This will be the first year that financial institutions will be required to submit to stress tests under the Dodd-Frank Wall Street Reform and Consumer Protection Act.  Previous stress tests, dating back to 2010, were required and conducted by the Treasury Department.  Like other financial companies with consolidated assets of more than $10 billion and which are regulated by a primary federal regulatory agency, Freddie Mac and Fannie Mae (the GSEs) will be required to take the tests to determine if they have the capital necessary to absorb losses as a result of adverse economic conditions.   

In addition to the Dodd-Frank requirements, the GSEs' regulator, the Federal Housing Finance Agency (FHFA) has continued its practice of working with the companies to develop forward-looking financial projections across three possible house price paths.  These are referred to as the FHFA scenarios and the GSEs, as in the past, have been required to conduct them in conjunction with the other stress tests.  This, however, will be the last year for this FHFA requirement.  

The projections from its recent stress test scenarios were published today by FHFA.  The agency said these are not expected outcomes but rather modeled projections in response to "what if" assumptions about GSE operations, loan performance, macroeconomic and financial market conditions, and house prices.  The key assumptions provided for the Dodd-Frank Act Stress Tests (DFAST) and the FHFA Scenarios are different and the GSEs used their respective internal models to project financial results based on the FHFA assumptions.  FHFA said that while the exercise achieves a degree of comparability between the GSEs it does not eliminate differences in their respective models, accounting differences, or management actions.

Results are all given relative to the eventual impact on the draws required by the GSEs against the outstanding balance of their Senior Preferred Stock Purchase Agreements (PSPAs) with the U.S. Treasury.  As of September 30, 2013 the GSEs have drawn a combined total of $187.5 billion from the Treasury and have a remaining commitment under the agreement of $258.1 billion.

These are the most severe DFAST scenarios and the three FHFA Scenarios.

House price paths influence projections of credit expenses through the mark-to-market loan-to-value (LTV) rations of guaranteed mortgages.  This impacts the probabilities of default, projections of loss through defaults and loss severities.  Assumptions about the prices of securities in retained portfolios affect mark-to-market losses and assumptions about the growth of the retained portfolios, and credit guarantee books influence projects of revenue.  The instantaneous default of the largest counterparty affects mark-to-market losses.

Under the most adverse DFAST test the GSEs would require increased draws from Treasury ranging between $84.3 billion and $190.0 billion depending on the treatment of deferred tax assets.  There would be  remaining funding available under the PSPAs of $173.7 billion without re-establishing valuation allowances of deferred tax assets or $68.0 billion assuming both GSEs reestablished valuation allowances.

Under all three FHFA Scenarios the cumulative Treasury draws remained unchanged at the current levels of $187.5 billion and available funds would remain at $258.1 billion.  Neither GSE would require any additional draws.  Under all three scenarios the GSEs would continue to pay senior preferred dividends to the U.S. Treasury which would range from $54.0 billion in Scenario 1 to $36.3 billion in Scenario 3.

Severe assumptions contribute to the higher losses under DFAST than under even the most stringent of the FHFA Scenarios (Scenario 3).  First there is a significantly more pessimistic house price path used by DFAST.  The decline in value of non-agency securities is also substantially higher and the DFAST scenario includes the default of a large counterparty not envisioned in the FHFA scenario.
House prices have been the major driver of GSE credit losses and while there are a wide range of future house price paths at the national and local levels, given the high levels of uncertainty about overall economic conditions in general and the U.S. housing markets in particular, FHFA directed the GSEs to project financial results using Moody's current baseline and Moody's "Second Recession" path as the downside alternative and Moody's "Stronger Near-term Rebound" as the upside alternative.  The DFAST scenario is significantly more pessimistic than any of the three FHFA alternatives.
Comparison of house price paths, page 9

FHFA says its annual projections of the GSE's performance under its scenarios has improved each year they have been published starting in October 2010 as exemplified by each year's projected cumulative Treasury draw.  In the first projection drawn from FHFA Scenario 3 in 2010 the starting draw was $148 billion and the incremental draw was anticipated at an additional $215 billion for a projected cumulative draw of $363 billion.  In 2011 the beginning and incremental figures were $169 billion and $142 billion for a projection of $311 billion and in 2012 the starting draw was $187 billion with only an additional increment of $22 billion expected.

The improvement has been driven by three factors.  First, the GSEs' exposure to single family credit has been reduced by improving credit quality.  Second, the house price paths have become increasingly less pessimistic, and third, in each year the actual price path has been much better than the most several path projected.



Going forward the FHFA Stress Tests will be replaced by the Dodd-Frank Stress Tests