Early in February, I expressed my "increasing concern" about the negative economic impact of "executive / fiscal policy errors", however, I concluded that post by noting that I was not currently on recession watch.
In early April, I
went on recession watch, but I'm still
not yet predicting a recession for several reasons: the U.S. economy is very resilient and was on solid footing at the beginning of the year, and perhaps the tariffs are not enough to topple the economy.
In the short term, it is mostly trade policy that will negatively impact the economy. However, there other aspects of policy that bear watching - especially immigration.
Here is some of the data I'm watching.
Housing: Housing is the basis of one of my favorite models for business cycle forecasting.
This graph shows the YoY change in New Home Sales from the Census Bureau. Currently new home sales (based on 3-month average of NSA data) are down 3% year-over-year.
Usually when the YoY change in New Home Sales falls about 20%, a recession will follow. An exception for this data series was the mid '60s when the Vietnam buildup kept the economy out of recession. Another exception was in late 2021 - we saw a significant YoY decline in new home sales related to the pandemic and the surge in new home sales in the second half of 2020. I ignored that downturn as a pandemic distortion. Also note that the sharp decline in 2010 was related to the housing tax credit policy in 2009 - and was just a continuation of the housing bust.
The YoY change in new home sales in late 2022 and early 2023 suggested a possible recession. I was able to look past the pandemic distortion and was able to predict a pickup in new home sales due to the low level of existing home inventory and because homebuilders could offer mortgage incentives that would somewhat offset the sharp increase in mortgage rates.
There are no special circumstances now, and if this measure falls to off 20% a recession seems likely.
Yield Curve: The yield curve is a commonly used leading indicator. I dismissed it when the yield curve inverted in 2019 and again in 2022. Both times dismissing the yield curve was correct (the recession in 2020 was obviously due to the pandemic, so we will never know if the yield curve failed to predict a recession in 2019).
Here is a graph of 10-Year Treasury Constant Maturity Minus 2-Year Treasury Constant Maturity from FRED since 1976. The yield curve reverted to normal a year ago and is currently positive at 0.59. If this inverts, this might suggest a recession is coming.
Click here for interactive graph at FRED.
Heavy Truck (and Vehicle Sales): Another indicator I like to use is heavy truck sales. This graph shows heavy truck sales since 1967 using data from the BEA. The dashed line is the August 2025 seasonally adjusted annual sales rate (SAAR). Note: "Heavy trucks - trucks more than 14,000 pounds gross vehicle weight."
Heavy truck sales were at 422 thousand SAAR in August, down from 442 thousand in July, and down 15.7% from 501 thousand SAAR in August 2024.
Usually, heavy truck sales decline sharply prior to a recession and sales have been a little soft recently.
Vehicle sales were over 17 million SAAR in March and April as consumers rushed to "beat the tariffs". Then sales were depressed in May and June.
This graph shows light vehicle sales since the BEA started keeping data in 1967. This is more of a
concurrent indicator than heavy trucks.
Light vehicle sales in August (16.07 million SAAR) were down 2.9% from the sales rate in July, and up 6.2% from August 2024. (Note: Sales in August were boosted due to the expiring EV credit).
Unemployment: Two other
concurrent indicators are the unemployment rate (using the "Sahm Rule") and weekly unemployment claims.
Here is a graph of the
Sahm rule from FRED since 1959. The Sahm rule is a measure of changes in the unemployment rate. It compares the three-month moving average of the unemployment rate (U3) to the minimum of the three-month averages from the previous 12 months.
The Sahm Rule was at 0.23 percentage points in August.
If this increases to 0.5 it will suggest a possible recession.
Note that this increased to 0.56 in August 2024, but Dr. Sahm (and I) argued this was not indicating a recession.
And weekly unemployment claims always rise sharply at the beginning of a recession (other events - like hurricane Katrina - can cause a temporary spike in weekly claims).
As I noted earlier, I'm not sure how to estimate the economic damage caused by these tariffs and other policies. The economy is clearly slowing due to policy.
There are also boycotts of U.S. goods and less international tourism based on both the tariffs and the inflammatory rhetoric of the current administration.
None of these indicators are suggesting a recession. For now, I'll focus on the leading indicators (especially housing) and I'll update this post monthly while I'm on recession watch.
Click on graph for larger image.
Click on graph for larger image.
Recent comments