Adapting to climate change is no longer a choice. The world has warmed by close to 1.5°C. That warming will stay with us for centuries, and the emissions of recent years have not yet had their full effect. I no longer see a realistic path that keeps us below 2°C: even if every nation delivers the cuts it has promised, we will pass that mark. Cutting emissions still decides how far beyond it we go. It no longer decides whether we have to adapt.
The hope, increasingly, is that markets will do much of the work. There is talk of an “adaptation economy”: companies selling the products, services and infrastructure a hotter and climate-disrupted world needs, financed by private investors who see a large and growing opportunity. I share part of that hope. I invest in such companies myself.
But I have doubts about how far it reaches. Adaptation has become a necessity. Will an adaptation economy deliver it? My answer is: in part, and least of all where the need is greatest. This essay sets out why, where I think the line runs today, and what might move it.
The future is uncertain, but certainly less liveable
In 2024 the global average temperature was about 1.55°C above its pre-industrial level, the first calendar year above 1.5°C. The UN Environment Programme estimates that the emission cuts governments have promised put the world on course for 2.3 to 2.5°C, and that the policies actually in place lead to about 2.8°C. Wherever we land in the longer term, I expect the first single year above 2°C in the 2030s.
Severe effects of climate change are not waiting for 2°C. Roughly half the world’s population already faces severe water scarcity for part of the year. In the twelve months to May 2025, about four billion people lived through at least a month of additional extreme heat because of climate change. Heat that lasts this long makes it hard to work and, in some places, hard to live. Food is already less secure. The IPCC finds that heat, drought and floods have reduced food and water security and exposed millions of people to acute hunger. And nature, on which water and food depend, is giving way. The mountain glaciers and snowfields that feed the rivers nearly two billion people rely on are shrinking. All of this worsens with every further tenth of a degree.
These are the effects we can estimate. The warmer it gets, the less reliable such estimates become, and we cannot estimate at all what happens once tipping points are crossed: thresholds past which a system keeps changing by itself and cannot be brought back. Scientists judge that warm-water coral reefs have already passed theirs. The Amazon rainforest, the polar ice sheets and the Atlantic currents that shape Europe’s climate are at risk at temperatures we are now approaching. At around 2°C we will be close to or beyond several of them, and self-reinforcing effects will then amplify the damage.
In 2024, in my closing words for The Shift conference on climate adaptation, I said that adaptation is no longer an option but a necessity, and that we are in this position because we failed at mitigation. I would now put it more precisely: we failed to mitigate in time. Every tenth of a degree we still avoid counts. But a great deal of change can no longer be avoided, and that is what we have to prepare for.
The promise of the adaptation economy
Much is expected of the adaptation economy. It means the companies, public bodies and rules that supply the goods, services and infrastructure a society needs to stay safe and prosperous as the climate changes. The numbers attached to it are large. A 2025 study by GIC, Singapore’s sovereign wealth fund, and Bain & Company put annual revenues from a selected set of adaptation solutions at about $1 trillion today and the investment opportunity behind them at $2 trillion. And the World Resources Institute, looking at 320 adaptation projects, found that every dollar invested yields more than $10.50 in benefits over ten years.
I have no reason to doubt these figures, and I have quoted them myself. But read the last one closely. The benefits it counts are losses avoided, jobs and productivity gained, and improvements in health and the environment. Those are real benefits, but most of the value goes to the public at large, not to whoever may put up the money. A ten-to-one return to society is not a ten-to-one return to an investor. That gap is the subject of the rest of this essay.
The money has not followed
Even the optimists expect governments to carry much of this. A 2025 study by BCG and Temasek says adaptation spending must rise to between $0.5 and $1.3 trillion a year by 2030. It expects the public sector to cover a significant portion, and notes that public actors provide most of what is spent today. Other data show the same.
The Climate Policy Initiative tracks climate finance worldwide. For 2024 it counted $2.06 trillion. Of that, $1.9 trillion went into cutting emissions and $64 billion into adaptation: about thirty dollars for mitigation for every dollar for adaptation. The sources of the money differ just as sharply. Private sources supplied more than two thirds of the money for mitigation. Of the money for adaptation, by the previous year’s count, they supplied less than a tenth.
These statistics miss a good deal of private spending. A farmer who buys drought-resistant seeds rarely appears in them. Nor does a company that raises its electrical equipment above flood level, or pays extra for passive cooling in a new building. Nobody knows how large this spending is. It may be substantial.
But consider what that spending buys, and for whom. The farmer protects their own harvest and the company its own site. That is sensible, and it is the business most adaptation companies live on today. It does little for the neighbours, the town or the river basin.
