The First-Mile Imperative

Executive Summary

Three pieces of work published in the past four months reach the same conclusion from three angles. They are J.P. Morgan's Agriculture, Soil and Trade: The Fates of Food in a Warming World (with data and visualisations contributed by Treefera); Treefera's The Blind Spot in British Food Security; and Treefera's The First Mile Problem: Food Security in a Disrupted Asia-Pacific. The historical patterns that financial institutions, supply chain operators and policymakers have used to price agricultural risk no longer reliably describe what is happening in the field.

Climate volatility, soil degradation and trade fragmentation have moved global agriculture into a regime where pattern-matching fails. The data infrastructure built to track that system was designed for a more stable era. It runs on annual or quarterly cadences, regional aggregates and post-harvest reporting. The decisions that depend on it (by banks stress-testing agricultural credit, by insurers pricing parametric cover, by buyers managing sourcing exposure, by governments managing food reserves) are being made on data that arrives months after the events it describes.

This paper makes three arguments. First, the regime shift documented by J.P. Morgan is now visible in regional production data. UK self-sufficiency in wheat fell from roughly 92% to 68% in a single year; APAC operates with a five-to-seven-month gap between physical ground truth and published crop estimates. Second, the binding constraint on responding is no longer technological. Radar satellite coverage that penetrates cloud cover, AI-driven stress detection calibrated to the biological window and field-level yield reconstruction are operational today. Third, the institutions that close this first-mile data gap first will earn a structural information advantage; those that wait will continue to misprice in a systematic and identifiable direction.

"Relying on historical patterns is no longer sufficient for decision-making." Dr. Sarah Kapnick, J.P. Morgan, Agriculture, Soil and Trade (2026)

1. The pattern that no longer holds

Agricultural markets have always been weather-sensitive. What has changed is the shape of the risk. J.P. Morgan's Climate Intuition series concludes that environmental threats and shifting trade dynamics are fundamentally re-shaping the risk landscape for agricultural commodities, and that the soil, water and climate stresses driving recent commodity price spikes are increasingly attributable to climate change rather than to ordinary cyclical variability.[1] The same paper finds that 88% of corporate agricultural assets measured by MSCI sit in regions with elevated soil condition risk[2], and that European soils, 60-70% of which are now degraded, are costing the EU an estimated EUR 50 billion a year.[3]

The regime shift is not gradual. Climate science has long established that crop yields respond non-linearly to temperature: decline above the optimum is significantly steeper than improvement below it.[4] Recent research finds wheat losses accelerating from 6.1% to 8.2% per degree Celsius once warming thresholds are crossed.[5] In the UK, autumn-sown crop area in the East Midlands fell 70% in 2024 because waterlogged soils prevented drilling entirely.[6] Soil that is 5% too wet still gets planted; soil that is 10% too wet does not. It is a step function, not a slope.

This is the conclusion that matters for risk management: portfolios, hedges and sourcing strategies built on the assumption that next year will resemble an average of recent years are systematically under-pricing the probability of step-change events. That assumption is the implicit prior in most agricultural credit models, in most parametric insurance triggers and in most procurement contracts. It is increasingly wrong.

2. Three regions, one signal

United Kingdom: a structural data gap

The UK typically produces around 92% of its own wheat. In 2024, that figure fell to roughly 68%.[7] Fresh vegetable self-sufficiency dropped to 53%, the lowest since 1988.[8] Wheat production fell to 11.1 million tonnes, the lowest in decades, and imports surged to a record 3.06 million tonnes.[9] Cumulative losses from recent UK harvest failures are estimated at £2.3 billion.[10]

The intelligence gap that compounds these losses is structural. The UK publishes crop production data once a year, with provisional estimates in October and final figures in December. There is no weekly crop progress reporting, no sub-county yield tracking and no quantified pre-harvest production intelligence. By contrast, the USDA publishes weekly crop progress reports from April to November and draws county-level yield data from a federal crop insurance programme covering around 85% of planted acres.[11] The EU's MARS bulletins provide monthly satellite-informed crop assessments. The UK sits behind both peers. Not marginally, structurally.

Asia-Pacific: a five-to-seven-month decision vacuum

Asia-Pacific is home to roughly 60% of the world's population and produces approximately 90% of its rice.[12] Three pressures are converging. Demand growth is narrowing every operating buffer; the region's urban population is projected to reach 70% by 2050, and 1.9 billion people in the region already cannot afford a nutritionally adequate diet.[13] Climate disruption is striking at biological windows. Severe floods reduce global rice yields by approximately 4.3% annually, around 18 million tonnes, with losses concentrated in eastern China, West Bengal, the Philippines and Indonesia.[14]Geopolitical exposure is converting energy events into food security crises within weeks.

