Financial analyst Marc Joffe is attempting to shake up Wall Street and the global financial markets with a new open-source methodology for rating sovereign debt called the Public Sector Credit Framework (PSCF). Joffe works for San Francisco based Public Sector Credit Solutions and his firm is hoping to recruit insurgent guerillas from the open source community to help perfect the simulation model.
Unsatisfied with the seemingly underhanded actions taken by major credit rating agencies who provide fixed income investors with analysis and guidance as they decide which debt instruments to buy or sell, Joffe says the problem stems from investors having "limited insight" into exactly how analysis is derived, so they are largely operating on faith.
NOTE: In terms of practice, major rating agencies often rate credits with proprietary, black box models; or in the case of government issuers, with no models at all.
Joffe asks, "Rating agencies played a major role in the 2008 collapse of the municipal bond insurance industry in the U.S. and are now key players in the European sovereign debt crisis — so are these agencies the best source of information when predicting something as critical as the next government default?"
"PSCF is hoped to enable analysts to perform multi-year fiscal simulations that go beyond traditional single-scenario analysis of debt-to-GDP ratios. Analysts can consider a variety of ratios (such as the interest expense-to-revenue ratios), and create budget projections that incorporate socioeconomic factors such as the impact of an aging population on social insurance costs," he said.
This approach supports more thorough analysis by making use of all the budget, economic, and demographic data and forecasts available from governments and private forecasters.
Will it work? Well, because PSCF is open source, this model is promised to be "completely transparent" and available free of charge. Unlike proprietary models used elsewhere, Joffe insists that there is no chance to fudge the numbers or quietly revise methodologies to cover up modeling errors.
