Rapid quote movement and the dislocation of stock prices remain a source of frustration for many institutions. Negotiated trades are longer in duration because the human element is central to the trading process, but in a market where the NBBO changes in the blink of an eye, a longer time horizon increases the trades’ exposure to a greater amount of quoting activity and the likelihood of a price move.
We designed our pricing method for negotiated trades to solve for this problem and we receive many questions from clients around quote behavior and execution performance. We examined some of the public data regarding recent quote and trade activity for the broader market and analyzed our trades for the first quarter of 2017. Unsurprisingly, market activity is high and we believe our results reinforce the benefits of our pricing model and the need for an effective method of trading blocks in today’s fast-paced marketplace.
Quote Activity and SEC MIDAS
Intermarket Sweep Orders (ISOs)
Luminex Execution Analysis
Every second, the market processes thousands of quotes at a pace so fast most traders are blind to the vast majority of them. Rapid and frequent quote changes are commonplace in an electronic market driven by speed and institutional traders struggle with having to keep pace with a market that most would agree is better suited for trading small orders.
Often, traders begin a negotiation only to find that during the process the NBBO has moved unexpectedly. Ultimately, traders must choose: decline the trade and risk missing valuable liquidity, or, opt-in and absorb the price move. More often than not, traders would opt for the liquidity, but because Reg NMS dictates the new NBBO is what ultimately matters, the trade – if it were to occur – must transact at a price that reflects the updated NBBO, leaving one of the parties disadvantaged. Yet, back when the majority of negotiations were handled by human traders – “upstairs” – in the instances when the quote moved, the execution price usually deemed most fair referenced the NBBO when the negotiation began.
Historically, executing “at the midpoint” was an indication of better execution quality because it reinforced the ability to remain anonymous, avoid market impact, and capture a portion of the spread. However, when it comes to negotiated trades simply executing “at the midpoint” isn’t as easy it used to be. In a market where venues and participants function at varying speeds and numerous physical locations, one definition of an NBBO, at a single point in time, is nearly impossible to determine.
In an electronic environment where high rates of quote activity are the norm, one could reasonably argue that negotiated transactions are more vulnerable to price moves simply because they take longer to complete. Therefore, is today’s fast-paced marketplace optimal for determining the execution price of a negotiated trade?
U.S. regulation is agnostic with respect to the type or method of trading one chooses to employ, so even though the length of time it takes to complete a negotiated trade – one that involves a human decision – is still comparatively an eternity, all trades must honor the NBBO. Because the U.S. has a “one-size-fits-all” structure, traders who wish to transact in negotiated form, must do so in a market primarily consisting of 100-share orders that update the NBBO in microsecond and nanosecond frequencies. Rule 611 is a frequent topic of discussion among market participants. Institutions find that the Order Protection Rule impairs their ability to trade blocks and the concept of a block exemption was deliberated as a potential solution at the April 5, 2017 SEC Equity Market Structure Advisory Committee meeting.
The truth of the matter is that several factors can cause erratic quote behavior, some of which have nothing to do with the natural force between supply and demand. Some reasons are well-known side effects borne out of the evolutions in technology, like the general automation of trading, technology glitches, and predatory trading strategies. However, there are other reasons, too, like the use of Intermarket Sweep Orders (ISOs) that can significantly impact the quote and the NBBO, but are far less noticeable simply because they are part of the normal course of trading.
* * * * * * * *
Quoting Activity – The New Normal
By now, the majority of participants are well aware of the pace of trading, but the actual number of messages transmitted often takes a back seat. Terms related to speed and time – microseconds and nanoseconds – have become part of the industry nomenclature, but they only tell us so much. Speed informs us of “how fast,” but doesn’t necessarily indicate “how many.” Market speeds are difficult to comprehend, but if contextualized by the actual number of quote changes – often in the millions – our view and perception of the market becomes clearer. For instance, small-cap stocks experience less quoting activity compared to large- or mid-cap stocks. Relatively speaking, this makes sense. However, the number of quotes, orders, and cancellations transmitted in the small cap universe is still extremely high, high enough that many times we are incapable of seeing it all with the naked eye.
One source for easy monitoring and viewing of quoting activity in the marketplace is the SEC MIDAS system. MIDAS collects data from the consolidated tapes and proprietary data feeds, and aggregates information regarding posted orders and quotes, modifications and cancellations, and executions on registered exchanges. The data is free of charge and made available on a quarterly basis.