In March 2026, disruption to shipping through the Strait of Hormuz sent urea prices up 46% in a single month. Approximately 30% of globally traded fertiliser transits this chokepoint.[15] The constriction is arriving across India's Kharif belt, Bangladesh and the Mekong Delta in the pre-monsoon window: the period when planting decisions substantially determine regional yields six months out. Official production data will not reflect those decisions until the harvest is in. By that point the intervention window has closed.

Global trade: concentration meets fragmentation

J.P. Morgan's analysis quantifies the same dynamic at the global level. Forty-two countries source more than half their total rice imports from a single origin.[16] Cocoa is 42% concentrated in Côte d'Ivoire; coffee is 31% in Brazil and 18% in Vietnam.[17] The microclimates these crops require are themselves shifting upward as temperatures rise; a finite adaptation, given mountains run out. Soybean self-sufficiency in China, the world's largest meat producer, sits at the lowest of any staple. In each case, a concentrated production base meets a fragmenting trade environment, and the result is a system in which a single regional weather event or policy decision is sufficient to move global prices.

Forty-two countries source more than half their total rice imports from a single origin. The biological signal that constrains the next harvest is detectable now. The intervention window closes well before the official data arrives.

3. Why the existing infrastructure fails

The data layer that financial institutions and supply chain operators rely on for agricultural intelligence has three structural weaknesses, each of which has become more consequential as the underlying risk has shifted.

Cadence. UK reporting is annual; FAO's Global Information and Early Warning System publishes quarterly.[18] For a crop under water stress during a six-week critical biological window, a quarterly cadence is retrospective by design.

Resolution. Regional aggregates obscure precisely the field-level variation that determines whether a given farm portfolio, sourcing region or insurance pool is performing better or worse than its peers. Without that granularity, climate-driven yield deviation cannot be separated from management underperformance. That is a structural gap that mis-prices agricultural credit and crop insurance in a systematic direction, over-pricing well-managed farms in difficult conditions and under-pricing poorly managed farms in favourable ones.

Coverage. Optical satellite imagery, the basis of most publicly available crop monitoring, cannot penetrate cloud cover. Key agricultural zones across Southeast Asia experience persistent cloud obscuration for four to six months during the monsoon growing season: the period of maximum agricultural activity and maximum risk.

At a recent conference on alternative data, private firms described generating more accurate real-time signals on economic conditions than central banks, using satellite imagery and AI-driven pipelines. The concept is called nowcasting: knowing what is actually happening before official statistics arrive. That is precisely what agricultural markets lack, at precisely the moment when volatility makes it most consequential.

4. Why this matters now: financial services and corporate supply chains

The regulatory and commercial environment has moved faster than the data layer. The Bank of England's Supervisory Statement SS5/25, published by the PRA in December 2025, requires geographic and sector-specific granularity in physical climate risk assessments, with bank compliance plans due June 2026.[19] The 2021 Climate Biennial Exploratory Scenario already identified agriculture as a sector where physical risk impairment rates would increase materially.[20] The EU Deforestation Regulation, now in force, has moved plot-level supply chain verification from future ambition to current legal requirement for palm oil, soy and coffee entering European markets.[21]

In December 2025, Lloyds Bank launched an Agricultural Transition Finance loan.[22] In the same month, the EU's first soil health law entered into force.[23] The U.S. announced a USD 700 million Regenerative Pilot Program built around public-private match funding.[24] The intent in each case is right. Effectiveness depends on whether the field-level evidence exists to underwrite, verify and measure outcomes.

For financial institutions carrying agricultural credit and crop insurance exposure, the practical implication is that the next regulatory cycle will require a level of granularity that the public data layer does not currently provide. For corporate buyers and processors, EUDR-style plot-level traceability is the minimum baseline for European market access. The same logic is moving through agricultural risk assessment, commodity pricing and capital allocation more broadly. The exposure is identifiable and quantifiable; the data infrastructure to manage it is the constraint.

5. The capability already exists

The tools required to close the first-mile data gap are operational. Synthetic-aperture radar satellites penetrate cloud cover. AI models calibrated to crop development stage can detect water stress, pest pressure and yield divergence weeks before conventional reporting. Field-level yield reconstruction from satellite signals, calibrated against ground truth, is producing pre-harvest estimates with measurable accuracy.

Treefera's analysis of Ghanaian cocoa produced yield estimates with an F1 score of 0.93 against actual outcomes, months before official publication. U.S. corn yield tracked final USDA figures to within one percentage point, four weeks ahead of official release. These are not pilots. The capability is in production, and it is improving on the cadence of the underlying AI stack. Model capability is currently doubling roughly every 200 days. Satellite constellations that once served defence and intelligence agencies are commercially available. The same tools that allow a technology company to nowcast retail footfall or logistics throughput can be pointed at a wheat field in Lincolnshire or a rice paddy in West Bengal.

The gap, in other words, is not technological. It is deployment at the right resolution, the right timing and the right analytical specificity for the decisions that depend on it.

