How To Connect the Financial Market to Environmental Violations: The Hidden Impacts of Fiagro on Indigenous Lands in Brazil

Investment funds issued to finance Brazil's agribusiness sector bundle various bonds; some of them labeled “green” are tied to socio-environmental violations. That was our hypothesis—but how could we confirm and prove it? We grabbed our pack of journalistic tools.

We dug into an investment fund when it was just in its infancy. Fiagro, a financial instrument created by the Brazilian parliament to mobilize private investment to fund agribusiness in the country, was entering its third year of operation when we decided to open it. 

The fund was successful in its start in the market: It reached a net worth of USD $8.1 billion on the Brazilian Stock Exchange (B3) from March 2023 to March 2025, while its major financial asset, Agribusiness Receivable Certificates (CRA, its acronym in Portuguese), increased in volume by 46% during the same period, reaching USD $28.3 billion. 

Fiagro is like a big box filled with dozens of hidden packages or financial assets—each one tied to a different company or project in the agribusiness world. To understand what’s really inside, you have to open each package and look into it closely. It takes time, tools, and knowledge, particularly in financial and geospatial analysis, to answer our investigation question: Is Fiagro financing land-grabbing and/or socio-environmental violations in the Amazon rainforest? With the support of the Pulitzer Center’s Rainforest Investigations Network, we unpacked the box and published our findings in the series Financial Market: A Black Box Over Green Areas.

---

As a nonprofit journalism organization, we depend on your support to fund more than 170 reporting projects every year on critical global and local issues. Donate any amount today to become a Pulitzer Center Champion and receive exclusive benefits!

---

Journalists working at O Joio e O Trigo had already uncovered that Fiagro was financing illegal agribusiness activities in the Cerrado. This Brazilian ecosystem, often considered the country’s “water tank” that supplies water to the entire country, is also threatened by the expansion of monoculture crops. We aimed to check if Fiagro had the same impact on the world’s largest tropical rainforest.

Hypothesis

Initial suspicions about Fiagro arose from its own regulations. The Brazilian Securities and Exchange Commission (Comissão de Valores Mobiliários/CVM), the government body responsible for regulating all financial instruments traded in the Brazilian capital market, has no rules preventing Fiagro’s resources from being accessed by farmers and companies engaged in illegal commercial activities.

Brazil has a history of socio-environmental violations financed by government rural credit and bank loans, but such abuses have been addressed and halted by regulations to exclude “dirty” farmers and companies. It was a lightbulb moment when we realized that there’s a lack of stringent, sustainable rules for Fiagro.

Hence the hypothesis that we set out to prove was that farmers and companies were borrowing money from bonds pooled by Fiagro and applying it to expand their operations and profit from illegal activities.

Methodology and data sources

After some interviews with fund managers and financial workers, we identified datasets that would be useful to test the hypothesis. The starting point was a database maintained by the Brazilian Stock Exchange, which contains mandatory regular filings for all public investment funds. These reports, available as PDFs, list the financial assets like CRAs included in each Fiagro fund. That’s what led us to take a closer look at the assets themselves.

The documents related to these assets are stored in private databases, but they are also publicly accessible through the online database of securitization firms—the companies responsible for issuing CRAs, the bonds backed by agribusiness receivables. That’s where the real digging began: We coded several scrapers using Python, a programming language, to scrape more than 10,700 PDF files from the websites of eight different securitization firms: True, Opea, Virgo, Ecoagro, Provincia, Canal, Vert, and Ceres. These PDF files contain detailed information of the Fiagro assets, allowing us to identify the amount of money involved, where it was going, and most importantly, the names of both individuals (such as farmers) and companies (such as those in the biofuel and meatpacking industries) connected to each bond.

We uncovered companies that accessed huge amounts of credits through CRAs to fund their operations, including meat giants JBS and Minerva Foods, as well as “green” ethanol producer FS Fueling Sustainability.

There was more to keep diving into: Some documents contain lists of rural producers—such as corn and cattle suppliers—paid with funds from Fiagro assets. As those lists surpass hundreds of pages in PDF, we had to make use of table extraction tools including Tabula (open source) and ExtractTable (paid) to extract the data from the PDFs into a spreadsheet for further analysis.

After organizing and cleaning the names, we cross-referenced them with a range of environmental crime datasets downloaded from the official database of the Brazilian environmental agency (Ibama). This gave us details of each crime, such as terms of embargoes, notes of infraction, and fines. The cross reference was done on RIN Data, an internal data platform developed by the Pulitzer Center based on Aleph, the open source data platform by the Organized Crime and Corruption Reporting Project (OCCRP) to support investigative journalism.

Another helpful approach was to connect with research and data organizations specializing in the data we were interested in. We formed partnerships with the Center for Climate Crime Analysis (CCCA) and Operação Amazônia Nativa (OPAN), which shared their databases and supported the investigation with their respective expertise in the field of meat supply chains and land conflict within Indigenous lands.

