‘Machine readable’ financial disclosures: a solution in search of a problem

Congress is considering legislation now on a fast track to require municipal issuers to file their disclosure statement with the Municipal Securities Rulemaking Board in “machine readable” format.

Unfortunately, the unique nature of the municipal market would make this plan expensive and likely unsuccessful. Benefits would accrue primarily to technology and data vendors who stand to profit from the initiative.

The National Defense Authorization Act for Fiscal Year 2023 (HR 7900), approved by the House on July 14, includes a provision that would effectively require the MSRB to establish standards for “machine-readable” issuer disclosure documents.

Presumably once established, the MSRB would specify that any disclosure filings made to the EMMA platform must meet the standards they set.

The municipal dealer community supports market transparency. The proposal before Congress, however, is poorly conceived, has few if any market supporters, and has the potential to be very expensive.

A machine-readable document is defined in law as data in “a format that can be easily processed by a computer without human intervention while ensuring no semantic meaning is lost.” In practice, this means financial disclosures in a format compliant with eXtensible Markup Language (XML), the general standard for machine-readable documents on the Internet.

For municipal disclosure documents, machine-readable would mean filings in something like eXtensible Business Reporting Language (XBRL), the data formatting standard for corporate issuers.

Machine-readable means data items in, say, an issuer’s CAFR would be “tagged” with pre-established, standardized labels. In theory, if the data labels in a document comply with the standard, an analyst would be able to quickly and easily pull data items from many different disclosure documents and compare financials across a broad group of issuers.

Unfortunately, implementation of a machine-readable standard for municipal disclosures would be full of problems.

A machine-readable disclosure scheme for municipal issuers would require a “taxonomy,” a collection of tags and definitions issuers would use in identifying data items in disclosure documents. In the corporate world, the developer of the XBRL taxonomy is the Financial Accounting Standards Board (FASB), the accounting standards setter for corporate issuers.

Under HR 7900, that role would be filled in the municipal market by the MSRB, raising the first concern about the proposal.

The MSRB has decades of experience in regulating municipal dealers and municipal advisors. They have also built an impressive data platform in EMMA. But the MSRB are not accountants. They don’t set accounting standards and have no expertise in financial disclosure other than hosting the documents.

Moreover, the municipal market does not even have a universal accounting standard.

We have the Governmental Accounting Standards Board’s published accounting standards. However, unlike in the corporate world, there is no federal requirement for issuers to follow GASB GAAP. Some states mandate compliance and some do not. Some states have set their own standards which may be only loosely related to GASB.

In short, unlike in the corporate market, there is no universal municipal accounting standard on which to base a taxonomy. 

In addition, Congress has not identified a need for machine-readable documents for municipal issuers. XBRL initially emerged among corporate issuers in 2008.

At the time, artificial intelligence and machine learning technologies were in their infancy, and there was no practical way for an analyst to automate the process of pulling data items from disclosure documents. Today it is possible to train software to identify and sort data items directly from searchable PDF disclosure documents, making machine readability unnecessary.

Also, the municipal market is dominated by thousands of relatively small, unsophisticated issuers. Compliance for these communities will be expensive and time-consuming. Local government accounting policy is generally set in state laws.

If the taxonomy for machine-readable disclosure documents does not comport with some states’ accounting policies, will those states have to change their laws?

This legislation would also represent a significant chink in the armor of the Tower Amendment.

Why should dealers care about this when issuers are the ones who will have to make these filings? Because the standards-setter under HR 7900 would be the MSRB. We all know that more than 90% of the MSRB’s revenue is derived from fees imposed on dealers. Any money the MSRB spends on this project would come primarily from the pockets of municipal securities dealers. And if FASB’s experience with taxonomy in the corporate world is any indication, this function will require an entire new team at the MSRB to develop and manage. It will not be cheap.

Perhaps most importantly, nobody in the market is asking for this. From what we at BDA see, the only ones pushing for this in Congress are data and technology vendors who stand to profit once it is implemented.

We do not see a call for this from investors — those whom this is presumably designed to benefit — or anybody else active in the municipal market. The lack of machine-readable disclosure documents is not holding back the market in any way.

HR 7900 is “must pass” legislation since its main provisions involve Defense Department spending. It has already passed the House, and Senate leadership have pledged that they will take up the legislation after the November midterms.

When they do, we should all urge our senators to abandon the machine readability provision applicable to municipal issuers. It is an expensive and needless proposal.