Investing in Timberland

In the spring of 2021 the price of framing lumber did something no commodity is supposed to do: it quadrupled in a year and then fell by three quarters in three months. Futures touched $1,733.50 per thousand board feet intraday on May 10, 2021, and by August the contract traded at $454. Every newspaper in the country ran the story as a windfall for people who own trees.

It was not. The landowners who grew the wood that became those record-priced two-by-fours were paid stumpage — the price of a standing tree, negotiated per ton at the roadside — and stumpage barely moved. The Alabama Cooperative Extension System published an explainer during the spike with the title the whole industry needed, “Why Stumpage Prices are Low Despite Historic High Lumber Prices,” and the answer was that sawmilling capacity, not fibre, was the bottleneck. The mills that converted logs into lumber captured the margin, because there were only so many of them and a great many acres of pine competing to supply each one.

Four and a half years later the arithmetic still looked like that. Lumber closed at $571 per thousand board feet on September 8, 2026, per Trading Economics. Southern pine sawtimber stumpage averaged $23.23 a ton in the fourth quarter of 2025 and pine pulpwood $5.96 a ton, both per TimberMart-South as reported by Yanshu Li in Southern Ag Today on January 26, 2026.

That distance between the lumber price you read about and the timber price you are paid is the first thing to understand about this asset, and it is not an aberration; it is the structure.

This guide is about what timberland has actually returned across a full institutional cycle, why the biology is the durable part and the price is not, how the listed and private routes differ, what an acre costs to own and to sell, how the US tax code treats a standing tree, and what 125 acres in the US South earns over ten years when you write down every fee. Investing in Farmland, the hub’s flagship, covers farmland’s $3.8 trillion market, and three sister guides carry the rest of it: Investing in Farmland Through REITs and Funds, Buying Farmland Directly and Investing in Water Rights. This one stays on the forest.

A timberland investment is two assets in one deed: the dirt, which has a value whether or not anything grows on it, and the standing inventory, which is a stock of merchantable wood that increases by itself every year. Nothing else in the alternatives universe behaves quite like the second half. A painting does not become a larger painting while it hangs on the wall. A barrel of whisky loses volume to the angels. A pine plantation on an average southern site adds several green tons an acre each year without a single decision from its owner, and the tons it adds are worth more per ton than the tons already there, because a tree that crosses a diameter threshold stops being pulpwood and starts being sawtimber.

That second effect has a name in the trade. Ingrowth, or product-class shift, is the migration of standing volume from a cheap product class into an expensive one as stems get bigger. In the US South the ladder runs pulpwood, then chip-n-saw, then sawtimber, and TimberMart-South’s south-wide averages for the fourth quarter of 2025 ran $5.96, $17.50–18.20 and $23.23 a ton respectively. A ton of wood that sits still for six years can be worth three or four times what it was worth without a single dollar of new investment, and the owner did nothing but pay the property tax.

The four regions and who is in them

The institutional market divides the United States into four regions, which is how NCREIF reports the index: the South, the Northwest, the Lake States and the Northeast. The South dominates — it accounts for roughly two thirds of the acres in the index, and for the large majority of its properties. That matters more than it sounds: when someone quotes you “timberland returns,” they are mostly quoting southern pine plantations on a rotation of roughly twenty-five to forty years, not Douglas fir on the Olympic Peninsula and not northern hardwood.

Ownership sorts into four groups. The federal and state governments hold enormous acreage that is largely outside the investable market. Vertically integrated forest-products companies once held the best of the private acres and mostly do not any more. Timberland investment management organisations, or TIMOs, buy and manage forests on behalf of pensions, endowments and sovereign funds; BTG Pactual’s Timberland Investment Group, one of the larger ones, reported 3.3 million acres and US$8.2 billion of assets and commitments as of March 31, 2026. And the timber REITs hold acreage inside a listed corporate wrapper that pays out its income.

Behind all four sits the largest owner class of all, and it is larger than the institutions by an order of magnitude. The USDA Forest Service’s National Woodland Owner Survey counts about 3.7 million family ownerships holding ten acres or more, together 253 million acres, or 34% of American forest land, in its 2018 round; count the smaller holdings too and the number of family owners runs to nine or ten million. Most of them will sell timber two or three times in a lifetime and have no idea what an internal rate of return is. They are the sellers institutional buyers meet, and they are usually the least advised party at the table.

How a price is discovered

There are two price discoveries here and they are not connected as tightly as you would expect. The first is the timber sale: a landowner marks a tract, a consulting forester cruises it to estimate volume by product class, bids go out to local mills and logging contractors, and the wood sells either as a lump sum for everything standing or per ton as it crosses the scales. The bid depends almost entirely on how many mills can reach the tract economically — wood is heavy and cheap, so hauling more than roughly sixty or seventy miles destroys the economics. This is why stumpage is a local price, and why two identical tracts a hundred miles apart trade at prices that differ by half.

The second is the land transaction: whole tracts change hands between investors, and the price embeds a view on both the standing timber and the bare land underneath. This market is thin, negotiated, brokered and slow. Forisk’s review of 2025 transactions described exactly the condition you would expect at a cycle top: values rising while volumes fell, because buyers could not make acquisitions pencil at the asking prices. When the quoted value of an asset rises while fewer and fewer people transact in it, the quote is doing more work than the market is.

The number the industry leads with is the NCREIF Timberland Index, and taken at face value it is spectacular. AcreTrader’s tabulation puts the index at 10.74% a year from its 1987 inception through the end of 2021. Beating equities over thirty-four years with a real asset that also produced a cash yield would be, if true in the form implied, one of the better trades available to a pension fund.

