Dark Pool Academic Papers – Less Filling or Tastes Great?


A paper was just published that examined the effects of fragmentation on the quality of our markets.  It is titled An Empirical Analysis of Market Segmentation on U.S. Equity Markets. It was co-written by three authors:

–          Frank Hathaway – NASDAQ OMX Group

–          Amy Kwan – Capital Markets Cooperative Research Center (CMCRC)

–          Hui Zheng – Capital Markets Cooperative Research Center (CMCRC)


You may recall from some prior Themis notes that the CMCRC is a cooperative institute that is funded by a variety of industry participants, including several stock exchanges (ASX, NYSE Euronext, Singapore Exchange, BATS CHI-X) as well as one of the global powerhouses in HFT – Optiver.


This paper concludes that in the U.S. Markets, dark pools have damaged price discovery and overall market quality in general, with the exception of large block transactions (our note- what a strange thing to exclude!). The crux of their conclusion is that market makers need to be able to trade against the “uninformed traders” to make enough money to offset their losses from trading against the “informed traders”. When the “uninformed” retail/index order flow is siphoned off and internalized, then the market makers lose their incentive to trade on the public markets. Said differently, HFT’s lose money when they are just trading against themselves. Here are the authors’ exact words:


“Our results show that dark venues successfully segment the market and attract uninformed order flow from the lit markets by offering sub-penny price improvement, leaving liquidity providers worse off on lit markets and consequently harming overall market quality.”


Hopefully it does not surprise you that a paper written by stock exchange employees and employees of a group funded by HFT firms and stock exchanges finds that trading in dark pools has harmed market quality.


An alternative viewpoint on the effects of dark pools on price discovery and market quality can be found in the work of MIT’s Haoxiang Zhu, whose 2012 paper – Do Dark Pools Harm Price Discovery? – demonstrated the opposite; adding a dark pool to exchange trading improves price discovery on the exchanges, as the order flow there becomes more informed.


Which do you believe (as an aside – wouldn’t a great industry or regulator panel include the authors of these two papers alongside an Exchange representative, and an ATS representative)?


We don’t think it is so cut and dry. We are pretty sure you do don’t, either. There is too much fragmentation, too many exchanges, and too many dark pools.  We are all very much like George and Billy – we advance arguments for our own various wants and agendas, but really all believe the same thing about the state of our modern markets.


And while exchanges have already started the consolidation process, as their profits diminish, eventually market forces will find their way to the dark pool industry as well. The one thing that regulators can do to speed that process along is to mandate improved dark pool disclosure (including their participants, order types, and matching practices). Once that happens informed end users will quickly decide what pools they want to swim in, as well as avoid.