Sustainable agriculture: Assessing nature loss and transition costs in emerging markets

In this article, we explore the evolving topic of sustainable agriculture and related developments. This article does not intend to comment on current credit ratings or the application of credit rating methodologies..

Maintaining agricultural resilience in emerging markets (EMs) is critical to global food systems. Brazil, for example, is the world's largest agricultural commodity exporter, and other EM economies, such as China, rely heavily on Brazilian agriculture exports to maintain their own food resilience (i.e. 65% to 75% of China's soybean imports, destined to its world-leading hog industry, came from Brazil over the past five years). Yet, according to the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services, global soil carbon, the organic matter that sustains agriculture, has declined by 8% since the 1970s.  

In the long-term, sustainable practices could strengthen the resilience of agriculture by enhancing soil quality and improving the ability of soils to withstand extreme weather. However, these transitions require significant additional investment beyond traditional methods. For instance, Brazilian soybean farmers who choose to transition face recurring cover-cropping costs of approximately $80 to $196 per hectare annually, while measurable agronomic benefits may take 3–5 seasons to emerge, with full value (carbon sequestration, soil health, and climate resilience) accruing over a 10–20-year horizon. In the current context of limited farmers’ affordability due to high fertilizer costs, transitioning to sustainable agriculture presents significant operational and financial challenges.

This exploratory research focuses on two key emerging markets, Brazil (soybeans) and India (rice), to estimate, under specific scenarios, what  the potential cumulative losses from continued reliance on conventional agriculture would be compared to the costs of transitioning to more sustainable, resilient production practices. We find that there are potential economic and resilience gains, but that important structural and other barriers limit their widespread adoption.

We use S&P Global Energy CERA Agriculture Sustainability data to estimate the costs of implementing sustainable agriculture practices. We apply Shared Socioeconomic Pathways (SSP) 2-4.5, a moderate climate change scenario. Costs are modeled across specific time horizons, with “short term” referring to 2021-2030 and “medium term” referring to 2031-2035.

To estimate the total economic output at risk to nature loss, we use the same model published in previous S&P Global Ratings research, Sustainability Insights: Nature Loss: A Potential Constraint To Economic Growth. The model simulates a stress scenario in which the functional capacity of five ecosystem services (biomass provisioning, water supply, water flow regulation, soil quality regulation and pollination) is fully compromised, resulting in the loss of the output directly tied to them. The same steps are applied: First, we calculate the dependency of a sector on the five ecosystem services. Then, we multiply the mean dependency by each sector’s value add, using 2020 as the base year (the sectors are “cultivation of oil seeds” in Brazil and “cultivation of paddy rice” in India). Exploring Natural Capital Opportunities, Risks and Exposure (ENCORE) dependency scores determine the extent that a sector is dependent on nature. Its use for this research is consistent with the S&P Global Sustainable1 Nature and Biodiversity Risk Dataset, S&P Global's published nature-related research, and other modelling approaches. Exiobase is used for the economic activity interdependencies for the same reason: consistency.

The model does not factor in the real dependency of each commodity and country-wide state of nature metrics; it is based on a sector-level proxy and is not location-specific. Furthermore, the nature loss estimates do not account for adaptation or potential price volatility resulting from supply shocks. They constitute a nominal cumulative exposure and are not estimates of expected losses. See the "Data and Approach" section for a detailed explanation.

The transition toward sustainable agriculture can be illustrated by the structural shifts occurring in production of rice in India and soybeans in Brazil. As the two largest single-country crops by harvested area, they represent the most significant land-use footprints in the global food supply chain (Figure 2). These agricultural commodities currently face two key pressures: tension between expansion and regulation, as Brazil seeks to meet export demand while complying with new deforestation standards such as the European Union Deforestation Regulation (EUDR); and tension between production volume and climate resilience. In India, a monsoon season characterized by weak or uneven rainfall could disrupt rice production (See "A Weak Monsoon Will Strain India's Rural Sector,” published July 2026). In Brazil, soybean planting, which normally occurs during the fourth quarter of each year, is expected to be delayed by heavy rains that will be brought by the 2026 El Niño.

Brazil accounts for around 40% of global soybean production as of July 2026, and the crop represents as much as 20% of the country’s total export value in nominal terms. Soybean meal is a fundamental feed in animal protein farming and therefore consumption. Given the long-term trend of rising animal protein consumption globally, any significant disruption to soybean yields could pose a systemic risk to global supply chains.

