Revisiting the Net Benefits of Freddie Mac and Fannie Mae
Freddie Mac and Fannie Mae Contributions to National Goals
The GSEs’ securitization and portfolio activities make the terms of mortgages more advanta- geous for borrowers. Although their effect on in- terest rates receives the most attention, the GSEs enhance other elements of mortgage contracts as well. They also improve access to mortgages with fixed interest rates and low downpayments. We discuss each of those issues in turn, and then discuss the GSEs’ contributions to macro- economic stability and review the benefits associated with homeownership.
Reducing Mortgage Interest Rates
That the GSEs reduce mortgage interest rates has been well documented in research on the difference between rates on loans above the conforming loan limit (from which the GSEs are excluded) and those just below the limit (which the GSEs can and do purchase). This is called the jumbo-conforming spread. The earliest study on this topic estimated the spread to be about 30 basis points (that is, 0.3 percentage point) in 1986.39 This study was based on loans originated by savings and loans in California. Cotterman and Pearce (1996) extended these results to lenders of all types in California and 11 other large states for the period 1989 to 1993. They found that the
spread varied widely over the period and concluded that the “core range” was 25-40 basis points.
The spread was in the high end of the Cotterman and Pearce range in 1989-91, when resolutions of insolvent thrifts may have disturbed the equi- librium in the mortgage market. The numerous studies conducted since then have estimated the spread at 30 basis points or somewhat lower.40 Unpublished and subsequent studies have gener- ally found the spread in the late 1990s and early 2000s to be 20-30 basis points.41 Exceptions include Woodward (2004), who, using a private data set, estimates the spread to be 42 basis points, and Passmore, Sherland, and Burgess (2005), who conclude that the spread is 16 basis points.
Interpretation of the spread is more complicated than its estimation. There are issues on at least three levels.42 First, the analyses may not hold constant all the respects in which conforming and jumbo loans differ. If jumbo loans are riskier on average than conforming loans, the spread will overstate the extent to which GSEs reduce rates on conforming loans. Among the factors not held constant in a typical study are differences in prepayments, house price volatility, and credit
Hendershott and Shilling (1989).
Studies published before 2002 are summarized in McKenzie (2002).
See Pearce (2000), Flannery and Lockhart (2005), McManus and Ramagopal (2005) and Roll (2005).
For discussion, see Pearce and Miller (2001) and McManus and Ramagopal (2005).
The Freddie Mac and Fannie Mae Contributions to National Goals