So private money does go into adaptation, mainly where whoever pays is also the one who benefits. Where the benefit is shared, it mostly stays away. That is part, but not all, of the explanation. Let me illustrate the rest with another example.
A company and a flood
Take a company that knows its risk. It has had its main operating site assessed and can put a number on how likely a flood is and how much damage it would do.
The company will want to reduce its exposure. Perhaps it can build a barrier around its site, or buy insurance. But in a growing number of places neither will be enough, or affordable insurance will no longer be on offer.
What would actually help is a drainage system for the whole area, a restored floodplain, or a different way of managing the river upstream. The company cannot buy that. It is too large for one company to pay for, and much of the benefit would go to its neighbours, who have paid nothing. Some of those neighbours could not pay if they were asked. Others do not yet believe the flood is coming. Often nobody is offering such a solution at all. And even once it is built, its value is hard to show: if it works, the result is a flood that does no damage.
For solutions like these, there is not much of a market yet. I see four reasons for this.
Four reasons
Who can be charged. Economists call this excludability: whether people who do not pay can be kept from benefiting. A solar panel is excludable. Whoever buys it gets the electricity. A dyke that protects a town is not. It protects every building behind it, whether the owner contributed or not. Where people cannot be excluded, they cannot easily be charged, and businesses rarely build what they cannot charge for. That is also why nobody was offering the drainage system in the example: suppliers do not develop products that no one can be billed for.
The same need can often be met either way. A flood warning can be a siren that everyone hears, or a monitoring service that only paying subscribers receive. Protection can be a barrier around one site, or a dyke for the whole town. Private money goes to the versions that can be sold. But the shared versions are often the ones that protect the most people, and they often do so at the lowest cost: the dyke, the restored floodplain, the trees that cool a street, the public warning.
Cutting emissions has had it easier, for two reasons. First, what is sold and what helps everyone are usually the same thing. A wind farm sells electricity to its customers and cuts emissions for all. In adaptation the two often come apart: the barrier around one site can be sold, but the dyke for the whole town cannot. Second, where the shared benefit needed a buyer, rules created one. A price on carbon makes emitters pay, and carbon credits pay those who keep a forest alive. That was possible because a tonne of carbon dioxide is the same wherever it is saved. Adaptation has no such common measure, because it means very different things in different places.
Who can pay. Even where a solution can be sold, the buyer needs money. Here adaptation runs into an inverse relationship: the people and places most exposed are often the least able to pay. Climate analytics, resilient building materials and specialised insurance are bought by large companies and wealthy property owners. A smallholder facing drought, a poor and densely populated city district without shade, or a community that has to move away from the coast is a different matter.
This is not only a problem of poor countries. In the European Union, about a quarter of households said in 2023 that they could not keep their home comfortably cool in summer. Among the poorest fifth it was more than a third. Among the richest fifth it was fewer than one in five. Between countries the gap is wider still. The UN Environment Programme puts the adaptation needs of developing countries at $310 to $365 billion a year by 2035. It estimates that private finance could provide only about $50 billion of that, roughly a sixth, and expects the share to be higher in upper-middle-income countries and much lower in the poorest. It adds that most such schemes change who provides the money, not who pays in the end: the cost returns to the countries and households concerned.
So commercially viable adaptation tends to serve those who need it least.
Who is paying attention. Demand needs someone who believes the risk is real and accepts it as their own problem. Many of those who would have to pay do not yet see it that way, or they see it but rank it behind more pressing concerns. Priorities usually change right after an event such as a flood, a heatwave or a failed harvest. Then people buy and governments act. But attention fades quickly. It will grow as climate change becomes more persistent and the events more frequent and severe.
For now, attention and demand lag behind the risk by years.
What the benefit looks like. But even where there is a buyer who can be charged, can pay and is willing to, a problem remains. What that buyer is asked to pay for is, in most cases, a loss that does not happen. It does not appear as revenue, and nobody knows when the protection will be needed. That is an awkward thing to sell.
A wall built for a flood expected once in fifty years may stand unused through the few years in which most investment rules ask for payback. When the water does come, the value of the wall is plain. For much else it is not. What a restored floodplain upstream, a heat warning or a drought-resistant seed has prevented can only be estimated against what would have happened without it, and such estimates are easy to dispute. And part of the saving often goes to an insurer or the state, who would have carried part of the loss.
Until avoided losses are shown with evidence a finance director or an investor accepts, a good deal of adaptation that works will remain hard to sell.
