Post date 14 Sep 2026 (original LinkedIn date)
Original LinkedIn version· 2,671 words
‘Movin’ on up’ by Primal Scream, released in 1991, on their album Screamadelica.
One of the most persistent explanations for why nature finance has failed to scale is that the market is not sufficiently standardised. There is something obviously right about this. Institutional capital cannot efficiently underwrite a succession of entirely bespoke transactions, each with different legal terms, metrics, reporting arrangements, payment mechanisms and approaches to underperformance. If every £3 million project requires six figures of legal work and a fresh investment committee education, there is not going to be a £30 billion asset class coming along any time soon!
But I am beginning to wonder whether we have interpreted the standardisation problem in the wrong way. Too often, the instinct has been to standardise the nature, which simply cannot be done. We look for a common unit, a methodology capable of producing a comparable score, or a credit that can be issued in one place and understood, traded or retired somewhere else. There are good reasons for doing that, particularly where the purpose is to create a commodity market. The difficulty (and beauty) is that ecology is stubbornly resistant to sameness.
A hectare of restored peat in Yorkshire is not performing the same function as a hectare of saltmarsh in Essex. A wetland reducing nutrient loading into a chalk stream is doing something different from a woodland slowing peak flows above a town. Even interventions that look superficially identical can perform differently because rainfall, soils, topography, land management, downstream assets and the wider ecological system are different. That is not a flaw in nature finance that better data will eventually eliminate. It is simply a property of the asset. So perhaps we should design the finance around it.
The food system makes the problem unusually obvious
Having been privileged to attend and speak at last week’s inaugural Food and National Security event, I have been thinking about this quite a bit over the last several days. There is a tendency in both policy and finance to talk about “food systems and natural capital” as though that describes a coherent investment category. But the truth is that it does not. Landscape restoration looks much more like long-duration infrastructure. Farm transition actually might often be more of a credit problem. Processing, storage and distribution may require growth equity or asset finance. Emergency food provisioning is largely a public good for which no logical private payer exists. Treating them as one investable category perhaps obscures rather than simplifies the financing problem.
Recent research makes the ecological equivalent of the same point. Bodirsky and colleagues modelled 23 different measures intended to transform the global food system across 15 environmental, health, social and economic outcomes. Every individual measure involved trade-offs. Combining measures could reduce those trade-offs and generate co-benefits, but there was no single intervention that simply improved everything everywhere.
A separate Nature paper published in July tackles the same problem from another angle. Modelling a transition towards healthier diets, higher productivity and lower food waste, the authors find not simply a smaller environmental footprint but a substantial restructuring of agriculture itself. Relative to their business-as-usual scenario, livestock production falls sharply while the value of vegetables, fruits, nuts and legumes rises, with significant differences between sectors and regions. The authors explicitly point to the political economy created by those uneven effects.
And a Nature Food perspective published only a couple of weeks ago makes an important related observation. Disagreement around food systems is not always evidence that we need better information until everybody converges on the same answer. Agriculture, food, health and the environment involve genuinely different values, interests and objectives. Productive engagement sometimes means making those differences explicit rather than treating them as noise that a sufficiently elegant model will remove.
And perhaps there is the lesson for finance. Complex natural systems are heterogeneous because the underlying functions, beneficiaries and trade-offs are heterogeneous. If we insist that institutionalisation requires us first to reduce them to a common ecological denominator, we may spend years trying to solve a problem that does not actually need solving. Finance has encountered this problem before, and it normally solves it somewhere else.
Finance rarely standardises the thing itself
Consider the UK Contract for Difference. An offshore wind farm is not the same asset as an onshore wind farm, solar installation or other low-carbon generating project. Their construction risks, output profiles, technology, operating characteristics and economics can differ substantially. Yet the financing architecture does not require the underlying power stations to become identical.
The current Allocation Round 8 documentation contains a set of standard terms and conditions alongside a front-end agreement containing the project-specific information required to tailor those terms to the individual project. The asset remains specific, and the contractual architecture is familiar to all.
Derivatives markets provide an even more striking example. The ISDA Master Agreement can govern transactions across different products and asset classes. Nobody imagines that an interest-rate swap and a foreign-exchange derivative have somehow become the same economic exposure because they sit within a common legal framework. What has been standardised is the relationship between the parties, including matters such as payment, representations, default and termination.
