MethodologyContact usSupportLogin
Fastmarkets launched the Fastmarkets Repair & Remodeling Index in early 2023 to meet a need in the market: timely and accurate data on repair and remodeling (R&R) activity. The surge in home renovation activity during 2020-21 coupled with skyrocketing inflation made an already-opaque market that much more difficult to gauge.
There are other ways to measure R&R activity in the US that existed previously. Two come from the Harvard Joint Center for Housing Studies (JCHS). One is the quarterly Leading Indicator of Remodeling Activity (LIRA), which includes historical data benchmarked using the American Housing Survey (AHS), modeled estimates for more recent history and a forecast for the near term. The LIRA captures improvement spending by homeowners.
The other datapoint that can be used to gauge R&R activity is an annual estimate of total spending on improvements and repairs by both renters and owners. This more complete spending estimate is released every two years by the JCHS in its Improving America’s Housing Report. The AHS itself can be useful, but it is also only released every other year. Additionally, the US Census publishes a Monthly Construction Spending survey, which includes private spending on residential improvements.
The Fastmarkets RRI is created using the principal components analysis technique to reduce three key variables to a single monthly index indicating the real volume of R&R activity undertaken.
The three variables are inflation-adjusted retail sales at building materials dealers, aggregate hours worked by residential remodelers and Google searches for terms related to do-it-yourself (DIY) home improvements and replacements. The sources of this data are the US Census, the Bureau of Labor Statistics (BLS) and Google Trends, respectively.
While the RRI is created from different sources of data than the previously mentioned ways to measure R&R activity, the correlation over time is generally high (0.9 or above). A more detailed explanation of how the RRI was created can be found in the report we published when we launched the index.
Shortly after we launched the RRI, the Harvard Joint Center for Housing studies re-benchmarked the LIRA using American Housing Survey Data. This benchmarking process in 2023 retroactively changed the LIRA in 2020-21 to show a general trend very similar to what our RRI had indicated, which was encouraging.
Now two years later, the LIRA has once again been updated with the latest American Housing Survey data. This time around, the Fastmarkets RRI and the LIRA told different stories about trends in R&R activity in 2022-23.
We have recently revised and improved the construction of the RRI to more fully capture the DIY side of the R&R market. The number of search terms included in the Google Trends portion of the index was increased from 26 to 48 and a handful of search terms were excluded.
The AHS and JCHS Improving America’s Housing Reports each break down R&R projects by type, and other than disaster repairs, all DIY project categories are accounted for in the new and improved Google Trends components of the RRI. Google Trends search terms are grouped into categories, with between one and five unique terms corresponding to each DIY project type. The average relative shares of total inflation-adjusted DIY spending based on the latest Improving America’s Housing Report and historical AHS releases were then used to weight the Google Trends data.
This is roughly the same process that we used to weight the Google Trends data in the original construction of the index, but now the number of project types included is more complete. With a new survey out, the sample used for weighting is also larger.
The Google Trends data is meant to capture interest in DIY R&R projects, so we also added a much more comprehensive set of filters to the data download to screen out, as much as possible, searches for contractors and companies to complete home projects. Still, the Google Trends data is not a perfect indicator, and in many cases the line between professional and DIY is blurry. The principal components analysis process should control for some of this noise in the data by revealing the common thread between the three key components.
The difference in the RRI before and after this change is minimal until the last three years. This improvement to the construction of the RRI lowered the year-to-date growth in the RRI for 2025 by about half. The new reading showing a 2.1% increase from the first half of 2024 to the first half of 2025 is more believable based on where wood products markets have been. Prior to this revision, the RRI was showing growth of over 5% in the first five months of the year.
It is worth exploring what the various indicators of R&R activity say about the sector over the past few years. We can then compare what each would imply for wood products consumption with industry shipments, trade data and Fastmarkets proprietary data to get a sense of which is the most trustworthy. Figure 2 shows the Fastmarkets RRI, spending on residential improvements from the US Census’s monthly Construction Spending Survey, and the LIRA from the Harvard JCHS. Monthly spending numbers and the LIRA are adjusted for inflation. For the sake of comparison, all three indicators are shown indexed to 100 in 2019.
