Economic Growth Likely to Accelerate – Rising Interest Rates a Risk, But Housing Expected to Remain Resilient
We continue to expect an acceleration in real gross domestic product (GDP) growth and employment in the coming months as vaccinations progress further, warm weather arrives, and the recently passed stimulus bill takes effect. Home sales are likely to be minimally affected by rising mortgage rates to date, though a further jump in rates is a risk. Revisions to our macroeconomic forecast this month were modest. Our 2021 GDP forecast on a Q4/Q4 basis is now 6.6 percent (down from a prior 6.7 percent), and our 2022 forecast moved to 3.0 percent from 2.8 percent. While changes to the topline were minimal, recent developments led to a series of partially offsetting dynamics that changed the underlying details of our forecast. Incoming data showed the consumer recovery in January was stronger than we expected, and some early indicators suggest further acceleration in early March. Partially countering this, however, were disruptions due to colder February weather across much of the country and related power outages in Texas, as well as semiconductor shortages hurting manufacturing output. Furthermore, the timing of some expenditures from the latest stimulus bill will likely occur later than we previously assumed, and the sharp rise in interest rates in recent weeks will likely dampen growth somewhat. Together, compared to last month’s forecast, we expect a modestly stronger consumer recovery this year, but a more drawn-out expansion of government expenditures and a modestly slower pace of private investment spending.
With a higher forecast for mortgage rates, we lowered modestly our forecast for home sales this year. We now expect total sales to be 6.2 percent higher in 2021 than in 2020, down from 6.9 percent in our prior forecast. The change was minimal due to our belief that the dominant factors determining home sales at the moment are largely COVID-related time displacement of homebuyers’ moving decisions and supply-side constraints on the availability of listings. We believe an ample number of homebuyers can absorb modestly higher mortgage rates in the near-term and, as such, an upwardly drifting rate will have only a minimal impact on the sales outlook. In contrast, the downward revision to our mortgage originations forecast was proportionately larger as refinancing activity is highly rate sensitive. We changed our total originations forecast for 2021 to $3.9 trillion from a prior $4.1 trillion, and our 2022 forecast to $2.9 trillion from $3.2 trillion.
Risks Remain as The Economy Opens Up Further
In recent weeks, the rollout of vaccines and COVID-19 case and hospitalization metrics continued to improve. However, the possibility of a more resistant virus variant emerging cannot be ruled out; the future path of the virus remains a near-term risk. The primary uncertainty over the next six months, however, is to what extent social distancing-restricted activities recover. If consumers and firms are reluctant, the strong growth we are expecting in the second and third quarters may not materialize. Alternatively, given the high level of built-up household savings that many consumers now have at their disposal, it is possible that growth in coming quarters exceeds our already robust expectations. Household checking accounts alone surged in Q4 to $3.2 trillion, about $2 trillion above the pre-COVID baseline. Combined with the next round of stimulus, consumers have the potential to drive growth even faster than our already somewhat-above-consensus expectation.
With the recent stimulus bill passed, attention will increasingly be turned to the next set of legislative activities. Uncertainties include the potential passing of additional infrastructure and tax bills later this year. Depending on the timing and composition, these on net could affect growth in either a positive or negative direction.
Lastly, we do not think rising interest rates to date, which are still historically low, are a major concern. Modest further increases in line with our forecast would likely be a reflection of a healthy, recovering economy. However, we continue to see as a major risk the possibility of stronger acceleration in inflation expectations leading to a more pronounced interest rate rise. While year-over-year inflation was subdued over the past decade, the Consumer Price Index exceeded 4 percent through much of the mid 2000s and corresponded with 30-year mortgage rates in the 5- to 6-percent range. Given the Fed’s intention to allow inflation to exceed its 2.0-percent target, and the unprecedented levels of expansionary fiscal policy, mortgage rates growing closer to these levels is possible if inflationary expectations become “unanchored” and move permanently upward. This could occur due to near-term growth acceleration exceeding our current expectations, or if the production capacity of the economy is revealed to have been significantly damaged by COVID-related restructuring. While the absolute level is still modest, market-based measures of inflation expectations, such as the 5-year TIPS/Treasury spread, have moved up considerably in recent months, indicating growing investor apprehension over rising prices in coming years.