The loan market has done much the same through standard documentation. Securitisation did not require every mortgage or borrower to be identical. Infrastructure finance does not require every bridge, airport or electricity network to perform the same physical function. Mature financial markets routinely standardise the wrapper while allowing considerable heterogeneity in the thing inside it. I truly believe that Nature finance might usefully learn the same trick.
What might the wrapper contain?
This is where I think the idea of a Resilience Purchase Agreement becomes more interesting than simply creating another environmental contract. This is my 18-month hope, Tom Pycraft. We have been arguing for a standard-form agreement through which an identifiable beneficiary could purchase a defined ecosystem service over a sufficiently long period to allow investment against the resulting cashflow. The attraction of a common contract is partly mundane but important: common definitions, payment mechanics, indexation, reporting, step-in rights, underperformance provisions and dispute resolution drastically reduce the amount that has to be reinvented for every transaction. But of course a standard agreement need not imply a standard ecological outcome.
Imagine instead a common contractual architecture with project-specific schedules sitting beneath it. The commercial part might define the buyer, the payment term, indexation, credit support, change of control and what happens if either party defaults. A performance schedule would define the service being purchased and the conditions under which payment is made. A measurement schedule would specify the baseline, the counterfactual, monitoring arrangements, uncertainty tolerances and what happens when observed conditions move outside the range envisaged when the contract was written. A site schedule would describe the intervention itself, its geography and the assets or activities expected to benefit.
None of those schedules needs to contain the same ecological metric from one contract to the next. An upstream catchment agreement might be concerned with the reduction in peak flow reaching a particular location. A wetland contract might focus on nutrient removal and avoided treatment cost, whereas a soil-restoration agreement could be linked to water retention, erosion or yield variability. Likewise a saltmarsh project might be evaluated against wave attenuation and the expected loss to coastal assets under different storm conditions. The common feature would not be that all four projects produce an interchangeable ‘unit of resilience’. Therefore it would become the case that an investor approaching any of the four encounters a familiar legal and financial architecture and can understand where the project-specific ecological risk sits within it. That feels like a very different form of standardisation.
Comparability does not require fungibility
This distinction matters because institutional investors need comparability, but they do not necessarily need ecological fungibility, and those things are frequently conflated. For portfolio construction, an allocator needs to know the expected cashflow, duration, counterparty quality, downside scenarios, conditions for payment, performance risk, default treatment, correlation with other assets and the extent to which actual outcomes can diverge from modelled ones.
All of that can be standardised in the way it is reported and analysed without requiring the physical interventions to be equivalent. Indeed, forcing equivalence may sometimes destroy useful information.
A generic unit that makes two projects appear comparable by suppressing their physical differences may make life easier for a spreadsheet while making the investment less intelligible. If the reason a wetland generates economic value is that it lowers a particular water company’s treatment cost, I would rather underwrite that causal relationship than convert the wetland into a more portable but less meaningful abstract score.
There is an important difference between making an asset legible to finance and making it fungible. We need much more of the former than the latter.
Aggregation without homogenisation
This should also change how we think about aggregation. Nature projects are usually too small for institutional capital. A £1 million or £3 million restoration scheme may be perfectly sensible economically and yet utterly irrelevant to a large pension fund whose minimum ticket begins at £25 million or £50 million.
The conventional response is therefore to aggregate projects. But fifty bespoke projects do not automatically become an institutional asset merely because someone puts them into the same fund. If they each contain different legal terms, reporting obligations, payment mechanisms and approaches to default, the complexity has simply moved upstairs.
A common wrapper could change that. If fifty transactions use the same underlying contractual architecture, then the investor can begin to analyse the portfolio rather than repeatedly analysing the documentation. Cashflows can be aggregated and counterparty exposures can be compared. Underperformance can be reported consistently, with the portfolio developing a track record whose meaning survives the addition of the fifty-first project. The ecology underneath can remain highly specific because specificity has been designed into the schedules rather than treated as an exception to the contract.
This is especially important because the appropriate unit of intervention is often larger than the land parcel on which the work happens. In the food-system work we heard about last week, we make the point that the investment unit may need to be the landscape because the unit of benefit already is. Water, pollination, shelter, soil processes and flood attenuation frequently cross ownership boundaries. Aggregation is therefore not simply a way of producing a larger financial ticket. It can be the structure through which an ecological externality is finally internalised. That feels much more interesting than bundling credits.