It is true, and it is also almost entirely historical. Forisk decomposed the same index over the thirty years from 1995 to 2025 and published the components in June 2026: an average 3.4% a year of EBITDDA — the industry’s cash-yield measure, earnings before interest, taxes, depreciation, depletion and amortisation — and an average 3.4% a year of appreciation. Those two sum to roughly 6.8% a year, which is a perfectly respectable return for a real asset and is nothing like 10.74%. The entire difference sits in the years the thirty-year window excludes: 1987 through 1994, when institutions were buying forests from paper companies at prices that reflected a wood market and not an asset-allocation market.

The direction of travel inside those components is the part worth sitting with. Forisk’s regional cut shows EBITDDA trending downward across every region over the thirty years, with the Northwest holding the highest cash yield at 4.8% and the South the lowest at 2.8%; by 2025 the regions had largely converged, most of them yielding between 1% and 3%. The South compensates with the highest appreciation, 3.8%, which is another way of saying that southern timberland has increasingly been priced as land that grows trees rather than as a business that sells wood. Timberland Investment Resources made the same point more bluntly in a June 2025 paper: inflation-adjusted timberland returns declined in each of the three decades it examined, while more money kept arriving.

NCREIF put the US timberland return at 4.9% for the year ended March 31, 2026, with the South — the leading region that quarter — at 6.56%. That is a respectable year for a real asset. It is also close to the thirty-year components and roughly half the since-inception number the industry still quotes, which is the point: the most recent print does not disturb the picture, it confirms it.

Which fees the index is gross of

This is the distinction the whole gap between the index and the investor turns on, and it is routinely stated backwards. The NCREIF Timberland Property Index is measured at the property level and unlevered, and it is inclusive of property-level management fees — the cost of the foresters, the roads, the taxes and the people who actually run the trees is already subtracted before the number is printed. What the index is gross of is the layer above that: the portfolio-level asset-management and investment-management fees a manager charges the investors whose money bought the properties. NCREIF runs separate fund and separate-account indices precisely so that before- and after-fee performance can be reported apart from the property series.

So the index return is what the forest earned after the cost of running the forest and before the cost of running the money. A TIMO’s version of that second layer is an annual management fee on committed or invested capital plus a carried interest above a preferred return; a REIT’s is corporate overhead and the manufacturing business bolted to it. Either way the investor’s net sits materially below the printed index, and the gap is a fee gap, not a measurement artefact. Everything else in this guide about the difference between the index and the investor is downstream of that one sentence.

What else the index leaves out

Three more separations sit between the index number and what an outside investor keeps, and all three cut the same way.

The first is appraisal smoothing. Private timberland is marked by appraisal, usually annually, and appraisals lag and dampen. A smoothed series shows less volatility than the underlying asset actually has, which flatters every risk-adjusted statistic computed from it and makes the correlation to equities look lower than the economics justify.

The second is survivorship and entry effects. The index contains properties that institutions chose to buy and continued to hold. It is not a census of American forests, and a fund that bought at the top of the 2005–2007 acquisition wave did not earn the index.

The third is the composition of the index itself. Two thirds southern pine means the headline is a southern pine number wearing a national label.

Strip out the price forecast and a timberland investment still makes money, which is the genuinely unusual property of the asset and the honest core of the pitch. The mechanism has three parts, and they compound on each other.

Growth is the addition of volume to existing stems. A managed southern pine plantation on an average site puts on something in the range of four to six green tons an acre a year across a rotation; Mississippi State Extension’s growth-and-yield work puts a moderately productive loblolly stand (site index 63 at age 25) at about 6.3 tons an acre a year, and the figure varies enormously with site index, planting density and whether the stand has been thinned. Over a decade that is roughly fifty tons an acre added to a stand that might have started with forty-five.

Ingrowth is the price effect described earlier: stems crossing diameter thresholds into higher-value product classes. On the fourth-quarter 2025 southern price ladder, moving a ton from pulpwood at $5.96 to sawtimber at $23.23 is a 290% gain on that ton, achieved by waiting.

Harvest timing is the option the two above create. Because the inventory is not perishable on any human timescale, an owner facing a weak market can simply not cut. Trees kept in the ground keep growing and keep migrating up the product ladder; the deferred harvest is not a lost harvest. This is the asset’s single best structural feature and it is genuinely rare.

Three cautions before you fall in love with it. Deferring a harvest is only free if you can afford the carry, so a leveraged owner or a fund facing redemptions does not really have the option. Growth is not linear: annual increment peaks in mid-rotation and declines afterwards, so a very old stand is a poor compounding machine. And the option has a cost you can measure — every year you defer, you finance the property tax, the management fee and the opportunity cost of capital out of pocket.

The cycle underneath the cycle

Timberland’s return alternates between an income-led regime and an appreciation-led one, and Timberland Investment Resources put numbers on that alternation in June 2025: since the index began in 1987 there have been nine income-dominant cycles averaging 7.7 quarters and ten appreciation-led cycles averaging 8.1 quarters. The appreciation-led run that ended around the first quarter of 2025 had lasted 14 quarters, nearly twice the historical average. TIR read that as evidence the asset was overdue to revert toward income-driven performance. The point that survives whether or not that call proves right is structural: when appreciation carries a real asset for three and a half years while its cash yield does not improve, the buyer of that asset at the end of the run is paying for something the trees did not produce.

Until the mid-1980s the best private timberland in America belonged to the companies that milled it. International Paper, Georgia-Pacific, Weyerhaeuser, Boise Cascade and their peers owned the forest as a raw-material hedge and carried it on the balance sheet at historical cost, which by the 1980s meant carrying it at a fraction of its value. Two things then changed at once: the conglomerate discount made a paper company’s undervalued land a target, and the passage of pension money into “alternatives” created a buyer that wanted exactly this — a long-duration, low-reported-volatility, inflation-linked real asset with a story about biological growth.