Rice cultivation in India is a matter of critical domestic food security and social stability. As the primary staple for the world’s largest and growing population, rice is a fundamental component of India's internal economy and its social safety nets. Globally, India is a relevant price setter given its leading rice production position, and any significant disruption impacts international rice markets.

Considering how critical these crops are to their respective countries and global food systems, they offer an important case study for evaluating the implications of scaling sustainable agricultural practices. The analysis that follows compares the estimated costs of transition with potential cumulative economic loss associated with certain nature-related risks, while also examining factors that may affect broader adoption.

We estimate that under a worst-case nature loss scenario, in which the functional capacity of key ecosystem services is fully compromised, the resulting cumulative loss in India’s rice economic output could total $54 billion by 2034 (using 2020 prices, see Figure 4). This estimate is based on a cumulative exposure model rather than a single-year expected loss. It reflects the aggregate impact of the gradual degradation of key ecosystem services over time. Our estimate is a static assessment based on 2020 gross output, representing a scenario in which key ecosystem services are fully depleted. For context, rice cultivation in India has generated an average annual gross revenue of $100 billion over the past six years (2020 base year), representing 1%-2% of the country’s gross output. Consequently, our analysis identifies a potential cumulative loss equivalent to roughly half of the sector's current annual output.

To quantify these potential production losses, using ENCORE data, our model assumes the degree to which the economic output of rice cultivation in India in 2020 was dependent on five key ecosystem services and these natural inputs: freshwater (approximately 52% of rice planting areas are rainfed (Mohapatra et al., 2024)); soil formation and nutrient cycling and erosion control (preservation of topsoil). Because long-term rice yields are intrinsically linked to the integrity of these biological processes, the long-term degradation of ecosystem health can translate into output loss.

Methodological limitations apply. Though the model may not factor in the local state of nature across India’s diverse rice-growing regions, we nevertheless are able to identify escalating risks across the ecosystem services evaluated in this research (see Table 1 and Figure 4).

India's transition to sustainable agricultural practices could save on average an estimated ~$65/hectare (ha) in the short term¹ and ~$87/ha in the medium term² across the two sustainable practices (DSR and AWD; see Figure 3). The estimates consider the impact of global temperature rise scenario SSP2-4.5 on yields and are nominal values.

AWD alone and the combination of DSR and AWD reduce total production costs from the first year of adoption while also increasing long-term resilience. In the short term, conventional rice production costs are estimated at $940/ha. DSR combined with AWD reduces costs by $70/ha (a 7% savings over conventional rice production costs), while AWD alone reduces costs by $60/ha (6%). This means a transition to sustainable agriculture results in cost savings. Table 2 compares the price of conventional and sustainable practices. Negative values represent cost savings.

The short-term economic case is strengthened by the impact on agricultural productivity during the initial adoption period. Adoption of both DSR and AWD and AWD alone result in a positive short-term yield impact of 1.2%³, translating to savings of $9/ha. This suggests that producers can experience productivity gains alongside their cost savings during the initial adoption phase. 

And the potential savings could increase in the medium term. According to the S&P Global Energy CERA Costs & Margin Analysis, conventional rice production becomes 15% more expensive by the medium term, with costs rising to $1,075/ha due to higher fuel and fertilizer prices.

¹ Short-term refers to the average for the years 2021-2030

² Medium term refers to the average for the years 2031-2035

³ Based on S&P Global Energy Emissions and Transition Cost meta-study

A key driver of savings across both time scenarios is the reduction in irrigation-related energy use. Fuel alone accounts for 11% of total conventional production expenses in Indian rice cultivation. Because DSR and AWD reduce water demand, they lower the energy required for irrigation, thereby reducing the energy share of total costs from 11% to 7%. Reducing this dependency is not only a cost-saving measure but also a gain in structural resilience for smallholder farmers, who have limited capacity to absorb fuel-price shocks.

Finally, carbon markets play a key role in improving the potential economics of a transition. Sustainable rice practices reduce methane emissions by changing water-management regimes, especially through intermittent flooding under AWD. These avoided emissions can be monetized through local carbon markets, creating an additional revenue stream beyond direct production-cost savings. As a result, the farmer's benefit is no longer limited to lower input and energy costs; it also includes potential revenue from the sale of carbon credits generated by emissions reductions.

Despite compelling short- and medium-term economic benefits, adoption of DSR and AWD remains limited. We observe key four obstacles: structural, agronomic, policy-driven and market-related (Figure 5).

The dominance of soybean monoculture in Brazil (see Figure 6) introduces stranded asset risks for the industry. The practice of leaving soil bare between cycles can accelerate soil erosion and increase fertilizer runoff, potentially leading to the systemic depletion of soil nutrients and the long-term deterioration of the land's productive capacity.