According to MIDAS, in the 1Q 2017, nearly 79% of quotes in large cap stocks were cancelled in less than 10 seconds. For mid-caps, the rate was 62% and for small-caps 48.3% (See below). This gives us pause because, naturally, we can’t help but think about how this may affect the negotiation process. Negotiated trades, by definition, are longer in duration because the human element is integral to the trading process. A longer time horizon in a rapidly changing market – for any trade – increases the trade’s exposure to greater amounts of quoting activity and the likelihood of a price move. For example, if a negotiated trade takes 10 seconds to complete, and the vast majority of quotes cancel within that same period, there is a good chance that the initial NBBO will not only become irrelevant, but also obsolete.
Intermarket Sweep Orders (ISOs)
An ISO is a Reg NMS order type designation that provides an exemption to Rule 611, the Order Protection Rule. (Rule 611, under Regulation NMS, promotes intermarket price protection of orders and requires that brokers and venues ensure they do not “trade through” the U.S. National Best Bid or Offer.) Generally, ISO orders are aggressive in nature and are intended to access significant amounts of liquidity via “sweeping” through multiple price points deeper in the order book. According to Reg NMS, “The intermarket sweep exception enables trading centers that receive sweep orders to execute those orders immediately, without waiting for better-priced quotations in other markets to be updated.”
One of the reasons the ISO designation came about was because market participants were concerned that the requirements of the Order Protection Rule would impair their ability to access posted liquidity deep in the order book. Under normal circumstances, if an institution were to route a marketable order to a trading center that is not displaying liquidity at the NBBO, the trading center would be obligated to check or route orders to those market centers with displayed, better-priced liquidity. However, doing so would take time and participants became concerned about the increased risk of missing liquidity, so, the SEC created the ISO designation to address the problem and alleviated trading centers of this obligation. (The obligation is transferred from the trading venue to the routing broker.) Therefore, in the example above, if the routed order includes the “ISO” designation, the trading center may execute the order at multiple price points in their order book and are not required to check away markets or route orders to access better-priced liquidity displayed on another market center.
Why is this important?
Needless to say, “sweeping the book” generates a flurry of activity around any remaining quoted liquidity and the NBBO, which would also affect executions in ATS venues. ATS venues reference the NBBO for pricing orders and trades, so if a significant move in the NBBO were to occur, marketable orders could match at a dislocated price. In fact, a 2012 research paper on Mini Flash Crashes stated, “Mini Flash Crashes are the result of regulation framework and market fragmentation, in particular due to the aggressive use of Intermarket Sweep Orders and Regulation NMS protecting only Top of Book.” The analysis found that 67.8% of the mini flash crashes in the data set were the result of ISO-initiated orders.
Most surprising was the high percentage of ISO volume present in the S&P 500 constituents. According to the Fidessa Fragulator, in 1Q 2017, ISO volume accounted for 27% of S&P 500 constituent volume. However, ISO orders only trade on-exchange, so upon examining a more detailed breakdown, the data shows that ISOs accounted for 41% of S&P 500 constituent volume executed on-exchange, including auction volume (see below). Taking it one step further, approximately 13.2% of all on-exchange volume in the broader market was from ISO volume in the S&P 500 constituents alone.
How frequently are ISO orders used?
For a variety of reasons, the use of ISO orders has grown significantly, which could imply a significant amount of “aggressive” orders are trading in the marketplace and potentially exacerbating changes in the NBBO. According to monthly order type statistics published by BATS, in April 2017, ISOs accounted for 28.24% of executed orders for BATS and Direct Edge combined. According to statistics published by NYSE, ISOs were responsible for 21.2% of their matched volume in April 2017. However, NYSE’s matched volume figure included auction volume, which was 17.91% of matched volume. ISOs only trade intra-day, so if you were to exclude auction volume, ISO’s contribution to intra-day volume is much higher.
Many well-intentioned market participants use the ISO designation, like liquidity providing market makers, institutions, and broker dealers, but some say that predatory firms are more prone to use them as part of or to enhance a particular trading strategy. In 2012, Haim Bodek wrote an article about some participants’ ability to gain an advantage using ISO orders. He stated, “In fast markets, [predatory] HFTs benefit from the slow SIP feed in a manner that may exasperate (sic) rapid price movement as they maneuver to avoid adverse flows — they are able to pull their unexecuted orders on venues before they trade against customers using marketable non-ISO orders that are rerouted or rejected due to phantom SIP prices.”