The disposals that followed were the largest transfers of private land in American history. The single biggest was announced on April 4, 2006, when International Paper agreed to sell approximately 5.1 million acres for approximately $6.1 billion: about 3.8 million acres across the South plus 440,000 acres in Michigan to an investor group led by Resource Management Service for around $5 billion, and about 900,000 acres in Louisiana, Texas and Arkansas to a group led by TimberStar for around $1.1 billion. The company called it the largest private forestland sale in US history and was probably right. On our arithmetic that entire portfolio changed hands at about $1,196 an acre, timber included.

The buyers were TIMOs and, increasingly, REITs, and the mechanism that made both work was tax. A vertically integrated forest-products company paid corporate tax on its timber earnings and then its shareholders paid again on the dividend; a REIT that meets the distribution tests pays no corporate tax on qualifying income, and timber gains qualify. Plum Creek converted from a master limited partnership on July 1, 1999 and was the first timber REIT. Rayonier followed in 2004, Potlatch in 2006, and Weyerhaeuser — much the biggest of them — elected REIT status with effect from January 1, 2010, clearing its accumulated earnings with a special dividend paid that September. Each conversion was, at bottom, a decision that the forest was worth more outside a C corporation than inside one.

The consolidation that followed ran the other way. Weyerhaeuser absorbed Plum Creek in February 2016, Potlatch and Deltic Timber combined on February 20, 2018, and PotlatchDeltic completed its merger with CatchMark Timber Trust on September 14, 2022. Every one of those deals removed a listed owner of American forests, and the process did not stop there. A sector that consolidates while its underlying returns fall is telling you where the remaining margin is: in scale, in manufacturing and in the corporate line, not in the trees.

The money kept coming and the returns kept falling

Nuveen estimated roughly $123 billion of institutional capital invested in timberland globally as of the second quarter of 2024. Set that against the return record and the story tells itself: the index compounded above 10% in the era when institutions were buying forests at $1,000-odd an acre from sellers who did not want them, and it compounds in the high single digits or below now that the same institutions own the forests and are trying to sell them to each other. Timberland Investment Resources’ framing of this — real returns falling in each of three decades while capital kept arriving — is the least flattering and most useful sentence written about the asset class.

There is a version of this that is good news. TIR also observed in the same June 2025 paper that since the global financial crisis, commercial real estate values rose 184% and farmland values 200%, while timberland values rose only 137%. On that comparison timberland is the cheap real asset, not the expensive one, and a buyer today is paying a lot less of a premium than a buyer of farmland is. The honest reading is that both things are true: the era of buying forests below their wood value is over, and timberland has not been re-rated the way its cousins have.

It is also worth knowing how few institutions actually do this. Nuveen’s 2025 EQuilibrium survey found that 53% of responding investors neither invested in timberland nor planned to, the highest such share of any private asset class it asked about, with 11% planning to increase an allocation. A thin institutional bid is part of why transaction volumes fell while appraised values did not.

Everything downstream of a standing tree is cyclical, and the cycle is housing. Framing lumber demand is single-family construction and repair-and-remodel; pulpwood demand is the paper, packaging and pellet mills; and both translate into stumpage only through the mill’s willingness to pay. The US Census Bureau and HUD reported privately-owned housing starts at a seasonally adjusted annual rate of 1,239,000 in July 2026, down 12.4% from June and 13.5% from a year earlier, with single-family starts at 808,000, down 9.9% on the month. That is the demand side of the timber price, and it was soft.

The supply side of the mill’s margin is capacity, and this is where the 2021 lesson lives. When lumber futures ran to $1,733.50 in May 2021 the constraint was sawmill throughput, not logs. Mills that could cut ran their margins to records; landowners with pine ready to sell found that the mill’s bid had barely moved, because there were still far more acres of merchantable pine within a hauling radius than there was capacity to saw it.

Alabama Extension’s account of the episode adds the other half. Sawmills had not rebuilt capacity after 2008, while the pine planted across the South in the late 1980s — much of it under the Conservation Reserve Program — began reaching sawtimber size just as the market collapsed, leaving a standing surplus that has not cleared since. Lumber is a manufactured good with a supply constraint; stumpage is a raw material in surplus. The two prices are related but the relationship is asymmetric, and it favours the manufacturer.

Sawtimber is below its 2019 price in nominal dollars

Alabama Cooperative Extension, citing TimberMart-South, recorded a southern pine sawtimber average of $23.47 a ton in 2019 and $23.35 in 2020. In the fourth quarter of 2025 the same series printed $23.23. Six years, no inflation adjustment, and the price of the thing a southern landowner actually sells is fractionally lower than it was — lower after a housing boom, a lumber spike that made national news, and fourteen consecutive quarters of appreciation-led returns in the index that values the land underneath. Every other number in this guide should be read against that one.

The quarter itself was unremarkable. Sawtimber at $23.23 was 6% below a year earlier and 10% below the early-2022 peak, ranging from roughly $17–21 a ton in Tennessee, South Carolina and Alabama to over $30 a ton in Florida and North Carolina; the third quarter of 2025 had been near $28. Sawtimber has held up far better than pulpwood, which is the only comfort on offer here.

The long view is worse than the quarter-to-quarter one. Stephen F. Austin State University’s East Texas stumpage-trend work shows prices climbing steadily from 1988 through the late 1990s, peaking, and declining since. One industry analysis, forestryandland.com, restates the pre-Great-Recession average of $40 a ton as $51.43 on the BLS inflation calculator, against regional averages near $22 at the end of 2025 — less than half, in real terms, of what an acre of pine was worth to its owner a generation ago. Whatever timberland has been for the last twenty-five years, it has not been a bet on the price of wood.

The pulpwood rung is impaired, not cyclical

The rung below sawtimber is worse, and its problem is not the cycle. Southern Ag Today, reporting TimberMart-South data on January 26, 2026, put fourth-quarter 2025 pine pulpwood at $5.96 a ton, down 22% year over year and 46% below its 2022 peak, with prices under $4 a ton in Arkansas, southern Louisiana, Tennessee and southeast Texas. The cause is stated in the same reporting: between 2023 and 2025, more than 10 major pulp facilities closed in the US South, removing over 25 million tons of annual fibre demand. That is the single most consequential development for southern landowners in a decade.

