Ten percent of US containerboard capacity was retired in thirteen months. The same event that restored the mills’ pricing power erased the market for six million landowners.
American containerboard producers permanently retired about 3.9 million tons of capacity, roughly 10% of the national total, between February 2025 and March 2026. Linerboard prices went up. South-wide pine pulpwood stumpage went to $5.96 a ton, down 46% from its 2022 peak, and in parts of Arkansas and Georgia it went to nothing at all. Those are not two stories. They are one supply cut read from two sides of a mill gate, and which side you stand on decides whether it made you money.
Bill Howard’s family has been cutting pine in southeast Georgia since 1898. His account of the last two years, given to the Wall Street Journal, is that pulpwood fetched $16 a ton and now fetches somewhere between zero and five dollars. “We went from the best pulpwood market in the nation to the pits,” he said. What is missing from that account, and from most of the coverage, is that the buyers who stopped paying him are having a decent year.
What Happened
International Paper closed its Savannah containerboard mill and a second mill at Riceboro in the autumn of 2025, taking out about 1.4 million tons of gross capacity and roughly a million tons net after a $250 million conversion at Riverdale in Selma, Alabama. About 1,100 jobs went with it across four Georgia sites. The Savannah mill opened in 1936 as Union Bag and Paper, employed more than 5,000 people at its height, and is now a 735-acre listing.
Forisk Consulting calculated that the two pulp mills represented 3% of US pulp capacity, and that together with the Riceboro sawmill the closures removed 4.6 million tons of annual wood use from southeast Georgia alone.
That is the demand side. Here is the price side over the same window.
| Measure | Direction | Detail |
|---|---|---|
| US containerboard capacity | Down about 10% | 3.9 million tons retired, February 2025 to March 2026 |
| North American linerboard price | Up | Net $100 a ton taken in H1 2026; a third increase of $140 a ton sought for September 1 |
| Linerboard, 42-pound grade | Up about 30% | Measured against year-end 2019 |
| South-wide pine pulpwood stumpage | Down 22% year over year | $5.96 a ton in Q4 2025, 46% below the 2022 peak |
| South-wide pine sawtimber stumpage | Down 6% year over year | $23.23 a ton, only 10% below its early-2022 peak |
Box demand did not recover to justify any of the price increases. The American Forest and Paper Association put Q1 2026 containerboard production down 8% year over year, with export shipments off 19%, and box shipments fell 1.9%. Bank of America’s George Staphos summarised the sector in July: “Containerboard is okay, but that’s because of pricing strength rather than demand.”
Pricing strength is the polite name for a supply cut. The mills made one. The landowners could not.
The Backstory
The southern pine economy was built on a demand-side invention. A Savannah chemist named Charles Herty worked out in the 1930s how to make paper from young loblolly and slash pine growing back over the cutover longleaf forest, and paper mills, unlike the portable sawmills that had chased the timber line west, could not move. So they taught the region to plant. Tens of millions of acres of exhausted cotton ground went into rows of pine, with federal erosion and price-support programs paying part of the freight.
The result is an ownership structure with no equivalent in American industry. More than six million people in the South own at least ten wooded acres. In Georgia, the top timber state, 88% of woodland sits with individuals and businesses rather than with the mills. That structure worked for ninety years because the buyer was permanent and close.
The thinning is where the arrangement paid. A pine plantation goes in at several hundred stems an acre to force straight, knot-free growth, then has to be thinned around year fifteen or the stand stalls and the stressed trees start attracting beetles. That thinning was the grower’s first cheque and the reason it was tolerable to wait twenty-five or thirty years for sawtimber. For a growing number of owners it is now an invoice.
Meanwhile the end market had been shrinking for a generation without anyone in the woods noticing. US corrugated box production peaked in 1999 at 405.1 billion square feet. Absolute shipments hit a record in 2021 during the stay-at-home boom, which disguised the underlying trend: per-capita box use is now more than 20% below that 1999 level. Recycled fibre took share from virgin pulp. Amazon and other shippers cut box weight, eliminated boxes inside boxes, and moved volume to poly and paper mailers. Eucalyptus pulp from Brazil, which grows in seven years rather than twenty-five, took the rest.
The Plan
The industry’s answer is to manufacture new demand for the fibre nobody wants.
