Autopsy of the Agrifood–Climate Tech Collapse

René Magritte, La Clairvoyance (1936)René Magritte’s La Clairvoyance shows a painter looking at an egg while painting a bird. Not as it is, but as it is expected to become. The gesture is calm, assured, almost inevitable. Nothing in the scene suggests doubt.That confidence, projecting a finished future onto an unfinished present, feels familiar. Over the last decade, much of agrifood and bio-industrial innovation was financed the same way: not for what it was, but for what it was presumed to become once scale, time, and capital did their work. The gap between those assumptions and physical reality is where this story lives. What follows is the consequence for that confidence: 113 companies. $18.5 billion in venture funding raised. Another $31 billion in public market cap destroyed. All within three years.I built the most comprehensive database of every agrifood-climate tech company that failed, collapsed, or entered severe distress between 2023 and early 2026. Vertical farms. Plant-based meat. Cultivated protein. Insect facilities. Mycelium leather. Ag robotics. Carbon platforms. By 2026, most of it had been written off.The $18.5 billion figure is funding raised by private companies that subsequently failed or entered distress. The $31 billion is public market capitalization destroyed: Beyond Meat down 97% from peak ($13.6 billion gone), Oatly down 98% ($12.7 billion gone), AppHarvest to zero, Tattooed Chef to zero, Danimer Scientific sold for pennies. If you add them up: roughly $50 billion in capital destruction across the agrifood-climate tech sector.This is the most detailed autopsy of the 2021–2025 bust available. Primary research: bankruptcy filings, administration notices, WARN Act layoffs, liquidation announcements, distress acquisitions.The failures were not isolated. They followed clear patterns. This post is about that.The failures started as a trickle. Two companies in Q1 2023. Five in Q2. Then the curve steepened: 7, 11, 17. Q3 2024 was the peak, seventeen failures in a single quarter as the 2021 vintage hit its wall. Companies that raised at peak valuations in late 2020 and 2021 had burned through their capital by mid-2024 with nowhere left to turn. The wave didn’t end there. It plateaued through 2025 at 7–11 failures per quarter, shifting from the spectacular collapses (Bowery, Plenty) to quieter shutdowns of $5–30 million companies that barely made the press.Where it hit hardest: alternative proteins account for 43 of the 113 failures and $6 billion in destroyed capital. Vertical farming: 18 companies, $4 billion. But the concentration is deceptive. Carbon tech lost $2.1 billion across just 4 companies. Precision ag lost $1.1 billion across 4. The damage per company was highest in the sectors that attracted the most generalist capital and the biggest cheques.Drill one level deeper—from sectors to subsectors—and the failure distribution sharpens. The 113 failed companies are spread across 70 distinct subsectors. But the casualties are radically concentrated. Plant-based meat alone accounts for 15 failures, vertical farming for 9, plant-based seafood for 6. Those three subsectors together represent more than a quarter of all failures in the database. Below them, 13 subsectors lost exactly 2 companies each. And the remaining 52 subsectors? One failure apiece. The pattern is clear: wherever the 2019–2021 capital wave created enough companies to constitute an actual competitive field, that field collapsed almost entirely. Plant-based meat didn’t lose a few players—it lost fifteen. Vertical Farming didn’t consolidate—nine companies went under. The subsectors that looked most like real markets produced the highest body counts. The subsectors with just one failure look better by comparison, but that may simply mean there was never more than one company to lose.Strip away the company names and something structural emerges. Remarkably few of these companies failed because the science didn’t work or the product was too bad. The failures sort into financing mismatches, broken unit economics, infrastructure shocks, and markets that evaporated. These are financing failures dressed up as technology failures. The distinction matters, because it changes what the next wave needs to get right.The failures sort into four patterns. They’re not independent. A VC timeline mismatch becomes fatal when unit economics aren’t proven, which becomes unfixable when demand shifts. Most companies in the database map to two or three. But separating them is useful because each one points to a different lesson for what comes next.“We just simply ran out of time.” — Arkeon’s CEO, precision fermentation, $13 million raised, filing for insolvency.Standard VC funds have 10-year lifecycles with pressure to show returns by year 7 or 8. The technologies being funded require 12–15 years to reach commercial viability. Everyone knew this. Everyone funded anyway. The implicit bet: someone else will buy the company before the math catches up. When rates rose in 2022, there was no one left to buy.Meatable had 86 scientists running regulatory processes in five countries at approximately $2 million each. That’s not a timeline compatible with venture returns. AgFunder’s data shows agrifood represented just 5.5% of all venture capital dollars in 2023, down from 6.7% in 2022 and 7.6% in 2021. The generalists fled. Some specialist agrifood VC firms failed to raise follow-on funds entirely. One investor surveyed by AgFunder: “Surprised me how quickly