APIs: Reshaping the landscape of e-FX trading

Using standardised industry protocols, almost everyone can learn how to conform to APIs – and as we move further into an era driven by AI, they also cut costs by reducing manual work and automating the processing of huge amounts of data.

“Consumption of data and systematic execution are the two main drivers of the evolution of trading firms’ API needs,” says Barry Flanigan, head of Asia Pacific at iSAM Securities. “Brokers want low latency APIs for pricing and execution to stay competitive and ensure they consume and distribute these products efficiently. Money managers typically need them for portfolio management across assets and connecting dots across the various asset classes, while prop and algo traders are looking for super low latency for their automated strategies.”

Traders are using APIs to run complex algo strategies which are becoming increasingly sophisticated thanks to vast and growing data accessibility. Flanigan refers to clients pulling live data to track market moves in FX and metals, building custom dashboards to see pricing in action and even back testing their models with historical data.

“Better pricing, faster connectivity, access to more markets and reduced human error have made the traders’ physical location irrelevant,” says Tomo Tokuyama, EVP managing director, FX at Trading Technologies.

The reason why most of the financial trading world was able to seamlessly adapt to pandemic lockdowns was they could trade virtually anywhere due to the power of API connectivity.

“We would not be able to build a best-in-class FX offering that rivals standalone platforms without APIs,” adds Tokuyama. “Having FX APIs allows us to build a multi-asset trading platform, which then allows clients to consolidate from multiple single asset trading platforms.”

Because the messages received via an API are captured for the most part in real-time, firms can use the data to view parameters such as spread widening, speed of price updates and the differences in spreads across LPs across different time zones, tenors and notional sizes.

Boosting multiple trading factors

APIs have dramatically accelerated the speed, accessibility and flexibility of e-FX trading, enabling instant access to real-time pricing, liquidity pools and execution engines and allowing firms to operate with precision and scale.

“Brokers, prop firms, corporates and asset managers now demand highly specific API functionality,” explains Finage co-founder, Remzi Gökhan Uçkan. “While brokers may focus on execution speed and uptime, corporates may prioritise stability and seamless ERP integration. We are seeing increasing demand for APIs that are modular, scalable and capable of aggregating data across multiple asset classes to support more sophisticated decision making.”

APIs allow traders to plug directly into execution venues, build bespoke algorithms and access tick-level analytics for back-testing or signal generation. There is also increased use of APIs for dynamic hedging strategies, spread monitoring and real-time liquidity sourcing.

Many of Finage’s clients deploy APIs to build proprietary dashboards or to automate risk management routines.

A growing number of trading firms are customising their workflows and developing applications instead of relying on third party applications that utilise the same APIs to interface with OMS and risk applications. 

“We have recently developed APIs that allow customers to build applications that utilise our rules engine and smart order router functionality,” explains Dave Weiss, chief technology officer Sterling Trading Tech. “These applications are crucial for our customers to manage their risk and order flow routing allocation.”

When it comes to assessing the respective merits of FIX, REST and WebSocket, Weiss observes that REST is ideal for interacting with static data (account balances, margin settings, buying power, risk controls) while WebSocket enables users to subscribe to and receive real-time data such as position updates, executions and market halts.

“FIX is the tried-and-tested standard API for real-time, order related transaction data,” he says. “Most of our customers use FIX to send order flow to us. All of our connections to exchanges and venues are FIX and most incoming execution messages into our risk system are FIX.”

A type for everyone

Tokuyama acknowledges that all API types have their pros and cons and the choice depends on a variety of factors such as latency sensitivity, technical skill of the developer and use case suitability.  

“RESTful APIs are generally not useful for subscription-based use cases,” he explains. “However, they are perfect for extracting trading data and middle office integrations. Given its design, FIX is most appropriate for those trading applications with moderate latency sensitivity. WebSockets APIs are similar to FIX but are generally more efficient and are also most appropriate for those trading applications with moderate latency sensitivity.”

Gökhan Uçkan agrees that FIX’s reliability and maturity make it ideal for low latency order flows. “REST is useful for fetching static or less time sensitive data,” he adds. “It is easy to implement and well-suited for onboarding and analytics. WebSocket is the go-to choice for real-time streaming and is particularly useful for price feeds, market depth and monitoring trading signals.”

Multi-asset and crypto trading is having a sizeable impact on API demand. With strategies crossing FX, crypto, commodities and equities, demand for unified, cross-asset APIs has surged. Many algo and quant firms now rely on a single API layer to stream normalised data across asset classes, simplifying both strategy development and operational risk.

“Through APIs, trading firms can ingest multiple liquidity sources, compare spreads in real-time and smart-route orders to optimise execution,” says Gökhan Uçkan. “APIs also support real-time analytics to measure slippage, fill ratios and market impact – crucial for liquidity optimisation in fragmented FX markets.”

These factors also help FX providers differentiate their customer experience by supporting tailored interfaces, faster onboarding and automated reporting. More importantly, they empower FX providers to offer data-rich user experiences from custom alerts to advanced charting.

“Algorithmic trading is driving demand for faster, more responsive APIs,” says Flanigan. “We have spoken a lot recently about the increase in volatility across markets, driven by geopolitical tensions and uncertainty. As a result, speed is more important than ever. Regardless of whether the chosen strategies are high frequency trading, there is a growing expectation to move and react as quickly as possible.”

