Economic Projections and Simulations

BACKGROUND

I developed a suite of econometric simulation models to project the path of the U.S. economy and its forest product markets. I use these models to conduct policy experiments that measure how the U.S. forest products sector will develop under alternative monetary and fiscal policy pathways and how it may respond to unexpected business cycle fluctuations. This page displays conditional forecasts of consumption, production, and prices for select forest products under point forecasts from both: 1) a three-stage regression model of the U.S. economy, and 2) the "consensus" macroeconomic outlook (which I represent as the median outlook from a sample of forecasts published by various banks, industry groups, consultants, and public agencies). Please note that this page is subject to change as new data is released and as model specifications are refined. This material is currently provided for educational purposes only and should not be misconstrued as investment advice. I am working to provide interested readers with complete technical documentation that includes a full summary of statistical results, model diagnostics, and additional scenarios. If you have questions, thoughtful feedback, or are interested in supporting this ongoing analysis, please contact me at: djr.analytics@gmail.com.

DATA AND MODEL

• Historical data on wood products trade and the U.S. economy from 1960-2025 was collected from public sources (see full list of sources in the table below). Data was primarily accessed from the FAOSTAT forest product statistics database and from FRED (the Federal Reserve Economic Database). 

• Forest products consumption and production are defined by groupings of national-level data in the FAOSTAT database. Prices are defined as averages of producer price indices defined in the BLS PPI reports. 

• Two macroeconomic scenarios were defined:

• 1. The first scenario is the baseline forecast generated from a three-stage regression model of the U.S. economy. The first-stage of the model relates economic policy variables to subsequent outcomes in the real economy. It simultaneously allows for the impact of economic outcomes to influence subsequent policy responses. Bayesian simulation with hierarchical prior selection is used to estimate the first-stage macroeconomic relationships. Each forecast from a first-stage simulation is then used in both a second-stage system of dynamic equations describing the financial sector and a third-stage system of dynamic equations describing the housing sector. The resulting means of the simulated forecasts define the macroeconomic outlook for this scenario. 

  2. The second scenario, referred to as the "consensus outlook," is based on the median of forecasted macroeconomic variables published by various public and private sector sources (see full list of sources in the tables below). 

• Reduced form equations of forest products consumption, production, and prices are estimated and used to generate conditional forecasts.

• The open source statistical software R was used to estimate economic relationships with standard econometric estimation routines installed (including BVAR, vars, systemfit, forecast, and dynlm). 

• Forest products consumption, production, and prices were simulated 5,000 times under each macroeconomic scenario using Monte Carlo procedures written in R. Results in the tables below show the median forecast from these simulations.

KEY RESULTS 

Updated Feb. 23, 2026. 

MACROECONOMY

• The three-stage model produces a baseline forecast with higher inflation and higher unemployment than the scenario defined by the consensus outlook. This can be viewed as a "stagflationary" scenario relative to the consensus outlook.

• Long term interest rates are projected to be higher under the baseline model. This is driven by higher inflation and expansionary fiscal policy. The effect of higher long term rates is a slowdown in the housing sector. 

• Monetary policy, as measured by the federal funds rate, is forecast to ease moderately in the near term under the baseline model. However, it pivots to a restrictive policy path as inflation begins to exceed 3.25%/year by 2028. In contrast, the consensus outlook suggests monetary policy will ease through 2030.

• Under the baseline model forecast, the federal debt expands to 123% of GDP by 2030, while annual federal deficits average 5% of GDP over the next five years.

WOOD PRODUCT MARKETS

• Higher mortgage rates forecast under the baseline model lead to lower annual housing starts and lower domestic consumption of wood products typically used for home construction (including lumber, plywood, LVL, and OSB) compared to modeled consumption under the consensus macroeconomic outlook. The consumption of these products is especially sensitive to a higher inflation, higher mortgage rate environment.

• Hardwood sawnwood consumption is measured to be less sensitive to GDP growth and inflation compared to softwood sawnwood consumption.

• The housing slowdown projected under the baseline model yields a 2.1%/year decline in softwood sawnwood prices and a 2.1%/year decline in plywood/veneer prices through 2030. 

• Lower mortgage rates and greater annual housing starts under the consensus outlook raise the prices of softwood lumber, hardwood lumber, and plywood (relative to the baseline model forecast).

• Despite a slower housing sector under the baseline model, millwork prices are still projected to expand due to a higher sensitivity to rising manufacturing prices.

• Under the baseline model, 47% of forecasted softwood sawnwood consumption in 2026 is lower than it was in 2024 (which is the last year of observable data available from FAOSTAT). By 2030, 68% of forecasted consumption is lower than it was in 2024 (however, it is only 23% under the consensus outlook).

• Under the baseline model, 67% of forecasted softwood lumber prices are lower in 2026 than they were in 2025. By 2030 this percentage is 74% (however it is only 62% under the consensus macroeconomic outlook).

PULP/PAPER/PAPERBOARD PRODUCT MARKETS

• A high estimated sensitivity to general price inflation pushes the hardboard, fiberboard, and particleboard price index higher by 2030 under the baseline model.

• Annual packaging consumption is modeled to increase at a slower pace of 0.5%/year through 2030 under the baseline model, but expands 1.2%/year under the consensus macroeconomic outlook. 

• Paperboard prices are forecast to decrease 0.2%/year through 2030 under the baseline model as a higher unemployment rate weighs on demand growth. Paperboard prices are modeled to grow by 0.3%/year under the consensus macroeconomic outlook. Supply expansions outweigh demand increases under the baseline forecast, leading to greater annual consumption of packaging and paperboard products, but lower prices.

• Newsprint consumption is forecast to be remain flat around 0.8 to 0.9 million tonnes/year under both the baseline model forecast and the consensus outlook. Newsprint prices decline by 0.7%/year under the baseline model forecast but decline 1.2%/year under the consensus macroeconomic outlook. This is a consequence of greater supply-side constraints and lower domestic production modeled under the baseline forecast.

• Under the baseline model, 43% of forecasted wood pulp consumption in 2026 is lower than it was in 2024 (the last year of observable data available from FAOSTAT) and 59% of forecasted wood pulp prices are lower in 2026 than they were in 2025. By 2030, these percentages are 34% and 47%, respectively.

TIMBER PRODUCT MARKETS

• The baseline model forecast produces a lower annual rate of domestic sawtimber and pulpwood production through 2030 (relative to modeled production under the consensus outlook). 

• Timber prices are forecast to decline under the baseline model forecast (falling 0.4%/year through 2030). This represents slower price growth compared to what is modeled under the consensus macroeconomic outlook (increasing 0.5%/year through 2030). This is driven primarily by slower rates of price growth for both softwood and hardwood lumber as well as less demand for softwood sawtimber.

MANUFACTURING COSTS

• Unit costs of manufacturing in both the sawmill/wood preservation sectors and the pulp/paper/paperboard sectors are expected to increase through 2030. Machinery prices are forecast to expand by 1.2%/year through 2030, while wage rates increase by less than 1%/year in all sectors through 2030.

FIGURES AND TABLES