By Huw Gronow, Head of Trading, Europe and Asia, Principal Global Equities*
Across the pond, the owners and operators of “dark” order books are coming to terms with the prescribed caps on activity in trades where there is no pretrade transparency under the Reference Price Waiver (RPW) in the original MiFID directive of 2007. What has been the outcome of this directive?
The characteristic that has evolved from the resulting competition for trading volume is the “dichotomy of dark”, where dark volumes are now either genuinely large in scale (in our opinion the original purpose of dark pools) or as small on average as those on the lit market (see figure 1); interestingly the same trend has evolved in the EU under the RPW. Might this latter category of trades as well be transparent on “lit” venues given the little or no price improvement especially on large cap stocks where spreads are thinnest? The EU seems to think so: activity without pre-trade transparency unless large-inscale will be capped from 2017, to the tune of 4% of trade per venue and 8% overall in any stock.
So, are the “lit” venues jumping for joy in anticipation of a flood of volume, and thus revenue, gushing back to their rightful place? Not necessarily. One could frame the landscape for the buy-side as being, for a number of years, stuck between the “block” – the serendipitous supersized transfer of inventory between institutional investors – and the “hard place” – the highly fragmented array of displayed and non-displayed venues, both exchange and competing alternative – which includes, in Europe, the broker internalisation pools, namely Broker Crossing Networks (BCNs) and Systematic Internalisers (SIs) – and where technology-levered market makers in most cases seek to provide the liquidity that is claimed that institutions immediately require, and where institutions are required to choose to navigate using complex algorithms.
The reality is slightly more nuanced, of course: the considerations of when and why institutions look to access liquidity is more complex than merely “lit” or “dark” at any given moment in the trade lifecycle.
Urgency is the key
How urgent is that institutional trade? This is the most important question where this scale of liquidity demand is usually incompatible with its availability at the currently accessible best bid and offer (the “top of book”). This is simply because the parent order is invariably many multiple hundreds, thousands or even million times larger in scale. This incompatibility has magnified over time as executed trade sizes have atomized in the fragmented ecosystem, and as asset manager growth and concentration, has resulted in fewer “natural” block crossing opportunities. The projected time horizon for what we term this “risk transfer” process is dependent on the alpha viewed in the idea – the key being to differentiate between short term alpha (the province of the HFT operators) and long term alpha. If research indicates that in the short term, the proposed trade does not benefit from aggressive liquidity capture, then patience in execution, including parking the block trade in a “large in scale” environment away from predatory short-term participants is warranted, as long as the opportunity costs are monitored and optimized, in order to avoid the associated frictional costs of continuous trading.
The choice of Dark versus Lit
In reality the soubriquet “dark pools” refers to a variety of non-displayed trade matching mechanisms which are variously accessed at different stages of the trade cycle: the large in scale venues generally at the outset of the trade lifetime, where the risk transfer process is nominally at its most urgent. The “small in scale” venues – be they ATS, broker BCNs, public or broker MTFs, in varying degrees and stages, down to the chosen strategy of the buy side trader and the chosen type of interaction with the dark venue, due to its chief differentiating factor of uncertainty of execution (by their nature).
Another consideration for buy-side use of the dark is how to proceed with uni-directional large orders in an elevated volatility environment. Intuitively, we infer that high-frequency liquidity providers tend to turn more into liquidity takers during spikes in intra-day volatility. Many academic studies are being produced on the impact of HFT and the aftermath of market dislocations: in a study of the 2010 Flash Crash, Kirilenko et al (2014) concluded that high frequency traders “can amplify a directional price move and significantly add to volatility”, while Menkveld and Yueshen (2015) confirmed the U.S. government’s and Kirilenko’s view on the Flash Crash. Further yet, Hasbrouck (2015) concluded that high frequency quoting significantly increased intra-day volatility, over a ten year period under study. Other studies, notably Brogaard (2012), have taken the opposite view, but what is clear among the noise is that there can be a liquidity premium in transacting in both lit and dark order books, in terms of footprint signalling and information leakage.
