IAPB · Methodology
A sawmill: what it is worth to pay for a longer rotation and habitat practices in the forests that supply it, and whether buying nature credits gives a better return than funding the same work directly.
This document sets out a method for working out the return a company can expect when it pays to improve how the forests in its timber supply chain are managed. It also tests whether buying nature credits gives a better return than funding the same work directly. It explains the question the method answers, the data it uses, how a change in the way a forest is managed becomes a change in what a company is worth, and which of its inputs are evidence and which are judgement.
The returns from protecting and restoring nature are documented for whole sectors and countries. There is little public analysis of the return to a single company, and none that measures what nature credits add to it. The method is meant for a company or investor to apply to their own business or client.
The document can be read without knowing the tool. Parts 1 to 6 describe the method. Part 7 says how far it can be trusted and what it leaves out. Part 8 describes what the programme does for nature. The appendices include what is needed to build the model again.
A company runs a sawmill and buys logs from a particular forest. How that forest is managed decides the size and grade of the logs the mill receives. A longer rotation, thinned with more care, grows bigger and better logs than a short one. That affects the company's finances and valuation.
The method answers two questions in turn. First, what financial value does better management of the forest offer the company and its business model? Second, does buying nature credits give a better return than funding the same work directly?
The answers are given as company value and cash flows, the measures a company uses for capital decisions.
| To | Read |
|---|---|
| Understand what the method does | Part 1 |
| Use the tool and read its results | Parts 1 and 6 |
| See how reliable the results are | Parts 1 and 7 |
| Build the model again | Parts 2 to 5 and the appendices |
| See what the programme does for nature | Part 8 |
For a short summary, read Part 1 and §7.2.
The model calculates the same forest and mill twice and reports the difference.
The first future is doing nothing. Nobody pays for a change. The owner manages the forest as they would anyway, on the shorter rotation that gives the owner the best return. The mill buys the wood that this produces.
In the second future the company pays for the programme, which is a different plan for managing the forest. The forest is grown on a longer rotation and thinned twice, not once. A set of habitat practices is carried out alongside. The wood that reaches the mill in the end is larger and of better grade, and the mill's accounts improve.
Both futures use the same land. The site describes the mix of land, such as how much is streamside or retained patches. The budget sets how many hectares are enrolled each year (§2.4). The company pays for the difference between the two futures, either directly or by buying credits.
To answer the second question, whether buying credits gives a better return than funding the work yourself, the model values the mill a third time, as a credit buyer (§1.4).
Figure 1 — Doing nothing and the programme, in the same model.
The return has a benefit side and a cost side.
The benefit comes through five channels. They do not count anything twice, because each acts on a different quantity.
At Reference the mill runs below its capacity ceiling, with 10% spare capacity (§3.3). So better recovery lets it saw more lumber from the same logs. Recovery, stem size and grade are therefore sales kept or added, because the mill saws more lumber and sells it at a better price. Operating leverage spreads the fixed costs over more output, so each unit of lumber costs less. It is a cost avoided. Recovery saves log cost only when the mill is at its ceiling.
Cheaper borrowing shows up in value created only. It never appears in extra cash, for the reason given in §1.3.
The cost of the programme has three parts. The owner's payment (silviculture) is what makes the owner willing to change how the forest is managed. Nature interventions are the habitat practices carried out alongside. They are not part of either management plan. MRV and verification cover monitoring, reporting, independent verification of the work, and the legal agreement that stops the owner taking the payment and then cutting early. Funding it yourself also pays a one-off coordination cost of USD 0.4m (§5.5).
When the company buys credits, the developer pays for the nature interventions and for MRV and verification before delivery. The developer's financing charge is added to those two. The owner's payment is made at delivery on both routes, so no financing charge applies to it (§5.3).
Figure 2 — How the return is made up.
The model reports the return in two ways. Both are divided by the same amount, what you commit: the company's own programme payments from 2026 to 2080, not discounted. They differ in what is divided.
value created = company value with the programme − company value doing nothing
(both modelled at the start of 2026)
extra cash = Σ free cash flow with the programme − Σ free cash flow doing nothing
(2026 to 2080, not discounted)
what you commit = Σ the company's own programme payments, 2026 to 2080, not discounted
times what you put in = value created ÷ what you commit (and extra cash ÷ what you commit)
Extra cash is free cash flow with the programme, less free cash flow doing nothing, summed from 2026 to 2080 and not discounted. Value created is company value with the programme, less company value doing nothing, at the start of 2026. Each is divided by what you commit to give a number of times what you put in.
The two measures answer different questions.
Extra cash measures the cash the programme produces, after tax and after paying for the programme. It leaves out any change in what the business is worth, so it leaves out cheaper borrowing entirely. It is used to compare the programme with other investments. It puts a nature credit on the same footing as any other capital project competing for the same budget.
Value created measures what the programme is worth to the company. It includes the effect on operations, the change in working capital, cheaper borrowing, and the value of the years after 2080, because the improvement lasts beyond 2080. It is used to value the programme. It says how much the programme adds to what the company is worth.
In forestry the two measures rank the routes differently at the default settings. Buying credits creates more value: USD 1.2m against USD 0.9m. Funding it yourself produces more extra cash: USD 14.2m against USD 12.3m (§6.4).
The main cause is the developer's financing charge. It adds about USD 6.4m to what a credit buyer pays. Extra cash is not discounted, so it counts the charge in full, about USD 4.9m after tax. In value created the credit payments come later and are discounted, so the charge costs about USD 0.6m in today's money. The duration premium (§1.4) adds a little more to value created, about USD 0.04m, and nothing to extra cash. The longer the wait for wood, the larger the charge and the wider the gap between the two measures.
The gap comes from how the two measures are built. Report both.
Each multiple compares a total over many years with the total committed over 55 years. It is the value created, or the extra cash, for each USD 1 committed. It is not a yearly rate of return.
Figure 3 — Value created and extra cash.
The same programme can be paid for in two ways, and the model values both.
If the company funds it itself, it pays the programme costs as they fall due and treats them as an expense. It also pays a one-off cost to organise the arrangement.
If it buys credits, it pays nothing until the enrolled stand is harvested and its wood is delivered (called delivery), 9 years after enrolment for loblolly. A developer borrows to pay for the work and is paid back on delivery. The cost of that borrowing is the developer's financing charge, and it is included in the credit price. The company treats the purchase as an asset and writes it off over 10 years.
Both routes pay for the same programme and get the same wood, so the benefit side cancels between them. What is left is cost, timing, and the differences below.
| Difference | Favours | How it works |
|---|---|---|
| Credit premium (duration) | Buy credits | A contract that runs for many years is assumed to lower the cost of capital by a further 0.1 percentage points, scaled by coverage, so about 0.02 percentage points at full coverage (§4.6). The general methodology also allows a premium for independently verified credits, the verification premium used in agriculture and water. Forestry does not use it |
| Coordination | Buy credits | Funding it yourself includes a one-off cost of USD 0.4m to set up the arrangement (§5.5) |
| Risk transfer | Buy credits | Credits are paid for only on the share of work delivered, plus a charge for the risk on the rest (§5.4) |
| Tax timing | Fund it yourself | An expense gets its tax relief at once. A credit written off over 10 years spreads it out (§4.2) |
| Developer's financing charge | Fund it yourself | The developer borrows to pay for the work until delivery, and the cost of that borrowing is included in the credit price (§5.3) |
The list is tested. Switching every difference off at once, and setting the developer's rate equal to the company's cost of capital, makes the two routes equal (§7.4). This shows that no difference is missing from the table.
Figure 4 — Buying credits or funding it yourself.
The physical measure is the difference in mature sawtimber yield between two whole management plans on the same ground.
For loblolly pine the comparison is between a 24-year rotation with one thinning and a 33-year rotation with two. The shorter plan yields 51.9 tons a hectare (US short tons, about 0.91 tonnes each) of sawtimber at final harvest. The longer plan yields 140.8. The difference of 88.9 tons a hectare (the sawtimber uplift) is what the programme adds. The whole benefit depends on this one physical quantity.
