It looks like a single number, but a lumber price actually is a calculation, the sum of a mental formula based on economic data, history, forecasts, intuitions, and X factors. But when the inputs become erratic, it gets harder to produce a good number. That’s why it’s more important than ever to understand housing and economic numbers and the forecasts that spring from them. Here are six factors that should make you look twice at data before you act.
1) Economic Analysts Often Are More Confident Than Accurate
The Federal Reserve Bank of Philadelphia regularly asks economists at big banks and corporations to predict whether a certain indicator will land in a certain bucket—for instance, whether the gross domestic product (GDP) will grow or fall by 1%, 2%, etc. A study by the University of California’s Haas school of business found that of 16,559 forecasts submitted, 53% of the economists turning in their numbers were confident they were correct. They were right only 23% of the time. Clearly, just because a big-time economist gives a prediction doesn’t mean they can read the crystal ball much better than you.
There’s a different kind gap—one featuring machine vs. man—in a report called GDPNow. This forecasting model from the Atlanta Federal Reserve takes all the elements that are used to calculate GDP growth and, as they come in, updates the formula with the recently arrived numbers. The GDPNow number will bounce around a bit as the data gets added, but over time it has done a great job showing weeks in advance what the GDP ultimately will be.
GDPNow has also won fans in recent years because it has consistently outperformed the forecasts by a Blue Chip set of economists. In fact, economists have performed worse than usual for a couple years, partly for reasons we’ll get into below.
2) Missing Data and Spotty Data
Former Defense Secretary Donald Rumsfeld famously talked about the “known knowns,” “known unknowns,” and “unknown unknowns” that bedevil planners. All three challenge the quality of data and data analysis.
The “known knowns” include reports we’ve relied on for years, such as housing starts figures and the unemployment report. But the federal government shutdown stopped many of those reports, forcing people as late as December to use housing data no later than from August. Those delays led Federal Reserve Chairman Jerome Powell to lament that, when it comes to using data to guide policies, “We’re sailing by the stars on a cloudy night.”
(You also can bet the Fed isn’t taking verbatim analyses written by Artificial Intelligence programs. There are too many examples of AI hallucinations to make those reports trustworthy on their face.)
Even when there’s no shutdown, challenges collecting the data have increased the number of subsequent revisions. That can lead to unpleasant surprises, such as the big change in the jobs report that angered President Trump so much he replaced the head of the Bureau of Labor Statistics.
There are lots of calls but no agreement on how to improve the federal government’s data-collection methods. Until those changes happen—if ever—your best bet is to not bet the farm on one monthly report, but rather wait and watch to see if a trend takes hold. You won’t be as nimble, but your odds of being right increase.
As well, keep time factors in mind when you look at the different ways the monthly housing figures are reported. Projects under construction are most relevant to lumber buyers because that involves current demand. A bit further down the pike is the data for projects that have just started. Even further into the future is permits data. Watch the typical number of weeks between pulling a permit and starting work; there are indications it’s getting longer.
3) Mood Swings
Another “known known” encompasses business sentiment polls like the NAHB/Wells Fargo Housing Market Index. The problem here is that such polls often overstate the prevailing mood. When you compare the Market Index with changes in housing starts, you’ll see that when starts tick downward, the level of pessimism drops much deeper than the actual drop in starts. Likewise, when starts go up, builders turn bubbly.
Readers of consumer sentiment polls run the risk of reaching big conclusions out of relatively small changes. A New York Federal Reserve survey that asks Americans to describe their economic situation a year from now found virtually the same percentage of people in January and November 2025 saying they expected their situation to be “about the same”—42.5% in January, 42.2% in November. What did change was a 10-percentage-point switch to pessimism from optimism. That 10-point change, not the plurality of unchanged prospects, is getting the attention.
4) Your Local Conditions Will Vary
A lumberyard owner in Maine once told this reporter, “Up here, we don’t have housing starts. We have housing start.” Meanwhile, in the Dallas-Fort Worth Metroplex, builders sell nearly 44,000 homes a year. Two markets, two entirely different needs to care about a data point.
Getting local goes deeper than state level; ehat’s happening in Dallas-Fort Worth differs markedly from housing conditions in Lubbock, or Lufkin, or Laredo. Too often, homebuilding and economic trends are analyzed with national numbers when it’s what’s happening inside a market that really matters.
You can find local numbers, but you have to be patient and look deeper. State unemployment figures come out one month after the national numbers, and local numbers even later. To get a sense of state and regional economies, turn to specialist sources. Many states have state demographers and/or analysts at the legislator who specialize in knowing their state’s economy. For example, Hawai’ians rely on the University of Hawai’i Economic Research Organization (UHERO) to get the lowdown for their state at a level of detail that national institutions can’t match.
5) The Known Unknowns
Mother Nature mucks up the supply chain so regularly that it’s a good bet to expect them even if a blizzard, hurricane, tornado, or hailstorm doesn’t happen in a particular year. Strangely, because damaging weather can prompt more sales, good weather the following year can make a company’s financials look bad in comparison. (Rain is the exception; showers in February can depress numbers and thus make a dry February the following year look extra good.)
This practice of judging current performance against the same month of the previous year gets bollixed up far too often by the weather. It would be far better to look not only at sales in a month but how that month’s weather compared with the long-term average. Companies like Planalytics now offer software to do that. Planalytics officials say their weather data makes it possible to toss out the weather wild card and get a truer idea of how well the company performed.
6) Surprise!
We’ve arrived at the “unknown unknowns:” influences warping our data that we didn’t expect or never imagined. You could put tariffs in this category, as you’d be hard pressed to find people outside of Washington who expected the Trump Administration would do all that it has—threatening tariffs, imposing tariffs, and suddenly lifting tariffs. Manufacturers’ and shippers’ responses to these tariffs have warped economic reports significantly. The same also is true with artificial intelligence. Aside from propping up the stock market, investment to construct data centers has become a huge part of our commercial building numbers.
Another potential surprise that most Americans haven’t thought about involves their next tax payments. The One Big Beautiful Bill Act reduced taxes on most Americans and made those reductions retroactive to Jan. 1, 2025. Most people haven’t adjusted their withholdings in response to the change, so there are likely to be millions of people who will get a bigger tax refund than they expected.
Will those people spend that refund? Pay bills? Put the money into savings? We don’t know yet, so whichever route they go, some forecasters will be surprised.
About the Author
Craig Webb is a nationally recognized expert in the lumber and building materials industry, known for his deep insights into dealer and distributor operations across the U.S. As President of Webb Analytics, he provides strategic consulting, custom research and industry analysis to help manufacturers, distributors and investors navigate market trends and challenges. With a journalism degree from Indiana University and decades of editorial leadership — including 12 years as editor-in-chief of ProSales magazine — Webb has built a reputation as a trusted voice through his writing, speaking engagements and creation of the ProSales 100 Conference. His career spans influential roles at The Wall Street Journal, McGraw-Hill, American Banker and UPI and he’s visited dealers in 49 states and multiple countries to stay closely connected to the industry’s pulse.