Alpha Group has announced:
that, subject to regulatory approval, Alpha ATS will implement a new trading facility – the Alpha IntraSpread™ facility, during the 4th Quarter of 2010. The Alpha IntraSpread™ facility, a set of new orders offered by Alpha ATS, allows dealers to seek matches within their firm without pre-trade transparency and with guaranteed price improvement upon the National Best Bid and Offer (NBBO) at the moment of the trade. The Alpha IntraSpread™ facility will be available to all dealers that are Subscribers of Alpha and for all symbols traded on Alpha ATS.
Some of the specifics are:
- Dark orders have no pre-trade transparency as information on Dark orders is not disseminated on any public data feeds.
- Price of a Dark order is calculated as an offset of the NBBO by adding the price offset to the national best bid for a buy order and subtracting it from the national best offer for a sell order.
- Price of the Dark order can optionally be capped.
- Price offset is calculated as a percentage of the NBBO spread with value expressed as 10%, 20%, 30% … 90%, but capped to one standard price increment. If either side of the NBBO is not set, or the NBBO is locked or crossed, Dark orders will not trade.
- Dark orders trade only with incoming SDL™ orders that are tradable at the calculated price of the Dark order. Dark orders do not trade with each other.
… while SDL orders:
- SDL™ orders are “immediate-or-cancel” – they trade with eligible Dark orders to the extent possible, and any residual is cancelled. Price can be market or limit.
- SDL™ orders only trade with Dark orders and do not interact with other transparent orders in the Alpha CLOB. In the first phase of the Alpha IntraSpread™ facility, SDL™ orders only match with Dark orders from the same Subscriber
I do not profess to be an expert on ATS marketting practices, but this appears to be an attempt by Alpha Group to forestall the creation of internal dark pools by its members (or pool the cost of such systems; Alpha is owned by the major dealers) by offering a sub-pennying mechanism in a manner that is smoothly integrated with extant trading systems.
Sub-pennying is a controversial element of the current market microstructure debate. There is an excellent comment letter from Bright Trading Systems that explains how it works (in the States!):
Statistics from the Commission’s Concept Release on Equity Market Structure, state that 17.5% of all trades are internalized by broker-dealers. A more alarming statistic from page 21 of the release states that, “a review of the order routing disclosures required by Rule 606 of Regulation NMS of eight broker-dealers with significant retail customer accounts reveals that nearly 100% of their customer market orders are routed to OTC market makers.” This means that almost every single market order placed in these retail brokerage accounts, is checked by the brokerdealer’s OTC market maker to decide if they can make money by trading against their customer. They can legally trade against their customers as long as they match or beat the National Best Bid and Offer (“NBBO”).
…
The only time the displayed order on the NBBO is filled from an incoming retail market order, is when the OTC market maker of the broker-dealer passes on the chance to trade against its customer’s order, and there are no undisplayed orders hiding in dark pools in front of the NBBO order. As a result the only retail orders getting through to the publicly displayed NBBO, are the orders that the first two market participants have passed on. If the first two participants have passed on the opportunity to trade against the order, there is a good chance that the incoming market order is on the right side of the market (in the short-term). Hence, the only NBBO orders that are filled are those that are more likely wrong (in the short-term). The displayed liquidity provider is “sub-pennied” when they’re right, filled when they’re wrong. As liquidity providers become discouraged, they will place fewer passive limit orders in the short term and ultimately leave the trading markets. This will lead to less depth in the market and larger spreads, both increasing the cost to investors in the long term.
In their recent update of the status of the market microstructure review the Canadian Securities Administrators stated:
Forum participants discussed the idea of price improvement in dark pools, as well as the concept of sub-penny pricing. Questions were raised whether dark pools should always be required to offer price improvement, how much price improvement is meaningful, and whether sub-penny price improvement is desired or even relevant. It was noted that sub-penny price improvement may only be meaningful for dark pools achieving block sized execution, but is of questionable benefit to the overall market or to the investors for small orders. Participants also discussed the fairness of allowing dark pools to offer sub-penny price improvement while transparent markets are not allowed to offer the same execution opportunities. Some participants felt that sub-penny quoting on visible exchanges would not be desirable, one reason being the impact of increased messaging due to sub-penny pricing and marketplaces’ technology infrastructure costs. We will examine the issue of sub-penny pricing with the goal of assessing how any changes in either printing or quoting in subpennies would impact both the market as a whole, and the individual participants. Additionally, we will consider both transparent and dark markets, and whether principles of fairness would allow both types of venues to offer sub-penny price improvement and printing or execution, or whether different market structure models necessitate different treatment.
Cost concerns are a red herring. If it costs more, charge more. Idiots.
Sub-pennying is annoying, but not a major issue. As previously noted, a value investor has a cost-of-capital advantage on the order of 12% over a high-frequency-trader … if you can’t kick-ass with an advantage like that, you deserve to be hungry, naked and homeless.