It matters far beyond the pulpwood cheque. Thinnings — the intermediate harvests that pay a landowner’s carrying costs and concentrate growth on the best stems — are mostly pulpwood sales. When pulpwood falls below roughly $6 a ton, a thinning in a marginal location stops covering the cost of the logging crew, and the stand does not get thinned. An unthinned stand grows more slowly, stays in the cheap product classes longer, and is more vulnerable to beetles and ice. The pulp mill closure two counties away is therefore not a price event; it is a change in the biology of your investment.

For most people the only practical way to own American timberland is to buy a timber REIT, and the menu got shorter on January 30, 2026, when Rayonier and PotlatchDeltic closed a merger of equals. Each PotlatchDeltic share converted into 1.8185 Rayonier shares plus $0.61 in cash, Rayonier paid roughly $3.25 billion to PotlatchDeltic holders, and the ownership split of the combined company was about 54% legacy Rayonier to 46% legacy PotlatchDeltic. The combined business trades as RYN — after a review the company announced on March 31, 2026 that it would keep the Rayonier name and ticker and change only its logo — and holds over four million acres of US timberland along with six sawmills, an industrial-grade plywood mill, and real-estate and rural land-sales operations. Its headquarters moved to Atlanta.

Weyerhaeuser is the other one, and it is much the larger: it owns or controls roughly 10.4 million acres of US timberland — the largest private holding in North America, on Forisk’s 2025 ranking of owners — plus long-term licences on Canadian public land. Its second quarter of 2026, reported July 30, 2026, produced GAAP net earnings of $162 million, or $0.23 per diluted share, on net sales of $1.9 billion, and returned $152 million to shareholders through the quarterly dividend. The base dividend is $0.21 a share a quarter, supplemented under a stated cash-return framework that targets 75–80% of annual adjusted funds available for distribution through a supplemental dividend, buybacks or both. Rayonier declared $0.26 a share for the second quarter of 2026.

What our tape shows

Our collection engine stored five sessions of closing prices for each of the surviving REITs at the end of August and the start of September 2026. Weyerhaeuser went $23.37, $23.38, $23.19, $22.83, $23.02 between August 31 and September 4, 2026 — down 1.50% across the week. Rayonier went $20.50, $20.81, $20.23, $20.32, $20.46, down 0.20%. On those closes the base dividends work out to a 3.65% yield on Weyerhaeuser and 5.08% on Rayonier, our arithmetic on the declared quarterly rates annualised.

That is also why our timberland sleeve is still marked awaiting in the composite at a 0.8% target weight, accruing history from REIT proxies, alongside farmland at 0.6% and water at 0.3%. The hub’s three land assets have the same problem in three forms: the thing itself does not have a daily price, and what does have one is a proxy for it.

Our PotlatchDeltic series is the instructive one. It holds exactly one observation, $41.73 stamped February 4, 2026, and nothing on either side of it. That is not a price. It is what a data vendor returns for a ticker that stopped trading, and it is a useful reminder that a series can look alive for months after the company behind it has ceased to exist. Anyone building a timberland comparison from a data feed should check that every ticker in it still has a company attached.

The scale comparison explains the liquidity. Weyerhaeuser returned $152 million through a $0.21 dividend, which implies roughly 724 million shares and, at our tape’s $23.02, about $16.7 billion of equity — that share count is our arithmetic on the company’s own disclosure, not a reported figure. The merged Rayonier was pitched at a pro forma equity market capitalisation of $7.1 billion and an enterprise value of $8.2 billion on the October 10, 2025 closing prices that framed the deal.

Twenty-odd billion dollars of listed timberland equity, then, against Nuveen’s estimate of $123 billion of institutional timberland capital globally as of mid-2024. The listed market is the small, liquid, publicly-priced corner of a much larger private one, which is why its shares move with interest rates and housing sentiment rather than with appraisals.

What you are actually buying in a timber REIT

Not a pure forest. Both companies run manufacturing — Weyerhaeuser’s Wood Products segment and Rayonier’s six sawmills and plywood mill — and a real-estate business that sells rural land and development parcels at prices well above timber value. That mix is a feature and a distortion. In a lumber upswing the mills capture the margin the landowner does not, which is precisely why owning the integrated company has beaten owning the trees in several cycles. In a downswing the mills lose money the forest would not have lost. And the real-estate segment means a share of the reported earnings comes from selling the asset rather than growing it.

The Rayonier quarter shows the integration plainly: Southern Timber acreage nearly doubled to 3.15 million acres from 1.69 million at year-end 2025, and segment adjusted EBITDA rose 85% year over year to $52.6 million on roughly 1.5 million tons of incremental harvest from the legacy PotlatchDeltic land. On August 5, 2026 the company also swapped ground with Resource Management Service in a like-kind exchange, selling about 36,000 acres in southwest Washington for $145 million and buying about 57,000 acres in Alabama and Texas for $146 million. That is roughly $4,030 an acre going out and $2,560 an acre coming in: a trade of expensive Pacific Northwest fibre for cheaper southern acreage, and the clearest recent read on what institutional timberland actually changes hands for in each region.

Institutions do not buy forests directly either; they hire a TIMO. The structure is a closed-end private fund with a ten-to-fifteen-year life, capital calls, an annual management fee and a carried interest above a preferred return, and a separate-account variant for investors large enough to want their own portfolio. The industry is concentrated: on Forisk’s 2025 ranking of North American timberland managers, six of the top ten are TIMOs, and those six run 12.9 million acres — 52% of the 24.8 million acres managed by the top ten firms.

