Reproducible References for Labelled-Bond Research
23% of labelled-bond disclosure documents vanish from their original URLs. Your citations shouldn't. XFIDs are permanent, content-addressed identifiers for the frameworks, SPOs and reports your research analyses — with a free, DOI-backed dataset behind them.
The Reproducibility Problem
Empirical work on green and sustainability bonds rests on disclosure documents hosted on issuer websites. Our own tracking shows those links rot fast: within four years, roughly a quarter are gone (Deleting the Receipts, covering $672 billion in outstanding principal). A footnote with a dead URL is a finding no one can verify.
An XFID is derived from the document's content, not its location. Cite the XFID
and your reference resolves at xframework.id/{XFID} for as long
as the registry exists — and the SHA-256 hashes let anyone confirm a copy is
the exact document you analysed, independent of us.
What You Get
Free Dataset with a DOI
The full registry — 8,229 documents from 6,225 issuers —
as CSV snapshots under CC BY 4.0, refreshed weekly, citable as
doi:10.5281/zenodo.21297903. Ready for your Data Availability Statement.
Citation Formats
Inline, parenthetical, bibliography and BibTeX formats for individual documents and for the dataset — see How to Cite XFIDs. Every document page has a copy-ready citation block.
Verifiable, Version-Locked Sources
Each record shows SHA-256 fingerprints (file bytes + canonical content), retrieval date, live source-URL status, and full version history — so your citation locks to the exact version you analysed, and silent revisions are visible.
A Research Corpus, Cross-Linked
1,171 academic papers on labelled bonds are indexed as first-class documents with abstracts, topics and a citation graph — linked to the very issuers and disclosure documents they discuss. Your paper can be part of the graph.
Free for Research
The dataset, the API, the identifiers and the citation infrastructure carry no academic licence fee. Academic and institutional email addresses get the dataset download link instantly by email — we just like to know who's building on the data. We ask one thing in return: cite the dataset DOI so the permanence flywheel keeps turning.
Start with the Data
Download the snapshot, read the data dictionary, or query the API directly.