Analytics - taking the guesswork out of FX execution

By David Cooney, Co-founder and CEO of MahiFX

What sort of key metrics are FX market participants now looking to extract from the vast amounts of trading data and associated information that’s becoming available to them?

Whilst sometimes intimidating, data is one of the most powerful things in today’s FX market, especially if used correctly. The overall goal is to enhance yield, decrease risk and identify different types of behaviour. Having detailed information at your fingertips leads to much greater control and differentiation of pricing shown to customers. MFX Echo is incredibly rich in functionality to assist clients with these goals. Among other things it can identify significant price moves to spot client latency arbitrage and sweeps. You can delve into per millisecond pricing charts to check system and LP performance, see how currency risk develops and decays from different sources, answer client queries and observe client behaviour checking who your best clients are and spot where spreads are too tight or need to be more competitive. Echo’s risk path also gives an accurate PNL to every trade based on its risk interactions. In essence, these sorts of metrics provide FX participants highly valuable insight and control over their business which is impossible without data. Having access to this type of advanced diagnostic tool is vital for today’s trader.

Regular analytic software cannot cope with millisecond data
Regular analytic software cannot cope with millisecond data

Is the growing demand for FX data analysis being driven mainly by the need to meet Best Execution compliance requirements or for developing more effective real-time trading strategies or is it a combination of both and other factors?

The primary motivation for most business owners is to make more money whilst reducing risk, they of course have an obligation to satisfy best execution compliance requirements but that doesn’t always inspire business owners. We are in data centred world now and it’s becoming increasingly important to have awareness and control of your business in a much more granular way as well as satisfying all your compliance requirements.

Access to analytical packages in FX has traditionally been limited to the biggest banks and some larger more sophisticated asset managers. Why is that?

The amount of data storage and computing power required to perform analytics has always been a huge outlay and even as technology improves this has not gone away. Regular analytic software cannot cope with millisecond data. This is where MahiFX has the experience and know how to provide analytics at reduced cost to all.

Typically analytical software has also been very technical in nature, which is not necessarily welcomed by all users, many of whom simply want to find key information quickly. We wanted to produce a product that brought together all the data in one place, but present it in a way that is meaningful for any user from a trader right through to a business owner, each with different goals, fundamentally making it accessible to everyone.  

MFX Echo was designed as a real-time analytics system
MFX Echo was designed as a real-time analytics system

MahiFX offers MFX Echo, a trade analysis and liquidity visualisation solution which is an extension of your existing algorithmic execution offering MFX Vector. How have you managed to overcome the challenges of delivering such a powerful tool like this so cost effectively?

Echo has been built as a cloud-native SaaS compliment to our Compass and Vector products (which are typically hosted in colocation with exchanges). Echo’s multi-tenant architecture is designed to securely partition each client’s data whilst leveraging shared infrastructure, allowing us to reduce operational costs. Echo leverages Amazon Web Services low-cost cloud storage as a big data store, and uses elastic scaling to ensure that capacity adapts to meet customer demand.

Many firms rely on very low latency and very speedy delivery of data for their use in their trading strategies. How have you achieved that with MFX Echo?

Echo was designed as a real-time analytics system from inception. A principal requirement was that data from a colocated Compass or Vector system should be made available for detailed analysis in under a second. This required us to design a proprietary “firehose” streaming mechanism able to quickly and efficiently replicate all available tick data from a customer’s trading system (for example leveraging our existing expertise in low-latency reliable messaging and efficient encoding of data). It was also necessary to give careful consideration to the backend storage mechanism and loading methodologies, as many of the shelf big data/data warehouse solutions are able to ingest large volumes of data but cannot do this in real-time.

How important is it to present graphical trading data in a more innovative way nowadays to help traders better visualise what’s happening in the market and how have you gone about designing that into the MFX Echo system?

Being able to visualize multiple price streams down to millisecond detail is a challenge as there is so much data to present. But this data provides real insight to market interactions and also exposes market sweeps and abuse.

Too much data can be visually intimidating and overwhelming for users and they typically have different requirements and interests in the way they wish to view information, so the key is giving the user full control to add and remove data from the screen as they see fit. This enables them to pursue a particular line of enquiry in a clean and clear way without it becoming too distracting.