Protect the Golden Goose: A Regenerative Infrastructure Thesis

As leader Stephen Covey described decades ago, real effectiveness is a balance between P(roduction), the eggs you produce, and PC, the productive capacity that lays them. Maximize P alone and you get a record quarter, but it will diminish your PC. That is a dead goose: a dry and desiccated slate of key resources. A world in turmoil.

The global economy has run a P-only strategy on its physical base for two hundred years, since the Industrial Revolution.

Soil: The Food and Agriculture Organization of the United Nations reports roughly a third of the world’s soils are degraded, and 1.7 billion people live where human-caused land degradation has cut crop yields by at least 10%.

Water: A global study in Nature found groundwater levels decreasing in most of the aquifers it tracked, with declines accelerating over four decades in about 30% of them.

Nature: The World Economic Forum estimates $44 trillion of economic output, more than half of global GDP, depends moderately or highly on nourished biodiversity. Yet UNEP counts $7.3 trillion a year flowing into activity that degrades it, against $220 billion into solutions that restore it. A 30 to 1 short position on our own life-support system. See below:

Weather: Insured catastrophe losses have topped $100 billion for six straight years. The January 2025 Los Angeles fires alone cost insurers about $40 billion, and Swiss Re– one of the biggest reinsurers in the world– sees losses growing 5 to 7% a year in real terms.

This is what killing Stella looks like, and if we do not move away from extractive processes, the consequences will be dire. Look at almost any global resource trend, and the assets that rebuild productive capacity stop looking like philanthropy and start looking like what they are: the most essential infrastructure on Earth, bought at a thirty-to-one discount.

Regenerative Infrastructure is the physical systems that get more productive as they work, and the people and capital that build them.

Every infrastructure asset of the last century was built with an expectation that it had a degradation timeline. Every pipeline, every highway and overpass, every coal plant. Each of these starts at peak value and decays toward zero, while drawing down the natural resources around it. We call this decaying infrastructure, or any that is not regenerative, “grey infrastructure.” Regenerative systems run the other way and take time to ramp up: a restored watershed filters more water in year ten than in year one. Rebuilt soil holds more water and grows more food every season.

Regenerative inputs, sun, soil and rain, cost near zero once built, where grey runs on fuel and chemicals with volatile costs. But the biggest difference is climate exposure. Grey infrastructure is built for the average year, but the average year is now a thing of the past. A levee is designed for the hundred-year flood that now arrives every decade. A reservoir is sized for rainfall patterns that no longer hold. Every extreme event shows up on the grey balance sheet as damage, downtime and a higher insurance bill.

Regenerative assets have the opposite exposure, because they are built from systems that evolved to absorb shocks. Healthy soil works like a sponge: it holds water through drought and absorbs it in a downpour, so the same field loses less in both. Solar and storage on the roof keep the lights on when the grid goes down, with no fuel supply to disrupt.

Most investments are a bet that the future looks like the past. Regenerative Infrastructure is a bet that it will not, and it pays either way.

Regenerative Infrastructure spans five physical systems. Each can be rebuilt for productive capacity, with a cash flow waiting to be financed and a workforce waiting to be trained.

WATER: Water can feel infinitely replenished, even when we’re told that levels are in record-low territory. But it isn’t, and we need to work to rebuild its supporting systems: Watershed restoration, aquifer recharge, nature-based treatment and efficient delivery all rebuild supply rather than mining it.

The proof case is New York City. Since 1997 it has spent about $2.5 billion protecting its Catskill and Delaware watersheds: paying farmers, buying land, fixing septic systems. In return it avoided a filtration plant estimated at up to $10 billion, plus hundreds of millions a year to run it. That is an avoided-cost contract, and every water utility facing drought can write one.

The work: water operators, hydrologists, watershed restoration crews, environmental engineers.

SOIL AND FOOD. The world’s largest infrastructure asset is the top foot of dirt that grows more than 95% of our food. We have been running it to failure.

Regenerative agriculture treats soil as living capital. Cover crops, reduced tillage, diverse rotations and integrated grazing rebuild organic matter, hold more water and cut the input bill. BCG’s work with the One Planet Business for Biodiversity coalition found a 15 to 25% return on investment over ten years once farms reach steady state.

The catch is a three-to-five-year transition valley before the soil pays back. And the buyers are waiting. PepsiCo has reached 4.7 million acres of regenerative practices against a 10-million-acre 2030 goal.