The fee stack is where the index return goes, and Forisk has published its shape. Total annual fees approach but do not exceed roughly 1% of assets under management — the Campbell Group has indicated 96 basis points and Hancock Timber Resource Group 95 basis points at the property level — charged on committed capital or net asset value. On top of that sits an overage: a carried interest of something like 20–25% of returns above a hurdle, with 6% a common hurdle.

Now stack it. The index’s EBITDDA is already net of property-level management costs, so the fees above are additional, and acquisition and disposition costs land at each end of a ten-year fund life. Apply all of it to a 6.8% gross component return and the investor’s net arrives in the low-to-mid single digits. That is the gap between what the index earned and what the person who bought the index earned, and it is the one number no marketing deck leads with.

Three consequences follow, and they are the practical reason a private individual should think hard before reaching for the private route.

• The minimums exclude you. TIMO funds are institutional products with commitments measured in millions and accredited-investor gates at the very least. The retail-accessible timberland fund is rare, and where it exists the fee load is usually worse, not better, than the institutional version.
• The lockup is real and the exit is a sale of forests. A ten-year fund that reaches the end of its life in a weak transaction market either extends or sells into a bid that Forisk’s own 2025 review described as thin. Timberland’s celebrated ability to defer a harvest does not extend to a fund’s ability to defer a liquidation.
• The marks are appraisals. You will receive a quarterly NAV produced by a valuer using the same discounted-cash-flow assumptions the manager gave them. That number will be smoother than reality and will lag any turn by two to four quarters.

A forest now has two customers that do not want the wood. One pays the owner to leave the tree standing; the other pays for the ground underneath it. Both are options rather than income, and both are priced in markets that publish no clearing price, which is why the figures below carry the name of whoever tabulated them.

Forest carbon pays a landowner to store carbon rather than sell it, typically through an improved-forest-management project that commits the owner to hold inventory above a baseline for a long period — a hundred years, under California’s compliance protocol — with a share of credits held back in a buffer pool against fire and disease. The mechanism is sound and the market is not yet trustworthy: additionality is genuinely hard to prove, baselines have been gamed, and the early-2020s wave of investigative reporting into voluntary credits damaged buyer confidence.

The prices say the same thing. Ecosystem Marketplace put the 2025 voluntary-market average at about $6.34 a tonne. MSCI’s Global Carbon Credit Price Index averaged $3.50 in 2025, down from $4.30 in 2024, while credits its methodology rates BBB or better averaged $6.80, up from $5.60. Trade trackers put improved-forest-management credits nearer $15 a tonne in 2026, with North American nature-based projects at a premium to the rest of the world. Every one of those is a vendor or index tabulation rather than a clearing price, and the spread between them is the point: quality, not tonnage, is doing the pricing. Treat any carbon revenue in a sponsor’s model as an option, not a coupon, and check whether the project’s obligations survive a sale of the land.

Solar ground leases are the other, and on the right acre they are transformative: an interconnection-adjacent tract with the right topography can be worth multiples of its timber value as a solar site. Lease brokerages publishing 2026 rates put the national average at roughly $500–700 an acre a year, with most leases between $250 and $1,000 and strong markets running $1,200 and up, on terms of twenty-five to thirty years with escalators of 1.5–3% a year and a nominal option payment of $10–40 an acre a year for the one-to-five-year window before the developer commits.

Read those numbers for what they are. They are published by the firms that broker the leases and earn a commission on signing them, they are not an index, and they describe the acres that win rather than the acres that apply. The rest of the caution is structural: the option is highly location-specific, and a lease that looks like free money can constrain harvest access, complicate title and trigger property-tax reclassification out of a forestry programme.

Direct ownership is a small business with a very long production cycle, and the work is front-loaded into the ninety days before you close. The sequence below is the one that loses the least money.

Choose the wood basket before the tract

The single largest determinant of what your stumpage will be worth is how many mills can economically reach your gate. Wood is heavy relative to its value, and hauling beyond roughly sixty to seventy miles consumes the margin. Before looking at a tract, learn what mills are within that radius, what they buy, who owns them and whether any of them has announced a curtailment. A tract with four competing sawmills and two pulp mills inside the haul radius is a different asset from an identical tract with one mill, even when the trees are the same.

Commission a cruise, and read it yourself

A timber cruise is a statistical inventory: a forester samples plots across the tract and estimates volume by species, product class and diameter, usually with a stated sampling error. This is the document that tells you what you are buying, and it is the one place where a buyer’s due diligence maps directly onto price. Published guidance puts a cruise at roughly $300 to $2,000 and up depending on tract size, terrain and sampling intensity — on a small southern tract that is something like ten to fifteen dollars an acre. The same guidance notes that landowners who sell without an independent inventory routinely leave 20–40% of the timber’s value on the table, which makes the cruise the highest-return line item in the whole transaction.

Expect the seller’s cruise, if one is offered, to be optimistic about product class. Verify the assumed product mix against the diameter distribution, because the difference between calling a stand chip-n-saw and calling it sawtimber is most of the price.

Hire a consulting forester, not the buyer’s forester

The consulting forester is the single highest-value professional in this asset class. They cruise, they design and administer timber sales, they run the bid process, they supervise logging and enforce the contract, and they file the paperwork for whatever cost-share and property-tax programmes apply. They are paid a management retainer per acre per year and a commission on timber sales.

Extension guidance puts the sale commission at 3–10% of sale receipts depending on the work, the size and the value of the sale, with full-service turnkey arrangements running 10–15%. There is no standard rate. The retainer — commonly quoted at something like $5–10 an acre a year, a convention rather than a published figure — is negotiated tract by tract. That commission is the best money you will spend: an administered competitive sale routinely beats a negotiated sale to the first logger who knocks, and the forester’s fee is a fraction of the difference.

