[ A L G O R I T H M I C T R A D I N G S U R V E Y ]
expected to continue into 2018 following the implementation of MiFID II was the consolidation of brokers and algo providers by the buy-side, and in turn the reduction of technology complexity and cost. Instead it seems the opposite is true and Figure 3 shows as much; the adoption of multiple algo providers has increased almost across the board, with only those on the smaller end – those managing between $ 0.25 and $ 0.5 billion – reducing the number of algos they use, albeit by a small margin.
From the very small( with up to $ 0.25 billion in AuM) to the larger players( more than $ 50 billion in AuM), long-only firms have been taking on more algo providers over the past year, even though most are still hovering around the two to three provider mark, possibly with an eye on the new requirements for best execution that have come into play. By increasing the level of access to algorithmic trading route options, firms will be better positioned to provide evidence of best execution practises, which would be limited by sticking to just one or two algo providers.
Even so, around one-third of long-only respondents to this year’ s survey are sticking with either one or two algo providers, according to Figure 4. This is consistent with the results from last year’ s survey, although there was a 5 % year-on-year decrease in firms using three to four algo providers, while the proportion of firms with more than five providers increased by the same amount. It would be surprising to see all but the largest buy-side firms continuing to use more than five algo providers as the effects of MiFID II bed in properly. Brokers will continue to assess where they are generating revenues from their buy-side clients, but in the new trading environment, the balance of power seems to reside with the asset managers.
Increasing popularity As algos become easier to use and advancements in technology march onwards, buy-side trading desks are becoming increasingly comfortable in letting algos do the work while human traders are able to focus on the finer details, although this will be an area of keen interest to regulators. As such, it’ s little surprise to see the proportion of trades conducted by algo increasing in the 2018 survey. The number of firms trading less than 30 % of their value through algos dropped significantly year-on-year, with just under onethird of long-only firms fitting into this category, down from almost half of respondents last year.
The most significant change occurred with firms trading between 30-50 % of their value with algos, which jumped from 12.57 % in 2017 to 26.21 % in this year’ s survey, suggesting that firms that had previously shunned algo trading are coming around to the benefits that automated trading can offer. There were marginal increases for number of algos used for 50-70 % of value traded, but a more noticeable spike in the 70-80 % range, where proponents of algo trading are investing further in electronic trading methods.
In terms of the types of algos that long-only firms are choosing to use, Figure 6 shows a distinct shift away from dark liquidity seeking, which has fallen dramatically since 2016, when 81.9 % of respondents were using these types of algos to just 54.27 % this year. Whether this is in direct correlation to the introduction of the DVCs under MiFID II remains to be seen but gains for the TWAP and implementation shortfall( basket)-type algos this year indicate that long-only firms are now choosing to move away from automated dark trading. One of the most historically popular types of algorithm, participation-based algos, saw consistent usage and the same is true for VWAP-type algos.
It seems then that even in the early days of MiFID II, certain trends are already emerging. Greater regulatory oversight has forced the buy-side’ s hand into learning far more about the algos they use
80 // TheTrade // Spring 2018