Tradeweb has been leading trade automation for almost a decade and the pandemic has only strengthened its partnerships with clients says Enrico Bruni, managing director, head of Europe and Asia business, Tradeweb.
How it started
From the early days of Tradeweb, we have listened to our clients and built many of our most successful innovations by trying to address their challenges and evolve their workflows. These trusted relationships have fuelled the expansion of our businesses across geographies and into new products. By working collaboratively with our longstanding network of clients and developing functionality with their needs in mind, we were able to set ourselves apart from our competitors right from the start.
A prime example of this type of collaboration is a tool we call Automated Intelligent Execution (AiEX), initially aimed at facilitating a faster and more efficient trading process. For many years trading desks have had to process more trades, more frequently and across more asset classes than ever before, which can be a challenge when operational budgets and headcount stay the same. AiEX was created to help clients handle an ever-increasing number of transactions in a time- and cost-effective manner.
The ability to simultaneously automate low-touch trades across several asset classes can be indispensable. For example, a large asset manager is likely to segregate low- and high-touch trades, often to dedicated desks. A client’s low-touch business may be across cash, credit, rates and derivatives, so the desk needs to account for the very different way trades are conducted. While managing the idiosyncrasies of each asset class, the trader must adopt a consistent approach.
During the early stages of AiEX, we mainly focused on automating the execution process for lower touch transactions, first in rates products and then for other asset classes. However, breakthroughs in technology and the integration of data and analytics within the trading workflow have meant that AiEX is now increasingly used to automate higher touch, larger size transactions.
Our partnerships with traders have allowed us to evolve AiEX and redefine the nature of the request-for-quote (RFQ) process with managing multiple trading protocols across a wide selection of asset classes as a key attributer.
Improvements in big data and machine-learning technology have also enabled us to help clients determine the most appropriate ways to trade. Ultimately, post-trade data proves whether the protocol achieved its objective. Customers use our award-winning transaction cost analysis tool to assess and, subsequently, improve their execution strategies. In the example of an index house, evaluating benchmark slippage may focus on market impact and encourage consideration of a firm price protocol to aggress prices.
While some clients may prioritise transaction costs, others may focus on fees or market impact. Deciding how many dealers to put in competition can be a critical factor. Going to a single dealer means there is very little information leakage, transaction costs are low and market impact is minimal – but it may be preferable to maximise responses. For different desks at certain times, this is a very different conversation. It is unlikely that any two clients will want the same set of rules.
Increasingly, the RFQ mechanism is morphing into an automated process with the trader just taking a last look. A spectrum is emerging among clients for the degree they wish to automate, depending on asset class and execution objectives. Beyond this, we are now seeing no-touch flow. This trend started in exchange-traded funds (ETFs), where rebalancing generates orders that may entirely bypass the trader unless it fails. Without a doubt, certain profiles will become exclusively no-touch.
Every month we see sizes getting larger and clients going down the liquidity spectrum, as traders quickly become accustomed to the benefits of automation. Clients first get comfortable in the most liquid sectors, however, the speed of adoption across asset classes and trade sizes surprises even the trading desks.
Our most recent AiEX enhancement allows traders to directly interact with their enquiry, transitioning from fully automated to hybrid or manual workflows to ensure their order gets done. Within the same screen, they are able to manage their AiEX parameters in real time and assess if they can still meet their execution strategies by bypassing some of the rules. The improvements we have developed are all built into AiEX and coupled with our building-blocks approach, they enable traders with almost no investment in technology to achieve their desired outcome far more efficiently.
The process is only as automated as traders want it to be; if not satisfied with the number of quotes they receive, they can interact wherever they want right up until the final decision. A trader may receive prices back from an RFQ – but a firm stream from an enabled counterparty outside the RFQ may offer a better price. We link protocols so, if the firm price is better, it automatically gives up the RFQ and aggresses on the firm price.
This model is particularly powerful in less liquid markets for example High Grade, High Yield and Emerging Markets, where the nature of the market means there’s more contextual information – inventory, price discovery and other factors – that determines how to trade. Our experience working across many asset classes has given us deep insight into what needs to be automated, to ensure traders have flexibility in all the most useful areas for that asset class. Working with buyside traders also helps us generate ideas and build better functionality.
How it’s going
Nearly 18 months since the pandemic and the resulting stressed market conditions, automation has significantly accelerated rather than stuttered. Looking at Q2 2021 data, we are seeing AiEX adoption rates reaching almost 60% of all tickets and 25% of all volume which proves that AiEX has become an integral part of our clients’ workflow, evolving into an extension of the trading desk. This is contrary to what we initially saw in the first two weeks of March 2020, when AiEX volume reduced. Its adoption continues to grow, spanning an ever-expanding profile of trading activity across multiple asset classes used by both the traders as well as the quant teams.
All the new technological developments mean that demands on buyside trading desks are changing. There is now a key role for quant-focused professionals with the skillset to determine what and how to automate. There is real value in creating a team with a synthesis of people dedicated to consuming data and traders with experience in the market.
The question used to be whether to automate; now, conversations are about how far we can go. Just like the shift from voice to electronic trading, the market will naturally continue its transition to automation and hybrid protocols. Given the sheer volume of transactions desks have to complete each day, I believe this transition can only accelerate.
©Markets Media Europe 2021