How has the importance of data in trading changed as result of the new regulatory landscape?
The new regulatory landscape’s requirements for best execution, reporting and transparency, as well as the rapid growth in technological innovation driven by fintech, is pushing the consolidation of new trading solutions, which as a result is increasing the relevance of data.
While reference data, pre-trade data and post-trade data were already available on Regulated Markets, in particular in equities, the most significant impact is apparent in the OTC fixed income space. Here a broader array of execution options is now available, including decisions on principal versus agency, direct market access and venue choice, execution method and algorithmic trading choice and where clean and standardised data is being requested by trading decision makers.
While post-trade data is now available through authorised APAs and a variety of other data aggregators, even if it is neither clean nor cheap, it is pre-trade transparency that is most critical for the buy side.
The increasing adoption of smart execution models in this multi-asset and multi-venue market is raising the industry’s interest in data mining, TCA-embedded systems and best execution monitors, which together allow for a quicker and more consistent evaluation of a firm’s execution policy, improved trading venue performance calculations and suggesting improvements in execution models. Navigating this new ecosystem requires skills, data and analytics to ensure the efficiencies are truly realised.
What impact does that have on the services buy-side firms need and the technology they use to connect to the market?
A properly functioning trading desk needs to be efficiently structured to enable high-touch and low-touch trading across multi-asset classes in a fragmented environment. That requires expert domain knowledge, strong data and well-defined processes so that even in high volatility scenarios the right decisions are being made. In European markets, post-MiFID II, there is a huge amount of data available. Real time data, data quality and accuracy are critical factors in the decision making process. To be effective the buy side needs to be able gain access to this data in a simple manner, eliminate the noise created by too much information and integrate this data into their EMS.
To ensure that a suitably high level of execution quality is being met, better tools are needed to assess the effectiveness of a firm’s best execution policy. This is where TCA applications play a role in measuring differences in execution quality. While TCA cannot replace best execution, it helps asset managers evaluate their best execution policy and assists with trading decisions in a multi-asset environment.
How can dealers respond to support those requirements?
As the nature of multilateral, fixed income, e-trading venue operators has changed, and as the number of venues has proliferated, additional services have provided venue operators and brokers with additional revenue sources, competitive differentiation or both. Looking beyond pure execution in fixed income, data offerings have become an increasingly important source of revenue.
Increasingly frequently, exchange operators are offering access to fixed income trading venues to both existing and new client bases, and as they do so the share of fixed income venues offering raw and aggregated data services has likewise increased, with the competition among market participants moving from execution to additional services. The implementation of MiFID II has also played a pivotal role in the provision of these additional services, with the aim of improving the client experience and sales proposition. For example:
Are there any risks for market participants as their way of trading changes?
Markets will become faster and more efficient, and therefore the information gap and asymmetry between the sell side and the buy side, between financial markets and dealers, and between global players and regional should diminish and create a different but more level playing field in trading.
The complexity of new execution protocols and multi-asset trading means both buy-side and sell-side dealing desks need to reassess the technology they have available in order to re-structure their workflows. Drivers such as structural market changes and new regulatory obligations, for example MiFID II, along with cost pressures on both the buy side and the sell side are forcing people to do things differently and encourage a more exploratory approach to data.
More data and better technological tools should allow human beings to make quicker and better decisions, but humans still need to be the control point. As the size and complexity of datasets increases, the demand for data scientists with skillsets that include not only statistical training and programming, but also business aptitude, is increasing. However, with the increasing availability of analytics support packages, data can be mined and modelled by anyone in a sort of “Democratisation of Data”, irrespective of their lack of programming skills. In this new environment, careful consideration is needed as to where these solutions are applied, who should have access to them and how financial firms’ data governance policies should be adapted and be appropriate.
What do market participants need to do in order take advantage of the new trading environment?
The buy side needs to preserve alpha, reduce costs and improve fund performance, while the sell side needs to shift its approach from manual to quantitative trading. To reach this goal, quantitative models are now designed to use all available patterns, trends, outcomes and analogies provided by fintech solutions. Algorithmic trading involves rapidly and precisely executing orders following a set of predetermined rules. This effectively may remove human error and the dangers of emotional decision-making, even while humans still need to play a central role in the setting up and monitoring of algorithmic trading. Thanks to big data analytics, opinion mining is combined with predictive modelling to complement financial analysis when making trading decisions.
However, it is important to keep in mind that big data analytics cannot predict market scenarios and behaviours perfectly every time. This is because of some of the limitations affecting the banking industry today. For example:
©TheDESK 2019