We estimate that approximately $14.5 billion of Brazil’s soybean economic output is at risk due to nature loss. This is a static, cumulative estimate based on the sector’s gross output and prices for 2020. The estimate considers a worst-case scenario in which the functional capacity of key ecosystem services is fully compromised, resulting in the loss of the output directly tied to it (see Figure 7). At around 3% of Brazil’s gross output over the past decade, soybeans are the country’s second-largest agricultural activity, behind animal protein farming.

Using ENCORE and Exiobase data, our model assumes the degree to which the economic output of soybean cultivation in Brazil in 2020 was dependent on five key ecosystem services. Under this worst-case scenario, producers would likely compensate for declining soil organic matter by intensifying fertilizer application, and disease suppression by applying more pesticides,  creating a feedback loop that further degrades soil biodiversity — ultimately undermining long-term stability and leading to economic output loss.

Methodological limitations apply. The Brazilian soybean analysis also has limited use of state-of-nature metrics. Table 3 reflects the ecosystem services data analyzed, while the map (Figure 7) shows production and nature risk data on a spatial scale.

Brazil’s soybean transition presents a more complex profile than India’s rice transition. Sustainable soybean practices do not generate uniform cost savings. Instead, the economics vary significantly by practice and by time horizon. In the short term, several practices reduce production costs and increase yields, making them attractive from both a cash flow and productivity perspective. By the medium term, however, most sustainable practices become more expensive than conventional production.

For Brazilian soybeans, sustainable agriculture amounts to an investment in soil capital. The main resilience benefits include reduced erosion, improved soil structure, better water retention, lower vulnerability to drought stress and protection against long-term decline in soil organic matter. These benefits may be particularly relevant to a production system exposed to nature loss, climate variability and rising input dependency.

In the short term, conventional soybean production costs are estimated at $1,158/ha. The cost of transition varies across practices, but cover crops, reduced tillage and the combination of reduced tillage with cover crops stand out as the lowest-cost options from a direct economic perspective. While these practices do introduce additional costs — notably on cover crop seeds, agrochemicals and fuel for extra field passes — they could reduce total costs by about $65-$69/ha relative to conventional production. This net saving is driven in large part by yield improvements associated with the practice change, which supplements income and more than offsets the additional input expenditure.

The main exception is no-tillage. In the short term, no-tillage adds $78/ha, 7% more than conventional production. Analysis shows that about 40% of this cost is driven by a short-term decline in transition yield of $33/ha, making no-tillage the most financially challenging short-term pathway because it combines a higher cost base with a yield-related revenue loss.

By the medium term, conventional soybean production costs rise by 6% to $1,225/ha. The cost of transitioning to sustainable practices, however, changes differently over time for each practice, depending on how the cost drivers interact.

According to the Brazilian Agricultural Research Agency (Embrapa), reduced tillage for soybeans is gaining traction, with around half of the country’s planted area having adopted the practice as of 2026. Meanwhile, the adoption of cover cropping remains limited. We observe that the transition to sustainable soybean production in Brazil is hindered by operational challenges and efficacy concerns (Figure 8).

Despite these hurdles, market-based mechanisms are beginning to emerge to encourage large-scale adoption. For instance, five key agricultural value chain players — Agrivalle, Bayer, BRF, GAPES and Produzindo Certo — formed a consortium called Reg.IA in 2024 that is offering a 2% price premium over spot market rates for a select group of Brazilian soybean farmers that follow sustainable practices such as cover cropping, high-quality no-tillage, organic fertilizer, and bio-pesticides.

Climate and ecological science research have identified potential resilience benefits associated with sustainable agricultural practices in the context of climate change and nature loss. However, known up-front investment needs in contrast to uncertain economic output loss related to environmental degradation over time could stand in the way of systemic adoption.

Climate risk integration has advanced in both political and farm-level economics. However, estimating the economic effects of nature-related risks remains challenging. Specifically, ambiguity persists regarding how the loss of ecosystem services translates into quantifiable output loss. This is compounded by inherent uncertainties and a lack of empirical evidence regarding long-term net economic gains of switching from traditional to sustainable agricultural models.

Despite the limitations, our approach illustrates that potential economic losses  from maintaining current practices could outweigh the cost of transitioning to more sustainable practices. Ultimately, our analysis suggests that the adoption of sustainable agriculture could depend on four main factors (Figure 9).

While regional demographic and structural nuances will influence the pace of transition, the long-term stability of the agricultural sector may depend on the successful integration of these four key drivers.