Aiding the decision-making process

These developments have fuelled demand for high performance APIs that can handle cross-asset trading without lag and seamlessly deliver data in-house for execution and analytics.

“APIs can pull liquidity from multiple sources like banks and ECNs to get the best prices for brokers,” observes Flanigan. “They also enable smart order routing to optimise relationships with their hedging/counterparty partners. APIs can be used to consume and analyse liquidity analytics, allowing brokers to make more informed decisions based on this data.”

For market makers, this enhances their ability to have a data-driven approach and provide highly customised pricing to a greater subset of clients who often have unique requirements and selling points.

Tokuyama describes direct API connectivity as the gateway to achieving the speeds necessary for many of his firm’s trading tools. “Our customers have various strategies that depend on our speed to markets,” he continues. “If they are trading any type of spreading strategies across markets, API connectivity is the only way we can support this.”

Liquidity curation is becoming more and more important for clients and is the foundation on which ECNs have been successful. The top ECNs have written to tens if not hundreds of different LP APIs, which allows them to curate different types of liquidity for different types of clients.

APIs allow LPs to provide bespoke pricing based on the client type since an LP may not want to provide a client with base pricing that is designed for a large set of clients.

A good example is full amount versus sweepable pricing. If a client’s strategy requires sweepable prices, an LP can curate a sweepable feed to a specific client or clients knowing that the flow will be sharper and can price accordingly. They can also provide another full amount API with tighter prices for more passive execution.

Supporting highly customised offerings

“The flexibility that APIs allow has enhanced the customer experience but has also helped LPs make better decisions as they can analyse customer trade data,” says Tokuyama. “For example, they can tighten currency pairs in which they have a decent yield but widen currency pairs they find difficult with certain clients. On the flip side, a client may curate their liquidity so that they price a group of currency pairs with one set of LPs that they believe have the best prices, but for another set of currency pairs they may have a completely different set of LPs.”

There will be a lot more automated trading in the future and those automated rules will be backed by data from APIs. Trading systems will one day automatically know that, for example, a trade wants to buy €200m at 10AM NY and will have pre-trade analytic models that suggest an algo is the best way to do that and will know that a certain LP’s algo or an internally built algo performs best for this type of order.

“Once completed, an automated transaction cost analysis report will be created with the performance benchmarks that are relevant to that trader or firm,” suggests Tokuyama. “Through machine learning and AI, the captured trading data will be analysed in real time so that the system continually improves on its decision making capabilities.”

Weiss refers to growing interest in high performance, cross-asset APIs and notes that his firm’s OMS API allows customers to set routing preferences to their outbound vendors.

“We plan on expanding this to make routing decisions based on such factors as venue latency or time of day liquidity statistics,” he adds. “Demand for our API functionality made us realise we had to get our API documentation and tools cleaned up.”

To this end the firm recently launched a new API library, a collection of pre-written code, tools and documentation that allows developers to interact with a specific software application or platform. “Accurate documentation and easy to use developer tools are essential in our industry where there is inherently a lot of financial risk,” says Weiss.

On the question of how APIs are facilitating the use of next generation technologies (such as AI) in the trading industry and where the impact of this is being felt, Weiss notes that although Sterling Trading Tech is exploring ways to integrate AI into its OMS and risk systems, there are privacy concerns.

“I believe AI will be used mostly by software used by asset managers and proprietary trading firms to make trading decisions based on AI-driven data,” he says.

Improving future risk management

So how will APIs shape the evolution of electronic trading as they become more embedded on trading desks?

“The flexibility and ease of creating new endpoints for customers is causing APIs to evolve very quickly,” suggests Weiss. “For example, we plan to offer endpoints for our next generation data warehouse, which will allow customers to request historical transaction and account-related data. This data can be ingested into our customers’ decision making applications. The evolving adoption of AI will also increase demand for historical trading-related data.”

The biggest impact of APIs is in algo trading and risk management where they feed real-time data to AI systems for tasks like predicting trades or managing risk.

“Real-time connectivity allows systems to respond instantly to market movements and optimise execution quality,” says Flanigan. “From a risk management perspective, AI-driven insights ultimately help improve decision making. As trading environments become more complex, access to real-time analytics is crucial for teams to make informed decisions to drive profitability.”

On the subject of future development, he adds that APIs are already embedded on trading desks where there is very little that happens without API connectivity, be it execution, pricing or analytics.

“The most sophisticated trading desks have become fully automated,” says Flanigan. “APIs connect FX with metals and crypto seamlessly, give desks real-time analytics to make smarter calls and let them build custom workflows.”

When asked how APIs are enabling next generation technologies, Gökhan Uçkan observes that they serve as the bridge between raw market data and AI-powered systems.

“Whether feeding data into machine learning models for predictive analytics or powering natural language processing-driven trading assistants, APIs are foundational,” he says. “We are seeing AI increasingly used in anomaly detection, trade surveillance and adaptive order routing, each depending heavily on high quality, API-delivered data.”

As APIs become deeply embedded in trading infrastructure, Gökhan Uçkan expects them to enable smarter, faster and more responsive systems.

“Their role will expand from data delivery to intelligence orchestration, powering everything from predictive risk systems to personalised trading interfaces,” he concludes. “APIs are not just shaping the future of e-FX – they are becoming the future.”