Check title, boundaries and access before price

Rural title is messier than urban title. Look for severed mineral rights, unrecorded easements, prescriptive roads, hunting-lease commitments, boundary lines last painted in the 1970s and, critically, legal access — a landlocked tract cannot be logged, and a right of way negotiated after closing is negotiated from a position of weakness. A boundary survey and a title opinion together are a rounding error against the purchase price and the commonest source of catastrophic buyer error.

Know the age class you are buying

A tract’s age class determines its cash-flow shape for a decade. Bare or recently planted land is cheap and produces nothing for fifteen years. Mid-rotation stands, roughly fifteen to twenty years old, are the classic institutional purchase: a thinning or two in the near term and a clearcut in sight. A mature stand priced to its standing inventory is a liquidation, not an investment, and paying full timber value for one leaves you owning bare land you paid a premium for.

The friction on a timberland investment is lower per year than on most alternatives and much higher per transaction than on anything listed. Both halves matter.

Round-trip transaction friction

Buying costs a cruise, a survey, a title opinion, legal work and closing costs — for a small tract, call it three to four percent of the purchase price, and note that most of it is fixed rather than proportional, so it hurts a 40-acre buyer far more than a 400-acre one. Selling costs a land brokerage commission, typically in the mid single digits, plus closing costs. A round trip therefore runs somewhere near eight to ten percent, which on a ten-year hold is roughly a point a year of drag before anything else happens.

Carrying costs

Property tax is the largest recurring line and the most variable, because every timber state offers some form of current-use or productivity valuation that taxes forestland on its capacity to grow wood rather than on its highest and best use. Enrolling is close to mandatory economically; the trade is a commitment period and a rollback tax if you convert the land to another use. Then a management retainer to the consulting forester, then the small permanent items: boundary maintenance, road and gate upkeep, firebreaks, liability insurance, and in the South the periodic prescribed burn. The worked example below assumes $17 an acre a year for all of it, which is a plausible southern number and an assumption, not a verified figure.

The costs the index does not charge you and the market does

Two more sit outside both lists. The first is the harvest’s own cost structure: stumpage is what remains after logging and hauling are paid out of the delivered price, so every increase in fuel, trucking wages or insurance is a direct subtraction from the landowner’s cheque, invisible in the lumber price. The second is reforestation: after a clearcut you must site-prep and replant, and that bill arrives in the same year as the harvest revenue. A model that shows the clearcut and forgets the replant overstates the exit by the cost of the next rotation.

The US tax treatment of timber is one of the genuinely favourable corners of the Code, and it is favourable in a specific way that most owners under-use. Nothing here is tax advice; it is the terrain, and rates, caps and elections change. Confirm each of the mechanics below with a tax adviser before relying on it.

Section 631 and capital-gain treatment

The default rule would treat timber sale proceeds as ordinary income. Section 631 overrides that in two ways. Section 631(b) treats the disposal of standing timber held more than a year — both the pay-as-cut contract that retains an economic interest and the outright lump-sum sale — as a Section 1231 transaction, which means long-term capital-gain treatment, no self-employment tax, and a net Section 1231 loss that can offset ordinary income. Section 631(a) lets an owner who cuts timber for use in a trade or business elect to treat the cutting itself as a deemed sale at fair market value on the date treated as cut, splitting the result into a Section 1231 gain up to that value and ordinary income beyond it.

Either election is reported on Form 4797 and Schedule D, and Form T (Timber) is filed when you claim depletion or make a 631(a) election. The practical consequence is that a landowner selling standing timber is generally taxed at long-term capital-gains rates rather than ordinary rates. That is exactly the treatment farmland gets and exactly the treatment art, wine and watches do not — those pay up to 28% as collectibles, as the hub’s flagship farmland guide sets out.

Depletion, and why basis allocation at closing matters

When you buy a tract you buy two assets, and you should allocate the purchase price between the land account and the timber account at closing, supported by the cruise. Tons removed later are charged against the timber account through cost depletion, which reduces the taxable gain on every harvest. An owner who never made the allocation has no depletion basis to claim and pays tax on the gross. This is a fifteen-minute exercise at closing worth several thousand dollars a decade later, and it is the most commonly skipped step in small-tract ownership.

Reforestation and the Section 194 deduction

Section 194 allows a current deduction of reforestation expenditures of up to $10,000 a year per qualified timber property — $5,000 on a married-filing-separately return — with the excess amortised over 84 months, that period beginning in the middle of the tax year in which the amortisable basis arises. A taxpayer with several qualified timber properties chooses how to allocate the cap among them. The point of the provision is that it converts the cost of the next rotation from a capitalised item recovered decades later into a near-term deduction.

Property tax, the estate, and what carries over from farmland

Current-use valuation is discussed above and is the single largest tax lever an owner pulls annually. On the estate side the machinery is the same as farmland’s, which the flagship guide fact-checked in September 2026: stepped-up basis at death wipes out a lifetime of appreciation for heirs; the federal estate-tax exemption stands at $15 million per person and $30 million per couple from January 1, 2026 under the One Big Beautiful Bill Act (P.L. 119-21); and Section 2032A special-use valuation, capped at a $1,460,000 reduction for 2026 deaths under Rev. Proc. 2025-32, can apply to timberland held in a qualifying family business.

Because timberland is real property it also remains eligible for a 1031 like-kind exchange, which the 2017 tax law preserved for real estate and removed for everything else. Rayonier structured its August 2026 swap of Washington acres for Alabama and Texas acres as exactly that, which is the clearest illustration available of what the provision is for: moving a timberland basis from one region to another without triggering the gain.

A ten-year hold of 125 acres of southern pine returns 4.49% a year before tax if the land compounds at the South’s own thirty-year appreciation rate, and 2.34% if it does not. Here is the whole calculation behind those two numbers: every fee, every carrying cost, a tax estimate, and the same tract run twice. All three timber prices are TimberMart-South’s published fourth-quarter 2025 south-wide averages; every other input is an assumption and is named as one.

