One of the best learning experiences in my career came early when I spent 7 years at Bloomberg. It was a time of massive expansion for Bloomberg as they strove to catch and then pass Reuters market share. Led by a visionary CEO and littered with serious talent across the company I was fortunate enough to be team lead as we launched fixed income electronic trading on the Bloomberg terminal. For some of this time the e-trading teams were seated alongside the data sales teams, and it was my first experience of the power of data and the revenue potential. This piece will take a closer look at the electronic FX (e-FX) data space and the potential for the next decade.
So often in e-FX the story has been told through the lens of trading volumes, liquidity providers (LPs), and ultra-fast matching engines. How many press releases from platforms seem to see Average Daily Volume (ADV) as the bellwether to success? However, in a world of spread compression and LP pressure on platforms to reduce costs there lies a quieter, far more profitable battleground.
Market data
While spreads in major FX pairs have reduced and trade execution fees have become commoditised, market data has become the asset class that gives e-FX venues most scope for contributing to EBITDA. It could be argued as one of the primary reasons major exchanges acquired FX platforms.
1. Why Market data matters more than execution
Unlike equities or futures, FX is decentralised, meaning there is no single consolidated tape or official pricing source. Every platform e.g. Bloomberg FXGO, LSEG FX All, Deutsche Boerse 360T, Cboe FX, FXConnect, FXSpotstream and SGX FX — contribute its own “window” into the market due to their pricing methodology and their areas of strength (Regional, Product or Client based).
This creates enormous value to data consumers because:
- data = transparency
- transparency = benchmarks
- benchmarks = trading decisions, risk, and valuation
Clients, which could be hedge funds, banks, OMS/EMS providers, exchanges, data vendors, brokers…the list is long, will pay a premium for the cleanest, deepest, and most representative price streams. In many cases, the data product is worth more than the trades themselves to these clients.
2. The Market data hierarchy: Why only a few venues win
Not all FX data is created equal. There are effectively three tiers of FX data quality:
Tier 1 — Primary or quasi-primary venues (e.g. EBS, Reuters matching, Cboe FX, 360T GTX)
These have:
- firm, anonymous liquidity
- large, diverse client bases
- long historical time series
- consistent quote and trade behaviour
Their data is used for algorithm calibration, TCA, risk models, valuations, and analytics.
These venues dominate the premium-priced data segment.
Tier 2 — Specialist or relationship-driven venues (e.g. FXAll, SGX FX)
These platforms produce valuable data for specific products or pairs (usually EM) or client segments, but not for the entire market and most prices here are not broad enough to be considered a benchmark.
Useful for:
- regional trading desks
- EM pricing models
- bespoke algo execution
Tier 3 — Single Dealer Platforms (SDP)
SDPs/ internalisers produce vast amounts of flow, but the data rarely qualifies as neutral or representative because:
- it reflects the LP’s own skew
- it is client-specific
- it doesn’t represent the entire market
This data is excellent for microstructure analysis and for client TCA but has limited external commercial value. For this article I am focusing on Tier 1 & 2 and leaving SDP data out.
3. Why data has become the real profit engine
Anyone active in the e-FX space or financial services in general will be well aware of the revenue opportunities in the data space. At a recent lunch with a Head of Enterprise Data at one of the world’s largest multistrategy hedge funds I was informed that the e-FX space is relatively naïve compared to other asset classes.
Many e-FX platforms still leave the data sales to the same salespeople who are selling the trading platform. I have seen this first hand where many salespeople either try and give the data away for free with no understanding of the unique proposition or others would attempt to extrapolate ridiculous charges for data available in many places for a fraction of the cost. I remember saving one such deal over the phone with a NY fund when a senior salesperson overpriced the data cost by 400% to a global hedge fund.
However, many e-FX platforms are still not seeing the real potential in data sales and the fact it needs a data sales specialist to sell the proposition. e-FX platforms can take some pointers from companies across multiple data products who do this well – CME in Q2 made nearly $200 million in revenue from data sales across their products. Bloomberg don’t report figures but back in 2023 the FT reported an estimated $2BN a year in data revenue.
