The Foreign Exchange Professionals Association (FXPA) now defines FX spread grids as indicative pricing tools, not firm quotes or contractual benchmarks.
A new FM Intelligence analysis takes that redefinition and asks the harder question: how far can the price a client actually gets drift from the published grid, and who absorbs the difference.
The full analysis is on the FM Intelligence DataLab portal.
The Gap Between the Grid and the Fill
FXPA published its guidance on June 29, a move FinanceMagnates.com reported at the time. FM Intelligence builds on it with a proprietary model, the Grid-to-Fill Gap, tracking the distance between the advertised grid and the spread a client realizes on execution.
Under normal liquidity, the firm puts that gap near 10%. Under stress, large size, or thin liquidity, its base case widens to roughly 200%, with a high case above 300%.
FM Intelligence describes these as illustrative modeled estimates, built from FXPA’s own statements rather than a transaction dataset, and revisable as data arrives.
A gap between an indicative grid and the fill is expected by construction, not evidence of wrongdoing. A published grid is a symmetric pre-trade reference, while execution is asymmetric.
Dealers internalize about 80% of spot orders, matching client flow in-house, according to the Federal Reserve Bank of New York, and apply last look and pricing skew of the kind the FX Global Code addressed when the GFXC revised its last-look section.
In FM Intelligence’s reading, divergence is the model working as designed.
Five Readings of One Document
The guidance is largely uncontested, FM Intelligence notes, yet liquidity providers, trading venues, the data-rich buy side, and conduct regulators each read it as a win for their own position. The analysis maps all five.
The simplest reading is the stated one, the firm adds: differing interpretations had caused disputes, and an industry body clarified definitions to reduce them.
FXPA’s membership spans buy-side, sell-side, venue and data firms, so a document that serves several at once fits a genuine consensus rather than capture by any single interest. None of the readings requires bad faith.
Where FM Intelligence sees latent risk is narrower, in the distance between a pricing representation and the fill, the same ground older enforcement cases turned on.
State Street settled for $382.4 million in 2016 over hidden FX markups paired with best-execution assurances, and Barclays paid $150 million in 2015 over its last-look engine.
Both involved undisclosed conduct, not a published grid, and FM Intelligence stresses that no action between 2024 and 2026 has targeted dealer pricing against an advertised grid. The analysis names no current firm.
Who Measures the Price Next
Every push to benchmark on realized data rather than the grid shifts attention toward the venues and analytics firms that measure fills, a shift visible in tie-ups such as TD Securities and Tradefeedr.
FM Intelligence estimates only about 25% of institutional FX participants run independent, multi-LP transaction cost analysis today, and projects the share of execution cost judged mainly on realized data rising toward 62% by 2027 in its base case, a scenario it presents with bull and bear ranges rather than as a certainty.
The full breakdown, including the Grid-to-Fill Gap model, the five-reading map, and the adoption scenarios, is on the FM Intelligence DataLab portal.
www.tradingview.com