Radar Perene / Archive / science

Who audits the auditor: the problem of an index no one regulates from outside

◦ Index methodology v2.2 (working papers with DOI). See the methodology.

Science

A bank has a supervisor. A fund has a regulator, an administrator, an independent audit. A listed company answers to a swarm of instances. Now consider an independent index provider — someone who calculates and publishes rulers over the market without selling any financial product: who, exactly, verifies that those rulers measure what they claim to measure? The honest answer is uncomfortable: in general, no one.

Independent index providers operate in a supervision vacuum: too small for the regulatory frameworks designed after the global benchmark scandals, and outside the supervisors' reach because they neither manage third-party money nor distribute products. What fills that vacuum — when anything does — is voluntary transparency: the decision to build, on one's own account, the accountability no rule demands.

The vacuum is not a hypothesis — it is the trade's normal state

The world learned to regulate large benchmarks the expensive way: after global reference rates proved manipulable, entire jurisdictions created frameworks for administrators of systemic indices. But those rules aim at the top — the indices on which contracts in the trillions depend. Below that line lives a population of private indicators, proprietary scores and market "thermometers" no framework reaches, published by analysis shops, consultancies and independent researchers.

For the reader, the practical problem is asymmetry: the index arrives ready, with a confident name and a well-drawn curve, and there is no external instance to ask whether the methodology was tested, whether previous versions were preserved, or whether errors were admitted. The seal that exists for funds and banks does not exist here. Each provider is its own auditor — which usually means no auditor at all.

What the house did with this problem

This house lives exactly in that vacuum — it publishes indices, sells no financial product, has no dedicated regulator — and the answer it built was to design for itself a voluntary auditability framework: a set of verifiable commitments standing in for the supervision that does not exist. The complete internal design is a working document, but its effects are public and checkable: studies deposited with DOIs and preserved versions, revisions admitted in the body of the papers themselves, data with integrity verification, and findings routinely passing through robustness tests before any publication.

The organizing principle is simple to state: everything an external supervisor would ask, the house tries to answer first, in writing and in public. Not because anyone demands it — precisely because no one does. The absence of external accountability is the argument for internal accountability, not the waiver of it.

Why voluntary transparency, and not an industry seal

The obvious alternative would be private certification — an "audited index" seal issued by some industry body. The house is skeptical of that path for the classic reason: seals paid for by the evaluated party inherit the conflict they meant to resolve. Voluntary transparency has a property the seal lacks: it transfers the audit to the reader. A DOI with preserved versions does not ask for trust in the house or in a certifier — it asks to be checked.

That does not make it sufficient. Voluntary transparency depends on the publisher's willingness, and disappears exactly where it would be most needed: among providers who prefer not to show. It works less as a system-wide solution and more as a screening criterion for the reader — whoever opens versions, data and errors is accepting a cost that whoever does not open has avoided. The size of that cost, and what it reveals, is the subject of the next article in the trail.

It is also an argument with a built-in deadline: voluntary commitments hold only while the record confirms them — which is why each one was designed to leave a public, dated trace, not a statement of intent. The reader years from now will be able to check whether the practice kept pace with the promise; that future exposure is deliberate.

Frequently asked questions

Are indices in Brazil not regulated?

The major exchange indices and reference rates have their own institutional frames, and administrators of systemic benchmarks answer to frameworks created over the last decade in several jurisdictions. The vacuum described here belongs to the proprietary indicators published by independent houses — below those frameworks' reach and outside the perimeter of supervisors of third-party money.

Isn't self-auditing a conflict of interest by definition?

It is — which is why the house's design bets on commitments verifiable from outside (DOIs, preserved versions, hashed data), not on self-declarations. The statement "we audit ourselves" is worth little; the public record any reader can check is worth more.

What can a reader check in an index with no regulator?

The existence of a declared methodology, of preserved previous versions, of errors admitted in writing and of verifiable data. The simultaneous absence of all four is sufficient information about the fifth item, which is the seriousness of the whole.

Does the house argue that regulation should reach small providers?

The article describes the vacuum and the answer the house built for itself; the optimal regulatory design for the sector is an open debate, with costs and benefits documented in the literature, and stays outside this piece's scope.

---

Continue the trail: Voluntary transparency has a cost: what the house decided to open and what it decided to keep closed →

House reading: today's reading is in the Diário; the archived episodes, in the Atlas.

The auditability framework the house designed for itself is a possible conversation with other index providers — at the bench, not in an article.