The increase in dark trading
Dark trading has existed for a long time in many stock markets. As noted above, it can
take place in many different forms including undisclosed orders on regulated exchanges,
trading on alternative trading platforms, off-order book trading on exchanges and other
OTC centres. Trading by using orders that do not appear in the visible order book has
traditionally been associated with the needs of institutional investors that want to reduce
the market impact of large orders. This need has become increasingly relevant as
algorithmic trading and high frequency trading (HFT) have increased in importance.
With respect to the argument that dark pools meet the needs to place large orders,
Table 4.1 shows the average trade sizes in ATSs in the United States in 2015. As seen in the
table, some ATSs execute large trade sizes of up to 500 000 shares. However, the top five
ATSs in terms of average trade size account for less than 3% of the total share volume
executed in ATSs. The top five ATSs in terms of volume traded, which account for 49% of
total share volume traded, had an average trade size between 153 and 233 shares. The
average trade size in all ATSs was 207 shares, which is very close to the average trade size
in stock exchanges for the same period (209 shares). These findings suggest that ATSs, with
respect to the execution of large orders, do not distinguish themselves from the regulated
exchanges whose listed shares they are trading.
The increase in dark volume, particularly in the form of dark pools, has raised
concerns about the efficiency of the price discovery process, the fairness of markets and
the level playing field among investors. For example, the US SEC in a recent release (US SEC,
2015a) on proposed rules with respect to Regulation of NMS Stock ATSs, expressed
concerns that;
In terms of trading, ATSs operate in a similar manner to securities exchanges with whom
they compete for business. However, unlike securities exchanges,
there is limited public
information available to market participants about their operations. These differences
may create a competitive imbalance between two functionally similar11 trading venues
that may trade the same security but are subject to different regulatory requirements.
This difference in operational transparency is to the disadvantage of market participants,
since it limits their ability to adequately assess the relative merits of many trading
4. CHANGING BUSINESS MODELS OF STOCK EXCHANGES AND STOCK MARKET FRAGMENTATION
OECD BUSINESS AND FINANCE OUTLOOK 2016 © OECD 2016 131
venues and to adequately discern how their orders interact, match, and execute on ATSs
and to find the optimal venue or venues for their orders.
There is a lack of transparency around potential conflicts of interest that arise from the
activities of the broker-dealer operator of the ATS and its affiliates in connection with
the ATS.
Recent US SEC enforcement actions highlighting the difference in regulatory
frameworks between national securities exchanges and ATSs have attracted public
attention in the United States. In 2011, for example, the US SEC settled an enforcement
action against a dark pool operator who advertised that no proprietary trading took place
in its dark pool while, on the contrary, the overwhelming majority of the shares traded on
its ATS were bought or sold by an operator’s wholly owned subsidiary. In addition, the
operator provided its subsidiary with certain access and information that improved the
subsidiaries ability to trade advantageously and were not known to other customers
(US SEC, 2011). In a more recent case, the US SEC has settled another enforcement action
against a dark pool operator who had also not disclosed that it was engaged in proprietary
trading within its ATS and gave its proprietary trading desk access to live feeds of highly
confidential order and execution information that were used to inform the desk’s trading
decisions (US SEC, 2015b).
