Why Measures of Existing Home Inventory appear Different
There is a significant difference between measures of inventory, and this is leading to some confusion.
The National Association of REALTORS® (NAR) reported this week that inventory was unchanged year-over-year:
Total housing inventory registered at the end of August was 1,280,000 units, a decrease of 1.5% from July and unchanged from the previous year.
And here is a graph from Redfin weekly data showing active inventory is at 2021 levels:
Here is a graph from Realtor.com showing inventory in August was 26.6% above 2021 levels and almost back to 2020 levels.
And on a weekly basis, Realtor.com reported inventory was up 28% year-over-year.
Active inventory continued to grow, and improved 28% above one year ago.
Altos Research also reported that inventory was up 26.5% from a year ago. Here is a graph of the inventory change vs 2021, 2020 and 2019. The blue line is the year-over-year data, the red line is compared to two years ago, and dashed purple is compared to 2019.
This shows inventory is up 26.5% year-over-year, and just down 3.7% from 2020 levels.
And my local market survey showed active inventory was up 25.0% year-over-year in August (I try to just use active inventory, but some sites probably include some pending sales).
Understanding the Differences
First, it appears the NAR is including some pending sales in their inventory total. Housing economist Tom Lawler noted this:
As I’ve noted before, the inventory measure in most publicly-released local realtor/MLS reports excludes listings with pending contracts, but that is not the case for many of the reports sent to the NAR (referred to as the “NAR Report!”), Since the middle of [Spring 2020] inventory measures excluding pending listings have fallen much more sharply than inventory measures including such listings, and this latter inventory measure understates the decline in the effective inventory of homes for sale over the last several months.
Here is a graph comparing the year-over-year change in Realtor.com’s active inventory, the NAR’s inventory, and Realtor.com inventory including pending sales. Note that the blue line (NAR) and dashed black line (Realtor including pending sales) track.
As Lawler noted, including pending sales understated the decline in active inventory in 2020 and 2021, and is now understating the increase in active inventory.
But what about Redfin? Redfin takes a very different approach. Their active inventory number for any month includes homes that came on the market and sold quickly during the month, whereas the other measures are a snapshot at the end of the month (or week). For August, Redfin is saying the total number of active listings (including homes that sold quickly during the month) was unchanged from August 2021, whereas - according to Realtor.com, Altos, and my local surveys - there are about 27% more active listings on the market at the end of August 2022 than in August 2021. Two different approaches to reporting the data.
Altos CEO Mike Simonsen also mentions immediate sales each week, but they aren’t included in active inventory at the end of the week.
Although tracking the level of immediate sales is important, I’ll be sticking to the Altos / Realtor.com approach to reporting year-over-year active inventory (snapshot at the end of the period).
CalculatedRisk Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.