Version 1.0 · March 31, 2026
Why we publish this: CurrencyBlazer is built on the principle of radical transparency. We disclose our full analytical methodology so that clients, their auditors, and their treasury teams can independently verify our calculations. If our math doesn't hold up to scrutiny, we don't deserve the engagement.
CurrencyBlazer's audit engine performs a transaction-level comparison between a client's executed FX rates and independent mid-market benchmark rates. The goal is to quantify the exact cost difference — in basis points and dollar terms — between what the client paid and what the market price was at the time of execution.
Our sole benchmark data source is Fixer.io, operated by Apilayer GmbH (Vienna, Austria). Fixer.io provides historical mid-market exchange rates sourced from the European Central Bank and a consortium of commercial FX data providers.
Key characteristics of this data source:
We chose Fixer.io over alternatives (Bloomberg Terminal, Reuters Refinitiv, XE) for three reasons: it provides genuinely independent mid-market rates (not dealer-quoted), it's accessible at enterprise scale via API without per-seat licensing, and its ECB-sourced rates are the same benchmark used by EU regulators for FX rate compliance.
Every transaction in a CurrencyBlazer audit is evaluated using three formulas:
This measures how far the client's executed rate deviated from the independent mid-market benchmark, expressed in basis points. One basis point equals 0.01%. For context, enterprise-tier FX providers typically advertise spreads of 20–40 bps; our audits frequently identify actual effective spreads of 50–140 bps.
This converts the spread drift into an actual dollar figure — the amount the client overpaid on a given transaction. The two-step calculation is critical: leakage is first computed in the target currency, then converted back to USD using the benchmark rate. This prevents currency-unit errors (e.g., reporting JPY amounts as USD).
Where transaction data includes both an order timestamp and a processing/settlement timestamp, we measure the time gap between them. Lags greater than 60 seconds are flagged, as the mid-market rate may have moved materially during this window — particularly in volatile corridors like GBP/USD or emerging market pairs.
The audit runs through the following stages:
The client's CSV is parsed, headers are normalised (with auto-mapping for provider-specific column names like Airwallex, Wise, WorldFirst), and required fields are validated. Missing or malformed data is flagged immediately.
Any personally identifiable information (names, emails, account numbers) is stripped from the transaction records. Each transaction is assigned a randomised internal ID.
For each transaction, we query Fixer.io's historical API for the mid-market rate on the exact date of execution, for the specific currency pair. Each transaction gets its own benchmark — we never average or interpolate.
The three core formulas are applied to each transaction individually. Results are classified by severity: CRITICAL (>$1,000 leakage), HIGH (>$500), MEDIUM (>$100), LOW (>$0), or CLEAN.
Transaction-level results are aggregated by currency corridor (e.g., USD/EUR, USD/GBP), by date range, and by volume tier. We calculate total leakage, average spread drift, and annualised projections based on the audit period.
A client-facing "Found Money" report is generated with executive summary, corridor breakdown, top flagged transactions, root cause analysis, and the Gain Share proposal. The report is delivered as a branded PDF.
| Severity | Leakage Threshold | Typical Cause |
|---|---|---|
| CRITICAL | > $1,000 per transaction | Large notional + wide spread drift, often on high-volume corridors |
| HIGH | $500 – $1,000 | Moderate spread widening, typically on mid-tier transactions |
| MEDIUM | $100 – $500 | Standard spread drift at enterprise volumes |
| LOW | $0 – $100 | Minor rounding differences, micro-timing lag |
| CLEAN | $0 or negative (client-favourable) | Rate matched or beat the benchmark |
Beyond severity, each transaction is classified by the likely root cause:
Annualised leakage projections are calculated by extrapolating the audit period's total leakage to a 365-day period:
This is a linear projection that assumes consistent transaction volume and patterns. We explicitly note in all reports that annualised projections are estimates, not guarantees, and that actual results depend on future transaction patterns, market conditions, and provider behaviour.
Clients are encouraged to independently verify our results. Every audit report includes:
Any client with a Fixer.io API key (free tier available) can independently look up the same historical rates and verify our calculations to the penny.
For questions about our methodology, data sources, or analytical approach:
CurrencyBlazer — Methodology
Email: client-inquiry@currencyblazer.com
Web: currencyblazer.com
CurrencyBlazer provides data analytics and benchmarking services only. Nothing in this methodology disclosure constitutes financial advice. Identified variances reflect mathematical differences between executed rates and publicly available benchmarks — they do not constitute evidence of breach of contract, regulatory violation, or misconduct by any FX provider.