The yield figures are sawtimber at final harvest. Wood removed at the thinnings is a separate product and does not reach the modelled mill. Thinnings are small stems that go to pulp, chip-and-saw or panel markets. The mill in this model buys sawlogs. §5.1 says how the thinnings are counted on the owner's side.
The model does not simulate growth. It takes the published yield comparison and applies it to the enrolled area. So the model does not include the uncertainty in how a stand grows. There is one yield figure for each management plan, not a range.
The two species work in different ways. The difference is in what the payment is for. The tool's settings are the same for both.
For loblolly, the thinnings are part of both management plans, so the thinning itself needs no funding. What the owner will not do alone is wait 9 more years, because waiting lowers the land's value at any commercial discount rate. So the credit pays the owner to wait. The payment is the difference in land value between the two management plans.
Historically the wood removed at the thinnings paid for them. At the defaults the model assumes the thinned wood cannot be sold (§5.1), so the owner earns nothing from it. It also assumes the thinning operation costs nothing. With no thinning income the longer plan is worth less to the owner, so the payment needed is larger. It is USD 499 a hectare, against USD 203 if thinnings could be sold. If the thinning had to be paid for, the payment would be higher still. The model has an input for this cost.
For eastern white pine the thinning itself is the extra work. The published field study records that white pine stands in the region are not often thinned. There is no local market for small stems, and thinning returns little to the owner. So the credit pays for the thinning, at its full cost of USD 988 a hectare, because no sales offset it.
This means the same credit does two different things in the tool, depending on the species chosen. The argument that the work would not happen without the payment differs too. For loblolly, a commercial owner would not wait 9 more years. For white pine, a commercial owner would not thin at all.
The two species do not have equal evidence, and the tool says so. The loblolly comparison is a published study of rotation economics. Four separate finance records from the same region and species support it. The white pine evidence is a single study of a site of about 1.1 acres. It was not repeated, and its authors say the results do not apply more widely. Eastern white pine gives a negative result at every setting tested. This is the expected result. At 9% a year for 25 years, whether that is the developer's rate or the company's own cost of capital, USD 988 a hectare grows to about 8.6 times its original value, and the extra timber cannot repay that.
The physical evidence compares two management plans under observed conditions. It is not a climate projection, and the method makes no claim about future climate.
This is a large difference from agriculture, which uses the same framework. There, the physical measure is a projection of how suitable the land will be for a crop under stated emissions scenarios. In forestry there is no climate setting and no choice of climate scenario. The method says nothing about how the stand would respond to a warmer climate. The benefit modelled is a growth and competition effect. It comes from changing how dense the stand is and how long it grows, and it depends little on the climate path.
So this method gives no forestry returns that depend on climate. The study of how the stand responds is based on observed conditions. Taking it beyond that range would weaken the result.
A site describes a mix of land, not a size. It has a total area in hectares. The habitat set-asides are entered as hectares within it. The productive share is worked out from them. It is the part of the site still growing timber, once each set-aside is weighted by how productive its ground is. The programme budget sets how much land of that mix is enrolled. Site area only sets the share of land set aside for nature. At the defaults nothing is set aside, so site area changes no result. Enrolment continues at the same rate every year to 2080, across the area the mill buys wood from.
There are two kinds of habitat practice. They behave differently, and the tool shows them separately.
Treatments are carried out across the area and take nothing out of production. Prescribed burning is the example. They cost money and leave the harvest as it is.
Set-asides take land out of harvest. Streamside buffers and retained tree patches are the examples. They cost money and reduce the wood that reaches the mill, so they reduce the return in two ways. The sawtimber uplift is multiplied by the productive share of the site (§3.5).
Set-aside land is not always as productive as the site average, and this differs by type. Streamside buffers lie along streams by definition, so they are on wetter ground with fewer trees. Retained patches are chosen for their old trees and habitat, not for the quality of the ground, so they are close to the site average. The model has a productivity factor for each type: 0.55 for streamside, 0.95 for retained patches, and 1.00 for other set-aside land. The ground of other set-aside land is unknown, so it gets no discount. Setting aside 5% of the area as streamside costs 2.75% of the volume, not 5%.
Using one factor of 1 for all three types would assume that set-aside land yields exactly the site average. That overstates the loss on the type of set-aside the model is most likely to be asked about.
Every set-aside costs wood, so each mix of practices has a set-aside share at which the return turns negative. The model reports that share. With streamside set-aside alone, counting the wood it removes but no extra cost, funding it yourself turns negative at about 31% of the enrolled area; buying credits does not turn negative within the 40% limit. Buying credits stays positive at higher set-aside shares, because of its advantages listed in §1.4. A validation test confirms the model finds the share accurately enough to quote to the nearest 1%. The tool's limits on set-aside area stay as they are, because they reflect what is physically plausible. It warns above 25% and treats results above 40% as invalid.
Ten sites are supplied. The sites are illustrative. For each species they cover the main growing regions and market conditions. The five loblolly sites are large industrial blocks of 850 to 3,600 hectares across the timber belt of the US South, in Texas, Mississippi, Georgia and Alabama. They include the area where the published thinning record was measured, and markets where wood is plentiful and where it is scarce. The five white pine sites are scattered family-forest tracts of 320 to 1,200 hectares. They are in the county of the field study and its neighbours in the southern Appalachians, plus two northern sites with different markets. No screening rule was applied.
The programme budget sets how much land is enrolled, and budgets can differ widely. That is why the tool offers to size the mill to the programme (§4.5).
Coverage is the extra sawtimber the programme delivers to the mill each year, as a share of the mill's yearly log intake. The model treats these extra tons as the upgraded wood. It is the model's most important figure. Every benefit channel scales by it.
The conversion ratio is the tons of logs the mill needs for each thousand board feet of lumber. It is 4.30 at the base.
mill intake = lumber output × base conversion ratio
delivered tons = hectares harvested that year (§3.5) × sawtimber uplift × productive share
× share reaching this mill × delivery rate
coverage = min( 1 , delivered tons ÷ mill intake )
At the default settings the mill uses 860,000 tons of roundwood a year. The programme enrols 2,472 hectares a year. Each hectare adds 88.9 tons of sawtimber, of which 64.2 reach the mill after the share sold elsewhere and the delivery rate. Coverage is 18.5% once the first 3 years of enrolment have been harvested (§3.5).
Coverage is worked out against a fixed base intake, not against actual output. Using actual output would make the chain circular, because output itself depends on coverage through the channels below.
A larger, better-formed log improves several things at once. These are the five channels of benefit in §1.2. They are independent because each acts on a different quantity. The formula below shows why they add up without overlapping.
MBF per hour = stems per hour × tons per stem × MBF per ton
(fixed by the (stem size (recovery
saw line) gain) gain)
Recovery raises the lumber the mill gets from a ton of logs. The model shows this as a fall in the conversion ratio, from 4.30 tons per thousand board feet towards a lower figure as coverage rises. Below the capacity ceiling the mill uses this to saw more lumber from the same logs, so recovery adds sales (§3.3).
Stem size raises the tons a shift can process. It is defined on tons an hour, not board feet an hour. Defining it on board feet would include recovery and count it twice.
Grade raises the price of the lumber produced.
Operating leverage turns extra output into margin, because a set share of conversion cost is fixed for each shift and does not vary with each unit. It spreads that fixed cost over more output, so each unit costs less.
Cheaper borrowing lowers the cost of capital used to discount the mill's cash flows. It scales by coverage, so a programme that upgrades nothing gets no reduction in the cost of capital.
A sixth channel exists in the model but is not shown in the tool. Where the upgraded wood is closer to the mill than the wood it replaces, haul cost falls. This channel is always set to zero, because its value depends on the area a particular mill draws wood from and cannot be shown in the tool.
The mill has spare capacity up to a ceiling. The ceiling decides which of two market states the mill is in.
potential output = base intake × (1 + stem size gain × coverage) ÷ conversion ratio actual output = min( potential output , lumber output × (1 + capacity headroom) )
Below the ceiling, better logs raise output and the mill buys more wood. At the ceiling, output cannot rise, so better recovery means the mill needs fewer tons for the same output and its wood bill falls instead.
At Reference the mill has 10% spare capacity (its capacity headroom, Appendix B). The programme raises output by 1.87% at the defaults, so the mill stays below its ceiling. The ceiling limits output only when spare capacity is below about 2%.
The user does not choose which situation applies. The ceiling decides it, and the tool shows the point at which it changes.
Every channel scales by coverage, and coverage counts only wood that reaches this particular mill. If the enrolled stands sell their timber to a different buyer, the company funds the programme and receives nothing through its own accounts.
The share of enrolled wood that reaches this mill is an input to the model. A right of first refusal in the sourcing contract makes a high share realistic. A company would not buy a nature credit from a forest it does not buy wood from. The Reference level is 85%, which allows for some wood that is sold elsewhere. The Log market setting runs it from 100% to 60% (§6.2).
The return is sensitive to this share. Below about 71%, funding it yourself no longer repays its cost. Below about 66%, buying credits does not either. At zero the company loses today's value of everything it agreed to pay.
In the general methodology, protection is the share of the physical effect that the company gets. In forestry, coverage (§3.1) does the same job. It scales every benefit and the cut in the cost of capital, and it builds up over time:
hectares delivering(y) = hectares enrolled in year (y − H) coverage(y) = min( 1 , hectares delivering(y) × delivered uplift ÷ mill intake )
delivered uplift = sawtimber uplift × productive share × share reaching this mill × delivery rate
H is the wait between enrolment and harvest. It depends on the enrolment age, the age of the stand when it enters the programme (24 years for loblolly by default). For loblolly the wait is 9 years. Enrolment builds up over 3 years. It is a third of the full rate in the first year, two thirds in the second, and the full rate from the third. So the first upgraded wood reaches the mill in 2035, coverage is 6.2% that year and 12.3% the next, and it rises to its full 18.5% in 2037. Before 2035 the mill gets nothing.
The delivery rate is the share of the work that succeeds. It is 0.85 at Reference. It is the forestry name for what the general methodology calls effectiveness. Like effectiveness in the other use cases, it reduces the wood that reaches the mill on both routes, and it also sets what a credit buyer pays (§5.4). A less successful programme lowers the benefit on both routes. It also lowers what a credit buyer pays, so buying credits gains ground as the delivery rate falls.
The company receives the benefit as a better margin on turning logs into lumber. It shows in its own income statement: better recovery, more output, better grade and lower cost for each unit.
This makes the benefit easier to check. A share of someone else's profit has no line of its own in a set of accounts and cannot be audited from them. A margin on turning logs into lumber can.
The owner's payment is a price. Its size is set by what the owner needs (§5.1), and it does not change with how much the mill benefits.
The model builds a full income statement, balance sheet and cash flow statement for each of three futures, for every year from 2026 to 2080.
A simpler model was possible. It was turned down because the questions the method answers are accounting questions. Whether a payment is an expense or an asset changes when tax relief arrives. Whether extra output raises stock changes working capital. Whether a benefit affects company value or only cash depends on where in the statements it is recorded. A model that reports a benefit without saying which line it appears on cannot be checked against a real company.
The balance sheet is a check. It balances to the limit of the computer's precision in every year of every scenario.
| Supply path | Cost treatment | |
|---|---|---|
| Doing nothing | Wood supply as it is | None |
| Fund it yourself | Upgraded share equals coverage | An expense, when it is paid |
| Buy credits | Same coverage | An asset written off over 10 years; paid at delivery, with the developer's financing charge |
The accounting treatment is fixed and has no switches. Any difference between the two routes must come from the amount of cash paid, the timing of tax, or an advantage of one route that is named in §1.4. A validation test confirms there are no other differences.
Recovery lowers the conversion ratio, so each unit of lumber needs fewer tons of logs. Below the capacity ceiling, the mill turns this into more output from the same logs, which raises revenue. In the worked example the log bill rises slightly, because stem size also lets the mill buy and saw more wood (§6.4). At the ceiling, output cannot rise, so the log bill falls instead (§3.3). Stem size raises output. That raises revenue and, through the fixed share of conversion cost, raises margin faster than output. Grade raises the lumber price the mill gets. Operating leverage is not a separate line. It is how the first two affect margin. Cheaper borrowing does not appear in the statements at all. It changes the rate at which the cash flows are discounted.
When the company funds it itself, the programme payment is an operating cost. It appears above profit before interest, tax, depreciation and amortisation. When the company buys credits, the yearly write-off of the credits is a charge below that line, and the cash paid for them is recorded in investing activities.
Interest depends on debt, debt depends on cash flow, and cash flow depends on interest. The model breaks the loop with a one-year lag. Interest in a year is worked out on the net debt at the end of the year before. This is a common modelling convention.
The model company is the stylised mill the method values. It is described by published industry figures where they exist and by stated judgement where they do not. Appendix B lists every parameter with its label.
The model works out company value. It does not take it as an input, so no large unsupported figure is used as an input.
The mill can be sized to the site or held fixed. The choice matters when sites are compared. With a fixed mill, a larger site is automatically a larger share of the same intake. So it shows a better return for reasons that have nothing to do with the quality of the programme. Sizing the mill so that the enrolled wood is always the same share of its intake removes that effect. Across an 8-fold range of programme size, the spread in return falls from 18.1 percentage points to 1.8, about ten times smaller.
When the mill is sized to the site, fixed assets and overheads scale in a straight line with output. This is reasonable only for mills of a similar size. Do not use this mode outside about a third to three times the reference mill.
Company value in a given year is the value, in that year's money, of free cash flow from that year on, including that year's own flow. Free cash flow here is before any interest or debt payments. The value of the years after 2080 is added using a standard formula that assumes cash flow grows at a steady rate (the Gordon growth formula). The discount rate changes from year to year, because it scales by coverage and coverage changes as the programme builds up.
With the programme, the cost of capital falls by 0.25 percentage points times coverage. At full coverage of 18.5% that is about 0.05 points. A credit buyer gets a further 0.1 points times coverage, the duration premium. Without cheaper borrowing (no cut in the cost of capital and no duration premium), value created would be USD 0.8m funding it yourself and USD 1.0m buying credits.
FCF unlevered = EBIT × (1 − tax) + credit amortisation − Δ inventory − credit purchase
Depreciation and the capital spending that keeps the mill running are set equal, so they cancel. This is the same as assuming the mill keeps its assets at the same level and does not grow them.
The cost of capital is built up step by step. The published cost of capital for the paper and forest products sector is 6.93%, measured across six US firms in January 2026. Those are listed companies spread across several businesses. The model company is a single unlisted mill in a cyclical trade, with no spread across products or places. So the model adds about 2 percentage points. The result is 9.0%. The cost of debt is set at 6.0% against a sector figure of 5.29%, for the same reason.
This part sets out the cost side of the return.
The programme pays for three things, whichever way the company pays for it. The tool lists them separately.
| Line | Loblolly | What it is |
|---|---|---|
| Owner's payment | 498.94 USD/ha | The payment that makes the owner willing to change how the forest is managed |
| Nature interventions | 177.21 USD/ha | Habitat practices carried out alongside, outside both plans |
| MRV and verification | 148.20 USD/ha | Monitoring, reporting, independent verification, and the legal agreement that stops the owner cutting early (§5.5) |
| Total | 824.35 USD/ha |
Cost per hectare decides how much land a given budget enrols, and so sets coverage:
cost per hectare = owner payment + one-off practice cost + MRV hectares per year = programme spend ÷ cost per hectare
At the default settings a USD 1.6m annual budget at USD 647 a hectare enrols 2,472 hectares a year. The recurring nature line does not enter this division. It is charged each year on the hectares under contract.
The budget sets how many hectares are enrolled. It is not what is paid each year. The owner's payment is made at delivery and is increased by the owner's 6% a year over the 9-year wait, to about USD 843 a hectare. The nature line is added each year. At full scale the programme costs about USD 2.9m a year, and the company's third is about USD 0.96m (§6.4).
The nature line at the default is prescribed burning at USD 19.69 a hectare a year across the 9-year wait. The owner's payment and the MRV and verification line are charged once for each enrolled hectare, although MRV and verification are spent over the wait (§5.5). The nature line builds up year by year.
Habitat practices are priced one by one, not as a bundle. A published figure of USD 38 to 83 a hectare a year exists for a package of four practices. But that package includes a longer rotation and streamside buffers, which are paid for elsewhere in this model. The source also says its figure cannot be split into separate costs. Prescribed burning alone is priced from a 2022 survey of southern practice across 85,268 acres, at USD 78.74 a hectare for each burn. On a 4-year cycle that is USD 19.69 a hectare a year. The cycle is a judgement. A 5-year cycle would give USD 15.75.
Streamside protection is charged only above the legal minimum. Landowners in the US South must already leave streamside buffers under best-management-practice rules. Charging the company for the whole buffer would bill it for something the owner must do anyway. That part would happen without the payment, so the credit cannot claim it.
The two set-asides are priced on different bases, and this decides what the programme costs. Streamside zones are published per hectare of zone. The source measures a 45-metre strip. So the figure is scaled by the zone's share of the site before it enters the programme cost. Charging it against every enrolled hectare would overstate it about 25-fold at a 4% buffer.
Retained patches are published per hectare of stand, so they are not scaled. The source quotes them by level: USD 321 to 401 a hectare for 5% of the area, and USD 601 to 761 for 10%. A figure per hectare of patch would not change with the level. The cost nearly doubles when the share retained doubles. That only makes sense if the figure is per hectare of stand. A separate check supports this. Taking 5% of the final harvest value under the longer management plan gives USD 293 a hectare, just below the published range.
A third kind of practice uses pipeline rights of way. Pipeline rights of way that cross a tract are already kept as open, low vegetation under required vegetation management. Sowing them with a native mix of wildflowers and grasses creates habitat on ground that never grew timber.
So the corridor costs money and takes no productive land. That puts it with the treatments, not the set-asides, even though it occupies area. Its only effect on the return is its cost. It removes no wood. It reduces coverage only because the money spent on it could have enrolled more hectares.
Establishment costs USD 865 a hectare of corridor. This figure is built up from the seed cost and the cost share below. It is not a published rate. High-diversity native mixes cost more than USD 2,470 a hectare in seed alone. Against that, the Conservation Reserve Program's CP-42 practice pays up to half of the cost of establishment and mid-contract management, and adds a one-time incentive of USD 370 a hectare.
The corridor is the only practice that returns money to the landowner. That money is taken off the programme cost, not added to the mill's benefit, because it goes to the owner and not to the mill. Two sources of income are claimed for it, and only one has a figure. Forage and cover for nesting birds make the land better for hunting. East Texas leases run at USD 5 to 10 an acre a year, against a statewide median of USD 10 to 18 for managed ground. The difference, about USD 6.50 an acre a year, is about USD 16 a hectare. That is the rise in hunting-lease income counted for the corridor. Native seed sold into a tight regional market is the larger commercial claim. It has no published figure, so it is set to zero.
The owner's payment is worked out from the land's value under the two plans. It is set from that comparison and entered in the model as an input.
Land expectation value is the standard measure for comparing management plans of different lengths. It values a plan repeated for ever. This is needed because a 24-year rotation and a 33-year rotation cannot be compared over one cycle without favouring the longer one for having had more time.
At the default settings the land is worth USD 635 a hectare under the shorter plan and USD 134 a hectare under the longer one. The difference, about USD 500 a hectare (498.94 in the model), is what the owner gives up. So it is what the payment must cover.
The owner's payment is entered as USD 498.94, from the land-value comparison in this section. If thinning income or the owner's discount rate changes, the payment must be updated to match. A check flags any gap (§7.4).
The default assumes that thinnings cannot be sold. Market evidence and recent regional price records both point that way. Pulpwood prices have fallen to or below the weakest of the four historical price sets in the source. Without thinning sales the owner earns less from the longer plan, so the payment needed is larger.
If thinning sales return to their historical levels, the same calculation gives USD 203 a hectare. It also reproduces the published finding that the longer plan's land value is 79% to 85% of the shorter one's. A validation test checks the calculation against that published range.
The three thinning-income figures move together. They describe the same small wood sold into the same markets in the same years. They should be changed together.
Costs are shared. The model company shares the cost with two co-funders, so there are three funders in all. The company pays a third of the programme. At the default settings this matters a lot. With a single funder, value created would be −USD 7.1m funding it yourself and −USD 6.8m buying credits (Appendix C).
The co-funders do not take this wood. This matters because the number of funders divides the cost and enters no benefit calculation. Raising it lowers what the model company pays and leaves the timber it receives unchanged. That only makes sense where the other funders pay for something other than timber. Examples are a government, a conservation body, or a brand that wants the nature outcome itself. If they were other mills, they would take a share of the wood. The share of enrolled wood reaching this mill would then fall to about one divided by the number of funders.
The model does not change either input to keep them consistent. The user must choose them together. The tool shows a note where this assumption applies at the default, and a warning where the number of funders is high enough that most of the return comes from other funders paying.
The owner is paid at delivery, not at enrolment. The payment is increased by the owner's own 6% a year for each year of waiting, so the owner is no worse off for being paid later. For the company, this moves the payment out of the early years and into the later years when the benefit arrives.
When the company buys credits, the developer borrows to pay for the nature interventions and for MRV and verification, and is paid back when the wood is delivered. The cost of that borrowing is the developer's financing charge, and it is included in the credit price. The credit price is the cost of each hectare, with interest on those two costs up to delivery:
credit price per hectare = owner payment at delivery
+ MRV, carried from the year it is spent to delivery
+ practice costs, carried to delivery
credit purchase(y) = hectares delivering(y) × credit price per hectare ÷ funders
× payment factor (§5.4)
The practice costs and MRV are charged with interest at the developer's rate up to delivery, the year the credit is paid for. The owner's payment is made at delivery. It is increased by the owner's own 6% a year for each year of waiting, so the owner is no worse off (§5.2).
The rate is 9%, because the developer lends for a long time and carries the risk that the trees fail or are damaged before the wood is delivered. The rate is a judgement with no published source.
The rate moves the result a lot. At 9% buying credits creates USD 0.3m more value than funding it yourself. At 5% the gap is USD 0.6m. At 12% it is less than USD 0.1m. At 15% funding it yourself is ahead by USD 0.3m. The advantage of credits disappears at about 12.5%.
The developer's rate equals the company's own 9% cost of capital at the default. Below that, part of the advantage of buying credits comes from the developer funding the work more cheaply than the company could. That part is cheap financing and is not a benefit of the credit structure.
Buying credits passes delivery risk to the developer. The company pays on delivery, and does not pay for what is not delivered.
payment factor = delivery rate × ( 1 + (1 − delivery rate) × share of risk repriced )
= 0.85 × ( 1 + 0.15 × 50% ) = 0.914
The payment factor is the share of the full cost a credit buyer pays. A credit buyer pays only for the 85% of the work that is delivered. The developer carries the risk on the other 15% and passes half of it back, raising the price of each delivered credit by 7.5%. So the buyer pays 91% of the full cost. The delivery rate and the share passed back are both judgements. The delivery rate also reduces the wood that reaches the mill on both routes (§3.5). So a less successful programme lowers the benefit on both routes, and buying credits gains ground because the credit buyer pays for less.
The model understates this benefit because of how it is built. A model that produces one central set of results counts only the expected payment that is avoided. The real value of passing on the risk is that it shields the buyer from results that are worse than expected, which the model cannot show. So the true benefit is at least the figure reported here.
MRV and verification cover monitoring, reporting, independent verification of the work, and the legal agreement that stops the owner taking the payment and then cutting early. It costs USD 148.20 a hectare, a judgement with no published source (§7.5). MRV and verification are spent year by year over the wait, not paid up front, so interest on each year's cost runs only from the year it is spent. That reduces the amount the developer's financing charge applies to.
MRV and verification are paid on both routes. If the company funds the work itself, it pays its share of MRV and verification as it is spent. If it buys credits, the developer pays for MRV and verification and recovers the cost in the credit price (§5.3), with interest to delivery and scaled by the payment factor like the rest of the price. For MRV and verification, the only differences between the routes are the developer's financing charge, the payment factor and tax timing (§1.4). MRV and verification are not counted as an advantage of either route.
Coordination is the one-off cost of setting up the arrangement: USD 0.4m in 2026. Only the fund-it-yourself route pays it, and it is not shared. USD 0.4m is a low estimate. Organising several companies to fund one programme needs contracts, governance and reporting, so the real cost is likely to be much higher, and part of it recurs each year. A higher cost would favour buying credits, because only the fund-it-yourself route pays it.
Each use case reads left to right in four lanes: the forest and the programme; what happens to the wood; what it does to the mill; what the mill is worth. Ten boxes each show the step without the programme and with it.
| Box | Step |
|---|---|
| 1 | Species and site |
| 2 | The nature programme |
| 3 | Harvest rotation |
| 4 | Stem size and volume |
| 5 | Log supply to the mill |
| 6 | Log bill |
| 7 | Lumber output and grade mix |
| 8 | Annual profit |
| 9 | What the mill is worth |
| 10 | Buy credits, or fund it yourself? |
Choices are decisions the buyer makes: the species and site, how the mill is sized (§4.5), the habitat practices and their areas, the programme budget, the enrolment age, and the number of funders. Whether to buy credits or fund it yourself is not a choice. Both routes are always calculated.
Scenario settings are conditions the buyer does not control. There are five, each set to Favourable, Reference or Adverse (§6.2). There is no climate setting (§2.3). The tool opens with every setting at Reference.
Each setting can be opened to show the inputs it changes. If you change one of those inputs, the setting is marked Custom, so the settings shown always match the inputs in use.
Favourable is good for the mill, Adverse is bad, and Reference is the point every other reading is measured from. Two settings describe the market; three describe how the programme performs.
| Setting | Group | What it moves | Favourable | Reference | Adverse |
|---|---|---|---|---|---|
| Timber price | Market | Lumber price, USD/MBF | 540 | 450 | 360 |
| Log market | Market | Share of enrolled wood reaching this mill; delivered log cost, USD/ton | 1.00; 38 | 0.85; 41 | 0.60; 45 |
| Programme cost | Programme | Streamside and patch-retention rates, USD/ha; burning, USD/ha/yr | 252; 321; 15.75 | 956; 361; 19.69 | 1,659; 401; 26.25 |
| Mill response | Programme | Recovery gain; stem-size gain; grade premium, USD/MBF; fixed share of conversion cost | 9%; 6%; 18; 0.50 | 6%; 4%; 12; 0.40 | 3%; 2%; 6; 0.30 |
| Credit terms | Programme | Developer's rate; delivery rate; duration premium | 6%; 0.95; 0.2 pp | 9%; 0.85; 0.1 pp | 13%; 0.70; none |
The programme-cost rates apply only to practices that are switched on. At the defaults only prescribed burning is on, so only the burning rate moves the result.
Each row moves one setting from Favourable to Adverse with every other setting at Reference. Loblolly pine, Neches River. Value created, at the start of 2026.
| Setting | Fund it yourself | Buy credits |
|---|---|---|
| All Reference | +USD 0.9m | +USD 1.2m |
| Timber price | +USD 2.4m → −USD 0.7m | +USD 2.9m → −USD 0.5m |
| Log market | +USD 2.0m → −USD 0.8m | +USD 2.3m → −USD 0.5m |
| Programme cost | +USD 1.1m → +USD 0.6m | +USD 1.3m → +USD 0.9m |
| Mill response | +USD 3.9m → −USD 1.8m | +USD 4.2m → −USD 1.5m |
| Credit terms | +USD 1.5m → −USD 0.03m | +USD 1.8m → +USD 0.4m |
| All Favourable → all Adverse | +USD 9.7m → −USD 3.7m | +USD 10.5m → −USD 3.3m |
Figures from Forestry_Mill_Model.xlsx at each setting.
What the table shows:
Every setting at Reference, every choice at its default: loblolly pine, Neches River (the model works on a 100-hectare representative site, because the site sets the mix of land and not the scale, §2.4), a mill of 200,000 MBF a year, a USD 1.6m annual budget shared among three funders in all (the model company and two co-funders), prescribed burning as the habitat practice, and enrolment at stand age 24.
Enrolment at age 24 means the stand is enrolled when the owner would otherwise cut it. The model assumes it then follows the longer plan, including the second thinning, and reaches the longer plan's yield (§2.1).
| Doing nothing | With the programme | |
|---|---|---|
| Rotation | 24 years | 33 years |
| Mature sawtimber at final harvest | 51.9 t/ha | 140.8 t/ha |
The sawtimber uplift is 88.9 t/ha. After the 85% that reaches this mill and the 85% delivery rate, 64.2 t/ha is delivered. Enrolment at age 24 leaves a wait of 9 years, so wood enrolled in 2026 is cut in 2035.
| Line | USD a hectare |
|---|---|
| Owner's payment (land-value gap, thinnings cannot be sold) | 498.94 |
| MRV and verification | 148.20 |
| One-off cost per hectare, which sets the hectares enrolled | 647.14 |
| Prescribed burning, 19.69 a year for 9 years | 177.21 |
| Whole programme | 824.35 |
| Owner's payment at delivery, increased at 6% a year for 9 years | 842.95 |
A USD 1.6m budget at USD 647 a hectare enrols 2,472 ha a year. At 64.2 t/ha that is 158,840 t of upgraded wood against a mill intake of 860,000 t. Coverage is 6.2% in 2035, 12.3% in 2036 and 18.5% from 2037.
A typical year once coverage is complete.
| Doing nothing | Fund it yourself | |
|---|---|---|
| Revenue | 90.00 | 92.13 |
| Log bill | 35.26 | 35.52 |
| Conversion cost | 40.00 | 40.45 |
| Gross profit | 14.74 | 16.16 |
| Overheads | 5.00 | 5.00 |
| Programme payment | – | 0.96 |
| Profit before interest, tax and depreciation | 9.74 | 10.20 |
The log bill and conversion cost rise because the mill buys and saws more wood.
| Doing nothing | Fund it yourself | Buy credits | |
|---|---|---|---|
| Company value, start of 2026 | 44.62 | 45.51 | 45.80 |
| Value created | +0.89 | +1.18 | |
| What you commit, 2026–2080 | 45.07 | 45.96 | |
| Extra cash, 2026–2080 | +14.15 | +12.34 | |
| Value created, against what you commit | USD 0.9m for USD 45m | USD 1.2m for USD 46m | |
| Extra cash, times what you put in | 0.3 times | 0.3 times |
Buying credits creates USD 0.3m more value than funding it yourself. But it commits USD 0.9m more cash (USD 46m against USD 45m). The developer's financing charge adds about USD 6.4m to what a credit buyer pays, which outweighs the lower payment factor and the coordination cost that credit buyers avoid. Buying credits also produces USD 1.8m less extra cash. After tax, the financing charge costs about USD 4.9m of extra cash, and the lower payment factor and the coordination cost that credit buyers avoid make up about USD 3.1m of it. The two measures rank the routes differently, for the reason in §1.3. Which one should decide depends on whether the decision is based on what the company is worth or on the cash it puts in and gets back.
The longer rotation and the extra thinning create the financial value, and they also benefit nature. To make the programme clearly nature positive, it also pays for prescribed burning, which brings in no revenue and is pure cost. Without it, value created would be about USD 1.7m funding it yourself and USD 2.0m buying credits, against USD 0.9m and USD 1.2m with it. Users can switch habitat practices off, or add others, in the tool (§2.4, §5.1).
The same example with eastern white pine gives −USD 3.9m on the fund-it-yourself route. This is the expected result. The thinning costs USD 988 a hectare and is paid 25 years before the extra timber arrives. Discounted at the company's 9% cost of capital, that timber cannot repay it.
Every input has one of four labels.
Sourced means a real external figure, dataset or published study, cited in §7.3. Derived means worked out from other inputs by a stated formula. Judgement means a stated choice with no external source. Structural means a modelling convention, not a quantity with a right value.
The label says where a figure comes from, not how much it matters. Some judgements move the result more than some sourced figures do, and §7.2 says where the two differ.
| Part | Label | Basis |
|---|---|---|
| Sawtimber uplift, loblolly | Sourced | Published comparison of rotation economics, US South |
| Sawtimber uplift, white pine | Sourced, weak | One field study, not repeated, about 1 acre |
| Conversion ratio | Sourced | Industry survey of 38 southern mills |
| Delivered log cost | Sourced | Split into the price paid for standing timber (stumpage), harvesting and haulage |
| Lumber price | Judgement | Mid-cycle judgement; see §7.5 |
| Recovery gain | Judgement | A cautious figure below an upper limit taken from a log rule, a formula that estimates the lumber in a log (§7.3, §7.5) |
| Stem size gain | Judgement | No public source |
| Grade premium | Judgement, bounded | Kept within what surveyed sawmills say they would pay |
| Operating leverage | Judgement | No public source |
| Cost of capital | Sourced, adjusted | Sector figure plus a stated premium for a single mill |
| Owner payment | Derived | Land-value comparison, checked against a published range |
| Nature intervention costs | Sourced, transferred | Priced in other forest types; see §7.5 |
| MRV and verification cost | Judgement | No public source |
| Developer's financing charge | Judgement | No public source |
| Delivery rate | Judgement | No public source |
| Share of enrolled wood reaching this mill | Judgement | Set by the Log market setting. A right of first refusal in the contract makes a high share realistic (§3.4) |
| Thinning income | Judgement | Set to zero because pulpwood prices have fallen below the source's weakest price set (§5.1) |
| Reward from banks and investors | Judgement | No public source |
| Duration premium | Judgement | No public source |
| Share of risk repriced | Judgement | No public source |
| Coordination | Judgement | A low estimate (§5.5) |
The physical benefit is based on published evidence. The way the model turns it into financial value is based on judgement. The yield comparison, the conversion ratio, the log cost and the owner's payment are all sourced or derived from sourced figures. The four channels that turn better logs into a better income statement are all judgement. Together they move the answer by about USD 5.0m, against a value created of USD 0.9m. The case depends on three of them. With recovery alone at zero, funding it yourself loses USD 1.8m and buying credits USD 1.5m. With grade alone at zero, the losses are USD 0.7m and USD 0.4m. With operating leverage alone at zero, funding it yourself loses USD 0.2m. Stem size is the only one that can be zero with both routes still creating value. With all four at zero, funding it yourself loses USD 4.2m (§7.5).
Some outside evidence supports these four channels. A survey of south-eastern softwood sawmills reports that they would pay a premium of USD 4.22 to 12.98 a ton for better pine sawtimber. It also reports that 57% of mills would pay a premium for wood grown on a longer rotation, which is what this credit produces. The model's own benefit is USD 8.96 a ton of upgraded wood, inside that range. So the total is supported by evidence, even though the channels are not each sourced. A validation test checks this.
Nature outcome (Part 8)
Physical and silvicultural
Mill and market
Financial
Willingness to pay
Habitat intervention costs
The three habitat sources are priced in forest types other than the two modelled here. §7.5 records this as a limitation.
The model has a set of 25 validation checks. Of these, 18 give a pass or fail result, and all 18 pass. The other 7 report a figure and have no pass mark. Every input is read by at least one formula, and no cell contains a formula error. The balance sheet balances to within 1.3 × 10⁻¹² in every year of every scenario.
These are the checks that test the method hardest.
The zero-effect test sets the sawtimber uplift to zero and requires both routes to give a negative result. A programme that upgrades nothing must cost money and produce nothing.
The decomposition test switches off every difference between the routes listed in §1.4 and requires the difference between the routes to fall to zero. It does so exactly. If there were an unlisted difference between the routes, the result would not fall to zero.
The spend test requires that changing the programme budget changes the answer. This guards against a cost that is shown in the tool but connected to no result.
The set-aside test requires that land taken out of production reduces the wood delivered to the mill, not only the budget available. The share-reaching-the-mill test requires that, when none of the enrolled wood reaches this mill, both routes give a straight loss.
The land-value test runs twice. At the published thinning incomes it must reproduce the source's land-value ratio of 79% to 85% and the payment that implies. It gives 83.0% and USD 203 a hectare. At the defaults it must match the owner's payment entered in the model. It works out USD 500 against the 499 entered.
The benefit test compares the model's own gain in profit before interest, tax, depreciation and amortisation, per ton of upgraded wood, with what surveyed sawmills say they would pay. It gives USD 8.96 a ton against a surveyed range of USD 4.22 to 12.98.
The mill-sizing test requires that sizing the mill to the site sharply narrows the spread in return across an 8-fold range of programme size. The spread falls from 18.07 percentage points to 1.75, with the habitat practices switched off.
A separate test replays all 21 fixed test configurations against the model and requires each to reproduce its recorded output. All 21 reproduce to within USD 0.0005m.
Ranked by how much each would move the answer.
What the model leaves out, why, and which way each would move the answer.
| Left out | Why | Effect on the result |
|---|---|---|
| Any climate effect on the stand | The evidence compares two ways of managing a forest under observed conditions. Taking it into other climates would weaken it (§2.3) | Either way |
| The spread of stand outcomes | The model uses one yield for each management plan, not a range | Either way |
| Lower haul cost where upgraded wood is closer to the mill | Depends on the area a particular mill buys wood from and cannot be shown inside the tool. The channel exists and is set to zero | The benefit is too low where it applies |
| Native seed sold from the pollinator corridor | The larger commercial claim for the corridor, with no published figure. Only the rise in hunting-lease income is counted (§5.1) | The corridor's net cost is too high |
| Wetland protection, lighter site preparation, dispersed legacy trees, snags and downed wood, a longer rotation on its own | Described for this use case but no defensible cost exists | Unknown |
| Biochar at replanting | Its cost is published only in 2011 USD and later analysis spans a net gain of USD 222 a tonne to a net cost of USD 584, so it is not clear whether it makes or loses money. Both claimed benefits would go to the owner or to a separate carbon instrument, not the mill | Unknown |
| Sudden losses: hurricanes, fire, pine beetle | The yield comparison is for stands that reached harvest. The delivery rate allows for some failure, but not for a large single event | Value created can be too high on both routes. A credit buyer would not pay for work lost to such an event (§5.4), so the advantage of buying credits is understated |
| The value of not bearing risk when buying credits | Buying credits makes the company's cost more certain, because it pays only for what is delivered. That certainty has value of its own. The model gives one central result, not a range of outcomes, so it cannot show this (§5.4) | Credits look worse than they are |
Parts 1 to 7 describe a financial method. Nothing in them depends on nature being measured. The model runs on stand growth and the size of what reaches the mill, and the financial value follows from that.
This part covers the separate question of what the programme does to the forest itself. It is descriptive. No figure in it enters the valuation, and changing it changes no result.
This separation marks a limit of the method. A method that priced biodiversity into company value would be claiming a link that does not exist here. The mill's return comes only from better wood. §7.5 makes the same point about the set-asides. A package with a lot of land taken out of production needs a buyer who values the outcome for its own sake.
The tool reports ecosystem integrity, from the Ecosystem Integrity Index.
The index scores a landscape by how far it is from what it could naturally support, across three pillars.
| Pillar | What it measures | Data |
|---|---|---|
| Functional | How much the land grows, against what this climate and soil could support | Copernicus Sentinel-3 productivity against a modelled natural potential, 300 m |
| Structural | How intact and connected the habitat is across the surrounding landscape | Global Human Modification v3, Theobald et al. 2025, over a 10 km neighbourhood, 100 m |
| Compositional | How much of the originally present native community is still there | Biodiversity Intactness Index, Impact Observatory, on the PREDICTS database, 100 m |
The compositional pillar is referenced to the community that was originally present, not to a count of what is there now. A species arriving because the stand has been disturbed does not raise it; a native species lost does lower it.
Sources are listed at §7.3.
The index takes the weakest of the three pillars, not their average, then reduces the result further where the other two are also poor.
index = weakest pillar × ( a + b − a × b ) a, b = the other two pillars, each from 0 to 1
The bracket is close to 1 when either of the other two pillars is close to intact, and it falls as both fall. Condition and change are worked out from this index (§8.4).
The reasoning is that a landscape which is productive and well connected but has lost its native community is not a healthy one, and an average would hide that. A very low score on one pillar limits the overall score.
Forestry benefits from this rule more than the other cases do. An even-aged plantation scores well on productivity and acceptably on habitat pattern, and poorly on composition, because it has many tall trees close together but few other species. So composition sets the score, and thinning and burning improve it. Thinning lets light reach the ground so plants grow back, and burning keeps the ground open. The programme improves the pillar with the lowest score, the one that limits the index. That is why the loblolly stands show the largest improvement of any landscape in the tool.
The rule also shows what would improve the score next. The programme addresses composition. The main forestry work does not change structure. Structure improves when land is taken out of production: streamside buffers above the legal minimum, and patches left standing through the harvest. As §7.5 says, those lines cost wood as well as money and will not repay themselves from timber benefit, so they need a buyer who values the outcome for its own sake.
The tool reports condition and change in words, not numbers, because the underlying values are reasoned rather than computed (§8.5). Each word matches a fixed range of values, so that the words mean the same thing on every landscape.
Condition, on the index value: Heavily modified (below 0.20) · Substantially modified (0.20–0.40) · Partly intact (0.40–0.60) · Largely intact (above 0.60).
Change, on the share of the gap to natural potential that the programme closes: Slight (under 3%) · Modest (3–7%) · Moderate (7–15%) · Marked (15–30%) · Major (above 30%).
Change is expressed as a share of the remaining gap, not as a percentage of the index value. A percentage of the value would reward degradation. The same action on a degraded landscape would score several times higher than on an intact one, because its starting point is smaller. Using the share of the gap treats all landscapes the same, whatever their starting point.
Every step of the model, in order, as built in Forestry_Mill_Model.xlsx. Inputs are in Appendix B. y is the year, 2026 to 2080. Steps 1 to 8 are worked out once. Steps 9 to 15 are worked out for each of the three futures: doing nothing (no programme, coverage 0), funding it yourself, and buying credits.
1. deferral H H = regime rotation − enrolment age (white pine: its rotation − its enrolment age)
2. uplift sawtimber uplift = regime yield − baseline yield, t/ha
3. productive share 1 − (streamside area × 0.55 + patch area × 0.95 + other set-aside × 1.00)
÷ site area × global set-aside multiplier
(the multiplier scales all three factors together; it is 1 by default)
4. delivered uplift sawtimber uplift × productive share × share reaching this mill × delivery rate
5. cost per hectare owner payment + one-off practice cost + MRV + one-off package cost
6. hectares a year programme spend × 1,000,000 ÷ cost per hectare
enrolled(y) hectares a year × min(1, (y − 2026 + 1) ÷ enrolment ramp), from 2026
delivering(y) enrolled(y − H)
7. mill intake lumber output × base conversion ratio
coverage(y) min(1, delivering(y) × delivered uplift ÷ mill intake)
8. conversion(y) base conversion ratio × (1 + basket drift × (y − 2026)) × (1 − recovery gain × coverage(y))
price(y) lumber price + grade premium × coverage(y)
output(y) min( mill intake × (1 + stem-size gain × coverage(y)) ÷ conversion(y),
lumber output × (1 + capacity headroom) )
9. cost of capital doing nothing: base rate
fund it yourself: base rate − reward × coverage(y)
buy credits: base rate − (reward + duration premium) × coverage(y)
10. payments fund it yourself, each year, divided by the number of funders:
owner payment at delivery, grossed up at the owner's 6% for H years
+ one-off practice and package costs at enrolment
+ MRV spread over the H years after enrolment
+ recurring practice cost spread over the same H years
+ coordination in 2026 (not divided)
buy credits, each year:
delivering(y) × credit price per hectare ÷ funders × payment factor
credit price per hectare = owner payment grossed up to delivery
+ one-off practice and package costs × (1 + developer's rate)^H
+ MRV, each year's share carried from its year to delivery
+ recurring practice cost, carried the same way
payment factor = delivery rate × (1 + (1 − delivery rate) × share repriced)
11. income statement revenue = output × price
log bill = output × conversion × (delivered log cost − haul saving × coverage)
conversion = lumber output × other conversion cost
× (fixed share + (1 − fixed share) × output ÷ lumber output)
profit before interest, tax, depreciation
= revenue − log bill − conversion − overheads − programme payments
(fund it yourself only)
credit write-off = each credit purchase ÷ 10, for 10 years (buy credits only)
interest = cost of debt × net debt(y − 1)
tax = 25% of profit before tax, and nothing in a loss-making year
12. balance sheet stock = (log bill + conversion) × stock days ÷ 365
fixed assets held constant; credit asset = previous + purchase − write-off
net debt(y) = net debt(y − 1) − (net income + write-off − change in stock
− credit purchase); check: assets − (net debt + equity) = 0
13. free cash flow operating profit × (1 − tax) + credit write-off − change in stock − credit purchase
14. company value value(2080) = ( FCF + FCF × (1 + g) ÷ (w − g) ) ÷ (1 + w)
value(y) = ( FCF(y) + value(y + 1) ) ÷ (1 + w(y))
15. results value created = value(2026, route) − value(2026, doing nothing)
what you commit = Σ payments 2026–2080, not discounted
extra cash = Σ FCF(route) − Σ FCF(doing nothing), 2026–2080, not discounted
Depreciation and the capital spending that keeps the mill running are set equal, so they cancel in free cash flow.
Tax in a loss-making year. The income statement charges no tax in a loss-making year (step 11). Free cash flow, which sets company value, applies tax to operating profit in every year (step 13). So in free cash flow a loss-making year counts a tax saving.
Inputs with no effect. Every input on the sheet is read by at least one formula. The scan for unused inputs finds none. Three inputs are read only by checks and affect no result. The first is the white pine thinning cost shown under species 2. The USD 988 that drives the white pine result is entered as the one-off practice cost (§2.2, §5.1). This copy is used only to check that the two agree. The other two are the white pine value gain and stumpage price basis. They feed a cross-check between the volume and value calculations. Anyone rebuilding the model should treat them as checks shown on screen, not as model inputs.
Guards. The Output sheet has guards for the horizon, the total set-aside share, the streamside share, the two overlaps between habitat bundles, the patch-retention cost against its level, the share of enrolled wood reaching this mill, the owner payment against the land-value block, whether the rotation fits the yields it implies, and the white pine settings.
Units.
Areas and rates per area are metric throughout the model, in hectares and tons a hectare. Mill quantities stay in the units the industry trades in: thousand board feet of lumber, short tons of log, and USD per thousand board feet. Sources published per acre are converted on entry at the factor the source paper uses, 1 hectare to 2.47 acres. The original figure is recorded beside the converted one.
