|
|
The section about off-exchange trading in MiFID II focuses on recalibrating the use of pre-trade transparency waivers. Trading under these waivers (i.e., in the dark) — specifically, the reference to price waiver and negotiated trade waiver — is allowed with some qualifications. In particular, dark trading under these two waivers is subject to a double-volume cap calculation. So, pre-trade price transparency must be provided in the following instances:
When a given dark trading venue hosts more than 4% of the total volume of trading in a particular security across all trading venues over the past 12 months.
When overall, trading in the dark in a particular security under these waivers is more than 8% of the total volume of trading across all trading venues in that security over the past 12 months. This second cap would, for example, bring onto lit markets a stock that has 3% of its volume on each of four different dark venues for a total of 12% dark share (i.e., not breaching the 4% per-venue cap, but breaching the 8% total-venue cap).
This double-volume cap has previously received criticism from Markus Ferber, a member of the European parliament (MEP) and rapporteur on the MiFID II dossier. He has noted that a lack of sufficiently accurate data across European markets may mean that the double-volume cap will be unworkable or have no impact; it will be very difficult to demonstrate that a particular trading venue in a particular stock has breached either of the percentage caps during any period. He also argued that the double-volume cap may simply act to direct more trading onto systemic internalisers, which also provide limited transparency.
Currently, dark pool market share in Europe as a whole is around 9% according to the Financial Times, with many individual stocks having a double-digit proportion of their trading volume in dark pools. This volume level implies that these stocks will break the double-volume cap on Day 1 of MiFID II. Ignoring any possible unintended consequences on spreads or depth, the implication is that many stocks may soon be barred from trading on dark venues (for six months after each breach), which would cause this activity to move to lit markets, which is actually an objective of the policy.
Although broker-crossing networks are not allowed under MiFID II, there is an SI regime that allows for bilateral trading between a principal (e.g., a market maker or broker) and a client. An SI is a firm that deals using its own account in an organised, frequent, and systematic basis by executing client orders outside of a regulated market (i.e., an exchange) or multilateral trading facility. But the SI regime is problematic in an ‘automated world,’ as Larry Tabb at TabbFORUM puts it. It is not difficult to imagine an SI having an automated quoting system that has quote feeds from various sources (including other SIs) going into an execution engine that automatically selects the most desirable execution outcome for the client order.
The concern is that this will serve as a loophole for market participants to register as an SI and create SI networks. By connecting to one another to create de facto multilateral trading venues, they could essentially recreate broker-crossing networks, which is not meant to be possible under MiFID II.
The attraction for HFTs to operate under the SI regime is the ability to execute transactions with selected clients in private. The allowed reporting delay under the SI regime would allow the HFT to hedge the position on an exchange before the original trade is made public. Some commentators worry this could cause a large shift away from trading in lit markets.