Administration Memo on reccovery.gov Falls Short

Asset Backed Securities Experience Holds Structured Information Lessons

On Dec. 22, 2004, eight months before I arrived at the U.S. Securities and Exchange Commission, the Commission unanimously adopted a rule formalizing staff guidance on asset backed securities, including the residential mortgage backed securities at the heart of today’s credit crunch. As published in the Federal Register on Jan. 7, 2005, the rule required “a new principles-based set of disclosure items” for asset backed securities. It continued:

We believe it would be impractical to provide an exhaustive list of disclosure items required for each asset class. Not only do we believe this approach would be impractical due to the many existing asset classes that are securitized today, it would not provide any effective guidance with respect to new asset classes that may be securitized in the future. Due to the dynamic nature of the ABS market, any such list would likely become outdated quickly. (Page 1531.)

It’s not surprising that standard disclosures seemed impractical to the SEC given what it knew at the time. Today, however, financial experts know XBRL can support non-standard disclosure, including the even more dynamic disclosures made by public companies. They also know that its extensibility helps investors by making non-standard information immediately obvious. Alas, XBRL was only being tested by the Commission in 2004 and had yet to be adopted by the FDIC for higher-level reporting.

The SEC proposed its ABS rule in May 2004. ABS debt was a fraction of its eventual size, yet the Commission took great pains to protect unsophisticated investors from perceived risks associated with its established policy of facilitating principles-based ABS disclosure. (One wonders if limiting the availability of ABS to certain investors inadvertently curtailed market scrutiny that would have revealed defects sooner.)

So-called “sophisticated” investors bought ABS, not retail investors. Once the market ramped up to its 2007 highs, a few short sellers paid to gather, structure, and analyze ABS information. The rest of the market soon made similar discoveries, but those who failed to sell or sell short suffered the grave consequences of delayed transparency.

The SEC’s 2004 conversion from staff guidance to formal rules for asset backed securities was justified in part because it helped “streamline” the securitization process, consistent with the Secondary Mortgage Market Enhancement Act of 1984 cited in the rule. Congress wanted mortgage securitization to grow and the rule was designed to be consistent with Congressional intent.

Simultaneously with development of the 2004 rule, the Commission was ramping up SOX interpretation, implementation, and enforcement. Highly structured U.S. GAAP disclosures were the gold standard and the focus of attention. Big public companies, not penny-stock-sized asset backed securities, were the priority.

Meanwhile, Wall Street investment houses, which previously built healthy revenue streams by understanding things about public companies that others did not, salivated over revenue streams from now relatively more opaque asset backed securities. (Investment houses or securities firms are often called investment “banks,” but what they did with ABS was far from traditional depository “banking.”)

While the SEC protected retail investors in the first instance, every regulator failed to contemplate the systemic risk to retail investors from the exploding popularity of ABS among “sophisticated” investors.

So what does this have to do with recovery.gov? Everything — particularly when you consider today’s headlines about the exploding popularity of issuing trillions of dollars of new Treasury backed securities.

On Feb. 18, 2009, Office of Management and Budget Director Peter Orszag issued a 62-page memo to every department and agency head covered by the stimulus bill outlining their responsibilities to recovery.gov in light of the federal government’s remarkable new debt creation.

As it implements recovery.gov, the Obama administration faces the same choice the SEC faced in 2004 regarding data disclosure on the old EDGAR system: Shall it provide for structured or unstructured data disclosure on recovery.gov?

While the SEC’s decision to let an up-and-coming asset class be disclosed in unstructured format can be explained in hindsight, experience to date would make an identical decision by the Obama administration to disclose the use of billions of dollars in taxpayer money in unstructured format inexcusable. Government spending today is growing at the speed of ABS as of a few years ago. There’s no substitute for accurate, complete, real time disclosure of what that government spending is being used to accomplish.

Unfortunately, preliminary indications from the Orszag memo are not promising. The memo purports to establish “a common framework for agencies to manage the risks associated with implementing Recovery Act requirements.” The government gets to protect itself with cross-agency standardization, but little hope is apparent for taxpayers.

The memo treats recovery.gov and agency Web sites as separate entities (1.1, bullet 2). This is a pure Web 1.0 approach.

Notwithstanding the popularity of creating a Web site to cure every ill in DC, the Internet is not merely a collection of Web pages. The Internet is the Internet. It is a system of data. And the reason Web 2.0 is better than Web 1.0 is that Web 2.0 recognizes this fact.

To be useful in a 21st century world, data must be capable of spanning sites and be made available in formats that make it easy to find and use. The memo’s resort to standardized URL’s — requiring each agency to add a “/recovery” folder to its base URL — is very 20th century.

The memo includes plenty of bureaucratese, but little call for the detailed transparency taxpayers deserve. If you can’t see the quality of mortgages in an asset backed security, what’s the use of the aggregate disclosure? If you can’t see the quality of the use of taxpayer money in a government program, what’s the use of aggregate disclosure? Without detailed data disclosure, trust in government could collapse as quickly this year as trust in ABS collapsed last year.

Publicizing contracts isn’t a new mission for the federal government. One of my first tasks in Washington, starting in 1987, was to translate contract announcements faxed to Capitol Hill from bureaucratese to English to determine how they affected our constituents. Recovery.gov should do more than outsource that work to Web surfers.

There’s a little good news on page 54, where the memo suggests agencies should use Atom 1.0 or RSS. But if those standards are merely used to feed things like TAFSs and CFDAs (don’t ask), they’ll be of little use to taxpayers or journalists.

Page 55 leaves less room for hope. There’s a pretty table suggesting a few “data elements” for agencies to report — but two of the four columns have exactly the same data in every single cell! Why create a column if it doesn’t differentiate a row? Two other tables are similarly silly. If this is how recovery.gov starts from the top, it’s worrisome to contemplate how it might end from the bottom.

What would be valuable would be a government-wide framework, in plain English, that can be used to classify every single expenditure, each of which in turn would require a concise plain English description of its purpose certified by the person responsible for approving the expenditure.

That framework, and the data within it, are what should be posted to the Web — not a hotch-pot of agency pages or unstructured information on recovery.gov.

Since it’s only spending, not financial reporting or ABS composition, this task is considerably easier than was the SEC’s task to make public company reporting comparable. It’s even easier than the task hedge funds assigned themselves to make ABS data comparable. The big question: Is it so easy that even OMB can do it?