Introducing the Fannie Mae Refinance Application-Level Index, a Leading Indicator for Refinance and Prepayment Activity

Fannie Mae's new Refinance Application-Level Index (RALI) provides weekly insights into refinance application activity, enabling the industry to enhance modeling and analysis of refinances and prepayments.

Fannie Mae receives mortgage applications every day from more than 1,900 lenders supporting single-family borrowers across the country, providing comprehensive and timely insights into mortgage activity and trends. We are now bringing some of those insights to the market with our new Refinance Application-Level Index (RALI). This weekly leading indicator of refinance application activity sources data from Fannie Mae's Desktop Underwriter® (DU®), the industry's most widely used automated underwriting system (AUS). We believe these early insights into refinance activity will support broader market analysis, including lender forecasts of origination volumes and investor forecasts of mortgage-backed security (MBS) prepayment activity.

From 2020 to 2021, historically low interest rates resulted in record levels of mortgage refinance applications and originations and, in turn, mortgage prepayments. Fannie Mae estimates that borrowers refinanced $5 trillion of mortgage debt in 2020 and 2021, resulting in 16.9 million Fannie Mae and Freddie Mac loans prepaying during that same time. Forecasting refinance originations and prepayments also became more challenging over the last few years, as the observed relationship between refinance incentive and prepayments has evolved. These changes have been attributed to shifts in origination channels, lender mix, and product mix.

Lenders throughout the industry use DU in the underwriting process not only for conventional loans delivered to the GSEs, but also for government, non-conforming, and jumbo loans. Based on our analysis of the Home Mortgage Disclosure Act (HMDA) data, DU processed 4.6 million closed refinance loans, worth around $1.3 trillion in originations, in 2020 – more than any other AUS in the industry. DU's unmatched scale ensures that users of the RALI benefit from an index backed by significant volumes of refinance applications from a broad array of lenders with exposure to different mortgage products. Further, because refinance applications are submitted to DU at the beginning of the loan origination process prior to the existing loan paying off, the RALI provides market participants a four-to-six-week lead on refinance origination and prepayment activity. We believe that sharing this data with the wider mortgage industry will improve transparency in the market and facilitate improved origination and prepayment projections. These modeling improvements should, in turn, support continued market liquidity without the need for supplemental loan disclosures (or related services) that risk compromising borrower privacy.

How We Constructed the RALI¹

The RALI is constructed using standard industry methodology for mortgage application indices. Specifically, we:

• Filter DU records to isolate legitimate refinance applications² that could result in new refinance originations (and the associated prepayments of existing loans).
• Create weekly application measures by aggregating applications from Saturday to Friday.
• Set the first week of 2004 as the base period, giving it a value of 100. Subsequent weeks reflect proportional changes from the base period. We provide the historical series of the RALI, back to 2004, to enable users to perform time series analysis.

We offer two RALI measures to meet the needs of users:

• RALI ($): Based on dollar volume measured by unpaid principal balance (UPB); and
• RALI (count): Based on loan count.

These two measures allow users to track growth rates in application volume and loan count.

The RALI is not adjusted for seasonality or for holidays, although we anticipate our weekly commentaries will note when volume is impacted by a holiday. Any revisions to the RALI, which we expect to be minimal, will be made on an annual basis in the second week of each year, and users will be notified through our website to ensure transparency and clarity.

How Does the RALI Perform?

The RALI is a highly effective measure of refinance activity, correlated with changes in mortgage rates (Chart 1), origination activity (Chart 2), and MBS prepayments (Chart 3).

As illustrated in Chart 1, the RALI ($) has been responsive to both decreasing and increasing mortgage rate environments. When there is a decline in mortgage rates, we see a corresponding increase in the level of refinance applications and, as a result, the RALI. In 2022, the RALI ($) has declined by 63 percent between the week ending January 7 and the week ending June 10. While refinance volume has declined in 2022 relative to 2021 and 2020, the RALI ($) may fall even further if higher interest rates persist, consistent with our May 2022 forecast. The latest reading of RALI ($), for the week ending June 10, showed that refinance applications will need to fall another 50 percent to reach the prior trough registered in the fourth quarter of 2018.

Chart 1. RALI ($) Performance in Various Interest Rate Environments

We find that the RALI ($) is highly correlated with our own estimates of total market refinance originations. A simple linear regression of the quarterly RALI (with one-month lag) and industry refinance origination volumes, as estimated by Fannie Mae, between the first quarter of 2007 and the fourth quarter of 2021 (Chart 2) yields an R-squared of 90 percent.

Chart 2. Closely Tracks Refinance Origination Estimates

We believe the RALI ($) will be particularly insightful for MBS investors, providing a view into prepayment activity that may occur in the next four to six weeks. We observe strong historical correlations between prepayments and the RALI ($) lagged by one month (Chart 3). Over the period of January 2004 through December 2021, the monthly correlation between the RALI and GSE prepayment was 92 percent. During the recent refinance wave of 2020 and 2021, when significant prepayment activity took place, the correlation remained high at 86 percent.

Chart 3. RALI ($) vs. Prepayment Volumes Since 2004

We designed the RALI to be a valuable tool for lenders, investors, and other market participants in the mortgage industry. We believe the additional transparency RALI brings to refinance market activity will improve prepayment modeling performance and investor confidence, and it has the potential to bring improved liquidity to the MBS market – ultimately benefitting homeowners through lower borrowing costs.

Fannie Mae’s RALI will be published on a weekly basis here according to this calendar. To learn more, please access the RALI methodology and FAQs.

 The authors thank Sonja Beaubien, Matt Classick, Hamilton Fout, Yasaman Hekmat, Tian Liu, Aleksandrs Rozens, Nicholas Sapirie, Daniel Schoshinski, and Kevin Vineski for valuable contributions in the creation of this commentary and the design of the research. Of course, all errors and omissions remain the responsibility of the authors.

Opinions, analyses, estimates, forecasts, and other views of Fannie Mae's Economic & Strategic Research (ESR) Group included in these materials should not be construed as indicating Fannie Mae's business prospects or expected results, are based on a number of assumptions, and are subject to change without notice. How this information affects Fannie Mae will depend on many factors. Although the ESR Group bases its opinions, analyses, estimates, forecasts, and other views on information it considers reliable, it does not guarantee that the information provided in these materials is accurate, current, or suitable for any particular purpose. Changes in the assumptions or the information underlying these views could produce materially different results. The analyses, opinions, estimates, forecasts, and other views published by the ESR Group represent the views of that group as of the date indicated and do not necessarily represent the views of Fannie Mae or its management.

¹ For additional background material on the RALI's construction, please review our methodology documentation on the Fannie Mae's Research and Insights web page.

² Because the RALI includes withdrawn or denied applications, users should follow standard methodology by considering potential changes in the "pull-through" rate, or the percentage of applications that result in origination and prepayment, to arrive at an appropriate estimate of refinance and prepayment volumes.