The tract and the assumptions

125 acres of planted southern pine at $2,000 an acre, or $250,000, bought at stand age 16 with 45 green tons an acre standing, split 15% sawtimber, 45% chip-n-saw and 40% pulpwood. Closing costs of $9,000 — cruise, survey, title opinion, legal and recording — bringing total capital in to $259,000. Carry of $17 an acre a year, or $2,125, covering current-use property tax, the consulting forester’s retainer, and roads, boundaries, firebreaks and insurance.

Growth of 5.5 green tons an acre a year. A thinning in year 4 removing 20 tons an acre at a blended $9 a ton, and a second in year 8 removing 18 tons an acre at a blended $12 a ton, each less an 8% forester’s commission — inside the 3–10% extension range quoted above. Sale of the whole tract at the end of year 10 at a 5% brokerage-and-closing cost, with the standing timber discounted 10% for a whole-tract sale rather than a competitive timber sale.

The three product prices are TimberMart-South’s published south-wide averages for the fourth quarter of 2025: sawtimber $23.23 a ton, chip-n-saw $17.85 (the midpoint of a reported $17.50–18.20 range) and pulpwood $5.96. The middle rung matters more than it looks, because chip-n-saw is 45% of the standing volume at purchase: every dollar a ton on it shifts value out of the appreciating land account and into the timber account, and the answer at the end of this section moves with it.

The $2,000-an-acre purchase price is an assumption, and a deliberately conservative one. Two dated transactions bracket it. Weyerhaeuser agreed in May 2025 to pay $375 million for about 117,000 acres of mature North Carolina and Virginia timberland, roughly $3,205 an acre. Resource Management Service took 57,000 acres in Alabama and Texas off Rayonier in August 2026 for $146 million, roughly $2,560 an acre. Both are institutional-grade tracts bought at institutional scale; a 125-acre parcel at $2,000 assumes you are buying something less good, more cheaply, which is the realistic case for an individual.

What the tract earns before you sell it

On those figures, the purchase price splits into $626 an acre of standing timber and $1,374 an acre of bare land. The year-4 thinning grosses $22,500 and nets $20,700 after commission; the year-8 thinning grosses $27,000 and nets $24,840. Against them run ten years of carry at $2,125, or $21,250. Net operating cash across the decade is $24,290, which is $2,429 a year against $259,000 invested — 0.94% a year.

That number is the honest headline of direct ownership and it is why the asset is institutional. Between harvests a forest is a bill. Everything else in the return is the standing inventory getting larger, and you cannot spend it until you cut it.

What the tract is worth in year ten

Ten years of growth at 5.5 tons an acre against 38 tons an acre removed in thinnings leaves 62 green tons an acre standing at stand age 26, now 70% sawtimber, 22% chip-n-saw and 8% pulpwood. At the same prices that is $1,281 an acre of timber, up from $626 — the biological engine doing exactly what it is advertised to do, more than doubling the timber value with no price appreciation whatsoever.

Base case. Let the bare land compound at 3.8% a year, the South’s own thirty-year appreciation rate in Forisk’s decomposition. Bare land reaches $1,996 an acre, the tract is worth $3,149 an acre after the whole-tract discount, and the sale grosses $393,605 and nets $373,924. Total returned across the decade is $398,214 against $259,000 in: a pre-tax internal rate of return of 4.49%, and roughly 3.89% after a flat 15% federal capital-gains assumption on the gain.

Downside case. Hold bare land flat in nominal terms and nothing else changes. The tract is worth $2,528 an acre, the sale grosses $315,944 and nets $300,147, total returned is $324,437, and the pre-tax IRR is 2.34%, about 2.00% after the same tax assumption. Note what did not happen in the downside case: no fire, no beetle, no hurricane, no mill closure, no forced sale. Land values merely stopped rising.

The comparison, and what it does and does not prove

$25,000 across the two listed REITs at our tape’s September 4, 2026 closes buys 543 Weyerhaeuser shares and 610 Rayonier shares, paying $1,090.52 a year in base dividends — a 4.36% cash yield, before any supplemental dividend or buyback under Weyerhaeuser’s cash-return framework, and before any share-price move in either direction. It requires no cruise, no forester, no property-tax filing and no buyer at the end.

The direct tract wins the base case by thirteen basis points: 4.49% against 4.36%, before tax, on a ten-year hold that requires a cruise, a forester, a property-tax filing, a decade of carry and a buyer at the end. That is the entire prize, and it is dwarfed by the one assumption the exercise turns on: the gap between the base and downside cases is 2.15 points of annual return — 4.49% against 2.34% — all of it the land-price assumption, and roughly sixteen times the margin by which the tract wins. Both of those margins are our arithmetic on the returns above. A fraction of a percentage point off that assumption erases the margin entirely.

After tax the tract does not win at all: 3.89% against a dividend yield that is itself pre-tax. And in the downside case, where nothing goes wrong except that land values stop rising, the tract loses by more than two points.

That is the finding. Direct timberland ownership is not, on these numbers, a cash-flow investment that happens to appreciate; it is a land investment with a modest biological subsidy and a very long settlement period. Anyone who tells you the trees pay for the land should be asked to show the carry line.

Timberland’s reputation for safety comes from a smoothed index and a long horizon, and both hide the tail. The losses in this asset class are not gradual.

Fire, wind and insects

A standing inventory is a combustible asset that cannot be moved. Wildfire in the West, hurricanes in the Southeast and the southern pine beetle across the South each destroy merchantable volume in days, and the market response makes it worse: a regional blowdown floods local mills with salvage wood, so the price collapses at exactly the moment every affected owner needs to sell. Timber insurance exists and is expensive relative to the value at risk. Diversification across regions is the institutional answer and is not available to someone buying one tract.