The work: regenerative farmers and ranchers, agronomists, soil scientists, farm managers, food brand operators.

LAND AND LANDSCAPES. Forests, grasslands and wetlands restored at landscape scale are Stella’s habitat. Every dollar invested in restoring degraded land returns an estimated $7 to $30 in food production, water quality and carbon storage.

The returns are real but scattered: the water utility, the insurer and the town downstream capture the value, while the landowner who restored it captures nothing. The investable model stacks those payments into one cash flow.

The work: restoration ecologists, foresters, wildland fire crews, GIS and drone mapping technicians.

OCEANS AND COASTS. Mangroves, kelp, seagrass, restorative aquaculture and living shorelines protect the most valuable real estate on Earth. Mangroves alone keep 15 million people out of floodwater each year, and in a 100-year storm they save $270 billion of property.

Yet ocean health is the least funded of all 17 Sustainable Development Goals, with less than a tenth of the money it needs. The natural counterparty is the insurer, whose loss curve these ecosystems flatten.

The work: marine biologists, restoration divers, kelp and shellfish farmers, coastal engineers.

DISTRIBUTED CLEAN POWER. Clean power is now the cheap choice and the fast one. IRENA found more than 90% of new utility-scale renewables commissioned in 2025 undercut the cheapest new fossil plant. Sunlight cannot be overharvested.

The most underpriced slice is solar and storage on the homes of working families. One in four low-income U.S. households spends more than 15% of income on energy, against a national median around 3%. With the homeowner solar credit gone and the third-party credit intact through 2027, the model shifts to investor-owned portfolios of thousands of roofs, underwritten on contracts rather than household credit scores. The family cuts its largest controllable bill and gains backup power in an outage, and the investor earns an infrastructure return.

The work: solar installers, electricians, storage technicians, lineworkers, wind technicians.

Whether you are looking for the light that will illuminate your career path or an investor looking to get ahead of a new asset class as it is defined in real time, three questions tell you which parts of this are worth your time or your money:

First, does it pay? For a career, that means real demand for the skill, at a wage you can build a life on. For capital, it means a contracted or saleable cash flow at market rates.

Second, does it regenerate? For a career, your work leaves soil, water, land or a community measurably better. For capital, the asset leaves its system more productive, measured not marketed.

Third, does it compound? For a career, that means skills that get more valuable with experience and cannot be copied by software. For capital, it means productive capacity that grows or holds, so time works for the owner.

Anything that passes all three is part of Stella’s recovery. Anything that fails the first is charity. Anything that fails the second is extraction. Anything that fails the third will be automated, depreciated or outcompeted.

The careers most insulated from AI are the ones that happen in a physical place, with your hands, under conditions that change. Regenerative infrastructure is almost entirely that kind of work. No model can download a harvest, recharge an aquifer or restore a coastline.

The evidence is already arriving. The entry rung of desk work is shrinking: Stanford’s Digital Economy Lab found that since generative AI took off, 22- to 25-year-olds in the most AI-exposed jobs saw a 13% relative drop in employment, while hands-on roles held steady or grew. That first job is exactly where the cuts are landing.

Physical work is the least automatable. Goldman Sachs estimated that AI could take over 46% of office and administrative support tasks, but only 6% in construction and extraction, and 4% in installation, maintenance and repair. That suggests a simple rule for choosing a career: the more of the job that needs you physically present, reading real conditions and making judgment calls, the more insulated it is.

The fastest-growing jobs in America are regenerative. The Bureau of Labor Statistics projects wind turbine technicians, up 49.9%, and solar installers, up 42.1%, as the two fastest-growing occupations from 2024 to 2034, against 3.1% for all jobs. These are small fields today, so treat the growth as a signal.

Meanwhile, the current workforce is retiring. In advanced economies, 2.4 energy workers are nearing retirement for every new entrant under 25, according to the IEA, and more than half of the companies it surveyed already report critical hiring bottlenecks. The work also pays without a mountain of debt: the median lineworker earned $92,560 in 2024, in a job that typically starts with a high school diploma and on-the-job training.

Desk roles in this field still matter, but they are AI-augmented, not AI-proof: AI will make each analyst and modeler more productive, which also means fewer of them. That does not make AI the enemy. Drones, satellites and soil sensors make field workers far more effective, and Stanford found employment grows where AI augments people rather than replacing them. The strongest position combines both: the hydrologist who can read the sensor data and walk the creek; the farm manager who can run the soil model and read the field.

