As the foreign exchange markets evolve, data and analytics have become the fuel that powers automated decision making of pricing, trading & risk management to drive increased profitability and customer retention.
As the foreign exchange markets evolve, data and analytics have become the fuel that powers automated decision making of pricing, trading & risk management to drive increased profitability and customer retention.
Yet while the growth of data and the continued electronification in the FX market is undeniable, analysis suggests that we are still in the very early stages of most firms becoming truly data-centric in how they manage their business.
Some of this is due to the significant cultural change needed for many of the more traditional participants in global currency markets. They are facing increasing competition from newer entrants, many of which are tech-driven startups, driven by a concept or process of ‘Continuous Intelligence’ where real-time and historical data are continuously combined for real-time analysis. The result is rapid, accurate, machine-driven decision making that is proving to be a game-changer for those who embrace it.
The challenge for incumbents doesn’t appear to be a lack of appreciation of the benefits of adopting technologies that drive continuous intelligence. According to recent independent market research commissioned by KX, 90% of respondents agreed that in order to remain competitive over the next three years, their organization will need to increase investment in real-time data management and analytics systems. Nearly two-thirds (64%) of organizations believe having access to real-time data is critical to making better business decisions, while over three-quarters (78%) say real-time data and insights are creating a competitive advantage for their businesses.
However, the study also revealed a number of significant barriers to increased adoption and usage, namely:
- Lack of technology needed to effectively capture data
- Shortage of people or skill sets needed to manage data analytics
- The technology to analyze data effectively for insights
- A lack of understanding of the value of the organization’s data
Getting Started
In foreign exchange in particular, we regularly see a range of data challenges including:
- The lack of a centralized marketplace and increased fragmentation
- 24 x 7 continuous operation
- Inconsistent liquidity across currency pairs, time zones, platforms etc.
- No centralized or consistent source of market data
- Continued pressure to increase the electronification of pricing and execution
The increased market fragmentation is further compounded by the blurring of the lines between institutional or wholesale and retail FX market participants. Historically there had been a fairly clear demarcation between the two but with a shrinking institutional spot market and continued growth in retail FX, there is a growing need on both sides to have the ability to either source liquidity or to capture data and trade flow from a broader range of channels. Many FX market participants now receive executable prices from traditional bank LP’s as well as non-bank market makers, aggregators, intermediaries and bridges that are simultaneously playing the role of both price maker and price taker.
The technology used to support today’s trading activities has to be both flexible and robust enough to handle the massively increased volume of market data flows with hundreds and thousands of price updates occurring in small fractions of a second driving near instantaneous machine-driven decision making.
Intelligent analysis of market and trade data can enable market participants to implement a flexible, rigorous and repeatable investment/execution process to optimize their price making capabilities, better manage their risk and increase their profitability while meeting regulatory requirements or adhering to industry best practice aligned to the FX Global Code.
The Rise of the Machines
It wasn’t that long ago that a large size FX transaction might be executed by voice using an intermediary such as an inter-dealer broker who would work the phones to discretely execute a handful of smaller transactions across a range of different banks to protect their client by avoiding the information leakage that could lead to a movement in prices that could impact their quality of execution. Today, both real-time and historic data is heavily relied on to construct and optimize automated execution and risk management trading strategies. That same large transaction today might be electronically filled as hundreds of separate trades using a smart order router to programmatically send orders to different electronic venues based on criteria such as expected fill ratios, latency, slippage, and transaction costs for each venue while also having size and time slices randomized to avoid detection.
Whether using execution algorithms to trade as optimally as possible, market-making algorithms to better manage risk, or more advanced opportunistic algorithmic trading strategies to generate alpha, many institutional investors have either increased or are planning to increase their use of algos in the coming year. In addition, tightening spreads, reduced margins and lower risk appetites have quant traders in search of alpha increasing their usage of cross asset and cross border trading and hedging strategies driving increased demand for both real time and historical data over longer periods of time and across a wider range of markets to drive their research and quant modeling tools.
The growth in quant trading is also driving the adoption of artificial intelligence in FX markets. With continued advances in big data and machine learning and the growth in use of both public and private clouds to store and analyze huge volumes of data, AI and machine learning techniques are being used for everything from pre-trade predictive analytics to determine how and where to execute transactions to post trade analysis of the quality of pricing received from liquidity providers such as spread and top of book analysis, as well as the behavioral patterns and profitability of clients looking closely at mark outs and P&L attribution.
The technical challenges facing most FX firms include keeping up with exponential growth in inconsistently structured and unstructured datasets, integrating modern technology with legacy systems, strategic mandates around incorporating the cloud into your data strategy and re-architecting your systems and processes to leverage all that cloud computing has to offer, and perhaps most important of all, achieving a level of confidence in the data and the machine-driven decisions it helps generate.
In addition to these technical challenges, market participants are also facing a range of business challenges such as rising regulatory burdens, conduct risk and margin compression while shrinking balance sheet capacity and reduced counterparty credit has banks focusing on credit optimization and central clearing of certain products.
Again, the findings of our research validate this view. Only half of those surveyed (52%) were confident that they will be able to do this with their current tools & resources and two thirds of respondents stated that their data challenges are more about the culture of their businesses than the data and data tools that were being used.
The breadth of these challenges requires the implementation of a data strategy that must be holistic and widely considered beyond the front office to also include additional areas such as regulatory reporting, consolidated surveillance of market abuse, AML, fraud detection and a range of other compliance led functions to satisfy the regulators, investors and key constituents of the organization.
The data required to overcome these challenges is all around us and with such ubiquity of data, capturing and optimizing the value of FX data has never been more critical in helping businesses derive intelligent insights to drive mission critical business decisions.
Becoming a data centric organization is a journey but building a culture that values and embraces data driven decision making supported by a robust data management ecosystem is certain to separate the market leaders and the market laggards of the future.