Weyerhaeuser, which holds about 6.6 million acres across eleven southern states, is spending roughly $500 million on a TimberStrand engineered-wood plant near Monticello, Arkansas, due to start in 2027, and has formed TerraForge Biocarbon Solutions with Aymium to turn wood fibre into a drop-in replacement for metallurgical coal, starting with a $100 million plant at McComb, Mississippi. At full scale that venture is designed to consume more than 7 million tons of wood fibre a year.
Lobbyists are pushing for statutes that would encourage burning wood pellets for electricity, which is worth holding against the record: Enviva built exactly that business at scale in this exact fibre basin and went through Chapter 11 in 2024. Cellulose-based jet fuel, battery materials and textiles are all being floated. A delegation from timber and from state and federal government is due to meet European Union officials next month to argue for lifting the bloc’s ban on sulfuryl fluoride, a fumigant, which would open a wood-chip export route to Europe.
Individual owners are not waiting. Superior Pine Products has put blueberries on part of its 110,000 acres and is looking at solar and carbon. Others have leased ground to solar developers outright, a demand source with its own subsidy dependency. Howard is planting fewer seedlings per acre to cut his own thinning bill. ArborGen, one of the largest loblolly seedling suppliers in the world, has seen southern volumes fall more than 20% across two fiscal years, which is the clearest signal available that growers are quietly deciding not to run the experiment again.
The Business Model Angle
Two things in this chain are being priced by the same event and only one of them can act.
A containerboard machine can be idled in a quarter, and idling it raises the price of everything the surviving machines make. A pine stand cannot be idled. It adds tons every year regardless of demand, and by the industry’s own account trees in the South are now growing faster than wood is being harvested. When one side of a market can withdraw supply and the other side is biologically forbidden from doing so, the entire burden of adjustment falls on the side that cannot. Growth, sold as the return for a hundred years, is functionally a short position.
The second mechanic is where the money actually goes, and it explains the zero.
Stumpage is not a price. It is a remainder. The mill posts a delivered price, the logging contractor takes cut-and-haul out of it, and the landowner receives whatever is left. Southern logging rates run around $25 a ton. A stumpage price of $5.96 therefore implies a delivered price near $31, and the 2022 peak of about $11 implies a delivered price near $36. Run that: the delivered price fell about 14% and the landowner’s share fell 46%. The grower holds a claim levered roughly three to five times to the mill’s number, and the leverage rises as the price falls, which is why the last stretch to zero happens so quickly. Push the delivered price a dollar and a half below the logging rate and you get exactly what Superior Pine’s chief executive described on an Arkansas thinning: a bill for $1.50 a ton. That is not a market failure. It is subtraction.
Then there is the geography, which is what makes each basin a monopsony wearing the costume of a market. Pulpwood has almost no value per unit of weight, so the economic haul radius is short. When a mill closes, the standing timber inside that radius does not get cheaper in any useful sense. It becomes non-tradable, and the displaced fibre floods the adjacent basins and knocks their prices down too. That is why the collapse is not evenly distributed: south Georgia stumpage fell 40% in a year while sawtimber, which has multiple buyers in range, fell 10% from its peak. Same tree, same region, same housing market, four and a half times the damage on the product with one customer.
Which brings us to the sourcing terms nobody has read closely. Forisk’s number for what southeast Georgia lost is 4.6 million tons of annual wood use. TerraForge, the largest single new fibre outlet announced anywhere in the US South, is designed for more than 7 million tons. It is bigger than the hole. It is also, per the announcement, supplied exclusively by Weyerhaeuser, and it is being built in Pike County, Mississippi, more than 500 miles from Statesboro. Monticello has the same shape: about 80% of its raw material will come off Weyerhaeuser’s own fee timberlands. Two flagship rescue projects, two disclosed sourcing terms, and a third-party share that rounds to almost nothing.
The largest timberland owner in the South is not building a market. It is building itself a customer. That is a rational response to free fibre and a completely useless one to the six million people who own the fibre and no mill.
Underneath all of it is a category error that has been sold to institutions for thirty years. Timberland is marketed as an asset class with the duration of infrastructure. It has none of infrastructure’s contracts. There is no take-or-pay, no minimum volume, no notice period, no counterparty obligation of any kind. International Paper announced in August and stopped buying in September. Joe Hopkins, who runs 75,000 acres of pine in southeast Georgia, told the Journal he struggles to net 1% for his shareholders and that a stockbroker could get him six. The NCREIF Timberland Index that institutions benchmark against is roughly two-thirds southern acreage, which means the region where the pulpwood market went to zero is the region the index mostly describes.
