Research published by Greenwich Associates and based on interviews with buy-side traders around the world indicated that 58% of FX trading desks were using TCA systems.
The research found that FX TCA was ahead of fixed income in terms of buy-side adoption, driven in part by a more easily accessible data set for analysis and the increasing usage of algorithmic trading and electronic trading venues.
John Halligan, Global
Yet Global Trading Analytics (GTA) president John Halligan suggests that the rate of deployment of TCA by FX market participants has not been particularly rapid, although he also accepts that investment in the technology is increasing.
The uneven quality of the product calls into question whether the promised cost reductions and trading certainty have been realised, says James Singleton, chairman and CEO of Curex.
“The jury is certainly out on measuring market participants’ use of TCA analysis to improve their executions,” he says. “Some have, but we assume that number to be small at this point.”
More widespread use of such analytics ought to result in more optimal execution performance over time. A move to fee-based orders should be providing more certainty around explicit costs, although it is important to also analyse implicit costs – such as market impact – and take an holistic view of overall performance rather than simply focusing on fees, observes Pete Eggleston, co-founder of BestX.
Locating weaknesses within the trading process requires accurate and intuitive analysis, explains GTA’s Halligan.
“The next step is to communicate those weaknesses effectively and consistently to all interested parties within the trading process,” he says. “The more the findings are communicated, the more costs will tend to trend downward over time and the easier it will be to maintain those improvements.”
According to John Crouch, founder of trading analytics and data science company Ideal Prediction, while data from single-dealer platforms is sufficient for some types of analysis, there are clear risks to using a single source.
It is important to gather as representative a picture of market prices and liquidity as possible, which is clearly not straightforward in a fragmented, over-the-counter market such as FX.
Indeed, regulations such as the PRIIPs’ (packaged retail and insurance-based investment products) costs and charges disclosure stipulate the need to estimate market mid rates based on as many market data sources as possible and applicable.
In defining TCA, the European Securities and Markets Authority went out of its way to rule out data coming from single banks or platforms, observes New Change FX managing director Andrew Woolmer.
Platforms would argue that they aggregate data, but that didn’t wash with the regulator for a number of reasons, he says.
“Firstly, if I know where your TCA data comes from, I will always execute your trade on that platform, which usually flatters the cost of the deal as the referencing is circular,” adds Woolmer. “Secondly, the platform is a component of the cost chain. Each platform charges the liquidity providers in some way, but that charge is not transparent to the client.
“In addition, some platforms provide client data to the market makers in ways that are unhelpful to the client. This means that the platform is part of the trading cost and you can’t measure a system from within the system.”
Providers who aggregate data from multiple venues and market participants are also subject to quality issues since that aggregated data set might include redundant liquidity, last look prices or flashing prices.
Leading banks have pre- and post-trade analytical tools – based on their own pricing data – which they can make available to their best customers.
However, the availability of pre-trade and real-time analysis for FX is relatively limited compared with equity TCA, reflecting the bilateral/full risk transfer nature of the institutional FX market compared with predominantly algorithmic execution in equities.
A further complicating factor is that these features need to be made available at the execution management system level to allow real-time execution adjustments and easier integration into the trader’s workflow.
This means it is necessary to distinguish between real-time TCA and on-demand TCA, which is simply an alternative to single-execution dealer reporting, says Ruben Costa-Santos, head of FX at ITG.
“Demand for real-time TCA and pre-trade analysis is increasing as buy-side firms look for consistent execution approaches across asset classes,” he says.
The definition of real time and the source of the data used to support these pre-trade tools should be a matter of thorough investigation by the user of those products, concludes Curex’s Singleton.