SEC Proposal: Use Python to Enforce Wall Street Transparency

The U.S. Securities and Exchange Commission published a 667-page proposal that, among other things, outlined the implementation and use of a Python program that would annotate asset-backed securities, allowing potential investors to have periodic access to an ABS's pool asset performance, which include the interest rate, level of prepayments, defaults and losses-given-defaults.

The Waterfall computer program, as it's called in the proposal, would provide potential investors with detailed information about an ABS. In doing so, the program would effectively reduce (but not eliminate) an investor's dependency on third-party analysis by providing updated information typically available only in the prospectus.

"By running the waterfall computer program in combination with other internally-developed or commercially available vendor interest rate, prepayment, default and loss-given-default models, cash flow engines, or computational services, investors should be able to promptly run cash flow simulations and generate present value estimates for ABS tranches," the proposal states.

The proposal requires that an asset data file be created every time an ABS and final prospectus is registered. The asset data file is then uploaded to the Electronic Data-Gathering, Anaylysis and Retrieval (EDGAR) system, the SEC's public database. Potential investors or the curious public can download the file and view it within The Waterfall computer program. Investors also can input their own predictions regarding pool asset performance.

Credit card master trusts would also be required to use the Waterfall computer program anytime an unplanned event affecting investors occurred.

The SEC chose Python specifically because it is open source and available freely to anyone. Its status as an interpreted language is also in line with the SEC's rule to keep executable code off of EDGAR for security reasons.

The SEC is not opposed to collaboration between the Waterfall program and other investment software, as long as users disclose, via a separately filed exhibit, the functionality.