BestX Unveils Pre-Trade API

While the story of BestX has without doubt been one of success – it is hard to describe it as anything else given its pool of data now covers more than $47 trillion of assets under management and contains billions of trades – if there was one gap in the firm’s inventory, it was in the pre-trade space. The platform’s data could indeed inform future execution policy, but in terms of actionable data close to the trade there was a gap – one that is now being filled with the release of a new pre-trade API.

As is the case with the firm’s other products there is a user interface available, but as Aled Basey, head of FX at BestX explains, the important aspect of the new service is that it enhances automation and doesn’t break the buy side trader’s existing workflow. “We wanted to empower the buy side trading process without interrupting existing workflow, and this is a good way of doing it,” he says.

The new API service can be pointed at the customer’s OMS/EMS and help provide more information to traders about not only which algo they could use, but also, if they are RFQ-ing, the optimal panel size and counterparty selection. Users are able to interrogate the BestX database – which, as well as the aforementioned AUM has over $110tri of traded notional opted into the pool – providing their given objective(s). This can be any of a number of standard objectives such as minimising signalling risk or slippage to arrival, and armed with that, the pre-trade API provides the top three algo strategies in the requested currency pair for the given objective.

The dataset used can either be from the client’s own trading history, or, if they have opted into the BestX pool, it can be from the broader community. The sample size is provided and users are also able to establish minimum thresholds for the number of trades in the dataset to generate a result.

Although algo wheels have not really established a footprint in the FX world, there is certainly more talk about them as a concept, but the BestX pre-trade API may actually mean the industry leapfrogs the algo wheel phase and goes to what could be termed “smarter selection”.

“This creates a meritocracy among providers,” Basey observes. “It is not about ‘whose turn is it’ or ‘this provider is my overall number one’, it is all about more granular decision making. If buy side traders are making more informed decisions on a case-by-case basis they will not only be rewarding best-in-class performance for individual strategies and providers, they will also have the opportunity to add their own Alpha.”

RFQ

The new service is not only about, possibly, helping the FX industry avoid the algo wheel, for the API can also be used to help customers create optimal counterparty panels and even select individual counterparties for certain RFQ trades.

On an individual trade basis, the user delivers the order details and their criteria (they are able to use BextX data and up-or down-weight certain aspects) and then, using data such as the liquidity profile of an LP, market footprint and BestX’s Par-Adjusted Win Ratio, the pre-trade solution provides a top three of counterparties based upon the analysis. “The API is FDC3 compliant, so we are responding to the demand for increased interoperability between platforms and ultimately providing the salient information back to the traders, at the point of trade,” explains Basey.

Equally, on a broader basis, the user is able to create their optimal panel of LPs by currency pair or size of trade, by interrogating the BestX database of their own trades, or again, if they have opted in, the broader universe of trades from all BestX clients.

This API weaponises TCA by providing actionable data that in turn reduces execution costs and builds more efficiency into the buy side trading process

For those clients that have opted into the pool, there is also the opportunity for what might be termed more efficient counterparty management. Previously, certain LPs might have been added to a panel through reputation or the fact that they won a certain percentage of trades in another currency pair – equally, some LPs would be excluded from a panel simply because the client knew little or nothing about them. By opting into the BestX pool, users will be able to see how other LPs, with whom they perhaps don’t have a relationship, are performing, and selectively onboard them, with less guesswork involved than is currently the case.

The obvious benefit for LPs is that superior performance will be recognised and they will have the opportunity to interact with a new counterparty, or increase their overall win ratio with existing clients.

“We have a huge amount of data that used to deliver analysis on an ad hoc basis, we are now automating more of the decision support process,” Basey says. “This API weaponises TCA by providing actionable data that in turn reduces execution costs and builds more efficiency into the buy side trading process. Another feature of this enhancement is to receive the ‘BestX Expected Cost’ for a trade ahead of the execution, which could then provide an additional input into RFQ optimisation on order acceptance – for example, auto-accept if within BestX expected cost estimate – bringing outlier detection into real-time.”