so many of the specialized agrifoodtech funds repositioned to become ‘climate funds.’”The most tragic failures in this database weren’t the overhyped ones. They were companies with viable technology that needed 18 more months. Nearly identical words from Inseco in South Africa ($15 million) and Bolt Threads ($300 million, biomaterials): “devastatingly close” to scale. The VC model’s clock ran out before the science did.When the clock runs out, companies pivot. In software, pivoting is celebrated. In hardware and industrial biotech, it’s usually a death spiral. Each pivot burns 12–18 months and resets regulatory clocks. MycoWorks ($185M) built the world’s first commercial-scale mycelium factory in South Carolina, then closed it barely a year later and pivoted to licensing its tanning technology. Iron Ox ($103M) went through multiple pivots, cut half its staff, and eventually ceased operations. Indigo Ag (~$2B raised) pivoted from microbial seed treatments to carbon credits, saw its valuation collapse 94% from $3.5 billion to $200 million, and restructured into separate business units.The 2021 vintage made all of this worse. In 2021, investors poured $51 billion into agrifood globally. Cultivated meat alone got $989 million. By 2025, cultivated meat funding had collapsed 93% to $65 million. Companies that raised at peak 2021 valuations got trapped: too expensive to acquire, too cash-hungry for bridge rounds, too diluted to raise again. Every company in this database that took a significant down round in 2023–2024 either shut down within 18 months or entered restructuring. The 2021 vintage of agrifood-materials deals may be the worst performing cohort in venture history for these sectors.Two companies in the database died of a failure mode nobody talks about: IP disputes. Motif FoodWorks ($345M raised) shut down because an IP lawsuit with Impossible Foods “drained resources” and made the company “uninvestable.” BIOMILQ ($24M, Chapter 7) had the same language: an IP dispute made it “uninvestable and unacquirable.” In both cases, the technology worked. The legal exposure scared every potential acquirer and investor away. IP disputes in deep tech don’t just cost legal fees. They make companies radioactive.The companies surviving right now have strategic co-investors (Cargill, ADM, DSM) who don’t need a 7-year exit. If your cap table is pure financial VCs, you have a structural problem regardless of your technology. The capital structure has to match the technology timeline. Patient capital, milestone-based funding, strategic investors with longer horizons. This isn’t a nice-to-have. It’s existential. And the bar for IP due diligence is permanently higher. Settle early or don’t raise on IP you can’t defend.“No more investors on the market. The more you run, the more the doors close.” — Antoine Hubert, co-founder of Ÿnsect, Europe’s most-funded agtech startup. $625 million raised. €656,000 in revenue against €80 million in losses before entering judicial liquidation.Too many companies raised on “we’ll figure it out at scale.” Scale didn’t fix the math. It magnified the losses.Vertical farms: $3.07/lb production cost versus $0.65/lb field lettuce. Energy represents 40–70% of operating costs. When European electricity prices spiked in 2022–2023, Infarm saw costs run 6x over budget in some facilities. The Oxford study found that even theoretical minimum energy costs for vertical farming ($0.50/kg) only approach field agriculture prices for high-water-content vegetables, and that’s before infrastructure, labor, or any other costs.Insect protein: 2–10x the cost of soy or fishmeal. The sustainability premium never materialized. Commodity buyers want cheap protein.Plant-based meat: 20–30% premium that consumers wouldn’t pay during inflation. Circana research shows 40% of European consumers opted for healthier traditional meat compared to just 9% choosing plant-based alternatives. And it wasn’t just the big names. Hooked Foods in Sweden: SEK 1.4M revenue versus SEK 6.2M loss. That’s cleaner evidence than Beyond Meat’s story because there’s no celebrity CEO or stock market noise obscuring it. Just: the math didn’t work.Ag robotics: FarmWise built $645K robots. Guardian Agriculture’s CEO said “$50M was too little” to compete against Chinese manufacturing. Small Robot Company ($12M): “Victims of the valley of death.” Monarch Tractor ($220M) lost its Foxconn manufacturing partnership and issued a WARN notice. The hardware cost curve couldn’t come down fast enough with venture-scale funding.An strange pattern from the database. Companies that trumpeted facility-size superlatives had a near-perfect failure rate. Ÿnsect (”Europe’s most-funded agtech”) to liquidation. Aspire Food Group (”World’s largest cricket facility”) to receivership, CAD$44M debt. Believer Meats (”World’s largest cultivated meat facility”) to shutdown. Jones Food Company (”Europe’s largest vertical farm”) to administration. AppHarvest (4 mega-greenhouses built before revenue) to full liquidation. “World’s largest” in a pre-commercial sector means you scaled capex before proving unit economics. Every single one failed.