Cold Weather Was Likely a Speed Bump, though Recovery Inflection Point Is Near
Much of February’s economic and housing data will likely be marked by the disruptions caused by unseasonably cold weather and the power outages in Texas. Early indicators suggested some slump in economic activity. Auto sales fell by 5.6 percent, refinery output and oil production fell, and mortgage application data declined sharply in the affected weeks. Our analysis of state-level mortgage applications and for-sale home listings suggested that the drop off in activity was correlated to regions with sharp declines in temperature; therefore, the slowdown was not isolated to Texas. We expect upcoming data releases of home sales and construction starts to show temporary weakness, and our forecast for first quarter consumer spending implies a modest pullback in February prior to resuming growth in March.
That said, our view of the economy beginning to strongly accelerate in the spring remains intact. Coinciding with further state-level relaxing of distancing measures, payroll employment growth accelerated in February to 379,000, with job gains concentrated heavily in the battered hospitality and food service industries. If not for the weather, we believe the payroll gains would have likely been larger. We expect gains to be in the 500,000 range in March and further acceleration thereafter.
Consumers seem eager to return to restricted activities following vaccinations and with warmer weather approaching. Internet search behavior related to airlines has jumped in recent weeks and searches regarding restaurant topics have also moved up, as well as an indicator of seating reservations. This coincides with a recent pick-up in gasoline demand and a jump of 6.2 points in the preliminary March reading of the University of Michigan Consumer Sentiment Index. Together with the next round of stimulus checks landing in household bank accounts, we believe economic growth will soon accelerate sharply.
Our Thinking on Interest Rates
Perhaps the most noteworthy development over the last month was the rapid rise in long-run interest rates. The 10-year Treasury rate as of this writing is 1.63 percent, up from 1.09 percent at the start of February. While the level remains modest compared to pre-COVID years (the 10-year Treasury rate averaged 2.32 percent from 2011 to 2019), the rate of increase has been rapid. We have noted in recent months our belief that a larger increase in interest rates and inflationary expectations are significant risks. Even if inflation expectations remain subdued, given our forecast for nominal GDP growth, it is entirely plausible that the 10-year Treasury rate could reach the 2.5 to 3.0 percent range by the end of 2022.
Despite this risk, our baseline view is that the recent rapid rise will not continue but that rates will drift only modestly higher over the remainder of this year. Essentially, we believe the Fed will keep policy accommodative for longer, not tightening until inflation clearly exceeds its 2.0-percent target for a substantial period. This view is consistent with current market measures, such as Fed Funds futures, not anticipating any rate hikes until 2023 and, even then, at a slow pace. If the Fed’s response to labor market tightening and accelerating inflation is slower than in the past, then long-run rates should remain low relative to what would have been expected in past cycles, under the expectation that it will take longer for the Fed to move the short-run rate higher. If inflationary expectations do not permanently move much higher, pushing up the long-run rate, this regime could be maintained well past our forecast horizon.
Also, U.S. interest rates do not operate within a vacuum. The world’s other major central banks appear to be on a similar page, which should limit the upside in U.S. yields. Some, notably the Bank of Japan and the Reserve Bank of Australia, have a policy of “Yield Curve Control” where they are explicitly targeting longer-run rates, keeping them suppressed. The European Central Bank, while not engaging in such an explicit policy, recently announced its intention to increase purchases of longer-dated bonds in response to the recent run-up in rates. Lower yields on foreign sovereign bonds act as a limiting factor on U.S. Treasury yields, as portfolio managers look for substitutes in search of higher returns. At least in the short-run, the slower pace of vaccine administration within the E.U. suggests that their recovery will be slower than in the U.S., putting comparative downward pressure on E.U. rates. The German 10-year Bund, for example, only moved up 14 basis points in the last month and is currently at negative 0.31 percent.
Compared to a given rise in the 10-year Treasury, we expect the 30-year fixed mortgage rate to move up somewhat less in the short-run. The roughly 55 basis point increase in the 10-year Treasury since the beginning of February, corresponded with about a 30-basis point increase in the mortgage rate. Over the past year, a surge in originations led lenders to build out operating capacity. Therefore, in the short run we expect originators to absorb some of the increase in funding costs to maintain production volumes. This is supported by Fannie Mae’s Q1 Mortgage Lender Sentiment Survey® (MLSS) which showed that the share of mortgage lenders expecting their profit margins to decline rose for the second straight quarter, with an even greater share expecting profit margins to fall further in the coming months. While the primary spread (30-year fixed mortgage rate minus 10-year Treasury) has already compressed significantly, we believe there is a modest amount of room still available in the near-term. We expect future drifts upward in the mortgage rate to therefore be somewhat less than movements in the 10-year Treasury rate.