The SEC is keenly aware of the kinds of predatory strategies and takes action against those engaging in predatory trading behavior. In 2015, the SEC charged Latour Trading, a proprietary high frequency trading firm, with violating SEC rules designed to ensure safe and efficient markets. While the order cited that the activity was due to a software coding change unknown to the firm, the violation was regarding millions of non-compliant ISO orders submitted to the market. According to the SEC order, “From October 2010 through August 2014, Latour Trading LLC (“Latour”) sent approximately 12.6 million Intermarket Sweep Orders (“ISOs”) that did not comply with the requirements of Reg NMS. These orders totaled over 4.6 billion shares, had a notional value of approximately $116 billion, and caused over 1.1 million trade-throughs and 1.7 million locked or crossed markets.”
In March of this year, the SEC filed a complaint against trading firm Avalon alleging they engaged in two manipulative trading schemes. One scheme, known as “layering” involves the submission of non-bona fide orders “with the intent of injecting false information into the marketplace about supply or demand for the stock.” According to the complaint, between December 2010 and September 2016 the firm engaged in “hundreds of thousands of instances of layering, involving hundreds of securities trades on numerous exchanges and other trading venues.” Technically, in this case, the violation was for submitting non-bona fide orders, not ISO orders. Having said that, there are instances where bona fide orders layer the book legally and, theoretically, could trigger the use of an ISO order by another participant.
Luminex Execution Analysis
Helping to mitigate the effects excessive quoting has on block trades is core to the Luminex philosophy and trading model, so we examined our executions in the first quarter of 2017 and compared them to changes in the NBBO. We wanted to study if the quote changed during the negotiation process, and if so, how Luminex executions performed. For context, we also measured the stocks’ “natural” behavior by examining the quote activity at the same time of day for the 1Q 2017 period. This allowed us to see if our execution results on trade-day differed from the stocks’ normal quoting and NBBO behaviors on the days we didn’t trade.
Luminex Trades vs. the NBBO Midpoint
According to our results, on average, the Luminex execution price was within 0.3 basis points of the NBBO midpoint at the start of the negotiation. Between the start and end of negotiation, on an absolute basis, the NBBO midpoint moved on average 2.2 basis points. This means that even though the NBBO changed over the course of the negotiation the execution price was closer to midpoint of the NBBO when the negotiation began, a result that shows our ability to protect trades from price moves that occur later in the window. (See below)
Contrarily, one could argue that when the NBBO moves, the participant who benefits has less of an incentive to transact at the original NBBO. This is true. However, under normal circumstances the Luminex negotiation would not execute outright at a price, let’s say, 10 seconds into the window. The midpoint ten seconds into the negotiation is calculated into the final execution price, but it carries relatively less weight versus the midpoint at the start of the trade. In other words, when choosing to transact at Luminex, the initial midpoint is most relevant when considering the potential outcome. Second, as we discussed earlier, quotes change for a variety of reasons unrelated to trader behavior and may not be an accurate representation of available liquidity, not to mention the fact that the contra has less of an incentive to trade there.
How did Luminex compare to the stocks “natural” movement?
For context, we measured the changes in the NBBO for all stocks in our data set at the same time of day for the 1Q 2017 period. Over the same length of time as the negotiation window, we found stocks “naturally” moved, on average, 1.7 basis points, a figure slightly less than the trade related data and an outcome we expected considering participants’ trading activity. The natural movement tells us that on any given day, one should expect some amount of quote movement, which is important when it comes to negotiating trades since quote movement is one of the biggest reasons behind fall-downs. The fact that we can expect some form of quote movement further reinforces the benefits of our execution pricing and effectiveness in protecting traders during the negotiation window so that no one participant is disadvantaged. (See below)
We realize that traders tackle many challenges on a daily basis and quote movement is just one of them. Trading in any capacity is proving more difficult as time goes on and while some of the rules central to U.S. markets are under review, changes in regulation take time. In the meantime, competition for order flow is strong and brokers – especially in the electronic and ATS space – must come up with intelligent solutions that can help modernize and re-define best execution. We believe Luminex is one such intelligent solution, and we look forward to continue helping you solve your trading challenge.
As always, if you have any questions, just let us know.