Without money, no adaptation
None of this would matter much if someone else were certain to pay. Nobody is. And what nobody pays for does not happen. That holds for a dam. It holds just as much for the larger part of adaptation, which is reorganisation: redesigning water and power systems, reshaping supply chains, building in reserves and back-ups, and preparing health and emergency services for what is coming.
Governments will have to carry a large part of the cost. They cannot carry all of it. Before the financial crisis of 2008, global public debt stood at about 60 per cent of world output. It is now at 94 per cent, and the International Monetary Fund expects 100 per cent by 2029. Adaptation competes for the same budgets as defence, pensions and health. And prevention rarely wins majorities. Governments find money after a disaster more easily than before one.
So the more adaptation private capital can pay for, the more adaptation there will be, and the more public money is left for what only public money can do. I have said before that we cannot rely primarily on business opportunities to drive adaptation. I still think so. But it would be just as wrong to leave unused the private capital that can be brought in.
One caution. Private capital does not make the cost disappear. Whoever provides it expects the money back with a return, paid by the users of what was built or by taxpayers over time. It helps most where there are users who can and will pay. Elsewhere it can spread a public cost over years, but it cannot remove it.
Where can private capital go today, and what might change that?
Private capital is not one thing. A property owner pays to protect a building. A bank lends where repayment is predictable. A venture investor backs a product that can grow fast. Each fits different solutions. But all of them need the same thing first: someone who pays in the end.
By that test, adaptation falls into three groups.
Investable today. The solution can be sold to a buyer who benefits and can pay. Climate analytics, resilient building materials, cooling, water-saving technology, drought-resistant seeds and some forms of insurance belong here. This is where the trillion dollars in revenues are earned, and where almost all private investment in adaptation goes, my own included.
Could become investable. The solution exists or is within reach, but one or more of the four reasons stands in the way. Examples are making power and water supply robust against heat and drought, protecting smallholder farmers, and restoring the wetlands and forests that hold back water. This is where I see solution providers struggle most. They often have a product that works and a problem that already exists. What they lack is a paying customer. Without one they grow slowly, which rules out venture investors, and they have no predictable income, which rules out banks.
Most likely a public task. Moving a community out of harm’s way, preparing health systems for heat and new diseases, and protecting the poorest are examples. It is hard to see a rule or an innovation that would produce a paying customer for these.
How large is each group? A 2026 study by Allianz sorted 61 adaptation measures by how far private actors can capture the benefit. Its categories are close to mine. It found 23 per cent to be primarily private, 39 per cent to need public and private money together, and 38 per cent to be primarily public. That is a count of measures, not of money, and where the boundaries fall is a matter of judgement.
The middle group is also where the most can change. None of the four reasons is a law of nature. Twenty years ago, wind and solar power were built almost only where governments guaranteed the price. Today most of it is financed privately. Guaranteed tariffs, a price on emissions and industrial policy made that happen.
Adaptation is harder, because it has no common measure as carbon does. But each of the four reasons can be worked on, in three ways.
Innovation can change what is possible. Sensors and satellite data can turn a shared warning into a service that can be sold. Insurance that pays out automatically when a measured threshold is crossed can be cheap enough to reach smallholder farmers.
Shared risk can bring a price within reach, through subsidies, guarantees or public co-investment. A public guarantee can make a bank willing to lend for a town’s flood defences.
Rules can create the buyer, through building codes, disclosure requirements, insurers and lenders that take protection into account, or credits that pay those who restore a wetland or a forest. On one stretch of the Mexican coast, hotels and property owners already pay into a fund that insures the reef protecting their beaches.
What to ask, and what comes next
Will an adaptation economy deliver the adaptation we need? In part. It will not do so alone, but its contribution can grow.
Going by the Allianz count, a little under a quarter of adaptation measures can be financed privately today. About two fifths more could be, if other stakeholders share the risk. The rest will most likely remain a public task, and it includes much of what protects the people with the least.
For anyone who wants to advance adaptation with the help of private markets, whether as an investor, a founder or a funder, three questions follow.
Where can capital go now? Who benefits, can they be charged, can they pay, and can they be shown what they are paying for? If the answer to all four is yes, there is a market.
What would open up the rest? Is there an innovation, a way of sharing the risk or a change of rules that would turn a no into a yes, and is anyone working on it?
And who has to pay where markets will not? If nothing will change the answer, this is a public or philanthropic task, and the useful question is how to get it funded.
These questions will run through much of what I write here. I will also look at single sectors and problems, sometimes as an investor, sometimes as an adviser, and sometimes simply as a concerned citizen. I expect to keep having more questions than answers. If you are working on any of them, I would like to hear from you.