute large block orders under the new pricing structure.
Early indications are that the pilot is having its intended effect. According to a report by Tethys Technology, bid-ask spreads in Whole Foods (WFMI) narrowed by 53 percent and the average trade size decreased by 24 percent, with trade volume rising 39 percent during the penny-quoting period. The volume in the underlying stock was up 26 percent during the period.
"The changes do not augur very well for upstairs market-making desks and electronic market makers," the authors of the report write. "Lower spreads, coupled with lower priority in the execution queue, will compress profit margins." Judging the pilot a clear victory for small and medium-size investors, Tethys also notes that this may be a positive condition for "large investors and hedge funds engaged in volatility-related strategies."
Although the pilot has positive implications for the retail investor, it's unclear that there would be a benefit to quoting all options -- particularly infrequently traded, low-liquidity options -- in pennies, according to Christopher Nagy, managing director, order routing sales and strategy, TD Ameritrade. "I would not anticipate that you will see a full-fledged rollout of every symbol out there," he says. "It is too early to tell whether there would be an investor advantage to that." Nagy also is the chair of SIFMA's options committee.
Data-Capacity Issues In any event, there are other items on the exchange and investor wish lists that will almost certainly force the data-capacity issue, regardless of the extent of the penny quotation.
Investors are becoming increasingly sophisticated and would like to see a wider range of strike prices for options, Nagy says. Most options are offered only in $5 increments, with some in $2.50 and $2 increments, and very few in $1 increments, he notes. Investors would like to see those increments narrow, just like the spreads, Nagy contends.
"The retail clients are interested in trading options at or near the money," Nagy says. "If you have a choice between a $35 or $40 call on a $37 stock, you are not near the money. There is a moratorium out on that right now, and that has a direct tie-in to the bandwidth requirements."
In addition, exchanges increasingly are interested in disseminating more-sophisticated data, via direct feeds to data-intensive trading firms. But such an effort must be accompanied by an effort to reduce the amount of unwanted data disseminated to more-run-of-the-mill investors, says the FIF's Wells, who formerly was an executive at Reuters and now is SVP of quality assurance and software releases at Exegy, a vendor of high-speed data-caching technology. Currently, exchanges only show the best bid and offer on options; the true depth of the book is usually hidden.
"Each exchange must itself have some sort of order-book system with the next-best price on the bid and offer sides," Wells says. "It seems inevitable they will want to present that to outside customers, who will want to understand it and have greater transparency."
The implication of these desires is that the options industry is not out of the woods yet. CBOE alone has 160,000 strike prices now and is looking at increasing the book depth available to investors, the exchange's Slocum says.
To accommodate some of these pressures, OPRA plans to convert to the FAST (FIX Adapted for Streaming) protocol in April, which should allow compression of tick-by-tick data packets to about 30 percent of their prior size. The vendor also will begin to break out its data offerings into unmitigated and selective-update streams.
"Many people do not need tick-by-tick, which changes 10 times a second -- if you are an algorithmic trader, you do," Slocum explains. "But Bob in Iowa wants last-trade information only."