6. The credit-bureau parallel

There is a precedent for the infrastructure this paper describes. Credit bureaus were not created by government mandate. They were built by private actors who identified a market need: the aggregation of dispersed credit data into a standardised infrastructure that lenders could use to make better decisions. The market used what they built. Regulation followed.

Agricultural risk intelligence is at an analogous point. The data to assess farm-level climate and production risk exists, in satellite imagery, in insurance records, in the yield histories sitting on farm management software across the country. What has not yet been built, at scale, is the aggregated, standardised, audit-grade infrastructure that makes that data usable by banks, insurers, regulators and corporate buyers at portfolio scale. Private actors have both the capability and the incentive to build it. Government's role is to create the conditions that make that investment viable and to be a credible consumer of the results.

7. What to do now

For financial institutions. Treat first-mile data as a credit and underwriting input, not a sustainability disclosure input. Field-level yield and stress data, refreshed in-season, materially improves the separation between climate-driven risk and management performance: the variable that determines whether a loan or policy is correctly priced. Use it to meet the granularity requirement implicit in PRA SS5/25 and to underwrite transition finance products on observed performance rather than assertion.

For corporate buyers and supply chain operators. Treat plot-level traceability as the new floor, not the ceiling. EUDR is the leading edge of a wider shift toward verifiable origin and condition data. The same infrastructure that enables compliance enables forward visibility into sourcing risk, weeks earlier than commodity desks or quarterly reports.

For policymakers. The Dimbleby National Food Strategy recommended a National Food System Data Programme in 2021.[25] Five years on, that programme does not exist in the UK. The infrastructure can be built privately and consumed publicly. Government's role is to set standards, fund the consumption layer and stop treating field-level intelligence as an aspiration.

The next difficult season is not a question of whether, but when. When it arrives, the decisions made in response (by farmers negotiating grain prices, by banks stress-testing agricultural portfolios, by buyers managing sourcing risk, by governments assessing food supply resilience) should be made on data that is current, granular and predictive. That capability exists today. The only question is how quickly we make it the standard.

References

[1]: Kapnick, S. (2026). Agriculture, Soil and Trade: The Fates of Food in a Warming World. J.P. Morgan Climate Intuition series, 16 January 2026. Data and visualisations: Treefera.

[2]: MSCI corporate agricultural assets overlay, cited in Kapnick (2026).

[3]: European Commission (2025). First EU Law on Soil Health, in force 16 December 2025.

[4]: Schlenker, W. and Roberts, M.J. (2009). Nonlinear temperature effects indicate severe damages to U.S. crop yields under climate change. PNAS 106(37): 15594–15598.

[5]: Tran, B.-L., Tseng, W.-C. and Chen, C.-C. (2025). Climate change impacts on crop yields across temperature rise thresholds and climate zones. Scientific Reports 15:23424.

[6]: AHDB (2024). Autumn drilling progress and crop establishment reporting.

[7]: DEFRA / AHDB (2025). UK wheat balance sheet data 2024/25.

[8]: DEFRA (2025). Agriculture in the United Kingdom 2024.

[9]: AHDB (2026). Early Balance Sheets 2025/26.

[10]: DEFRA / industry estimates (2020–2025). Estimated cumulative losses from UK harvest disruptions.

[11]: USDA NASS Crop Progress; USDA RMA Summary of Business 2025.

[12]: United Nations DESA (2024). World Population Prospects 2024; FAO (2026). Food Outlook.

[13]: UN DESA (2024). World Urbanization Prospects 2024; FAO (2021). Regional Overview of Food Security and Nutrition in Asia and the Pacific.

[14]: Flood damage to global rice production. Science Advances, 14 November 2025. DOI: 10.1126/sciadv.adx7799.

[15]: CSIS (2026). Chokepoint: How War with Iran Threatens Global Food Security; International Fertilizer Association (2026).

[16]: USDA Foreign Agricultural Service (2023). Rice: World Markets and Trade.

[17]: Kapnick (2026), § The dangers of geographically concentrated and extremely climate-sensitive crops.

[18]: FAO (n.d.). Global Information and Early Warning System (GIEWS). fao.org/giews.

[19]: Bank of England Prudential Regulation Authority (2025). Supervisory Statement SS5/25: Managing the financial risks from physical climate change.

[20]: Bank of England (2022). Climate Biennial Exploratory Scenario: Key elements of the 2021 CBES.

[21]: European Commission (2023). Regulation (EU) 2023/1115 (EU Deforestation Regulation).

[22]: Lloyds Bank (2025). Agricultural Transition Finance product launch.

[23]: European Commission (2025). First EU Law on Soil Health enters into force.

[24]: USDA (2025). Regenerative Pilot Program announcement, December 2025.

[25]: Dimbleby, H. (2021). National Food Strategy: The Plan.