“Energy prices have escalated, doubled across Europe, which puts a lot of additional pressure on our business and seriously impacts our cost of production in affected markets.” — Erez Galonska, co-founder and CEO of Infarm, in an internal email to staff, November 2022. $604 million raised. Within a year, Infarm entered administration.Climate tech got undone by “close to climate” infrastructure. Vertical farms needed cheap electricity during an energy crisis. European companies faced energy costs nobody modeled.The smaller companies reveal something the big failures obscure: location-specific risks killed companies that might have survived elsewhere. Inseco in South Africa collapsed after 4-hour rolling blackouts. You can’t run a BSF facility without continuous climate control. Sustenir Agriculture in Singapore ($22M) couldn’t compete with field-grown produce prices in tropical Southeast Asia. Konscious Foods in Canada ($32M, founded by Gardein’s creator) cited US-Canada trade barriers as a factor.Technologies designed to address environmental problems were killed by environmental and infrastructure volatility. Match your technology to your geography. Don’t build energy-intensive operations in energy-unstable regions. Don’t build premium indoor farming where tropical field agriculture undercuts you 5:1. This sounds obvious. 18 companies and $4 billion in vertical farming losses say it wasn’t.“We built incredible technologies, but the voluntary carbon market just got a lot smaller in the last nine months. We were building this for a growing market and all of a sudden it was shrinking.” — Marty Odlin, CEO of Running Tide, $54 million raised. Shut down June 2024.Plant-based meat sales peaked in 2021 and declined every year since. The “ultra-processed” backlash hit at the same time as inflation. Consumer willingness to pay sustainability premiums evaporated.“Cultivated meat is broken.” That’s not an activist. That’s Jeff Tripician, CEO of Meatable, a cultivated meat company, explaining why he was winding down operations. The carbon market told the same story at industrial scale: the voluntary market collapsed 61% from $1.9B (2022) to $723M (2023). This was the fastest sector collapse in the database. Five companies, $1.28B raised. Running Tide ($54M): demand simply wasn’t there. Nori ($17M): “challenges of stagnant Voluntary Carbon Market.” Indigo Ag ($1.1B) pivoted away from carbon entirely. These companies died of market risk, not technology risk. The science often worked. The buyers disappeared.Beyond Meat trades at 1x revenues, far below the 3–5x multiples typical for CPG acquisitions or 8–10x for tech. Analysts describe Chapter 11 by 2027 as the “base case.”The plant-based bust wasn’t just a Beyond Meat story. At least 20 smaller brands died with barely a headline. Bolder Foods ($3M), Choppy! ($3M), Beastly Brews ($2M), Olala! ($4M), Willicroft ($6M), Neat/Plantcraft ($5M), PlantMade ($8M), Planetarians ($5M). These were consumer brands that couldn’t get shelf velocity. The sector’s problem isn’t “one or two companies overpromised.” The entire consumer-facing plant-based category collapsed.Companies built supply for demand that existed in pitch decks. Any company building a business model that depends on a policy-driven or voluntary market (carbon credits, sustainability premiums) rather than direct economic value should study this period carefully. Those markets can vanish in 18 months. And if you’re pitching a consumer brand in this space in 2026, you need to explain why you’re different from the 20 that just died.The problem climate tech was built to solve hasn't gone anywhere. The crops still need to be grown, the proteins still need to be made, the materials still need to change. What died was a specific version of how to get there fast, venture-funded, and consumer-facing. That version is over. For five years, climate tech ran on conviction. The conviction that consumers would pay more, that scale would fix margins, that the market would arrive on schedule. The market and the consumer didn’t agree. That version is also over.What does remain? The survivors share a profile: asset-light, B2B-first, backed by patient capital. Precision fermentation ingredients selling to food manufacturers rather than consumers. Plant-based protein repositioned as a health ingredient rather than a meat replacement. Whether that profile holds or becomes the next set of assumptions this database will eventually record is an open question. B2B doesn't automatically fix unit economics. But the margin for error is narrower now, and the capital is less forgiving. That might be enough.The database methodology: private company failures and severe distress events between January 2023 and February 2026. “Failure” includes bankruptcy, liquidation, administration, distress acquisition at >75% discount, shutdown, or receivership. “Severe distress” includes major restructuring (>50% headcount reduction), production halts, or public warnings of insolvency risk.Private companies: 113 companies, $18.46 billion in funding raised, 10 sectors.Public companies: 8 companies, $31.4 billion in market cap destruction. Beyond Meat ($13.6B lost), Oatly ($12.7B lost), AppHarvest ($1.2B lost), Atlantic Sapphire ($1.4B lost), Tattooed Chef ($1.0B lost), Danimer Scientific ($871M lost), Kalera ($375M lost), Local Bounti ($250M lost).Total: ~$50 billion in capital destroyed.These numbers are conservative.Share The New BioeconomyA note to founders: this is not an indictment. Building in agrifood and bio-industrial company means living with uncertainty: technical, regulatory, and commercial. Failure is part of the work. What’s described here is not primarily founder error, but the collision of structural constraints, external shocks, and misaligned incentives. And despite all of it, we need people willing to build, because without builders, there is no progress, and no impact worth investing in.