This analysis identified individuals and companies linked to land invasion and conflicts, Indigenous people’s expulsion from their territories, and illegal deforestation in the Amazon region. Using Brazilian government databases including the Land Management System (Sigef), Rural Environmental Registries (CAR), and Integrated Information System on Interstate Transactions with Goods and Services (Sintegra), we were able to identify the locations where these actors were carrying out commercial activities, such as raising cattle and growing crops. Such information guided our field reporting trips. It was crucial to visit the locations to verify what was happening on the ground and talk to affected communities.

In Rondônia, the first Amazonian state we visited, we witnessed cattle being raised inside an embargoed area overlapping the Tanaru Indigenous land. That was a clear violation of the law, as all commercial activities are prohibited in embargoed areas. Using GTAs (cattle transport permits), we tracked the cattle from the embargoed area to another legal farm owned by the same family to be "laundered" before being sold to JBS and Minerva Foods, the two largest meatpacking companies accessing Fiagro’s funds to purchase cattle.

In another Amazonian state, Mato Grosso, Indigenous members of the Kawaiweté/Kayabi people told us they tried to return to their ancestral land in the early 2000s after they were expelled in the 1960s. However, they were threatened and warned to leave the territory by producers of soy, corn, and cotton who operated farms on the recognized Indigenous land. Two of them are shareholders in a biofuels company that profited from Fiagro assets.

Data sources used in this investigation:

• CVM files: The Agribusiness Bulletin, published every three months by the Brazilian Ministry of Finance, provides information on key issues and recent legislation related to agribusiness bonds;
• Fundos.net: a database of public investment funds and bonds, directly connected to the Brazilian Stock Exchange;
• Anbima: the Brazilian Association of Financial and Capital Market Entities displaying compiled financial data and finfluence reports;
• Securitizers websites: gives details of a bond’s content. Eco Securitizadora was responsible for the first public issuance of CRAs (Agribusiness Receipt Certificates) in the Brazilian financial and capital markets;
• Edgar/SEC: gives an overview of American companies’ annual and sustainability reports. It only works for publicly-listed company information;
• Financial reporting of several news outlets;
• Ministry of Agriculture and Livestock: agri-financial instruments reports that are published monthly;
• Sustainalytics, Climate Bonds Initiative, ICMA: reports and international standards for bonds issuances;
• Social media: YouTube and LinkedIn were the main platforms to find both financial actors and investors who became our sources;
• Instant messaging groups: Telegram groups and channels are a gateway to enter the communities of financial analysts, finfluencers, and investors;
• Sayari, CruzaGrafos, Aleph: sources of customs data, trade flows, and company information;
• ONR, Sicar, Sigef: Brazilian land registries and documentation;
• Sintegra: a compilation of information on interstate transactions involving goods and services;
• Federal Regional Court system, JusBrasil, prosecutors press-office service, and a partially-paid database of court cases involving individuals and entities;
• Prodes/INPE, Mapbiomas: deforestation datasets and alerts;
• Funai and ICMBio: geospatial data of Indigenous territories, protected areas, and conservation units;
• Ibama: The Brazilian Institute of Environment and Renewable Natural Resources, supervised by the Ministry of Environment (Ibama), is a regulatory agency in Brazil responsible for embargoes, notes of infraction, and fines;
• Fala.br or Freedom of Information Act: requirements to access official and non-public data, such as details of embargo terms;
• Planet Explorer, Sentinel Hub EO Browser, Google Earth Pro: satellite imagery;
• Imazon reports: slaughterhouse locations;
• GTAs: In Brazil, every head of cattle must be transported with a document called a GTA. It contains: the origin of the animals and their owner; place of departure; date and time of departure; intended destination; and expected duration of the journey. There are some leaked lists of cattle transport permits.

Limitations of methodology

One of the biggest limitations in the financial data was the lack of detail about where the money actually ends up. Only a few investment funds and bonds include lists of entities who receive the credit raised from the capital markets. When those lists are available, they are a gold mine, as they often include names and identification numbers like CPFs for individuals or CNPJs for companies. But issuers are not required to publish this information, which makes it harder for journalists to trace specific companies or rural producers who benefit from the loans.

One key lesson we learned is that we should have looked at the bigger picture from the start, especially by examining the underlying Fiagro assets earlier in the process. In the beginning, we focused heavily on the Fiagro documents themselves. We later realized the real value was actually hidden in the documentation of the assets, which sat on the sidelines of the main Fiagro documents. In fact, much of the data I ended up scraping midway through the yearlong investigation came from those asset-related files.

Finally, financial jargon can be a barrier for those unfamiliar with the language of the capital markets. Still, with time and effort, it’s possible to break through and make sense of it.

How this methodology could be replicated for other investigations

Journalists and newsrooms can replicate this methodology to help improve transparency in Brazil’s capital markets and beyond. It offers an investigative path that connects two key points. On one side are companies or farmers accessing credit through the stock exchange with little or no public scrutiny. On the other side are investors seeking financial returns through bonds and funds that may be linked to threats on Indigenous lands or protected territories.

The tools and datasets mentioned earlier can support investigations into other companies and supply chains, not only those involving meat and corn, as shown in this case. They can also serve as a starting point for uncovering other underreported aspects of the financial market.