What Compounds in Climate

The Inflation Reduction Act, signed earlier this month, commits $369B to clean energy and climate infrastructure over the next decade. Coverage of the bill has focused on its short-term political implications. The more important question for capital allocation is structural: what does a decade of policy-stabilised demand do to the investment case for climate technology, particularly as the broader macroeconomic environment turns?

The short answer is that the case strengthens.

Two regimes converge

Climate technology has been operating across two largely separate regimes. The first is environmental — climate as moral imperative, sustainability as ESG mandate, decarbonisation as long-horizon civilisational project. The second is industrial — climate technology as productivity infrastructure, where battery storage, alternative energy, and resource efficiency are bought because they produce better unit economics than the alternatives they replace.

The convergence of those two regimes is what changed this year. The IRA price-stabilises demand for clean infrastructure at scale. Russia’s invasion of Ukraine recasts energy independence as a security argument rather than a moral one. Drought conditions across India, the American Southwest, and southern Europe make food and water resilience operational concerns rather than abstract risks. Climate technology stops needing the environmental argument to clear capital diligence; it now clears on grounds that hold even if the environmental argument were ignored entirely.

Food systems, looking out

The longer-horizon thesis is food. Roughly 80% of global agricultural land currently supports animal husbandry. In some markets, 80% of cultivated grain feeds livestock. Globally, livestock contributes around 18% of caloric intake while consuming the majority of agricultural land, water, and feed. The structural inefficiency is well documented; the question is whether any of it changes at scale, and on what timescale.

Two things are changing simultaneously. Alternative-protein cost curves — fermentation-derived, cell-cultured, plant-protein — are following classical learning-curve dynamics. Performance is improving and costs are dropping by predictable annual percentages. Demand patterns are shifting independently of the cost curves, particularly among younger cohorts whose purchasing decisions are already differentiated from prior generations.

The combination is what matters. Cost curves alone produce nothing if demand resists; demand alone produces nothing if cost remains prohibitive. Where they meet, categories form. Several of those categories are forming now.

What survives a downturn

The case for investing through a recession is not that the recession will not happen — by most signals, it likely will — but that the categories with structural tailwinds compound through it. ClimateTech now has three: policy stabilisation (IRA and equivalents in Europe), security relevance (energy independence as strategic priority), and learning-curve economics (technologies that get measurably cheaper with deployment volume). All three are independent of capital-market sentiment.

The companies that get built during downturns in capital-intensive categories tend to be the ones that look obvious in retrospect. They survive the period when capital is most disciplined; they emerge into a more abundant funding environment with operational track record and customer traction that better-funded competitors from the prior cycle never had to develop. Battery cost compression, grid-scale storage, alternative proteins, and water management all sit in this pattern now.

Reading

The next decade of climate investment will not look like the last. The last was largely funded on environmental conviction; the next will be funded on industrial economics, with environmental impact as alignment rather than primary motivation. That shift broadens the addressable capital pool and tightens the operational discipline applied to portfolio companies. Both effects strengthen the category. The downturn is the entry point, not the exit signal.