Research: does a usable parallel/street-rate data source exist for NGN?

> Written to the standard set by #33 — read that issue first for the expected level of detail. ## What Research whether a usable parallel/street-rate data source exists for NGN (and other target currencies), separate from the official-rate provider already in use. Document what is found — a source, a cost, an access method, or a documented absence — rather than building anything. This determines whether "gap to official rate" and "gap to parallel rate" can both be reported as separate dimensions. ## Why it matters Nigeria has a well-known gap between the CBN official rate and the parallel market rate. If the providers used by `refrate/` track the official rate, the loss-floor finding is measured against a rate that may not reflect what a user can actually get. If a parallel-rate source exists, the project could report both gaps — giving readers a more complete picture of the cost landscape. But this is a research question first. If no usable parallel-rate source exists (or if the ones that exist are unreliable, expensive, or require API keys the project does not have), the answer is "no source found" — and that is worth stating rather than assuming one exists. ## Which layer **Layer 1 — observable fact.** This is research into what data sources exist, documented with evidence. No code changes. ## Exactly where the code goes - **New file:** `docs/parallel-rate-research.md` - The document should list: which sources were investigated, what each one provides, what it costs, and whether it is usable for the project's purposes - If no source is found, the document states that plainly ## Where the data comes from - `refrate/` package — which providers are currently configured - Public APIs for parallel/exchange rates (e.g., abokiFX, parallel.ng, other Nigeria-specific sources) - General exchange rate APIs that may offer parallel rates - The providers' own documentation ## What "unmeasurable" means here If a source claims to offer parallel rates but does not document its methodology, that is a finding: "the source exists but its rate is unverifiable." The project cannot use a rate it cannot defend. ## Acceptance criteria - [ ] At least 3 parallel-rate sources are investigated (even if all are rejected) - [ ] For each source: name, URL, cost, access method, and whether it documents its methodology - [ ] A clear recommendation: use source X, or no usable source found - [ ] If no source is found, the document explains what would need to change for one to become available - [ ] No code changes — research and documentation only - [ ] The document is linked from `docs/corridor-measurements.md` or the README ## Out of scope Do not change: - `refrate/` code — this is maintainer-owned - The verdict thresholds - Any corridor's reference rate - Building or integrating a new provider ## How to verify locally ```bash cat docs/parallel-rate-research.md ``` No build or test needed — this is a research task.