we note their good argument against a trade-at rule, which begins on page 12 of their letter as such:
The Concept Release asks whether, if commenters believe that the quality of public price discovery has been harmed by non-displayed liquidity, the Commission should consider a “trade-at” rule. Such a rule would prohibit any trading center from executing a trade at the NBBO unless the trading center was displaying that price at the time it received the incoming contra-side order. The trade-at rule would require a trading center not displaying at the NBBO at the time it received an incoming marketable order either to execute the order with significant price improvement (e.g., the minimum allowable quoting increment), or route intermarket sweep orders (“ISOs”) to the full displayed size of NBBO quotations and then execute the balance of the order at the NBBO price.
SIFMA strongly opposes the concept of a trade-at rule. Initially, and in response to the Commission’s threshold question, such a rule is not warranted given the health of our markets (described above) and, importantly, the absence of compelling evidence that non-displaying trading venues are impairing public price discovery. A trade-at rule would likely lead to a deluge of additional message traffic and increased incidence of flickering quotes. The added costs to trading centers and broker-dealers would likely be significant and it is not clear that the anticipated benefits of additional quotes at the inside would outweigh them.
We also believe that a trade-at rule would have significant adverse consequences for investors, and retail investors in particular. Competition with respect to other best execution factors – such as market depth, reliability, and liquidity guarantees – would fall largely by the wayside under a trade-at rule that effectively dictates the manner in which broker-dealers must trade. For example, broker-dealers executing orders internally currently may provide a customer with faster executions along with opportunities for price improvement. By contrast, a trade-at rule might instead require that same order to be routed out, both slowing the execution of the customer’s order and, potentially, causing the customer to miss the market and lose the opportunity for price improvement. In addition, a broker-dealer routing an order to an away trading center may well incur additional costs in the form of fees for accessing the liquidity of the away market. These fees, ultimately, may be passed on to customers. Price competition among trading centers would be significantly hindered by a trade-at rule. A trade-at rule would require certain quotes to be hit in various trading centers, which in turn would reduce the incentive for trading centers to provide lower cost executions by, for example, lowering access fees.
More fundamentally, a trade-at rule would stifle innovation, making it less feasible for new business models that have been introduced into the markets during the last decade to exist, to the detriment of all investors. For example, the rule would significantly impact the ability of investors, including long-term investors, to use non-displaying trading venues to handle sensitive order flow. The requirement that such a trading venue offer price improvement at least in the amount of the minimum increment to execute orders when the operator of the venue is not quoting at the NBBO would be difficult to meet given that many stocks trade in penny increments. Alternatively, the routing of ISOs to the full displayed size of NBBO quotations would subject such venues to access fees in away markets and significantly reduce the ability of non-displaying venues to offset customer orders.
Routing under a trade-at rule also might increase the chance of information leakage, signaling to other market participants the possibility of additional order flow at the non-displaying trading venue, thereby disrupting attempts of institutional investors to reduce implicit costs associated with large orders. While order routing is required in some circumstances under the OPR, the risk of information leakage is ameliorated somewhat by the promotion of the regulatory policy of not allowing a better priced limit order to be bypassed, and thus the fact that the routed order receives a better price as a result of the routing. In addition, investors who prefer not to have their orders displayed or routed could miss execution opportunities should potential contra-side liquidity have to be routed away to comply with a trade-at rule.
In sum, a trade-at rule would have detrimental effects on the speed and cost of executions, the liquidity currently available in the market, and the ability of investors to control their trading interests. It would undercut best execution by dictating a particular manner of trading, which we think is unnecessary given the recent performance of the equity markets. In doing so, the rule would extend well beyond even the OPR in its clear preference of investors who display orders over investors who decide it is in their best interest not to display some or any of their orders – even if they may be willing to execute at the same price as the displayed markets. In this respect, a trade-at rule comes very close to a consolidated limit order book or “CLOB.” Both would negate the competitive benefits of dispersed order flow and unnecessarily impede investor choice. We note that the SEC has considered a trade-at rule or CLOB in the past and determined that such restrictive trading measures were unnecessary.
They make a nice argument. Last week we highlighted the article in Traders Magazine, which highlighted several industry opinions on the Trade-At rule. Most of those opinions were against such a proposal, and we disagreed with some of those arguments, and stated so in out blog entry:
We include SIFMA’s argument here as it makes us think, and in turn we wish to continue to encourage the debate about such structural issues.