With the construction spending data, the LIRA and the RRI all indexed to 2019, it becomes even apparent that the construction spending data is not only more volatile than the RRI and the LIRA, but that it shows a massive shift in the level of R&R activity over the last three years. Additionally, both the construction spending data and the LIRA show a large jump in R&R in 2022 that does not appear in the RRI.
This raises a few questions. Has R&R activity since 2022 been 40-50% higher than it was in 2019, as indicated by the construction spending data? Or, as the LIRA indicates, has it been 15-20% higher than it was in 2019, following a surge in 2022 that puts the 2020 spike in activity to shame? Or is R&R activity closer to what the RRI readings tell us, about 6% higher than it was in 2019 and still below the 2021 peak?
R&R accounts for a significant share of wood products demand. The exact proportion varies by product type and by year, depending on the relative strength of different end-use markets. Over the last five full years (2020-24), R&R has accounted for 42% of domestic US softwood lumber consumption, 27% of domestic structural panel consumption and 35% of domestic nonstructural panel consumption. The percentages are similar for the preceding five-year period (2015-19). With R&R demand for wood products being such a large proportion, we can use what we know about total wood products consumption to triangulate which method of measuring R&R activity is most accurate.
One way to validate our modelled wood products demand numbers is to compare them to apparent consumption (production plus imports minus exports), which is a more concrete number based on industry statistics and trade data from official sources. Over time, domestic demand and apparent consumption will be equal, with the only difference between them being inventory fluctuations. For the sake of demonstration, the chart below shows that US lumber demand and apparent consumption have generally not differed much from each other in any given year. However, if the R&R portion of lumber demand had trended with either of the other two indicators of R&R activity since 2020, overall demand would have been several billion board feet higher than apparent consumption for several consecutive years, which is all but impossible.
The pattern in the above lumber analysis holds true for structural and nonstructural panels as well. Across lumber, structural and nonstructural panels, R&R end-use demand estimates using the LIRA’s trend in place of our current estimates lead to total consumption figures 7-10% higher than apparent consumption in each of the last three years. The delta between apparent consumption and implied demand according to the construction spending census data is even larger.
Product prices also contain a lot of embedded information, and as long as we understand the supply side of the market, we can also use movements in price to get a sense of what is happening with demand. The year that stands out as having the largest difference in trend between the RRI and other market indicators was 2022: both the LIRA and construction spending data from the census showed a 17% gain in R&R, while the RRI showed a 3% decline. In that same year, inflation-adjusted building materials sales declined more than 7%, which is enough to view the alleged 17% increase in R&R with some skepticism. Additionally, the implied growth from a 17% increase in that one sector would have been enough to move total consumption up significantly for all wood products. Not only was apparent consumption flat or down, but 2022 was the year in which prices corrected across the board. The fall in wood products prices from the start to the end of 2022 was of historic proportions. Based on how inflated prices were at the start of the year, some price correction was inevitable, but the rapid and severe price collapse fits much better on a backdrop of declining R&R than on one of historic expansion.
As a reminder, this past month we fortified the RRI by expanding the scope of the Google Trends component of the index and improving the precision of the weighting of the various included search terms. A sudden high degree of interest in a small subset of home renovation project types was skewing the index upward beginning in January 2025. By broadening the scope of the terms that constitute the Google Trends component of the RRI, we have made the RRI less prone to biases that arise from short-lived or temporary swings in a small group of search terms. We have taken a proactive approach to improve the reliability of the RRI. This ensures the RRI remains an up-to-date indicator of R&R activity and a key element in understanding demand for wood products.
Capturing trends in the R&R space is absolutely crucial for understanding wood products demand. This retrospective highlights the difficulty in doing so, especially in a timely manner. The monthly data released by the US Census Bureau is unreliable in part because it is collected in a survey not designed for that purpose. The data released by the JCHS is more reliable but available far less frequently. The Fastmarkets RRI and these two alternative indicators of R&R activity each tell a different story about the past five years. Hard data on wood products production and trade strongly indicates that the Fastmarkets RRI is the most accurate way to gauge the total volume of R&R activity in the US, at least as it pertains to wood products demand. This is the case whether we are focused on lumber, structural panels or nonstructural panels. With the additional benefit of being updated monthly, the RRI is a must-have for understanding the full picture of these markets.
For an expanded look into the recent revision to the RRI as well as other insights, read the full analysis on the Fastmarkets Dashboard here. No account? Speak to one of our experts to get started.