The biggest mistake in nature finance was asking people to accept less.

For two decades, impact was sold as a trade: give up some return, get some good. That capped the market at the size of philanthropic budgets. It is why private investors supply just $23 billion of the $220 billion that flows to nature-based solutions. Fiduciaries cannot buy a concession. Nor should they.

Regenerative Infrastructure makes no such request. Its cash flows carry the properties institutions already pay up for in core infrastructure: essential services, long-dated contracts, inflation linkage and low correlation to public markets. Core infrastructure has returned 8.8 to 9.2% for three straight years; farmland has delivered roughly 10% annualized with low volatility. Regenerative assets are those same categories with two extra edges: no fuel cost, and a productive base that improves instead of wearing out.

Nature is not unprofitable. It just sends the invoice to the wrong address. The work is structuring: find the beneficiary, write the contract, finance the cash flow. Each of the five systems already has a natural buyer: the utility, the food company, the insurer, the household.

The capital is already moving. Brookfield closed a $20 billion transition fund in October 2025, the largest on record, with pensions and sovereigns underwriting to market returns. On nature, the TNFD’s 2026 status report counts 802 adopters representing $26.6 trillion in assets, yet finds nature data barely shaping portfolio construction. That much capital now measures nature risk and has almost nothing to buy that reduces it. That is the opening.

Four forces are converging, and each one raises the value of a healthy and nourished physical world. and of the people who tend to it. Two we have covered: capital rotating into the physical as AI makes software abundant, and a generation looking for work that lasts just as the physical workforce retires.

Geopolitics has made self-sufficiency a security asset. The Iran war is the latest reminder that imported fuel is a strategic liability. Communities that generate their own power, grow their own food and secure their own water are less exposed to someone else’s conflict.

The natural infrastructure is visibly failing, with insurers retreating from whole regions. When the systems we took for granted stop delivering, people start putting a price on them.

None of these forces are waiting for permission or subsidy. They are the ordinary mechanics of capital and talent moving toward what is scarce and essential.

Of course, the transition to regenerative presents a set of key risks. These are the ones that matter.

Measurement and integrity. The voluntary carbon market famously sold credits that did not hold up. If “regenerative” becomes a label rather than a measured outcome, this category repeats that collapse. Answer: underwrite on cash flows first and credits last, and make every claim auditable.

The transition valley. Soil, forests and watersheds take years to pay back. Answer: match patient capital to patient assets, structured like construction finance.

Policy whiplash. Incentives come and go, as U.S. clean energy credits did in 2025. Answer: deals that work on contracted revenue alone, with subsidies as upside. For workers, choose skills that serve every system, not one subsidy program.

Robots come for physical work too. Automation will reach farms, factories and job sites. Answer: it moves slowest in messy, outdoor, ever-changing places, which is where regenerative work lives. Learn to run the machines rather than compete with them.

Climate risk inside the asset. A restored forest can burn; a reef can bleach. Answer: diversify across systems and places, insure what can be insured, and price permanence honestly.

Illiquidity and complexity. These assets are hard to build, measure and sell quickly. That risk is also the moat. This is an operator’s game. Returns and careers go to those who can do the work. Complexity is why the discount exists, and why it will not last.

We spent two centuries optimizing for eggs. We got very good at it, and we nearly killed the bird. The next century belongs to those who rebuild productive capacity, not just production.

If you are sixteen, pick up something you can only do in a real place: a trade, a science, time outdoors. The world will always need people who can grow food, move water and keep the lights on.

If you are twenty and choosing a major, pair it with the physical world. Engineering with fieldwork. Biology with restoration. Finance with a farm, a forest or a power project behind it.

If you are early in a career and worried about AI, look to work with it rather than against it. Your analytical skills are worth more when applied to these problems with the helping hand of next-gen tools. Your environmental consciousness and concern will soon sync with economic progress in incontrovertible ways.

If you manage capital, buy what grows back. Pay the forest that fills the reservoir. Finance the farmer through the valley. Let the insurer buy the reef. Put the cheapest electron on the roofs of the families who need it most. Earn a market return every time.

The last century built assets that wore out and careers that software could copy. This one will build assets that grow back. At the forefront of that growth are the careers that will last.

Soon, society will stop eating Stella, and begin to nourish her. The careers and fortunes of the next century belong to those who help drive that transition forward.