(continuing from the front page …) On the other hand, landlords have more pricing power. However, new landlords (aka. property investors) in 2023 are confronted by the fact that property prices around them have significantly increased and at current rents (rents that existed before their purchase of Zthe property) that their existing rates of return have fallen. Thus, there is an incentive for landlords to raise rents in the period after home prices have risen and maybe lower rents if home prices fall. This introduces five questions: 1) does rent appreciation (depreciation) mimic home price appreciation (depreciation) in the months following home price changes, if so, 2) what is the maximum magnitude; 3) what is the lag; 4) is the relationship identical across cities; 5) does a fall in home price appreciation lead to rent appreciation slowing 12 months later or is more going on and this impact owner’s equivalent rent (OER) and the consumer price index (CPI)?
Section 2: Do rents (and OER) follow home prices?
A renting household can either rent a single-family detached property or an apartment unit in a multi-unit building. Here, we discuss (1) median rents charged each month for three-bedroom single family residential properties and (2) median rents charges on apartments in multi-unit buildings. Rent data by construction consists of rents that were newly contracted that month combined with all the rents which were contracted the prior 11 months. Let rental rate appreciation be defined as RRA3bd = %ΔR where R is the median rent in given city for a three-bedroom property and RRAapt = %ΔR where R is the median rent for an apartment in a multi-unit structure (of any bedroom size) in a given city. Also let home price appreciation be defined as HPA = %ΔP where P is the median price of a three-bedroom home. My rent data (R) is a rolling average of all prior month’s contract rate whereas the sales price data (P) reflects sales that transpired that prior month (a spot rate).
Table 1 indicates that rents are highly correlated with home prices. This initially looks strange because the HPA reflects the spot rate at time of purchase. Renters generally sign a contract with landlords. One would expect that rent changes might initially have little to do with home price changes. This is the idea of sticky rents. Table 1, however, suggests that rents are not sticky.
Data from the U.S. Census Bureau’s Property Owners and Managers Survey in 1995 (single-family and multifamily units, “excluding data not reported or for rent-free units) showed that 44.4 percent of all units had annual leases, 4.0 percent had leases longer than one year, 36.1 percent had leases less than one year, and 15.5 percent had no leases.” Even if ½ of all tenants had an annual lease it is surprising that 3 months after home price rise, rents rise also.
We can think of across time or across CBSAs. Chart 1 shows the month-over-month increase in rents in 20 cities (RRA3bd, or core based statistical areas, CBSAs) for renting three-bedroom properties as of May-23. The chart points out that median rents have risen month-over-month in every CBSAs.
Chart 2 shows the month-over-month increase in rents in 20 cities (RRAapt, or core based statistical areas, CBSAs) for renting an apartment in a multi-unit building (all bedroom sizes) as of May-23. This chart points out that median rents have risen month-over-month in every CBSAs.
Both Charts 1 and 2 seem to indicate that RRA is following MOM HPA which has rebounded despite very high mortgage rates. But other forces impact rent appreciation.
One must ask, “why would rent appreciation fall or go negative just because home price appreciation has fallen or gone negative?” It is widely perceived that there is a shortage of homes for renting (vacancy rates are low) and a shortage of homes for buying (inventory-to-sales ratios are very low).
The point is: the cause of MOM RRA rising again is not obvious. Why might rent appreciation increase or decrease? There have been discussions that household formation has slowed (individuals are moving in together again to share the costs of renting). This seems like an unlikely explanation as to why markets have adjusted so quickly. It could also be a dead cat bouncing. That is landlords bumped up rents so quickly that renters now say, “I am leaving if you increase my rent” but now landlords are increasing the rents again. There is also the notion that new rental supply is coming online. Indeed, in cities where there is very little available land to build single family properties, building up with multi-unit structures makes sense. The construction of new rental buildings, however, takes time.
Section 3: Do rents (and OER) adjust to relative prices?
Table 1 above shows that RRA does follow HPA prices near concurrently, but other forces are impacting RRA3bd and RRAapt.
A renter essentially has three choices: (1) pay the-going rental rate (2) buy a home or (3) move back with mom and dad. We can compare the costs of the first two options. Here buying a house refers to the cost of owning for a month (principal repayment and mortgage interest) rather than the wider definition that also includes the costs of mandatory services and charges, regular maintenance and repairs, taxes, and utilities.
Chart 3 shows the two different costs (the cost of renting (R) and cost of owning for a month (B) as a ratio of rent over buying (RVB). If the two goods are substitutes, then a household should be indifferent between renting or owning (a three-bedroom property) back in Nov-22. What we see is that in every one of the 20 CBSAs in Chart 3, renting is cheaper than owning. The same would be true if we looked at the relative cost of renting an apartment. Through most of 2023, buying a home in almost every city in the nation is prohibitive to the median income family. There is little economic incentive to buying a home during Jan-23. Thus, renters are left with the single option of renting. Because the rents used to calculate RRA3bd are contract rates, the impact of RVB on RRA3bd can take a long time.
If it is cheaper to rent than to own a year ago, then rents will go up and home prices will go down. There is a mechanism built into markets which will push RVB closer to 1.0. Empirical studies show that after 12 months RRA3bd is also being driven by RVB being different than 1.0.
Section 4: Which force is stronger?
We initially talked about how rents might mimic home prices in Section1. Table 1 highlights the strong correlation between HPA and RRA. There is statistical evidence that HPA impacts RRA in the short run (3 to 6 months).
After accounting for this serial correlation and the fact that lower interest rates impact HPA with a nine-month lag, we find that RVB impacts RRA in the longer run (12 to 15 months). This suggests that YOY RRA will slow but not drop to the 2 to 3 percent range in the 2nd half of 2023.
Section 5: Impact of rent appreciation on CPI going forward:
How would this show up in CPI OER. Mortgage rates now (during Jul-23) are roughly 360 bps higher than at their recent minimums (Sep-21). Higher mortgage rates and higher home prices have made buying a home unaffordable to the average homeowner in almost every city in America. This shows up in Chart 3 with the denominator of RVB increasing and RVB falling to record lows below 1.0. Empirical evidence shows that after 12 to 15 months of low RVB, this in-balance between relative costs pushes RRA up (again this long lag is due to RRA3bd being measured using existing and new contracts).
Without even having to forecast RRA3bd, Chart 4 uses the information that increases in rental rates (RRA3bd) enter into the Federal Reserve’s CPI measure of owners’ equivalent rent with a lag of about 18 to 24 months because of the way renters are surveyed, to project OER. Thus, as RRA3bd did in early 2023, we might expect growth in OER to slow initially, but then stay in the 4 to 6 percent range through 2023. Here, p_lnRRA3bd is simply historical data pushed forward and serves as a projection of OER.
Notes:
A city is measured as a Core Based Statistical Area, or CBSA in this report.