May-24: Can Higher Mortgage Rates Cause Rent Inflation? (Section IV, Semi-Elasticities)

28 May 2024

 

Can Higher Mortgage Rates Cause Rent Inflation?​​ 

(Section IV,​​ Semi-Elasticities)

 

  • Introduction

In Sep-21 mortgage rates began increasing due to an increased demand to buy a home. Then in response to rising inflation, the Federal Reserve raised the federal funds rate by 25 bps on March 17, 2022. This took the benchmark rate from 25 bps to 50 bps. The federal funds rate currently is 5.50%. In June 2022, it began reducing its balance sheet gradually (known as quantitative tightening, or QT) by not reinvesting all the proceeds of maturing securities. Higher demand to own a home and this series of actions raised mortgage rates from 2.84% in Aug-21 to​​ roughly​​ 7.00% by the end of May-24.​​ A semi-elasticity is a numerical value​​ of​​ the​​ impact mortgage rates​​ have​​ had on the​​ year-over-year​​ (YOY)​​ rent appreciation for single family properties (Δ​​ RRA / Δ mortgage rate)​​ where RRA = %Δ rents. Table 1.a shows the estimated semi-elasticities at different​​ time​​ lags for 20 CBSAs.

In Table 1.a, out of 20 CBSAs, the highlight cells in red indicate which markets had a positive, and statistically significant, rent appreciation after​​ mortgage rates​​ changed at different lags,​​ in the time period from Jan-17 to Mar-24.​​ The cells in yellow indicate which CBSAs experienced a​​ negative relationship between appreciation​​ and mortgage rates​​ at each respective lag.​​ We see in Table 1.a, higher mortgage rates​​ 24 months​​ earlier would cause​​ ​​ rent appreciation​​ to accelerate​​ in 12 of the CBSAs. At a lag of 30 months, the rate of rent appreciation accelerated​​ following higher mortgage rates​​ in 15​​ of the CBSAs.​​ Only New York, NY had a negative response to mortgage rates.

 

  • Semi-Elasticities:

Let​​ year-over-year​​ rent appreciation, RRA = %Δ rent ​​ = (Δ rent / rent). We are interested in how RRA speeds up or slows down due a change in the level of the mortgage rate, or Δ RRA for a given Δ in mortgage_rate. The speed (or magnitude) of how rent appreciation changes ​​ ΔRRA / Δ mortgage_rate is called the semi-elasticity.​​ At a lag of 12​​ months the​​ average​​ semi-elasticity for​​ Chicago, IL​​ is​​ about​​ 0.60​​ in Table 1.a.​​ 

 

We can piece-together​​ a calculation of a​​ semi-elasticity over​​ one​​ single year for Chicago, IL in Table 1.b​​ to gain an insight into the calculation. Over the 12 months ending 3/1/2023 the 30-year mortgage rate increased by 237 bps​​ (column 2). This 237 bps change in 3/1/2023 impacts renters rent in 3/1/2024.​​ At a lag of 12 months, the 237 bps higher mortgage rate is followed by rent appreciation jumping by 280 bps​​ (column 4).​​ For that one year,​​ the semi-elasticity​​ in Chicago, IL would be​​ 1.18.​​ 

A screen shot of a graph

Description automatically generated

In this example for Chicago, IL, the very large jump ​​ in mortgage rates (237) would have been followed 12 months later by the very large jump in rent appreciation (280 bps).​​ 

 

Having said that, (1) why should mortgage rates impact rents?​​ (2) why with a lag of any length? (3) are markets efficient across time periods?​​ 

 

Renters do not take out a mortgage. They should not care about what happens to ​​ mortgage rates.​​ However, as mentioned in​​ an​​ earlier blog,​​ Dias and Duarte (2019)​​ and Haidorfer (2024)​​ demonstrate that a contractionary monetary policy raises rent. They conjecture that monetary policy affects rent-versus-own housing tenure decisions. As the cost of homeownership rises, renters on the margin substitute away from purchasing, pushing up rents.​​ There would then be a positive relationship between mortgage rates and rent appreciation.​​ 

