The catchment doesn't average

“I don't know if you can see, the changes that have come over me In these last few days, I've been afraid that I might drift away So I've been tellin' old stories, singing old songs That made me think about where I came from That's the reason why I seem so far away today”

"Caledonia" sung most memorably by Paolo Nutini at the Garage in Glasgow in July 2006 (and originally written by Dougie MacLean).

I make no claims to be the world’s biggest football fan, but it’s been 28 years, so hopefully you’ll forgive me the above nod to the bonnie lads in blue over in North America?! I still would prefer it as our national anthem – Flower of Scotland is too dirge-like for my tastes. But to the real business at hand, this is part deux!: two water reports from the same bank, 24 days apart. Then a third, from a man who used to run it.

Goldman Sachs Asset Management's April note on water stress was product-distribution: an investable theme, one dollar in five of listed-market revenue exposed, calibrated rather than catastrophist, and a pretty clean physical-risk analysis that any major bank has put out this year. The report that followed 24 days later was different in kind. It carried Jared Cohen's geopolitical signature - think Iranian drone strikes on Bahraini desalination, the suspension of the Indus Waters Treaty, the leverage dynamics on the Yarlung Tsangpo - and was built to travel into sovereign meetings and allocator conversations rather than product decks. The Goldmans framing moved from investable theme to strategic concern with an investable response. That is a different thing to put in front of a mandate committee.

And behind both sits the plumbing. The Paulson Institute (let’s not forget that Henry Paulson ran Goldman before he ran the US Treasury) co-authored the AIIB/EBRD framework for nature as public-private infrastructure, with worked examples such as Bank of Jiangsu lending against projected gross ecosystem product as partial collateral for a wetland restoration loan. Surely these are not just three publications that happened to share a quarter? (who doesn’t love a good conspiracy theory?!) I think they might well be three layers of one construction: the product, the strategic rationale that lifts it to priority, and the delivery mechanism through which the capital would flow.

And what’s more, I think that their direction is broadly right: markets are made, not found, and this is competent market-making. But it is worth noting and setting beside them two documents from the same window that answer to none of those incentives - one from the UK Parliament and one written expressly to make water investable.

The first is POSTnote 768, the UK Parliament science office's peer-reviewed synthesis on nature-based flood and drought resilience (refs below). No theme, no buckets, no allocator in mind, and read alongside the market-making, it documents why the asset everyone now wants to finance is so hard to price.

The finding that runs through it: effectiveness depends on where you put the intervention. The same leaky dam attenuates a flood peak in one sub-catchment and worsens flooding near a river outlet or on a steep tributary (“right intervention, right place”, I can hear Neil Entwistle chirping in my ear!). Field-corner bunds in the Scottish Government's Eddleston study cut discharge peaks by up to 55%; the same measure elsewhere does very little. This is the physical-science version of the El Niño point I made last time. A balance sheet is not exposed to the planetary mean, rather it is exposed to one orchard, one aquifer, one river basin at a time. The catchment is the resolution at which the value becomes very real, and the catchment does not average.

The report is also unusually honest about measurement. Projects are frequently not monitored, baselines often do not exist, and hydrological variability swamps the signal from any one intervention. And we know full well that you cannot mark an asset you cannot measure. The Jiangsu mechanism only works if the ecosystem-product gain is measurable, verifiable and durable over the life of the loan, and POSTnote 768 is in effect an account of how hard each of those is at catchment scale. Insurers told the same inquiry they cannot yet be confident these measures cut flood risk to most properties. In essence they said that you cannot underwrite what you cannot attribute.

Which is where the second document picks up the thread, from an unlikely direction. This month the Global Climate Finance Centre, the Abu Dhabi body established at COP28, working with Water.org, published its Water Finance Framework for the UN Water Conference, built on 192 real projects across 40 countries. It is not a critique of any one, instead it is a manual for making water investable, and its governing discipline is that capital must follow service reality, not ambition. Read against the market-making, that turns out to be quite a high bar.

Its organising principle is that financing follows a credible payer (“start with the end in mind, who’s the buyer?”, asked Robert Gardner !). Capital can shift when a cost is paid; it cannot invent who pays it. Where the payer is unclear, fragmented or politically unreliable, repayable finance does not hold, and risk instruments do not rescue it. Guarantees, insurance and blended structures move risk around; but they cannot conjure bankability the underlying service does not support. Apply that to the two things the Goldmans and Paulson work is really about. Water ecosystem conservation sits in the framework's lowest, non-commercial band unless a downstream beneficiary is contracted into paying for the service, the water-fund logic of Upper Tana–Nairobi or the Xin'an River compensation scheme in China; and the moment that payer appears, the concessional money exits. Flood and drought protection, the thing every resilience pitch is selling, is the one archetype it says never reaches commercial finance at all. Across those 192 projects the commercial share tops out near 9%, described as a ceiling rather than a stepping stone. Protection that benefits everyone downstream can be billed to no one.

There is a distributional sting on top, and I think it is the one that matters most. Work in Communications Earth & Environment (Hill et al., 2025) finds market-based instruments for nature-based flood management can disproportionately benefit affluent areas: funding follows the landowner who can host an intervention cheaply, not the catchment where flood risk and social vulnerability coincide. Left to itself, the market allocates resilience to the people who already have it (or can best afford it). You can call that the difference between fixing a market and shaping one. There is a more unsettling way to read it, though, and it lands on my own argument as much as on Goldman's.

