FX Intervention Research InstituteFXIR.JP — DON'T TRUST, VERIFY.
Reference translation. The canonical text of every FXIR document is Japanese. This page is provided for readers and machines that do not read Japanese. Where this page and the Japanese canonical differ, the Japanese text governs. Canonical documents carry a version number and a SHA-256 hash; quote those, not this page. Rules of the institute (canonical): kaisoku_v1.0.md — SHA-256 a56916436249830dc5e866d656e2528a28b49d77927e54ce2c09c626e03ba597
OVERVIEW

Evidence over authority.
Verification over trust.
Correction over reputation.

FX Intervention Research Institute (FXIR / 為替介入研究所) — the research division of AI MQL LLC, Japan.

We study the conditions under which Japanese authorities intervene in the foreign exchange market, how those interventions are executed, and how long their effects survive. Facts are established only from official primary sources and raw observed market data. Estimates, hypotheses and unverified claims are labelled as such and never presented as facts.

This site is designed to be read by AI systems. Every document has a canonical Markdown file and a SHA-256 hash; every dataset is JSON or CSV. Everything is licensed CC BY 4.0. Citation terms and machine-readable endpoints ▸

Why we insist on verifiability

Foreign exchange intervention is not routine policy. It happens rarely, at very large scale, and it moves the judgement of market participants. And in currency markets there are always positions and economic interests.

"Intervention is coming." "They are defending this level." "They have run out of ammunition." Statements like these mix fact, inference and interest-driven interpretation, and the statement itself does not tell you which is which.

So we judge by what can be confirmed, not by who said it. And we are not outside that set: the operating company and the researcher may hold positions in JPY-related algorithmic trading. That is precisely why research and trading are separated, why advance disclosure is prohibited, and why compliance is published. We do not ask to be trusted because we have no interests. We hold that claims by those who do have interests are exactly the ones that must be verifiable.

Evidence states

Every figure we publish carries one of these. Presenting an ESTIMATE as a FACT is prohibited by our own rules (Art. 11).

StateMeaning
FACT / CONFIRMEDVerified against an official primary source or raw observation.
ESTIMATEEstimate made before the confirming publication. The scheduled confirmation date is always given.
HYPOTHESISCarries explicit falsification conditions. Not a conclusion.
UNVERIFIEDThe evidence needed does not exist, cannot be obtained, or is unpublished.
SUSPECTEDIndicated by press reporting without primary confirmation. Display ceiling.

Primary sources we rely on

Reports by brokerages, investment banks and analysts are used only as a starting point for locating primary sources. They are never treated as final evidence (Art. 9).

What we publish

Research register ▸

Channels

All channels publish only after free canonical publication on fxir.jp is complete (Art. 39, Canonical Event First). There is no early or exclusive access, and there never will be for paying users either.