 

In regard to​​ the second issue,​​ most renters sign rent contracts that limit their ability to move in the current time period. If renters cannot do “anything” for a year,​​ at minimum,​​ without breaking the rent contract and paying a penalty,​​ this precludes changes in​​ mortgage rates​​ leading​​ to​​ changes in​​ RRA in the initial 12 months.​​ It could certainly be the case that landlords​​ are impacted by HPA and might break the rental contract and raise rents, but I do not know this apriori.

 

In regards to the third question, the evidence in Table 1.a indicates​​ three​​ things (1) in the first six months​​ following ​​ a mortgage ​​ rate​​ change, rent appreciation (RRA3bd) generally​​ moved inversely with mortgages (the results show that 12 CBSAs had a statistically significant negative relationship); ​​ (2) following​​ the 12th​​ month​​ after a mortgage rate change,​​ the statistical relationship between mortgage rates and the percent change in rents​​ -​​ the semi-elasticities​​ -​​ become statistically positive in a growing number of CBSAs; (3)​​ as rental markets adjust​​ following​​ changing mortgage rates 30 months earlier,​​ 15 CBSAs respond​​ positively​​ to higher mortgage rate.​​ 

 

  • Why different semi-elasticities?

In addition to the co-factors specified in Equation 1 below, what​​ might​​ distinguish the CBSAs​​ in which rental appreciation rates moves versus those that did not? Chart 1 below shows the semi-elasticities of RRA3bd​​ for a lag of 24 months​​ for​​ those 12 CBSAs with statistically significant semi-elasticities from​​ the 30-year FRM​​ on the y-axis.​​ The​​ credit score is on the x-axis.

A chart with numbers and a line

Description automatically generated with medium confidence​​ 

It appears that following an increase in mortgage rates 24 months earlier, rent appreciation jumps in those CBSAs with lower average credit scores by more than those CBSAs with high scores. Renters in less affluent CBSAs have less market power and are not able to buy, switch neighborhoods, or switch CBSAs. ​​ 

 

  • Methodology:

We measure the impact of mortgage rates twelve months earlier on rent appreciation​​ as:​​ 

 

RRA3bd = ​​ α ​​ + β1​​ L3RRA3bd + β2​​ L3HPA3bd​​ 

 ​​ ​​ ​​​​ + β3​​ L12RVP_3bd + β4​​ L12mortgage_rate​​ 

 ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​​​   ​​ ​​ ​​​​ + β5​​ employment + β6​​ llordc_shr​​ 

 ​​ ​​ ​​​​ + β7​​ L12vacc + β8​​ dum_cov + ε  ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​ ​​​​ (1)

Here:​​ 

  • HPA3bd is home price appreciation..

  • RVP_3bd is the monthly rental rate versus the price of buying.

  • Mortgage_rate is the monthly average 30-year fixed rate mortgage, from Freddie Mac.​​ 

  • Employment is the year-over-year change each month in payroll employment, from Bureau of Labor Statistics.

  • Llordc_shr is the share of properties that are rental properties in a city, from Department of Census.

  • Vacc is the rental property vacancy rate, from Department of Census.

  • Dum_cov is value of 1 from 2020-03-01 to 2022-12-01 and 0 otherwise.

Table 1.a​​ shows the estimated β4​​ coefficient for each of 20 cities​​ (the estimated semi-elasticities)​​ using Equation​​ 1​​ at​​ different months of lag​​ on the mortgage rate.​​ 

 

Bibliography

Dias, Daniel A. and João B. Duarte (2019) “Monetary Policy, Housing Rents and Inflation Dynamics”. International Finance Discussion Papers 1248.

Haidorfer, Anton (2024) “The Dynamic Impact of Monetary Policy Over Short Horizons on Local Rental Markets”, Submitted to the AREUEA 2014 June conference.

 

 

2