Basis risk, which is the quiet one

A stumpage price is local. When the pulp mill or the sawmill inside your haul radius closes, your asset repriced overnight and no index recorded it. The southern pulpwood collapse is the working example: more than ten major pulp facilities closed in the US South between 2023 and 2025, taking over 25 million tons of annual fibre demand with them, and pine pulpwood fell to $5.96 a ton in the fourth quarter of 2025, 46% below its 2022 peak. Owners two counties from a closure and owners two counties from a new mill experienced completely different assets.

Illiquidity, and appraisal illusion

There is no bid for your tract on a Tuesday. A private sale takes months, the buyer pool for a small tract is local, and the marketing period lengthens exactly when prices fall. For fund investors the analogous risk is the appraisal: a quarterly NAV is a valuer’s opinion informed by the manager’s own cash-flow assumptions, and it will be smoother and later than the truth. Forisk’s 2025 review of the transaction market — rising values, falling volumes, buyers unable to underwrite at the asking price — describes the condition in which appraised values and clearing prices diverge.

Fraud, and the fact that nobody can count trees from a desk

The asset’s defining verification problem is that its value is an inventory estimate of an object you cannot see from the road. That is an invitation, and the industry’s largest scandal took it. Sino-Forest, a Toronto-listed Chinese forestry company, was accused by Muddy Waters in a report published on June 2, 2011 of massively exaggerating its assets.

The chronology from there is short. The Ontario Securities Commission issued a temporary cease-trade order on August 26, 2011. The company obtained creditor protection under the Companies’ Creditors Arrangement Act on March 30, 2012 and was delisted from the Toronto Stock Exchange on May 9, 2012. In July 2017, after one of the longest hearings in the regulator’s history, an OSC panel found that the former chief executive and four other senior managers had engaged in deceitful or dishonest conduct over the company’s standing-timber assets and revenues. Six years from allegation to finding, on an asset nobody could count from a desk.

The lesson generalises to every scale: the cruise is the asset. Commission your own, from a forester you hired and who is not paid by the seller, and read the sampling error.

Concentration and the single-tract problem

An institutional timberland portfolio spans regions, species, age classes and mill markets. A 125-acre tract spans one of each. Every statistic quoted about timberland’s low volatility and low correlation is computed on a diversified, appraisal-smoothed index, and none of it describes what happens to one tract when one mill closes or one storm lands.

1. Decide whether you want the asset or the exposure. If the answer is the exposure, stop after step 2. The listed route captures most of the economics for a fraction of the work, and the worked example above says so in dollars.
2. Buy the listed REITs first, and hold them for a year before doing anything else. Two names, a small position, and twelve months of watching how the shares behave against lumber prices and housing starts will teach you more about the cycle than any amount of reading. If the volatility of a listed wrapper bothers you, that is useful information: the private version has the same economics with the price hidden.
3. If you still want land, pick the wood basket before the tract. Map the mills inside a sixty-mile radius, learn who owns them, and find out what has closed or curtailed in the last three years. This single step determines more of your return than the negotiation will.
4. Hire a consulting forester before you make an offer, on your side of the table. Ask for references from landowners rather than from brokers, and agree the retainer and the sale-commission schedule in writing at the start.
5. Commission your own cruise and read the product-class assumptions. Check the diameter distribution against the claimed sawtimber share. If the seller declines an independent cruise, you have your answer.
6. Run title, boundaries and legal access before you run the numbers. Severed minerals, missing access and unrecorded easements are deal-enders, not price adjustments.
7. Allocate the purchase price between land and timber at closing, in writing. Set up the timber account and the depletion schedule on day one; enrol in the state’s current-use property-tax programme in the first year you are eligible.
8. Write the ten-year cash-flow model before you close, with the carry line visible. If it does not clear your hurdle at prevailing regional stumpage with no price escalation, do not close.

Five readings tell you whether the case for this asset is improving or decaying. Each has a threshold and a source you can check yourself.

Southern pine sawtimber stumpage

The benchmark. TimberMart-South’s quarterly southern average was $23.23 a ton in the fourth quarter of 2025. A sustained move back above roughly $30 a ton would mean sawmill capacity in the South has finally caught up with fibre supply and the landowner is capturing part of the lumber margin for the first time in two decades. A move below $20 would mean the sawtimber rung is following the pulpwood rung down.

Pine pulpwood and mill announcements

Pulpwood at $5.96 a ton in the fourth quarter of 2025 is the impaired rung, and its recovery depends entirely on new fibre demand replacing the 25 million tons of annual capacity that closed between 2023 and 2025. Watch mill announcements rather than prices: a new pulp, pellet or mass-timber facility inside a haul radius is worth more to a local owner than any national statistic.

Housing starts, especially single-family

Single-family starts were running at 808,000 a year in July 2026 against total starts of 1,239,000, per the Census Bureau and HUD. Single-family construction is the lumber-intensive component; a sustained recovery through 1.0 million single-family starts a year would be the demand condition that lets sawmill capacity fill and stumpage bids rise.

The NCREIF Timberland Index’s income component

The cash-yield component, not the total return. Forisk’s thirty-year average is 3.4% a year and falling. If the index’s income component turns up while appraised values are flat, the asset is re-rating toward cash and a buyer is being paid to wait. If total return holds up on appreciation alone while income drifts lower, the appraisals are doing the work — the condition Timberland Investment Resources flagged after fourteen straight quarters of appreciation-led returns through the first quarter of 2025.

Transaction volume against appraised value

Forisk’s transaction reviews are the tell, and the 2025 review is stark: 46 completed US deals covering 940,000 acres, against a longer-run average of about 2.8 million acres a year since 2015, with the South taking roughly 56% of the volume. Per-acre values have compounded at about 6.8% a year since 2021 while that volume shrank. Rising appraised values with falling deal volume means the marks and the market disagree. Convergence in either direction is a real signal; a widening gap is a warning that the exit you are modelling does not exist at the price you are modelling it.