THETRADETECHFX DAILY from the floor
How the buy-side is rethinking FX algo performance
What’ s the best way to empirically assess the latest in algo performance metrics? The industry is becoming increasingly data driven and as a result the trading analytics space is becoming saturated. While this breeds competition and innovation, it can also lead to difficulty finding the most useful offering. When I think of assessing the latest in algo performance metrics, the highest priority is having quality data. We do not have adequate sample size for significant conclusions in algo trading, so I want to make sure there is clean peer data to draw statistically significant conclusions and build robust models. In addition to data, metrics need to help in a pre-trade situation rather than evaluate post-trade performance. Post-trade analytics, albeit essential, do not push the needle forward or add value to my decision-making process at the point of trade. Finally, I need something easily digestible and integrated into my workflow because time is of the essence when getting trades to market.
The TRADE catches up with ERIC BROWN, FX trader, T. Rowe Price, delving into the increasingly data-driven world of FX trading and unpacking how buy-side firms are reassessing how they evaluate algo performance.
others achieve cost savings during less liquid times when internalising flow and passive trading is key. In addition, each counterparty has differentiated franchise flow. At T. Rowe Price, we think this is the single biggest factor that distinguishes offerings. If we are trading a large Scandi pair, we want a broker who can help minimise our market impact and naturally offset that flow. If we only had one provider, we would not be able to take advantage of these matches in the same way. In the future, we see the potential to add additional providers and lean into each of their strengths.
How can this effectively influence algo selection, both in-house and third-party? At T. Rowe Price we exclusively use thirdparty offerings. Since we do not know everything happening under the hood, it is essential for us to understand how each strategy will behave. This allows us to select the best algo for each trade that we execute. We are striving to become more systematic and consistent in that selection process. With improved algo performance metrics, we can make an informed, datadriven decision. These metrics also help us decide if an algo is not the best execution method, there are times when liquidity or volatility conditions justify trading RFQ / RFS or going voice. Overall, data and performance metrics should be at the heart of each trading decision. The FX markets are ever evolving and as traders, we need to have the most up to date information to execute in today’ s market and not the market of yesterday, last week or last year.
The TRADE’ s recent Algorithmic Trading survey found a clear bifurcation in the industry between long-only firms who use five or more algo providers, and those who use just one – what are your thoughts on this? We use five or more algo providers. We find each provider has strengths and weaknesses, so it is beneficial to us to have a varied panel. Some providers perform better in volatile markets, while
What is the buy-side most craving when it comes to algo providers? What should they be focused on? From my perspective, the buyside is craving transparency and differentiation when it comes to algo providers. Transparency is key when deciding which provider to select because I know exactly how that strategy is going to behave in the market. Data and analytics providers are helping this effort by offering the same data feed to both the buy and sell side, which facilitates concrete, data driven conversations. Differentiation is the second piece. This can come in the form of liquidity, logic, or even style. Providers need to find a way to distinguish themselves and be able to show how their algos add value to our process.
The official newspaper of TradeTech FX US 2026 TheTradeNews. com 21