It also gives uncertainty somewhere to go
There is another advantage. One of the more uncomfortable realities of nature-based investment is that the relationship between intervention and outcome is rarely deterministic. Weather changes and baselines move. Ecological responses take time, and an extreme event can overwhelm an intervention that performs perfectly well under more ordinary conditions. Modelled benefits may prove larger or smaller than the benefits observed. Standardisation cannot remove that uncertainty and attempts to create a universal metric sometimes risk concealing it.
A good contractual wrapper can instead allocate it. The parties can agree what constitutes acceptable performance, how often the baseline is revisited, which risks sit with the project and which sit with the buyer, whether payments operate within a corridor rather than at a single threshold, and what happens when an event lies outside the range for which the intervention was designed. In financial language, we can make basis risk visible rather than pretending it does not exist.
That seems particularly important if nature is to be treated as infrastructure. Nobody expects a conventional infrastructure asset to be risk-free. What investors require is clarity about which risks they are accepting, which have been transferred, which are retained by counterparties and how they will be compensated for bearing them. Nature should not be held to some imaginary higher standard in which uncertainty must first disappear.
Perhaps we have confused standardisation with sameness
This, I think, might have been our mistake. The nature-finance sector has often behaved as though scale requires us to discover a unit in which ecosystems become sufficiently alike to trade. For some markets, that may be exactly right. Carbon markets require a tonne of carbon dioxide equivalent to mean something reasonably consistent across transactions. Biodiversity markets designed around statutory obligations need rules that establish what counts towards discharging those obligations.
But resilience is just a bit different. Its economic value is often relational. A restored floodplain has value because of what lies downstream. A wetland matters because of the water body into which it drains and the treatment costs it changes. The value of healthier soil depends partly on rainfall, crops, management and the system surrounding it. The same intervention in another place can therefore have a completely different economic effect. Trying to strip away that context so that the asset becomes tradable may also strip away the causal relationship that creates the revenue. So the better route may be almost the reverse ie preserve the ecological specificity, but make the contractual claim around it familiar.
The opportunity for government
This is why I think there is a relatively simple intervention available to government that could have disproportionate effect. Rather than spending another few years trying to devise a universal nature asset, DEFRA, HM Treasury and any other government/quasi-govt group could convene the market around a standard-form contract for purchased ecosystem services.
There is nothing particularly exotic about that proposition. Government already publishes hundreds of pages of standard terms for the electricity CfD precisely because contractual certainty helped convert highly capital-intensive, long-duration assets into something finance could repeatedly underwrite. The latest CfD continues to combine generic standard terms with project-specific information.
A resilience equivalent would clearly need to evolve through use. Different service types would develop different schedules. Insurers, water companies, network operators, local authorities and food businesses would want different performance provisions. Some contracts would fail. Terms would be refined as claims were tested in the real world. That is what market development looks like.
The important thing is that the legal architecture could evolve collectively rather than being rediscovered privately behind the doors of every transaction. And that matters because the problem facing nature finance may no longer be primarily an absence of evidence that nature has value. We have accumulated an extraordinary amount of evidence on that question, and it clearly does. The harder problem is converting a particular ecological function, in a particular place, for a particular beneficiary, into a claim sufficiently clear that somebody can sign it and somebody else can finance against it. And that does require standardisation, but just not necessarily where we have been looking for it.
Nature is variable, contextual and interconnected. Food systems are full of real trade-offs rather than optimisation problems waiting for one correct answer. Landscapes provide different services to different beneficiaries at different scales. None of those characteristics is likely to disappear as our measurement improves, and nor should we want them to.
The financial system does not scale because every underlying asset becomes identical. It scales because the rights, obligations and risks around those assets become sufficiently legible that capital can compare them, price them and combine them. Perhaps nature finance has spent too long trying to standardise nature. We might make considerably more progress if we standardised the wrapper and left the ecology alone.
References
Bodirsky, B. L. et al. (2025), “A food system transformation pathway reconciles 1.5°C global warming with improved health, environment and social inclusion”, Nature Food, 6, 1133–1152.
Gibson, M. et al. (2026), “Food systems transformation would reshape global agriculture”, Nature, 657, 156–163.
Jackson, L. E. et al. (2026), “Engagement in food system-related controversies can strengthen science–policy–society interfaces”, Nature Food, published 2 September 2026.
Department for Energy Security and Net Zero (2026), Contracts for Difference Allocation Round 8: Standard Terms and Conditions and Generic Agreement.
International Swaps and Derivatives Association, Legal Guidelines for Smart Derivatives Contracts: The ISDA Master Agreement.
Conference on Food & National Security, London, 8 September 2026.