If Further Rate Increases Remain Modest, Effects on Housing Should Be Minimal
We continue to forecast a slowdown in home sales over the course of 2021 (though the yearly total will likely be higher than 2020). However, we do not expect this deceleration to be primarily rate driven, but rather due to waning timing effects of homebuyers’ delaying or moving forward purchases due to COVID-19 and an extremely tight supply of homes for sale limiting transactions. The rise in mortgage rates over the past month, and our upwardly revised rate forecast, led to only a slight decline in our sales outlook.
However, if rates move up more aggressively than our baseline forecast, it’s useful to examine the last period of significantly increasing mortgage rates as a comparison. In 2018, the 30-year mortgage rate rose a little more than 100 basis points in a 14-month period. On a quarterly basis, total home sales fell by about 8 percent peak-to-trough, despite employment and incomes continuing to expand. If something similar were to occur over the next year, there are reasons to believe that the drag on sales would be considerably smaller:
- The “lock in” effect will be weaker – When mortgage rates rose in late 2018 to 4.9 percent, the highest level since 2011, most potential repeat buyers likely purchased homes or refinanced their notes at rates lower than the going market rate, creating a disincentive to move to a new home. In contrast, even if mortgage rates rise to about 3.5 percent going forward, that would still be lower than any period prior to this past year. Even at 3.7 to 4.0 percent the rate would be lower than during most of the last decade. Most homebuyers do not move within a year or two of purchasing, and given the fixed costs of refinancing, few people intent on moving within the next year refinance. Thus, we expect the lock-in effect to be comparatively muted for the next couple of years if rates do not move beyond roughly 4 percent.
- Rates are still historically low – Even with the rapid appreciation in home prices this past year, mortgage payments remained comparatively affordable relative to typical incomes. Using a 30-year fixed mortgage at 90 percent loan-to-value (LTV) ratio, the ratio of principal and interest payments on a median priced home relative to median family income is currently well below the 2018 peak. To reach a similar point today, the 30-year rate would have to be about 3.9 percent.
- More homebuyers likely able to absorb higher payments – Household checking accounts alone in Q4 2020 were about $2 trillion above end of 2019 levels. Credit card balances were also paid down by $118 billion over the past year. While there are distributional aspects to these figures, credit/debit card spending data suggest that a heightened saving rate was common this past year among a broad swath of potential buyers. In addition, given a large majority of home purchases are by couples, the typical homebuyer will have received at least $6,400 in stimulus checks in about a year’s time (thousands more if the family has children). The combination of paying down other debts and building up savings suggests the potential buyer pool has lower back-end debt to income (DTI) ratios, stronger credit scores, and a greater ability to make larger down payments compared to 2018.
- Sales are currently limited by inventory – The limited supply of homes for sale is likely holding back transactions. Many would-be buyers are likely unable to find a suitable listing or are being outbid. Even if some people drop out of the market due to rising rates, we think there is an ample “reserve” of buyers able to fill their place in the near-term. Home price appreciation would likely soften as fewer bidding wars occurred, but the effect on transactions would be limited. Additionally, homebuilders are currently struggling to keep up with demand, suggesting they would continue a brisk construction pace even if foot traffic cooled.
Taking these factors into account, and given the uncertainty over the future path of interest rates, one of our alternative forecast scenarios this month was for mortgage rates (relative to our baseline) to rise an additional 50 basis points by the end of 2021 and 85 basis points by the end of 2022, but assuming an otherwise materially similar macroeconomic environment. This translated to a 30-year fixed mortgage rate of 3.7 percent and 4.3 percent, respectively. This is a range more typical of the pre-COVID period previously discussed.