Trust, but verify
The self-evident lack of transparency in some dark venues (in terms of types of participants, matching engine, IOIs and other considerations) is an issue between the asset manager and the broker as intermediary and operators of these types of venues where applicable. In order to mitigate this, due diligence documents are proliferating between buy-side and sell-side as a best practice, but is no substitute for independent third party transaction cost analysis as the route to verification of the execution experience.
In addition, the increased publication of Reg ATS forms for US venues is welcome additive information, moved forward in no small part by IEX, the US dark pool which catalysed the new trend to increased US dark pool disclosure. We at Principal Global Equities*, since the fragmentation of electronic markets, have set parameters and excluded certain pools and types of intermediary as a result of this ongoing process; with the work of the FIX Trading Community Investment Management Working Group on Execution Venue Analysis, we believe that in the end, asset managers across the spectrum will have the tools to become “judge and jury” for themselves. But this all adds to the complexity of the individual routing decisions where urgency and volatility are key variables – and where signaling risk and information leakage is a primary concern – even in the dark.
The Australian and Canadian models
An evolution of the notion of dark pool provision under MiFID 1 was the establishment of choice of interaction in broker crossing networks (BCN); in Australia, where there is no cross-border competition, broker dark pools dominate the non-displayed landscape with over 20 dark pools operated by 16 brokers, in addition to those run by the ASX and Chi-X Australia. In 2013, the regulator ASIC introduced two important enhancements to address the issue of transparency and effect on price formation: that trades executed in non-displayed environments had to show meaningful price improvement, or routed to a displayed order book.
Similarly, Canada introduced a comparable change in 2012, successfully reducing dark activity where there was no significant price improvement, according to research by the University of Sydney this year. In addition, ASIC provided for a minimum execution size should this measure not have the desired effect – a measure that many asset managers, including ourselves, employ as a matter of course. These contrast with the controversial “caps” on the smaller dark trades proposed by the EU – the desired effect (greater transparency) may give rise to unintended consequences (larger trades executed outside of the public exchanges and alternative venues).
The changing relationship
The inclination towards more discontinuous trading by asset managers is indicated by some of the largest asset managers throwing their weight (and presumably intent to direct their order flow) into new large in scale venues, such as the proposed Plato Partnership vehicle, and Turquoise Block Discovery Service; and surely others will evolve. One of the largest asset managers in Europe, Norges Bank Investment Management, has already indicated in a recent paper that they are shifting towards more patient, block-type executions for lower market impact costs. Recent scandals involving dark pools, particularly in the US, have served to have asset managers even more closely verify the activity and interaction with these venues and in many cases punish them via exclusion from their routing decisions. Perhaps, on the buy-side, all patience has been lost: the energy behind the formation of Luminex in the US in particular, a buy-side only environment where block trades are non-negotiable and entirely exclude HFT firms, is telling. This contrasts somewhat with the largest comparable mechanism operated by Liquidnet, where a negotiation usually precedes execution, leading to potential information leakage in the event of an unsuccessful trade. Nevertheless, this suggests that the buy-side appetite for genuine, committed transfers of sizeable inventory, while avoiding market impact, is not diminishing.
Asset managers have faced a daunting surge in market complexity and, by and large, through their broker intermediaries, have been able to successfully navigate this through the acute understanding of the proliferation of lit and dark venues allied to their measurement of and modification in their interactions with them.
However, in the face of regulation, scandal and increased research into the effects of high speed markets and price volatility, there appears to be an inclination to defragment the market in terms of a return to block trades; in effect, putting Humpty Dumpty back together again.
It is intriguing to contemplate the extreme inference of asset managers across the globe choosing to eschew exchanges and streaming liquidity venues in order to avoid the continuous liquidity premium. Liquidity providers can’t just provide liquidity to each other ad infinitum. Will HFT eat itself? Whether that is fanciful or not; and whether through regulation or desire, asset managers’ relationship with the dark may be moving more towards the “block” in favour of the “hard place” – an interesting step back in time.
We’d love to hear your feedback on this article. Please click here
By Huw Gronow, Head of Trading, Europe and Asia, Principal Global Equities*