Academic journal article AEI Paper & Studies

Jumbo Rates Are below Conforming Rates: When Did This Happen and Why?

Academic journal article AEI Paper & Studies

Jumbo Rates Are below Conforming Rates: When Did This Happen and Why?

Article excerpt


The government-sponsored enterprises Fannie Mae and Freddie Mac (the GSEs) are the dominant institutions in the US residential mortgage market. An important question is the extent to which their presence has led to lower mortgage rates for borrowers. A specific measure of this benefit is the jumbo-conforming spread, the difference in mortgage rates between larger, jumbo loans, ineligible for purchase by the GSEs, and conforming loans that are eligible for GSE purchase.

Pre-crisis estimates of the jumbo-conforming spread, utilizing a variety of methodologies, ranged from 10 to 25 basis points (Passmore, Sherlund, and Burgess, 2005; Sherlund, 2008; Kaufman, 2014; An and Yao, 2016; DeFusco and Paciorek, 2017). Similar results were obtained from both survey data and loan-level regressions that controlled for credit quality and other factors that influence mortgage rates. Some of these loan-level estimates were obtained with data from the Monthly Interest Rate Survey (MIRS), which was recently discontinued but previously had been conducted by the Federal Housing Finance Agency and its predecessor, the Federal Housing Finance Board. More recent evidence has come from loan-level data found in proprietary datasets with more comprehensive market coverage and a broader array of loan characteristics.

Before the financial crisis, the government guarantee for the GSEs was only implicit. In 2008, Fannie Mae and Freddie Mac were placed into conservatorship, and the GSEs were backstopped by the US Treasury. The implicit backing of the US government became more explicit, although uncertainty regarding the future state of the GSEs remains. Given this change, one might have expected the jumbo-conforming spread to increase. However, in the post-crisis market, it has in fact decreased, and as discussed below, has been negative since 2013.

This study re-examines the jumbo-conforming spread in light of these changes, utilizing loan-level data over 2000-2017 from MIRS and from Black Knight and CoreLogic. The latter data sets allow us to identify not only whether the loan amount is above or below the applicable conforming loan limit, but also whether the loans below the applicable limit were actually purchased by the GSEs. We can also observe whether the jumbo loans are placed in securities or held in portfolio. Hence, we are able to estimate not only the conforming versus jumbo rate differential as in prior studies, but also to compare the rate differential between conforming loans actually purchased by the GSEs and jumbo portfolio loans, which comprise nearly the entire post-crisis jumbo market.

With the estimates of these spreads in hand, the next step of the analysis is to explain their dynamics. Four hypotheses are tested. First, did the increase in GSE guarantee fees (g-fees) lead directly to a decline in the spread? Pre-crisis g-fees tended to average about 20 basis points. Now, they average closer to 60 basis points. The analysis tests whether the timing of the increase in g-fees coincides with the decline in the spread.

Second, did the increase in bank demand for jumbo loans lead to a reduction in the spread? The banking system as a whole has had abundant, low-cost deposits in the post-crisis environment, and many banks may have used these deposits to fund the retention of mortgages on their balance sheets. Many banks also have sought wealth management business from affluent customers, which may have increased their willingness to hold jumbo loans. The analysis tests whether strong deposit growth has influenced the spread.

Third, did the change from a national conforming loan limit to market-specific limits result in a decline in the spread? In the years immediately before the crisis, the national conforming loan limit was $417,000 for 1-unit properties, with higher limits only for Alaska, Hawaii, the U.S. Virgin Islands, and Guam. As a result of a sharp reduction in the availability of jumbo credit during the crisis, Congress expanded the conforming limit to as high as 150 percent of the national limit in designated "high-cost" areas. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.