chadlupkes on Nostr: I've been working on a securities database for long-horizon investing, and I keep ...

I've been working on a securities database for long-horizon investing, and I keep running into a problem that seems much bigger than my own project.

We have reasonably good open infrastructure for identifying legal entities: LEI/GLEIF.

We have open infrastructure for identifying financial instruments: FIGI/OpenFIGI.

What I have not been able to find is the equivalent open reference layer **through time**:

Which instrument carried this ticker during this date range?

What was it called before that?

Did a merger, bankruptcy, conversion, redomiciliation, spin-off, or other corporate action terminate one instrument and create another?

Which historical identifier belongs to which lineage?

This sounds mundane until you try to maintain a 20- or 30-year investment history. Then it becomes foundational. A wrong identity relationship quietly contaminates everything built on top of it.

What surprised me most is that the industry appears to have recognized the problem already. FINOS launched a Securities Reference Data project in 2018 around essentially this problem: institutions repeatedly maintaining their own mappings, duplicating work that provides little competitive advantage.

That effort never produced the shared data layer it was trying to create, and the repository has now been archived.

That made me wonder whether the problem is not primarily the data model.

Maybe it is the coordination model.

I'm exploring a different architecture: instead of building a centrally governed "golden database," publish small, immutable, signed attestations.

"This ticker belonged to this instrument from date A through date B."

"This corporate action connected these two instruments."

"This dividend was declared on this date."

Each assertion carries its provenance and the identity of whoever made it. Different researchers do not have to agree before their work can coexist. A consumer decides whose determinations it trusts.

The interesting property is that this can be useful with **one participant**. It doesn't require a consortium to exist before anything gets built.

Eventually there may also be a way to compensate people at very small scale for doing the tedious reconciliation work — perhaps using Lightning micropayments — but that part is only a hypothesis. I don't know yet whether those economics work.

A lot of this is unproven.

I haven't built the distributed network.

I haven't solved reputation, spam resistance, conflict resolution, or payment-on-use.

And I certainly don't want to imply that I've discovered something nobody else has considered. I may simply not have found the people already working on it.

But the gap itself looks real enough that I think it's worth putting the idea into the open.

So this is the question:

**Is anyone working on an open, provenance-preserving historical identity and corporate-actions layer for securities?**

And if not, is anyone interested in talking about what it would take to build one?

I'm especially interested in hearing from people who work in securities reference data, corporate actions, market-data infrastructure, financial open source, identifier systems, or distributed data.

I'd much rather discover the fatal flaw in conversation than spend two years carefully building it.

Document: TGI — Atomic Reference Data: A Proposal
Version: v0.1.1
Date: 2026-08-30
Licence: CC BY 4.0
HTTP: https://totalgrowthinvesting.com/proposals/
TGI_Atomic_Reference_Data_Proposal_v0.1.1.pdf
IPFS CID (.pdf): QmQkRdvY4JWCZQhYTiaBZUcTmHvnBpoXcHL1Awck1azSXv
SHA-256 (.md): 7c23ab6b6703785d4f6d9d960fd523410b8f1e1bd87be4705748827a1094d261
SHA-256 (.pdf): 7d7c0629c7c6291036c9f19359cedf301cf221f6a18b9ba3390f02e7f58cc7c2