Dark Horses Down Under


As the issues surrounding dark liquidity grow more and more contentious, Steve Grob, Fidessa’s Director of Group Strategy, looks at the Australian trading landscape, how this type of trading has evolved and what the dark future holds.
Over the past few years, much attention in Australia has focused on the competition between ASX and Chi-X. But, in fact, multi-venue trading (via dark pools) was available to Australian investors long before Chi-X ever opened its doors back in November 2011. The term ‘dark pool’, however, has now become a catch-all phrase that describes a variety of activities that match trades but do not distribute pre-trade prices. Partly because of their name and partly because they are misunderstood, dark pools have attracted great controversy worldwide, and Australia has been no exception to this. So, what does the Australian dark landscape really look like, who is in it and what value does dark trading really bring to the market?
In 1997, ITG launched POSIT, a pre-market VWAP cross, Australia’s first alternative venue. However, it wasn’t until 2008 – anticipating the end of the ten-second rule1 – that Liquidnet followed with its buy-side crossing network, and finally Instinet with its BLX crossing network in 2011.
Likewise, broker dark pools were also around before Chi-X was born. UBS PIN (Price Improvement Network) launched in 2009 with other major brokers quickly following. These days most of the big brokers in Australia operate dark pools, although UBS PIN remains the biggest.
Finally, ASX’s own dark pool, Centre Point, was launched with much fanfare in 2010, and it is thriving (see diagram 1); it in fact has greater market share than Chi-X. To illustrate the complexity of these issues though, at a conference in Melbourne in May 2012 ASX chief Elmer Funke Kupper warned of the dangers dark pools presented to the Australian trading landscape, even though ASX earns 0.5bps on the AU$100 million-odd worth of trades executed on its venue every day.2

Genesis of Dark Pool Trading
Institutional buy-sides have always preferred to cross their order flow with other buy-sides that have the ‘natural’ other side in the stock they are looking to buy or sell. The reason for this is anonymity, or not divulging their trading intention to the market at large. The larger the order size the greater the value placed on this anonymity. Dark pools provide exactly this – but have they always been available or are they a new luxury brought about by the wonders of electronic trading?
In the days before widespread computer usage, if an institution wanted to move a big line of stock, and didn’t want to send it down to the exchange for fear of moving the market, they’d call their broker. Often the broker would call their internal counterparts to see whether they, or a client, was buying or selling a similar block and would execute an ‘upstairs’ trade at an internally negotiated price. Depending upon the skill of the broker, this provided a high degree of anonymity and certainly more than would be achieve by just dumping the order on a lit market for all to see.
It was a natural step then for brokers to employ computers to automate this upstairs activity so as to increase its effectiveness and reduce their reliance on human memory. The result was a dark pool that matched different client orders away from the public gaze of the lit markets.
When is a Dark Pool not a Dark Pool?
The terms ‘internaliser’ and ‘dark pool’ are often thrown around synonymously when talking about non-lit trading, particularly in the Australian media. Correctly speaking, however, an internaliser describes a firm conducting a specific activity where the broker is looking at incoming client orders and choosing to cross them against its own book. This can be contrasted with a broker crossing network, which is anonymous to both the clients and the broker – neither can see the orders inside the pool until they have been executed. A third category of non-lit activity includes dark books operated by venues that provide a similar service. And, finally, some venues permit dark order types (where part of the order is hidden) to intermingle with their lit liquidity.
Not only are there many types of non-lit or dark venues but, to make matters worse, each is defined and treated differently by various regulators around the globe. No wonder then that dark pools tend to be perceived suspiciously (see diagram 2).

The Dark Side of Australia
Dark pool operators would, of course, argue that dark pools make the world a great deal better for Australian institutions. They facilitate anonymous trading in a way that was never possible before they arrived, and allow price improvement (as trades may be executed at prices better than those available on the exchange). If lit markets are all about price discovery, then dark markets could be seen as being all about size discovery.
Enter the regulators. While many have accused dark pools of being unregulated, this is far from the truth. The firms that run dark pools are heavily regulated, as are the dark pool structures themselves; but just how to do that has been, and continues to be, a thorn in the regulators’ sides.
While ASIC certainly doesn’t want to see prices moving around as institutions throw huge blocks of stock on to the ASX, they also take their lead on dark pool regulation from IOSCO, which says that national regulators should try to preference lit venues over dark ones in their thinking.
This has led to an uneasy relationship between dark pool operators and ASIC in Australia. ASIC said in 2012 that it would not impose a proposed minimum value for dark transactions and indeed would allow trades below block size, as long as they were executed within the national best bid and offer. The media squealed that the big institutions had forced ASIC out of a rule that would protect retail investors, conveniently ignoring the fact that many big buy-side institutions are aggregators of retail money anyway.
The problem now troubling regulators and market participants is that the average trade size in dark pools is shrinking. These venues were created, so we are told, to allow institutions to trade big blocks of stock. Yet the average trade size in Australia is consistently falling. However, the average trade size on the lit market is falling too, not just in Australia, but around the world. Most likely this is a result of two things: smarter algorithmic trading, chopping large trades into smaller chunks to aid the hunt for liquidity; and high frequency trading, which is the other bane of the regulators.
ASIC still maintains its right to insist upon a minimum trade size and is making the operators of non-lit venues provide it with monthly statistics on their volumes. Unfortunately for ASIC there is little hard information available from abroad. Anecdotally, however, European and US buy-sides moan that the sheer number of dark pools makes it less and less likely they will find the other side to their block orders without lifting the lid on each and every one of them. This takes both time and money and so the dark pool phenomenon risks becoming a vicious circle. More dark venues mean that firms are increasingly tempted to ping them with small orders to see if there is any relevant liquidity there. This then decreases dark trade size further and sets the alarm bells ringing with regulators (see diagrams 3 and 4).

Nevertheless, the fact remains that market ecosystems have evolved and now there is no turning back. As each country-specific or regional regulator grapples with the issue, they run the risk of encouraging regulatory arbitrage, where firms will simply move out of markets where regulation is too prescriptive and set up shop in the places they can most easily do business.
Until (if ever) a harmonised approach arrives to address what is a global phenomenon, regulators and trading institutions will need to deal with dark liquidity in their own backyards. Technology exists to piece the liquidity landscape – lit and dark – back together, and while the rulebook is being written and re-written, those who focus on simply getting the best trading results possible will, as always, come out ahead.

1 Under ASX regulation, trades had to be displayed on the ASX for ten seconds before they could be crossed.
2 Based on AU$2.425bn worth of trades executed in May 2012.