(A) Spread compression has killed fee-based revenue
Why is data now an important battleground? As LP competition increased, spreads on major pairs reduced and as such e-FX platforms who would generally charge spot $10-$20 per M in 2010 are now around the $1-$5 per M charge….and it’s only heading one way. During this period the cost of operating a venue did not fall proportionally especially if the platform needed to get on board with regulatory pressures and establish regulated entities (especially in the UK/EU). This forced platforms to look for non-transactional, high-margin revenue — and data was the obvious answer.
Market data typically has:
- >75% gross margins
- subscription-style recurring revenue
- low operational cost
This makes it one of the most desirable and scalable revenue lines in the entire FX stack and one with the most potential. Even if you were to strip down the data potential to one subset of clients (hedge funds) there is probably at least $4-5BN of data sales available globally with a fair portion of that in FX. As one hedge fund said to me ‘if it makes us money, we will buy it!’.
(B) Fuelling data sales: Algorithms need data — lots of it
FX algos have become more mainstream across the buy-side now. In my experience running sales/revenue generation at e-FX platforms the growth of algos from bank internalisers. hedge funds, asset managers and more recently corporate treasurers has seen impressive growth and adoption.
Algos are data-hungry and require a minimum of 5 criteria (below):
- historical tick data
- venue-specific fill data
- order book dynamics
- time-of-day seasonality
- volatility and liquidity heatmaps
e-FX venues realised this and anyone who has worked with an exchange at a strategic level knows this was one major reason exchanges bought FX platforms. It was a subtle point that could be cross utilised in the exchange group as well as an area offering tremendous potential.
(C) Regulation increased the need for transparent data
Exchanges were quick to understand that regulatory pressure that would move into the FX market would increase the need for data. Even though spot FX is not a MiFID II instrument, a range of regulatory and risk pressures have pushed firms toward TCA, Best Execution policies, Audit trails, Vol and mark to market validation and risk governance. Quality data is essential for all of these.
Exchanges recognised this early and built entire business lines around FX data licensing. Once an FX venue is inside an exchange group, data monetisation becomes dramatically more efficient.
4. Exchanges apply a three-step strategy to data sales:
Step 1 — Package & rebrand the FX dataset
Turn raw tick data into indices, consolidated feeds, analytic products, real-time and historical bundles and risk/vol dashboards.
Step 2 — Cross-sell to existing exchange customers
Think about an Exchanges client list. FX data can be sold to equities clients, futures traders, corporate users, risk departments, index providers and clearing members
The distribution is already built — FX data simply plugs into it.
Step 3 — Add FX data into enterprise contracts
Large banks and funds often buy all the data from an exchange which creates high-margin, multi-year bundled agreements with FX being a part of this.
5. The future: The rise of FX “Data Utilities”
e-FX platforms will see new monetisation channels beyond simple “data sales.” Some possible trends or stories that will continue to grow:
The consolidated tape never took off, and some companies tried and failed to get the market to buy in to the story but there is no doubt with so many venues and no official benchmark, regulators and large buy-side firms could re-float this idea.
The big buzzword in all markets is AI. AI is data-hungry and LLMs, predictive models, and microstructure AI tools will need clean tick data and venue specific data. The winners will be venues with deep, long-term history and flexible data delivery tools.
Selling bespoke market data. Clients increasingly want per-client TCA, customised liquidity heatmaps, venue-specific execution probability models and counterparty-specific skew behaviour
6. Data vendors/Analytics providers and their role in the e-FX space
We haven’t even touched on dedicated data and analytic vendors and their impact in this space. Companies like BestX, Tradefeedr and New Change. Perhaps just for now it is worth mentioning some e-FX platforms who have taken a strategic approach to aligning themselves with specialist providers in the space such as LSEG FXAll strategic alignments with Tradefeedr and FXConnect’s collaboration with Best X on BestXecutor (both State Street owned). May be this will need to be a follow up piece.
Conclusion: Market data is becoming one of the Crown Jewels of e-FX
Trading venues are no longer just matching engines. They are data factories feeding algos, analytics, compliance, valuation models, risk systems and macro strategies. In an environment of compressed spreads and low transaction fees, data is the most strategic — and profitable — asset FX venues own.
The war behind the scenes in e-FX is no longer about trading volume. It’s about who controls the data that defines the market itself.
Readers can see more of John McGrath’s articles on his Substack page:
