Carbon removal is beginning to look like an infrastructure market
For years, voluntary carbon markets were dominated by a relatively simple idea.
A company emitted greenhouse gases somewhere in the world and purchased a credit representing an emissions reduction somewhere else.
The credibility of that model eventually came under severe pressure.
Investigations raised questions about whether some projects were genuinely additional, whether forests would really have been cut down without carbon finance, whether claimed emissions reductions were exaggerated and whether credits represented climate benefits that lasted long enough to justify the claims made by buyers.
A different part of the market is now attracting increasing amounts of capital.
Instead of paying primarily to avoid future emissions, buyers are increasingly seeking carbon dioxide that can be demonstrably removed from the atmosphere.
That can mean engineered approaches such as direct air capture, biochar and enhanced rock weathering.
It can also mean rebuilding forests on degraded land.
In Brazil, Mombak is attempting to turn that second category into an institutional-scale business.
The carbon-removal company announced in September 2026 that it had reached the first close of its second Amazon reforestation investment vehicle and is targeting as much as $150 million for the fund.
At the same time, Salesforce joined the company's expanding group of carbon-removal buyers.
The development provides a useful snapshot of how the carbon market is changing.
The next phase is increasingly about financing projects years before their credits exist, measuring removals more rigorously and building assets capable of delivering verified carbon removal at scale.
Mombak wants to raise $150 million for Amazon Reforestation Fund II
Mombak's new Amazon Reforestation Fund II is targeting $150 million to finance restoration projects in the Brazilian Amazon.
The fund will also have access to a 200 million Brazilian real credit line, worth roughly $39 million at the time of the announcement, from Brazil's Climate Fund through development bank BNDES.
The combination of equity capital, debt and long-term carbon-credit purchase agreements illustrates how reforestation is starting to resemble a conventional infrastructure-finance business.
Capital has to be committed before forests mature.
Land has to be acquired or secured through long-term agreements.
Native species have to be planted and maintained.
Fire protection has to be established.
Carbon accumulation has to be measured.
Independent verification has to occur.
And buyers may wait years before the removals they purchased are finally delivered.
That financing gap historically made high-quality reforestation difficult to scale.
Long-term carbon-removal contracts are beginning to change the equation.
The first fund already raised $120 million
Mombak's first Amazon reforestation fund raised approximately $120 million from investors including institutional capital connected with AXA, CPP Investments and Bain Capital.
According to the company, the fund has financed restoration across 15 farms in the Amazon.
Nearly 15 million native trees have been planted.
The strategy concentrates on degraded agricultural and cattle-pasture land rather than clearing intact forest or building monoculture plantations.
Mombak buys or partners around degraded land and then restores native forest through active planting and assisted natural regeneration.
The investment thesis is unusual.
The physical asset is land.
The productive asset is the recovering forest.
And part of the financial output is measured in tonnes of atmospheric carbon dioxide removed as biomass accumulates.
Salesforce has become the latest large buyer
Alongside the new fund, Mombak announced a multiyear carbon-removal purchase agreement with Salesforce.
The size and financial value of the transaction were not publicly disclosed.
Salesforce joins a buyer base that has included Google, Microsoft, McKinsey, McLaren Racing and other organizations purchasing or contracting for Mombak removals.
This matters because the carbon-removal market has remained unusually dependent on a relatively small group of technology companies.
Big Tech became an early source of demand partly because companies such as Microsoft and Google adopted ambitious climate commitments while simultaneously facing rapidly increasing emissions associated with data-center and artificial-intelligence expansion.
Whether demand can move beyond those early corporate buyers is becoming one of the central questions facing the carbon-removal industry.
Mombak believes it can.
Its management expects interest to expand into industries including mining, steel and oil, sectors where eliminating every residual emission may be significantly harder than in parts of the digital economy.
Google has quadrupled its Mombak purchase
Google is one of the most prominent buyers supporting Mombak's scale-up.
In November 2025, Google announced an agreement to buy 200,000 tonnes of carbon dioxide removal from Mombak.
Google said that was four times the size of its earlier purchase from the company.
The removal will come from restoration of degraded land in the Brazilian Amazon using biodiverse native forests.
Google has also said it intends to use DeepMind's Perch artificial-intelligence technology to help quantify biodiversity benefits created by restoration.
That is important because forests can provide value beyond carbon.
A restored Amazon ecosystem can create habitat, improve hydrological systems, reduce erosion and reconnect fragmented landscapes.
Those outcomes are difficult to capture in a single carbon-credit price.
Microsoft helped establish demand at much larger scale
Microsoft has been another major Mombak buyer.
The Symbiosis Coalition says Microsoft had already signed an approximately 1.5-million-tonne Mombak offtake before Mombak became the first project formally selected through the coalition's project process.
Isometric said in April 2026 that Mombak had secured offtakes for more than 1.8 million tonnes of carbon removal across buyers including Microsoft, Google and McKinsey.
An offtake agreement is important because it commits a buyer to future production.
A company building a reforestation project does not have to wait 10 or 20 years for trees to mature before proving there will be a market for the carbon they store.
A credible long-term buyer can provide the future revenue visibility required to attract investors and lenders today.
This mechanism is increasingly central to carbon removal.
The carbon-credit market itself is not suddenly booming
It is important not to confuse investment commitments with current credit-market revenue.
MSCI estimates the primary global carbon-credit market remained around $1.4 billion in 2025, roughly unchanged for a fourth consecutive year.
About 202 million tonnes of credits were retired during the year.
Only around 10% of retired credits represented removals, while the majority continued to come from emissions-reduction projects.
At first glance, that does not look like a rapidly expanding market.
But underneath the headline number, capital allocation changed substantially.
Higher-quality credits became more valuable.
Nature restoration and carbon engineering attracted increasing attention.
Meanwhile, lower-quality categories lost market value.
The market is therefore not simply becoming larger.
It is becoming more differentiated.
Future carbon supply attracted a record $22 billion
The clearest sign of change appears in forward investment rather than current spot trading.
MSCI says capital committed or deployed into future carbon-credit supply reached a record $22 billion in 2025.
That was 72% higher than in 2024 and more than five times the level recorded in 2021.
Carbon engineering accounted for approximately $10.3 billion of activity.
Nature restoration accounted for another approximately $10.1 billion.
Together those two categories represented 93% of investment and offtake activity tracked by MSCI during the year.
That provides critical context for Mombak's new fund.
Investors and buyers are moving away from simply purchasing whatever cheap credits happen to be available today.
They are increasingly financing specific future projects expected to meet much higher standards.
Forward contracts are changing how carbon removal gets financed
MSCI says pre-purchase and forward agreements represented roughly two-thirds of carbon-market offtake value in 2025, compared with less than one-quarter in 2022.
This is a major change in market structure.
Historically, much of the project-development risk stayed with the carbon developer.
A company would build a project and hope there would eventually be buyers for the resulting credits.
Under a long-term offtake, the buyer assumes part of the delivery risk by promising to purchase future removals.
That future contract can make it easier for the developer to raise equity or secure debt financing.
It is effectively creating project finance for atmospheric carbon removal.
Mombak has now issued actual reforestation certificates
One of the biggest risks in carbon removal is the difference between a promise and delivery.
Many projects sell credits years before the underlying removal occurs.
Mombak crossed an important threshold in July 2026 when Isometric issued the first nature-based carbon-removal certificates under its registry.
The issuance covered 21,771.53 tonnes of removal from Mombak's Reforesting the Brazilian Amazon Project 1.
Isometric described them as the world's first native-species reforestation certificates issued using its dynamic-baseline methodology.
Mombak subsequently began delivering credits to buyers including Google, McKinsey and McLaren Racing more than two years earlier than some contractual deadlines.
Delivery is important because the carbon-removal market has been repeatedly challenged by projects that promised large future volumes but delivered late or failed to materialize entirely.
Mombak expects a much larger issuance before the end of 2026
Mombak expects another reforestation issuance of approximately 80,000 metric tonnes by the end of 2026.
If achieved, that would represent a substantial increase over the company's first verified reforestation issuance.
But these numbers remain small relative to global emissions.
Human activity emits tens of billions of tonnes of carbon dioxide annually.
Even millions of tonnes of contracted removal are tiny in comparison.
The climate importance of the current market is therefore less about present scale than about whether the industry can develop financing and verification systems capable of eventually operating at much larger scale.
Reforestation credits are different from avoided-deforestation credits
Carbon-market terminology can obscure important differences.
A reforestation project removes atmospheric carbon dioxide as newly growing trees accumulate carbon in trunks, roots and soils.
An avoided-deforestation project generally issues credits based on emissions that would supposedly have occurred if existing forest had been cut down.
Both approaches can support forests.
But their accounting questions are different.
Avoided-deforestation projects, commonly associated with REDD+, have faced particularly intense scrutiny over whether developers exaggerated the amount of forest loss that would have occurred without their projects.
Reforestation starts from a more physically visible change: degraded land is converted into growing forest.
That does not eliminate accounting risk.
But it changes the nature of the problem.
Carbon removal does not automatically mean high quality
A tree absorbing carbon is real atmospheric removal.
A carbon credit representing that tree still requires assumptions.
How much additional forest exists because carbon finance was available?
How much carbon would the land have stored anyway?
How accurately has biomass been measured?
How much soil carbon should be counted?
What happens if a wildfire destroys the forest?
Could illegal logging reverse the storage?
Does restoring one property push cattle ranching or deforestation somewhere else?
How long must the carbon remain stored?
These are questions of additionality, baseline design, leakage, measurement and permanence.
They are the foundations of credit integrity.
Isometric uses dynamic baselines for Mombak
Mombak's first reforestation project was certified under Isometric's Reforestation Protocol.
The system uses dynamic baselines rather than assuming once, at the beginning of a project, what would have happened to the land without intervention.
Isometric says its methodology also accounts for factors such as albedo, the amount of sunlight reflected by the changing landscape.
This may sound technical, but it matters.
A forest's climate impact is not determined only by the carbon contained in trees.
Replacing bright land with darker vegetation can change the amount of solar energy absorbed locally.
High-integrity carbon accounting increasingly attempts to capture effects that older methodologies sometimes ignored.
Forest carbon can disappear again
Carbon stored underground through mineralization can potentially remain there for thousands of years.
Carbon stored in a forest faces a different risk profile.
Trees can burn.
They can die during drought.
They can be illegally logged.
Disease can spread through a population.
Political changes can affect land protection.
This is known as reversal risk.
Isometric currently calculates an average reversal risk of around 10.5% for Mombak's first registered Brazilian Amazon reforestation project.
The project contributes a portion of certificates to buffer pools designed to compensate if some stored carbon is later lost.
No reversals have been recorded for the project to date.
The buffer mechanism does not eliminate physical risk.
It attempts to insure the accounting system against it.
Climate change makes permanence harder
There is a paradox at the center of forest carbon removal.
Forests are being financed partly to address climate change.
But climate change itself can make forests more difficult to protect.
Higher temperatures, drought and extreme weather can increase wildfire and mortality risk in some regions.
That means a credit representing a tonne of CO2 stored in a forest today carries different durability characteristics from a tonne mineralized permanently underground.
High-quality markets increasingly recognize this difference instead of treating every tonne as interchangeable.
High-quality reforestation commands a premium
Price differences across carbon credits have become enormous.
Low-cost avoided-deforestation credits have sometimes traded for less than $10 per tonne.
Reuters reported that newer Brazilian reforestation projects have secured prices above $50 and in some cases more than $100 per tonne.
MSCI has separately found a growing premium for higher-quality afforestation, reforestation and revegetation credits.
That premium is important for the economics of restoration.
Planting native forests properly is expensive.
Land, seedlings, workers, fire management, monitoring, maintenance and verification all require capital.
A market that values every forest credit at a few dollars per tonne may simply be unable to finance high-quality restoration.
Mombak is restoring a forest, not building a tree plantation
The biodiversity component of the company's strategy is central to the value proposition.
The Symbiosis Coalition says Mombak's approved project manages more than 20,000 hectares of degraded pastureland.
The restoration strategy includes more than 80 native Amazonian species, including endangered species.
That is materially different from planting a single fast-growing commercial tree species across a landscape.
A biodiverse native forest can provide habitat and ecosystem services that a monoculture plantation cannot replicate.
For carbon buyers, that creates a potential co-benefit.
The same capital used to remove CO2 can also finance ecological restoration.
The Symbiosis Coalition is trying to create a quality floor
Google, Meta, Microsoft and Salesforce launched the Symbiosis Coalition to build demand for high-quality nature-based carbon removal.
The coalition later expanded its membership.
Its goal is to contract more than 20 million tonnes of nature-based removals by 2030.
Mombak became the first project formally selected after the coalition evaluated 185 projects across more than 40 countries.
Google and McKinsey purchased 215,000 tonnes from the project under their coalition commitments.
The purpose of an advance market commitment is not merely to buy credits.
It signals to developers and investors that future demand will exist if projects can meet the required quality standards.
Carbon removal remains highly dependent on a small number of buyers
The industry's biggest weakness may be concentration.
MSCI found that Microsoft accounted for an exceptionally large share of publicly announced carbon-removal offtake value in 2025.
Microsoft's strategy spans both engineered carbon removal and nature-based projects.
Its long-term contracts have helped developers finance projects that might otherwise have struggled to reach construction or implementation.
But concentration creates risk.
A healthy global market cannot rely indefinitely on one or two technology companies to buy most of the product.
Salesforce joining Mombak's buyer base is therefore directionally significant even if the disclosed contract is much smaller than Microsoft's overall carbon-removal activity.
The critical next step is diversification into aviation, steel, cement, mining, oil and gas, financial institutions and eventually regulated demand.
AI is creating an uncomfortable connection to carbon removal
Technology companies are simultaneously helping build the carbon-removal industry and creating new emissions pressures.
Artificial-intelligence infrastructure requires enormous numbers of accelerators, increasingly large data centers and growing electricity supply.
Google, Microsoft and Meta have all acknowledged the challenge of meeting climate goals while their computing infrastructure expands.
That is helping create demand for carbon removal.
The relationship is not a substitute for decarbonization.
Buying forest removals does not eliminate the need to make data centers more efficient or procure cleaner electricity.
But residual emissions are becoming a powerful demand source for removal contracts.
Reforestation is competing with engineered carbon removal
Nature-based and engineered approaches solve different parts of the carbon problem.
Direct air capture can potentially provide extremely durable geological storage, but it remains expensive and energy intensive.
Enhanced rock weathering can convert atmospheric carbon into chemically stable forms but requires large-scale mineral deployment and complex measurement.
Biochar can lock carbon in stable biomass-derived material.
Reforestation is biologically straightforward and can provide biodiversity benefits, but its permanence is lower and land requirements are substantial.
A mature carbon-removal portfolio will likely contain several pathways rather than one technological winner.
Microsoft already buys across multiple categories.
Mombak itself has also expanded into enhanced weathering alongside reforestation.
Mombak has become both a nature-based and engineered-removal supplier
In April 2026, Mombak generated its first enhanced-weathering carbon-removal credits through Isometric while registering its reforestation projects on the same platform.
Isometric said this made Mombak the first supplier on its registry with both engineered and nature-based carbon-removal projects.
That diversification shows how the carbon-removal business may evolve.
Companies that begin with one method could eventually build portfolios of different removal technologies, each with distinct cost, durability and ecological characteristics.
Brazil has enormous restoration potential
Brazil sits at the center of the nature-based carbon opportunity because it combines enormous biological productivity with extensive areas of degraded agricultural and pasture land.
Restoring degraded land can increase forest cover without requiring the conversion of intact food-producing or natural ecosystems.
The Amazon also creates unusually high biodiversity value.
But Brazil carries equally significant risks.
Land tenure can be complex.
Illegal deforestation remains a threat.
Fire can reverse stored carbon.
Political attitudes toward forest protection can change.
Local communities and Indigenous Peoples can be affected by land-use decisions.
High-integrity investment therefore requires much more than calculating how quickly trees grow.
Land rights may become one of the market's biggest issues
As carbon stored in ecosystems becomes more valuable, ownership of carbon rights also becomes more economically important.
Who owns the land?
Who owns the carbon accumulated by the forest?
Who receives revenue from the credits?
Were local communities properly consulted?
Do restoration projects alter access to resources traditionally used by nearby populations?
Could rising carbon values encourage speculative land purchases?
These questions are becoming increasingly important as institutional capital enters nature restoration.
A project can measure carbon accurately and still fail broader tests of social integrity.
Carbon markets are being forced to mature
The previous generation of voluntary carbon markets often emphasized quantity.
Cheap credits allowed companies to claim that emissions had been offset at relatively low cost.
The newer carbon-removal market is moving toward a different model.
Quality is expensive.
Monitoring is expensive.
Long-term stewardship is expensive.
Verification is expensive.
Permanence risk has to be priced.
The buyer increasingly wants to know exactly where the tonne came from and how it was calculated.
That makes the product less fungible but potentially more credible.
The total market could become much larger
MSCI estimates the carbon-credit market could grow to approximately $5 billion to $20 billion by 2030 under its current scenarios.
By 2050, the range expands dramatically to roughly $60 billion to $270 billion depending on demand and the availability of high-quality supply.
Removal credits are expected to capture a growing portion of market value because they tend to trade at higher prices and align more closely with long-term net-zero strategies.
Those forecasts remain highly uncertain.
Policy could accelerate the market.
Corporate climate ambition could weaken it.
New accounting rules could exclude certain credits.
Technological advances could reduce removal costs.
And future scientific scrutiny could invalidate methodologies considered credible today.
Carbon removal is therefore an emerging asset class rather than a mature commodity market.
The biggest transition is from offsets to assets
The Mombak story illustrates a deeper structural change.
Old carbon markets often sold certificates generated by existing projects.
The emerging removal market increasingly finances assets before they exist.
An investor provides equity.
A bank provides debt.
A corporate buyer signs a long-term offtake.
The developer restores the land.
Independent systems monitor carbon accumulation.
Verified removal certificates are issued over time.
That begins to resemble renewable-energy project finance far more than the old image of buying a cheap offset online.
Reforestation could become an investable natural infrastructure class
A restored forest is not infrastructure in the conventional sense.
There is no turbine or transmission line.
But economically it can share several characteristics with infrastructure.
Large upfront capital expenditure.
Long-duration operation.
Ongoing maintenance.
Measured productive output.
Contracted future revenue.
Environmental and regulatory exposure.
Long asset lives.
If carbon markets continue maturing, large restoration portfolios could increasingly be financed using structures borrowed from infrastructure, forestry and real-estate markets.
Mombak's combination of investment funds, bank credit and corporate offtakes is an early example of that transition.
Success will depend on whether the tonnes remain credible
The carbon-removal market cannot scale on capital alone.
It ultimately depends on trust in the unit being sold.
One credit is supposed to represent one tonne of atmospheric carbon dioxide removed or avoided according to the applicable methodology.
If buyers stop believing that claim, the market loses its economic foundation.
That is why measurement, reporting and verification are becoming competitive advantages rather than administrative obligations.
Projects able to prove additionality, quantify uncertainty and manage reversal risk can command higher prices.
Projects unable to do so increasingly risk becoming stranded carbon assets.
Mombak is testing whether Amazon restoration can become a repeatable business
The first fund demonstrated that institutional capital could be assembled.
The first farms demonstrated that restoration could be implemented at meaningful scale.
The first verified certificates demonstrated that at least part of the promised removal could be independently issued.
Early deliveries demonstrated that credits could reach buyers.
Fund II now tests the next question.
Can the model be repeated and expanded without sacrificing integrity?
That question matters well beyond one Brazilian startup.
If the answer is yes, degraded tropical land could become a new destination for climate capital around the world.
If quality declines as deployment accelerates, the industry risks repeating the credibility crisis that damaged earlier carbon markets.
Carbon removal's future will be determined by quality, not simply volume
The carbon market spent years attempting to make one tonne look interchangeable with another.
The evidence increasingly points in the opposite direction.
A tonne avoided through a renewable-energy project is not economically or scientifically identical to a tonne removed through a forest.
A tonne stored in a tree is not identical to a tonne mineralized underground for thousands of years.
A biodiverse native forest is not identical to a monoculture plantation.
A credit measured using conservative dynamic baselines is not identical to one relying on optimistic assumptions.
The future market will likely price those differences increasingly aggressively.
The Amazon is becoming part of a new climate-finance experiment
Mombak's $150 million target is small compared with global capital markets.
But it represents something larger than one fund.
It is an experiment in whether atmospheric carbon can become a financeable output of ecosystem restoration.
Google, Microsoft, Salesforce and other buyers are creating long-term demand.
Investors are financing restoration before credits exist.
BNDES is bringing public climate finance into the capital structure.
Independent registries are tightening measurement.
And actual verified reforestation removals are beginning to be delivered.
The carbon-removal market remains young, concentrated and vulnerable to questions about permanence and integrity.
But it is becoming more financially sophisticated.
For the Brazilian Amazon, that could create a new economic proposition for degraded land: a forest may eventually be worth more restored than left as exhausted pasture.
If that proposition can survive scientific scrutiny, wildfire risk, changing politics and the demands of local communities, reforestation could move from philanthropic conservation into a genuine climate-infrastructure business.
That is the market Mombak's second fund is now trying to build.
Reader questions
Frequently asked questions
What is Mombak?
Mombak is a Brazil-focused carbon-removal company that restores degraded Amazon land with native forest and generates verified carbon-removal credits as the recovering forest absorbs atmospheric CO2.
How large is Mombak's new Amazon reforestation fund?
Amazon Reforestation Fund II is targeting up to $150 million.
How much did Mombak's first fund raise?
Mombak's first reforestation fund raised approximately $120 million.
How many trees has Mombak planted?
Mombak says nearly 15 million native trees have been planted across projects financed by its first fund.
Who buys carbon-removal credits from Mombak?
Mombak's buyers have included Microsoft, Google, McKinsey, McLaren Racing, Salesforce and other organizations participating in the developing carbon-removal market.
How much carbon removal is Google buying from Mombak?
Google announced an agreement in November 2025 to purchase 200,000 tonnes of CO2 removal from Mombak, four times the size of its earlier agreement.
How much carbon removal has Microsoft contracted from Mombak?
The Symbiosis Coalition says Microsoft had previously contracted approximately 1.5 million tonnes from Mombak.
Has Mombak actually delivered carbon credits?
Yes. Mombak's first major reforestation issuance occurred in July 2026, with 21,771.53 certificates generated through Isometric, and the company subsequently began delivering credits to buyers.
How is reforestation carbon removal different from REDD credits?
Reforestation removes atmospheric CO2 by growing new forest on degraded land. REDD-type credits generally represent emissions avoided by preventing existing forests from being destroyed.
Why do reforestation credits cost more than some traditional carbon credits?
High-quality native-forest restoration involves land, seedlings, labor, long-term management, monitoring, fire protection and independent verification. Higher-integrity credits can therefore command substantial premiums.
Can forest carbon be permanent?
Forest carbon can remain stored for long periods but faces reversal risks including wildfire, drought, disease and logging. Carbon registries use mechanisms such as buffer pools to help account for those risks.
What is a carbon-removal offtake agreement?
An offtake agreement is a contract in which a buyer commits to purchasing carbon removal that a project will generate in the future, providing the developer with greater revenue certainty.
How big is the global carbon-credit market?
MSCI estimates the primary global carbon-credit market was worth just over $1.4 billion in 2025, although forward project investment and offtake commitments reached a much larger $22 billion.
How large could the carbon-credit market become by 2030?
MSCI's current scenarios suggest approximately $5 billion to $20 billion of annual market value by 2030, depending on demand and the availability and quality of supply.
Why are technology companies buying carbon removal?
Companies including Google and Microsoft have ambitious climate commitments while also facing rising emissions associated with expanding cloud and AI infrastructure. Carbon removal is intended to address residual emissions that cannot yet be eliminated directly.
Does buying carbon removal eliminate the need to cut emissions?
No. High-integrity climate strategies generally treat carbon removal as complementary to direct emissions reductions rather than a substitute for decarbonizing energy, operations and supply chains.
What is the Symbiosis Coalition?
Symbiosis is a group of major corporate buyers designed to create advance demand for high-quality nature-based carbon removal. It has committed to contract more than 20 million tonnes by 2030.
What are the biggest risks in Amazon reforestation carbon credits?
Key risks include inaccurate baselines, lack of additionality, wildfire, drought, illegal logging, leakage, land-tenure disputes, social impacts and failure to maintain stored carbon over the promised period.
Could Amazon reforestation become a large investment business?
Potentially. Long-term corporate offtakes, institutional equity and project debt are creating financing structures that could allow restoration portfolios to operate more like infrastructure assets, although the model still depends on credible verification and long-term forest protection.
Nexuswild welcomes factual corrections. Email [email protected] with evidence and the article URL.