Areas and rates per area use the same unit throughout, so no conversion happens inside the model. A rate times an area gives the same answer whatever the unit of area. So the choice of unit does not change any result or validation test. The 21 fixed test configurations reproduce their recorded outputs under either unit.
Model company.
| Parameter | Value | Label |
|---|---|---|
| Lumber output | 200,000 MBF/yr | Sourced, typical mill size |
| Lumber price, mid-cycle | 450 USD/MBF | Judgement |
| Base conversion ratio | 4.30 tons/MBF | Sourced |
| Delivered log cost | 41 USD/ton | Sourced |
| Other conversion cost | 200 USD/MBF | Judgement |
| Overheads | 5.0 USDm/yr | Judgement |
| Fixed assets | 110 USDm | Judgement |
| Depreciation life | 25 yr | Judgement |
| Stock, days of cost of sales | 45 | Judgement |
| Net debt to profit before interest, tax, depreciation and amortisation | 2.0× | Judgement |
| Cost of debt | 6.0% | Sourced, adjusted |
| Tax rate | 25% | Judgement, stated basis |
| Cost of capital | 9.0% | Sourced, adjusted |
| Terminal growth | 2.0% | Structural |
| Basket drift, the long-term change in log quality in the area the mill buys from | 0.0 | Sourced |
Programme.
| Parameter | Value | Label |
|---|---|---|
| Regime rotation | 33 yr | Sourced |
| Baseline rotation | 24 yr | Sourced |
| Mature sawtimber, regime | 140.79 t/ha | Sourced |
| Mature sawtimber, baseline | 51.87 t/ha | Sourced |
| Owner payment | 498.94 USD/ha | Derived |
| MRV and verification | 148.20 USD/ha | Judgement |
| Prescribed burning | 19.69 USD/ha/yr | Sourced |
| Programme spend | 1.6 USDm/yr | Structural, user choice |
| Funders, the model company and its co-funders | 3 | Structural, user choice |
| Enrolment ramp | 3 yr | Judgement |
| Enrolment age | 24 yr | Structural, user choice |
Benefit mechanism.
| Parameter | Value | Label |
|---|---|---|
| Recovery gain | 6% | Judgement |
| Stem size gain | 4% | Judgement |
| Grade premium | 12 USD/MBF | Judgement, bounded |
| Capacity headroom | 10% | Judgement |
| Fixed share of conversion cost | 40% | Judgement |
| Haul cost | 0.14 USD/ton-mile | Sourced |
| Haul miles saved | 0 | Structural, not exposed |
| Share of enrolled wood reaching this mill (called mill offtake in the workbook) | 0.85 at Reference | Judgement, set by the Log market setting |
| Reward from banks and investors | 0.25 percentage points, scaled by coverage | Judgement |
Buying credits.
| Parameter | Value | Label |
|---|---|---|
| Write-off period | 10 yr | Structural |
| Coordination cost, fund it yourself only | 0.4 USDm | Judgement |
| Developer's rate | 9.0% | Judgement |
| Duration premium | 0.1 percentage points, scaled by coverage | Judgement |
| Delivery rate | 0.85 | Judgement |
| Share of delivery risk repriced | 0.50 | Judgement |
Land-value block.
| Parameter | Value | Label |
|---|---|---|
| Owner discount rate | 6.0% | Sourced |
| Establishment cost | 741 USD/ha | Judgement |
| Baseline thinning income | 0 USD/ha | Judgement |
| Regime thinning income, first | 0 USD/ha | Judgement |
| Regime thinning income, second | 0 USD/ha | Judgement |
| Baseline final harvest | 4,935 USD/ha | Sourced |
| Regime final harvest | 5,854 USD/ha | Sourced |
| Thinning operation cost | 0 USD/ha | Structural |
Species 2, eastern white pine.
| Parameter | Value | Label |
|---|---|---|
| Sawtimber uplift, standing merchantable volume | 21,146 BF/ha | Sourced, weak |
| Standing tons per thousand board feet | 5.0 | Judgement |
| Rotation | 70 yr | Sourced |
| Enrolment age | 45 yr | Structural, user choice |
| Thinning cost | 988 USD/ha | Sourced |
There are 21 fixed test configurations, which the tool also offers as presets. Each states how it differs from the defaults and has a recorded output. All 21 are replayed against the model as a test that it still gives the same answers.
They fall into five groups. Base and cautious cases show the default and the lowest result, with both claimed physical gains switched off. Mill-state cases cover the mill at its capacity ceiling, a programme sized to the mill, a single funder, and enrolment on bare ground. Habitat cases cover the smallest and the full set of practices, the set-aside priced above the legal minimum, and a site rich in nature. Species and market cases cover eastern white pine, none or part of the enrolled wood reaching this mill, no market for thinnings with and without a thinning cost, and a comparison of the two species with the same wait for wood. Financing cases cover a smaller payment gap, an expensive developer's financing charge, and a case where wood is scarce in the area the mill buys from.
All 21 reproduce, so the presets in the tool match the model. In the single-funder case, value created is −USD 7.1m funding it yourself and −USD 6.8m buying credits.
| Rate | Value | Source |
|---|---|---|
| Owner payment, loblolly | 498.94 USD/ha | Derived from the land-value comparison |
| Owner payment, loblolly, thinnings can be sold | 203.21 USD/ha | Same calculation, published thinning income |
| Thinning operation, white pine | 988 USD/ha | Field study |
| Prescribed burning | 19.69 USD/ha/yr | 78.74 USD/ha per burn, 2022 survey of southern practice across 85,268 acres, on a 4-year cycle (judgement) |
| Four-practice habitat package | 38–83 USD/ha/yr | Florida slash pine, not separable |
| Streamside zones, gross | 252–1,659 USD/ha | Eastern hardwoods |
| Patch retention, 5% of area | 321–401 USD/ha | Eastern oak-hickory |
| Patch retention, 10% of area | 601–761 USD/ha | Eastern oak-hickory |
| MRV and verification, including the legal agreement that binds the owner | 148.20 USD/ha | Judgement |
| Establishment | 741 USD/ha | Judgement, US South range 618–865 |
| Pollinator corridor, establishment | 865 USD/ha of corridor | Built up: seed above USD 2,470/ha, less CP-42 cost share and incentive |
| Pollinator corridor, owner revenue | 16 USD/ha/yr | Rise in hunting-lease income only; native seed revenue has no published figure and is set to zero |
| Harvest | 11.24 USD/green ton | Published |
| Haul | 7.42 USD/green ton over 53 miles | Published |
Streamside and patch-retention rates are gross figures priced in other forest types. The streamside line in the model is the part above the legal minimum, not the gross cost, for the reason given in §5.1.
How the tool presents the figures in this document. None of it changes a result.
The tool's summary shows both routes side by side, funding it yourself and buying credits, with every figure signed.
| Card | Figure | What it is |
|---|---|---|
| You commit | Cash, 2026 to 2080 | The company's own programme payments, not discounted (§1.3) |
| What the mill is worth | Doing nothing | Company value with no programme, start of 2026 |
| You create | Per route | Value created (§1.3) |
| Times what you put in | Value created ÷ what you commit, under the display rule below | |
| Cash | Extra cash before tax and timing | The change in operating cash before tax, less what you commit |
| Extra cash | As defined in §1.3 | |
| Times what you put in | Extra cash ÷ what you commit | |
| Credits line | Buying credits adds … | Value created by buying credits less value created by funding it yourself, and how much less or more cash credits need. Has a Judgement badge, because it depends on judgements: the developer's rate, the delivery rate, the share of risk repriced and the duration premium (§5.3, §5.4, §7.2) |
| Nature | Ecosystem integrity | Condition and change in words, marked Indicative (Part 8) |
At the defaults, at Neches River, You commit shows USD 45m and USD 46m. You create shows USD 0.9m and USD 1.2m.