The vertical-farm bubble — why a multi-billion-dollar dream collapsed one by one

In 2021, vertical farming was at its peak. Stack LEDs and hydroponic racks in urban warehouses and buildings, save 95% of water, use zero pesticides, be unaffected by weather, and produce leafy greens year-round right next to where they are eaten — investors poured billions into this story, and several companies reached unicorn valuations above a billion dollars. Yet within just two or three years, many of them were driven into bankruptcy, liquidation or drastic downsizing.

This was not a run of bad luck but a structural failure faced by an entire industry category. And crucially, the collapse is not a repudiation of 'growing vegetables in cities' in general. What failed was a specific capital-intensive, energy-intensive business model — not the soil-and-people community garden. Not confusing the two is the first step in learning from this failure.

Discussions of urban farming often foreground a story of 'solving the food problem with cutting-edge technology.' The vertical-farm bubble taught, painfully, exactly where that story collides with reality. Three walls — energy, capital and crop constraints — could not be scaled by enthusiasm or innovation. Knowing this case is a practical vaccine for governments, investors and operators before they leap at 'the next vertical farm.'

The expectations vertical farming attracted had real grounding. Yields per unit area were dramatically higher than open-field farming, and water use far lower. Almost no pesticides, and short food miles because production sits near consumption. Independent of weather, season and climate change, it could supply uniform-quality vegetables year-round. It looked like a silver bullet for urban food self-sufficiency and for the food security of import-dependent nations.

This story married the abundant venture capital of the ultra-low-interest era. SoftBank, Google-affiliated funds and celebrity investors piled in, and companies like AeroFarms, Bowery, Plenty, Infarm and AppHarvest each raised hundreds of millions to over a billion dollars. Precise like a factory, scalable like a tech company — the market was blanketed by the expectation that agriculture was finally becoming an 'industry.'

The collapse was a chain reaction. AppHarvest, which even went public on its giant Kentucky greenhouses, went bankrupt in 2023. AeroFarms, a symbol of vertical farming, filed for Chapter 11 in 2023 (later rebuilding at reduced scale). Publicly listed Kalera, the 'automation standard-bearer' Fifth Season, indoor-hydroponics Iron Ox, the Chicago pioneer FarmedHere, and France's Agricool were all driven into liquidation or shutdown.

Germany's Infarm, among Europe's largest, announced a major retreat from its European operations in 2023, cutting sites and staff heavily. New York–based Bowery Farming, which drew celebrity investors, shut down in 2024. And even Plenty — called the next unicorn, having raised over a billion dollars — filed for Chapter 11 in 2025. A cohort of companies described only a few years earlier as 'the future of farming' was all but swept away.

Research points clearly to the root cause. As long as sunlight is replaced by artificial lighting (LEDs), vertical farming is inherently energy-intensive. Studies of indoor agriculture warn it could become a new driver of global energy demand. The electricity cost for LEDs, air conditioning, dehumidification and pumps kept squeezing revenues. From the moment the 'free inputs' of soil and sun were replaced by the 'paid inputs' of electricity and equipment, the outcome was, in a sense, foreordained.

And even the environmental edge is doubtful. Studies comparing the carbon footprints of open-field and urban agriculture show that the life-cycle emissions of controlled-environment agriculture (CEA) and vertical farms can, under some conditions, exceed those of open-field farming. If the power comes from fossil fuels, the very premise of being 'eco-friendly' collapses. A body of work on vertical-farm economics has likewise repeatedly noted weak viability in competitive markets and the indispensability of high-value crops.

Decisive is the crop constraint. Profitability skews to a narrow set of items — lettuce, herbs and leafy greens that are light, fast-growing and high-priced. Staple grains, tubers and heavy vegetables do not command prices that match the cost of the light and space they need. The story that 'vertical farms will feed the world' was in reality confined to the narrow niche of salad greens. A study calculating benefit-cost ratios in Australia and Nepal likewise concluded that the model is only marginally viable with high-value crops, with land cost a key barrier.

Not all of vertical farming was pointless. For high-value herbs and greens, seedling production, pharmaceutical and research uses, and regions where extreme climates or import dependence make open-field farming difficult, controlled-environment growing still works reasonably under limited conditions. What failed was less the technology than the overblown story that 'this replaces all of agriculture and scales explosively,' and the capital expectations built on it. Selling a niche tool as a universal solution is the essence of the bubble.

This distinction matters. A technology simply has situations where it works and situations where it does not; it is neither good nor evil. The question to ask is always concrete: which crop, on which energy source, for whom, at what cost. Skip that question and move investment or policy 'because it's futuristic,' and you will repeat the vertical-farm bubble's mistake.

Ironically, the vertical-farm bubble's collapse threw the value of the ordinary, soil-and-sun community garden into paradoxical relief. A community garden's chief value is not, in fact, its vegetable yield. Human connection, food education, welfare, biodiversity, disaster resilience, green-space function — the 'relationships' and multifaceted public goods that billions of dollars of LEDs never managed to produce, the humble garden produces from the start. Lumping the two under the single phrase 'urban agriculture' badly distorts how we evaluate them.

Three implications for practice. First, when evaluating an urban-farm project, think separately about capital- and technology-intensive models versus community- and soil-based ones: depending on the desired outcome (food production or connection), the right answer differs entirely. Second, in the budget, look at how well the project harnesses 'free inputs' — soil, sun, rain, volunteers. Third, do not let 'newness' sway public funding or grant decisions. Had even a fraction of the capital poured into vertical farms flowed instead to countless soil-based gardens, far more public good would have been created — the heaviest question this bubble leaves behind.