The result was only a modest reduction to the pace of home sales relative to our baseline of about 1.0 to 2.0 percent in 2021. Declines in 2022 were somewhat larger at 4.0 to 5.0 percent (again relative to baseline), as some of these previously mentioned factors diminish, but the softening is still modest compared to the past. That is, given the other COVID- and macroeconomic-related dynamics in play over the next couple of years, we do not believe the primary uncertainties regarding the level of home sales are around the future mortgage rate path, at least not within this moderate range. A greater future interest rate rise would create more uncertainty around the general economic environment, but in our view the risk that rising rates will have a major direct impact on our home sales forecast is whether the upward movement meaningfully exceeds 4.0 percent (e.g., the 2018 peak of about 4.9 percent).
Taking into account the macroeconomic drivers associated with a potentially faster rate increase, if the movement in rates is due to stronger realized or anticipated growth, then rising incomes and employment levels would likely help counteract the drag. In contrast, if rates rise because of a sudden shift in inflationary or monetary policy expectations, such as the taper tantrum in 2013, then the effect on demand could be comparatively greater.
Refinance Mortgage Originations Fall on Rate Increases
Consistent with our view of home sales only being modestly affected by the recent rate rise, our forecast for purchase mortgage originations was little changed from last month. We forecast $1.8 trillion in purchase mortgage volumes for 2021, representing 13 percent growth from 2020’s volume. This growth is driven by our forecast for higher full-year total home sales in 2021 compared to 2020, as well as continued price appreciation.
In contrast to purchase activity, refinance originations are far more rate sensitive. Due to rising rates and the upward revision to our baseline outlook, we made a corresponding change to our refinance originations forecast. We now expect 2021 volumes to be $2.1 trillion, a 6 percent downward revision from our February forecast. Application activity suggests refinance volumes will stay elevated in the first half of 2021 before retreating over the second half as the mortgage rate is now projected to rise faster than we previously thought. The forecast for 2022 refinance volume was also revised downward by about $240 billion to $1.1 trillion.
While higher rates led to a downward revision, even at the current 3.1 percent mortgage rate, we estimate that 48 percent of all outstanding mortgage balances have at least a half percentage point incentive to refinance. This is still a large pool of potential borrowers and is consistent with continued refinancing strength in the near term. However, in the case where rates rise further, refinance activity would significantly decline. Returning to our alternate scenario where the 30-year fixed rate is 3.7 percent by year-end 2021 and 4.3 percent by the end of 2022, we would expect refinance originations to be about 14 percent lower in 2021 relative to our baseline forecast, and about 44 percent lower in 2022.
Multifamily Starts Revised Up on Suburban Market Strength
We continue to expect multifamily housing construction to be comparatively soft relative to single-family, but recent developments led us to meaningfully upgrade our multifamily starts forecast. We now expect starts in 2021 to be 392,000, up from a previous forecast of 339,000, and starts in 2022 to be 387,000, up from 364,000 previously. Multifamily housing permits surged in January, jumping 28.1 percent to the highest level since mid-2015. This series is notoriously volatile, and we expect that some of this activity reflects developers pulling forward planned projects in anticipation of rising interest rates. We thus predict a near-term pullback but believe that investor and developer optimism has grown recently. New construction activity within the urban cores of large, expensive metro areas will likely still be subdued going forward, but interest in many suburban markets appears to be stronger than we anticipated. Rent metrics show high single-digit year-over-year increases in many of these suburban markets, even as some urban cores experience annual declines. The former appears to be inducing sufficient activity to outweigh much of the drag from the latter. Furthermore, we understand that weakness in office and retail markets is causing a shift in focus among many commercial real estate developers toward multifamily housing.
For more on multifamily market conditions please see the March 2021 Multifamily Market Commentary.
Economic & Strategic Research (ESR) Group
March 12, 2021
For a snapshot of macroeconomic and housing data between the monthly forecasts, please read ESR’s Economic and Housing Weekly Notes.
Data sources for charts: The Federal Reserve Board, Bureau of Economic Analysis, Freddie Mac, Moody’s Analytics, National Association of REALTORS®, Federal Housing Finance Agency, Fannie Mae ESR Group.
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.
ESR Macroeconomic Forecast Team
- Doug Duncan, SVP and Chief Economist
- Mark Palim, VP and Deputy Chief Economist
- Eric Brescia, Economist
- Nick Embrey, Economist
- Rebecca Meeker, Financial Economist
- Richard Goyette, Business Analyst