The Risk
The strongest case against all of this is that it is a trough, not a structure.
Sawtimber only fell 10% from its peak, and the reason is housing, which is cyclical by definition. Containerboard capacity has already been cut hard enough that mills are taking price against falling volume, and if demand returns against a smaller base, machines restart and fibre demand comes with them. Prices are also intensely local: north-central Florida, south Georgia, eastern North Carolina and Virginia were still clearing $9 to $14 a ton in Q4 2025 while Arkansas and southeast Texas sat under $4. Regional zeroes are not a national zero.
The second objection is that the value did not disappear, it moved. Cheap fibre is a straightforward transfer to whoever owns conversion capacity, which is why Weyerhaeuser’s projects pencil at all. The owners being hurt are precisely the ones who never integrated, and the remedy for that has always been cooperatives, long-term supply agreements or selling out to somebody with a mill. None of it is new.
Third, the land is doing fine even where the timber is not. Land values have risen against what timber earns, which is the whole reason selling parcels for weekend places is tempting. Solar leases, blueberries, pine straw and carbon are all repricings of the same acre from a fibre-yield basis to a land basis, and for many owners that is an upgrade dressed as a crisis.
Fourth, it self-corrects. ArborGen’s seedling volumes falling more than 20% is the supply response working. Less planting now means a tighter market later.
The honest counter to that last point is the clock. A grower who declines to replant in 2026 is making a decision that clears around 2051. Nobody in this chain has a balance sheet that patient except the largest institutional owners, which means the self-correction is also a consolidation, and the people who bear the cost of it will not be the people who collect the benefit.
One more asymmetry is worth flagging because it is policy rather than markets. Washington has put 25% tariffs on Brazilian paper while exempting wood pulp. The import that competes with the American mill is taxed. The import that competes with the American tree comes in free.
Quick Questions
Is pulpwood really worth nothing? In some basins, yes, and slightly less than nothing in others. The south-wide average was $5.96 a ton in Q4 2025. Arkansas, southern Louisiana, Tennessee and southeast Texas were under $4. Where the delivered price has fallen below the cost of logging and hauling, the landowner pays to have the wood removed.
Why did sawtimber hold up so much better? More buyers within range. Sawtimber goes to sawmills, which are more numerous and more geographically distributed than pulp mills, and it carries enough value per ton to travel further. Pulpwood cannot pay for its own freight, so it is captive to whatever is nearby.
Did International Paper close Savannah because demand collapsed? The company’s stated reason is competitiveness. The plant needed roughly $300 million in repairs including a new roof, and its output was going into lower-margin export markets rather than to IP’s own box plants. The $250 million Riverdale conversion was chosen instead. Demand weakness is the context; the decision was a capital allocation call.
Are the new demand projects real? The capital is real. Weyerhaeuser has committed around $500 million in Arkansas and is a partner in a $100 million first plant in Mississippi. Whether they help independent growers is a separate question, and the disclosed sourcing terms suggest they mostly will not.
Should landowners stop replanting? That is exactly the decision the industry is worried about, and it is the one decision an individual owner can make that permanently changes the regional supply curve. It also takes a full rotation to reverse.
The Business Model Analyst Take
The most useful thing in this story is not the price of pine. It is the demonstration that a hundred-year-old asset class can turn out to have had no contract underneath it.
Six million people were told they owned a crop. What they actually owned was proximity to somebody else’s capital budget. The trees were never the asset; the mill was, and nobody ever paid for it, insured it, contracted for it or even counted it. When it closed, the thing they owned did not become cheap. It became unsellable, which is a different condition entirely and one that no timberland pro forma models.
The rule generalises well past forestry. If your product’s freight cost is a large fraction of its value, you do not have a market. You have a neighbour. Count the buyers inside your economic haul radius, and if the answer is one, your price is a courtesy rather than a quotation, and it can be withdrawn on a quarter’s notice by a capital committee that has never heard of you. Poultry growers, dairy farmers, aggregate quarries, scrap yards and rural biomass suppliers all live at this address.
The metric to watch is not pulpwood stumpage. It is the third-party fibre share in the next round of mill announcements. Read every rescue project for one line: how much of its feedstock does the sponsor already own? Weyerhaeuser’s two answers are roughly 80% and effectively all of it. Until somebody announces a plant that has to buy its wood from strangers, the South has not found new demand. It has watched the largest owner build a private one.
