Future Proofing Trade Execution Systems
A GlobalTrading Roundtable Discussion.
Regulation, technology advances and behavioural changes mean that trade automation will increase, and systems and their providers need to be flexible, agree speakers at a GlobalTrading roundtable. But, the degree of standardisation and the merits of outsourcing are controversial.
The technology and business functions within buy- and sell-side firms can no longer be siloed; they are interdependent. A trading desk can only be successful if it incorporates and prioritises technology that, if applied properly, will ensure a firm’s survival, agreed speakers at a GlobalTrading thought-leadership roundtable sponsored by Itiviti and hosted by the LSE on 22 November 2017.
Regulatory headwinds are an important catalyst for technology improvements that are creating new data sets and the opportunities to use them. The proliferation and fragmentation of liquidity and how to use it is a key issue for dealing desks.
For instance, the restrictions on dark pools imposed by the Markets in Financial Instruments Directive (MiFID) II will propel the automation of block trading to find alternative sources of liquidity and firms will need a cost-effective way to do this – which means making technology core to their businesses. Without the adoption of technology in a central role, firms will struggle to compete.
Furthermore, there is a danger that firms might re-invent the wheel for different asset classes, so it makes sense to move towards multi-asset platforms and integrated operations. Avoiding duplication reduces costs.
But, too often in the past, technology has been applied like sticky plaster, patching up gaps in systems and salving deficiencies. Instead, future proofing technology needs a clear sighted integration of the business strategy with the technology architecture. It might be expensive for smaller firms to implement, and consolidation within the industry might be an inevitable consequence. Clearly, those firms that have already invested in technology as a key component of their business strategies are at an advantage.
Most topically, if a firm implements MiFID II correctly, then it should be well-positioned to adjust to new technologies and future regulations. This means applying the requirements of the legislation diligently across asset classes to ensure accurate transaction cost analysis and best execution. As data is fundamental to achieving best execution, greater automation is inevitable in order to tap into new sources of information and optimise processing. The scalability of data collection is essential.
Some asset managers have already unequivocally embraced an automated trading process. This includes integrating platforms and aligning data processing facilities in a systematic way with the objective of enhancing access to liquidity and maximising operational efficiency. The result, within just a few years, will be almost no human involvement in the process which instead will be managed by artificial intelligence and machine learning, according to a speaker.
However, even if this were an established trend, inertia would likely prevent its unanimous adoption within the industry, noted another speaker.
Of course, there is a danger that the perspectives of business strategists and technical experts gets lost in translation. Their disciplines have different languages, so perfect integration and mutual-understanding is perhaps an ideal benchmark rather than an attainable target.
Standardisation and customisation
Some speakers argued that firms should build internally the systems that give them a competitive advantage, such as data analytics, trade algorithms and content distribution, but buy – or more accurately rent – systems for their more basic activities. Another way of expressing the model is “to customise round the edges but install a core provided by vendor”. Basically, the firm has a standardised workflow system surrounded by smart algorithms.
However, the distinction between what should be proprietary and what is generic is not always clear. For example, some firms, especially if they are large and complex, might build their order management system in-house, others might conclude that it can be standardised. Vendors need to make various predictions about the direction of automation, and there can be logistical and cost problems trying to align those bets.
A significant insight that arose from the discussion about the buy/build or hybrid models was recognition that the “buy” option is now more often a “rent” option. Firms increasingly hire systems rather than own them, which allows them to be more flexible to advances in technology and the imposition of future, unknown regulation. This characteristic is likely to predominate among the younger generation entering the financial industry, who are more comfortable with the concept of renting and sharing, but still causes nervousness among veteran industry participants uncomfortable without fixed, secure data centres and copyright-protected trading systems.