Writing in Nature the same month, Walter Radermacher, a statistician and former head of Eurostat, no romantic about these things, argues that pricing nature cannot address the worst risks at all, and that treating it as capital gets the relationship backwards. His objection is not that the numbers are hard to produce. In essence, for shared resources like the atmosphere, the oceans and the wider biosphere, two questions have to be answered before any price can mean something: how much pressure the system can take, and who is entitled to how much of that capacity, across which countries and generations. Those are political decisions, not economic ones, he rightly says. Settle them and scarcity is defined and a price can do real work. Skip them and the price is modelling assumptions with a decimal point (more model, less on the ground fact). He calls the line where that happens the monetisation possibility frontier, and he is clear it is not a data gap to be closed with better monitoring but a question of sequence. Scale and distribution come before allocation, or the allocation is fiction.

This is the strongest objection to the series I have been writing. My argument has been that nature is mispriced, that the substrate sits at zero and the accounting will have to catch up. Radermacher's is that for the part which matters most there is no right price to catch up to, because the thing was never a line on a balance sheet. The economy is not a system with nature inside it; it is contained by nature and wholly dependent on it. Capital accounts, he allows, can handle the low-risk, local cases, the standing timber, the water a single forest regulates, but they cannot hold the commons. And of course, he is right … but I think I am too!

That concession is the interesting part, because it draws the line roughly where this whole argument has been circling. The catchment, the orchard, the aquifer: the asset-level resolution I started from is exactly where Radermacher allows a price can still mean something. The planetary average, the indexed commons, is where it stops. And that line turns out to be a familiar one.

The commons he wants settled politically first, the atmosphere, the biosphere, the aggregate pressure on the Earth system, is the territory of mitigation: limiting the drivers, deciding who may emit and consume how much. He is right that none of it can be priced into shape before the scale and the sharing are agreed, and right that allocation without that agreement is fiction. But the agreement is not arriving and we have already spent three decades trying to assemble the political will for it and yet still we do not have it, and we certainly do not have the time to wait while biodiversity drains from the system. His framework does not reach the other half of the problem tho. Adaptation, coping with the floods and droughts already locked in, is local by nature, it is urgent, and it cannot be sequenced behind a global settlement that shows no sign of coming. At the catchment scale, for adaptation, a price and a payer are not a category error. They are the only instrument moving at the speed the water is, and the catchment is where you buffer the flood and rebuild the biodiversity at the same time, settlement or no settlement.

Which is the real test for the instrument none of these documents quite names. We already have Biodiversity Net Gain: a measured unit of habitat, credited, traded, mandatory for most new development in England. The next coinage almost writes itself: Resilience Net Gain (thank you Neil Entwistle , Alistair Maltby and Ludovic Pittie ), a catchment's capacity to buffer flood and drought, quantified and sold to whoever carries the downstream risk. The Paulson framework is the plumbing it would run through and the Goldmans reports are the demand signal. Restoration finance is just the early product.

We know that success needs three things all at once: a unit you can measure, which POSTnote 768 suggests may dissolve at the resolution where the risk actually lives; an offtaker you can contract, which the Global Climate Finance Centre, whose whole job is to find one, classes as structurally absent for flood and drought; and it would have to stay on the right side of Radermacher's frontier: defensible as the buffering of one catchment for downstream beneficiaries, a category error the moment it is aggregated, indexed and sold as a commons-scale asset.

So I will give Radermacher the commons. His is the better theory of mitigation, and a price was never going to stand in for the political settlement that limiting planetary pressure requires. But theory is a luxury of people who are not about to be flooded. For adaptation the catchment is the only scale where anything is actually moving, and the absence of a global agreement is not a reason to wait, rather it is the reason to build what works locally, now.

The catch sits in the gap between the two. If cracking on at the catchment becomes the way those who can afford resilience quietly buy their own and leave the distribution question for someone else, then Hill is right and the market has just protected the people who already had cover. That is why I believe that Resilience Net Gain is worth building. Whether it buys time for the harder politics or buys the comfortable an excuse to skip it is not something the instrument decides.

I'll leave that one there. C’mon Scotland!

---

References

Goldman Sachs Asset Management, Finding Investment Opportunities in the Global Response to Water Stress (April 2026). https://am.gs.com/en-us/advisors/insights/article/2026/finding-investment-opportunities-in-water-stress

Goldman Sachs Global Institute, Securing and Financing the Future of Water (2026). https://www.goldmansachs.com/insights/articles/securing-and-financing-the-future-of-water

AIIB, EBRD and the Paulson Institute, Unlocking Private Finance for Nature as Infrastructure: A Public-Private Partnership (November 2025). https://www.aiib.org/en/news-events/news/2025/_pdf/Unlocking-Private-Finance-for-Nature-as-Infrastructure11.15.pdf

Ben Cutting and Jonathan Wentworth, Nature-based flood and drought resilience, POSTnote 768, UK Parliament Office of Science and Technology (1 May 2026). https://doi.org/10.58248/PN768

Global Climate Finance Centre with Water.org, Water Finance Framework (June 2026). https://gcfc.com/wp-content/uploads/2026/06/Water-Finance-Framework.pdf

B. Hill et al., 'Market-based instruments to fund nature-based solutions for flood risk management can disproportionately benefit affluent areas', Communications Earth & Environment 6, 714 (2025).

Walter J. Radermacher, 'Treating nature as capital cannot save the planet', Nature 653, 984 (28 May 2026). https://doi.org/10.1038/d41586-026-01628-z