U.S. home prices climbed to new highs in 2021, primarily due to low interest rates, supply-demand imbalances, and pandemic effects, such as remote work and school and the desire for more living space. Certain regions saw substantial home price appreciation (HPA), with cities such as Austin, Boise, and Phoenix experiencing annual price gains of roughly 30%. Although some price-driving factors reverted to pre-pandemic norms in 2021, population dynamics continued as 30 of the 50 U.S. states experienced net migration-related population changes, according to the 2021 Atlas Van Lines Migration Patterns Study. While housing-related macroeconomic and lifestyle factors continue to support HPA, income growth has generally failed to keep up. As a result of these factors and their impact on the affordability metric S&P Global Ratings considers in its over/undervaluation analysis, we now view U.S. home prices as approximately 15% overvalued in aggregate. On a regional level, the proportion of overvalued metropolitan statistical areas or divisions (which we broadly refer to as MSAs) has increased to 88%, up from 50% as of our last assessment (see "The Shift From Under To Overvalued: What It Means For U.S. RMBS," published Oct. 29, 2021).

We view home prices as overvalued or undervalued based on how much the ratio of price to income (PTI) for a specified region (e.g., MSA or state) is above or below its long-term (15-year) average. Our regional inputs are the Federal Housing Finance Agency (FHFA) price index (which increased roughly 9% in the second half of 2021 and is the driving force behind the broad increase in overvaluation) and income per capita.

On April 1, 2022, we updated our over/undervaluation measures as well as the FHFA index values using fourth-quarter 2021 data. We use these values as they relate to certain U.S. residential mortgage-backed securities (RMBS) in our loan evaluation and estimate of loss system (LEVELS) model (see "LEVELS Model For U.S. Residential Mortgage Loans," published Aug. 5, 2019), which provides loan- and pool-level calculations of default likelihood (foreclosure frequency), loss given default (loss severity), and loss coverage. Depending on the pool diversification and location of the underlying mortgaged properties, overvaluation means our loss severity assumptions will tend to be higher (because a greater correction in home prices could occur under adverse scenarios). However, since our property value index now reflects fourth-quarter 2021 FHFA data, the indexed valuation will generally be higher, depending on when the valuation was performed and given the HPA between second- and fourth-quarter 2021.

Housing Is Roughly 15% Overvalued Nationwide

While we believe housing is about 15% overvalued across the U.S. (up from our prior assessment of approximately 5%), there is a substantial degree of regional variation, and a skew toward overvaluation in terms of the number of states and MSAs and the extent of the overvaluation (see charts 1 and 4). However, the recent mortgage rate increases should soften HPA (according to the March 31 Freddie Mac survey, the average 30-year fixed-rate mortgage rate is 4.67%--a year over year increase of about 150 basis points). Inflationary effects could also nudge income growth upward and correspondingly tighten the PTI ratio.

Chart 1

The Driving Forces

The state-level overvaluation shown in chart 1 is part of a pattern we have been observing over the past several years, which, in many cases, results from inter-state population dynamics. Indeed, some of the increases in local HPA have resulted from the flux of people leaving high-cost areas and settling in regions where home prices are more affordable. This effect is amplified by the increase in remote-work options, which have allowed some homeowners to sell their properties and migrate to more affordable areas without switching jobs, resulting in reduced monthly mortgage obligations and, in some cases, larger living spaces. In addition to the upward price pressure caused by out-of-state buyers, population growth across the U.S. continues to drive demand for housing.

It is likely that pandemic-related migration has already peaked, and segments of the population may be returning to densely populated areas. Such a trend-reversal could create renewed demand in urban areas while increasing home inventory in sparsely populated ones. In any case, remote work and geographic dispersion have yet to reach a new steady-state, and whether there will be any corresponding secular shift remains unclear. In our 2022 housing and mortgage outlook report, we highlighted this correlation between population migration and HPA (see "U.S. Housing And Mortgage Outlook 2022," published Jan. 27, 2022). In our analysis for this report, we also looked at HPA levels for the top 10 and bottom 10 states against annual population change (data from the U.S. Census Bureau) (see chart 2), and annual HPA by state (plus Washington D.C.) and our corresponding over/undervaluation assessment (see chart 3).

Chart 2

Chart 3

The Most Over/Undervalued MSAs

Certain MSAs have much higher overvaluations than the 15% national average. Chart 4 shows the 10 most overvalued and undervalued MSAs, and the distribution of over/undervaluation for 399 MSAs (based on our latest assessment) for use in evaluating mortgage pools and corresponding U.S. RMBS. Previously, overvalued and undervalued MSAs across the country were almost evenly split by count. Now, there is a strong skew toward overvaluation, with some states having multiple overvalued MSAs. For instance, Idaho continued to experience high immigration levels in 2021 and contains three of the top 10 overvalued MSAs, with the Boise region having the highest overvaluation level, at almost 70%. Arizona and Florida also have multiple overvalued MSAs in the top 10.

Chart 4

The Impact On Residential Mortgage Pools

Our over/undervaluation measure provides information about affordability in terms of deviations from a long-term average. Assuming the long-term average represents a steady state, overvaluation could influence how much property prices decline under certain scenarios. To account for this when rating certain U.S. RMBS, our assessment of over/undervaluation applies to our market value decline (MVD) assumptions when calculating the loss severity on a loan, based on our criteria (see "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018).

Under a 'AAA' rating stress, we assume that (for a given region) 50% of the overvaluation amount of a mortgaged property will factor into the MVD, with a corresponding value of 20% at a 'B' rating level. At the national level (assuming a 15% overvaluation), our 'AAA' MVD assumption is now approximately 53%, up from approximately 48% (when we last assessed national overvaluation at approximately 5%). This assumes the additional decline in a property's value under a 'AAA' rating stress would reduce the property's liquidation proceeds by approximately 5% and, therefore, increase the loss severity assumed for a given loan.

The significant rise in HPA has resulted in both an increase in our general overvaluation assessment and an update to the FHFA index we use for properties with valuation dates before fourth-quarter 2021. When indexing property values, we apply 50% of cumulative upward movements and 100% of cumulative downward movements, based on our criteria. Higher property values resulting from this indexation could reduce the probability of defaults and temper loss severities. Overall, we believe the credit impact of high HPA and overvaluation levels on U.S. RMBS will depend on the geographic distribution of the mortgage pools, as well as the valuation dates of the properties backing the loans in the pools.