ML powered Credit risk analytics system of Ethiopian Farmers

There exists an enormous supply-demand gap in credit in rural areas of Ethiopia where 80% of 120 million of the population lives. Studies show banks have only given credit to 300,000 customers (since banking started in the country!)[1] while MFIs are assumed to have 5 million customers[2], leaving more than 66.7 million adults without access to credit. In Africa, 70% of the population works in the sector, but only 3-5% of commercial credit goes to the sector[3]. According to the Findex report, 1.7 billion people don’t have access to credit word wide.

The financial products, and services offered by MFIs, which are tasked with filling the supply-demand gap in rural areas, are too expensive for most farmers. Due to a lack of proper customer risk analysis, rural farmers are charged high-interest rates[4] compared[5] to investors and urban businesses[6] to offset a false perceived risk. The high inaccurate perceived risk is given as the reason why most MFIs charge a 21.5-25.5% interest rate to farmers with the total cost being higher when transactional, service, and other fees are included, pushing the cost of borrowing to 2 to 3 times that of big businesses in the Addis Ababa.

Private companies working in the agriculture sector prefer to work with big investor farmers as they lack the capacity and knowledge to conduct thorough risk analysis of the rural populations. Thus, they shun direct credit sales to smalls scale rural farmers and prefer to work with MFIs. Products offered through MFIs like agricultural machinery and solar home systems have investment costs & interest rates unreachable as a majority (80%) of farmers have an annual discretionary spending which is less than 660 USD. Additionally, strict regulations and requirements of the MFIs push prices of goods, excluding the cost of borrowing, further up with at least 15-20% seen, with the farmers expected to cover the cost. The high capital requirement for equipment means unless they access these products for cheaper credit, most will be stuck in the perpetual cycle of low production and low income.

Research shows high-interest rates discourage the appetite for borrowing, with an increase of 1% in interest rate decreasing the probability of accessing formal credit by 19%[7]. Others argue due to high interest[8], rural MFIs trap households in a vicious cycle of poverty, negatively impacting healthcare expenditure, household income, asset accumulation, nutrition, health, child labor, and empowering women.

[1] https://www.linkedin.com/posts...

[2] https://addisfortune.news/awas...

[3] 2022 Africa Agribusiness Outlook

[4] MFI characteristics and loan preferences of farmers: Household-level evidence from rural Ethiopia Fekadu Gelaw

[5] https://addisfortune.news/awas...

[6] NBE Quarterly report shows a commercial lending rate of 14.5%

[7]: Shewit Kiros & Getamesay Bekele Meshesha | (2022) Factors affecting farmers' access to formal financial credit in Basona Worana District, North Showa Zone, Amhara Regional State, Ethiopia, Cogent Economics & Finance, 10:1, 2035043, DOI:

[8] Coleman, B.E. (1999), “The impact of group lending in northeast Thailand”, Journal of Development Economics, Vol. 60 No. 1, pp. 105-141.

We have built Ethiopia’s first machine learning-powered credit risk model that uses the credit history of rural farmers and their unique socioeconomic & cultural variables to build their risk profile allowing us to accurately predict the behavior and default probability of those who don’t have any history, increasing the financial inclusion of rural farmers who have been left behind by the current system.

Accurately quantifying the risk profile of rural farmers will allow Ethiopian businesses with diverse products and services like importers of agricultural capital goods, agrochemicals, green energy technology, and other services providers to offer these products and services to rural farmers by credit, incentivizing them to diversify their customer base from big agricultural farmers (investors) to small-scale farmers who represent nearly 80% of the total working-age population. This will help farmers access necessary goods and services that will increase their productivity, income[1] and quality of life.

Allowing Ethiopian agribusinesses to sell directly to farmers helps bypass the tedious regulatory requirement saving businesses and farmers money and time, decreasing costs. Businesses will avoid regulation-induced costs like 10-15% revenue saving in a closed account of MFIs for years, compliance costs; financial costs; and other indirect regulation costs. Consumer side costs like transaction, service, and origination fees can be removed, further decreasing the cost associated with borrowing. The total cost of purchase of goods directly from agribusinesses instead of MFIs is 25-38% cheaper for the rural farmer. The time saved will also be high as companies don’t need to get approval from regional and federal regulatory bodies before they sell their products through MFIs. Farmers also don’t need to wait for months before they get access to credit.

Large-scale implementation of such systems will decrease the risk companies take by distributing the risk across different farmers with different risk profiles. With the Ethiopian capital markets in the process of implementation, we will be in a pole position to move risk from private companies (by selling the debt) to the public markets, increasing the funding pool for farmers.

Compared to manual risk assessments, data-based decision-making will reduce costs by improving operational efficiency, reducing costs and time, reducing administrative costs per unit, and the number of personal interactions with borrowers. It would also reduce discrimination and corruption, synonymous with the Ethiopian credit industry, and encourage private companies to make loans to those who can repay them, maximize their profits, and control a larger market. Our model enables fast, efficient, profitable, and farmer-centric operations. Market-led solutions like ours have a good chance of succeeding as previous solutions that were reliant on government interventions tend to be corrupted and subsidy dependent.

[1] Factors influencing access to formal credit by small-scale women tea farmers in Kenya: A case of Thika district, Kiambu County (Doctoral dissertation, Master’s Thesis. Nairobi University)

Our primary service will be properly quantifying the risk associated with selling products to rural farmers, allowing them to purchase necessary products and services from companies with cheaper credit. Such accurate risk analysis will help companies sell their products & service to a wide range of customers according to their risk appetite directly creating a platform for a rural credit market.

Farmers can purchase machinery, solar home systems, agrochemicals, fertilizers, and other services in a pay-as-you-go method with a long- or short-term loan or by lease financing. By using our services, farmers can leverage their collective data, our analytics & models, and their collective purchasing power to negotiate prices and buy services and products by credit. Hence our primary service is connecting farmers & businesses. 60% of these farmers have annual discretionary spending of 60 -75 USD, 20% 75-660, and 14% more than 660 USD. These farmer households in total are 13.6 million with each household having an estimated 5.2 to 5.8 members.

By using our models, farmers will save up to 25-38%. Money saved will be used to increase productivity, education, health, and other necessary goods and services. Any additional money made or saved by rural farmers is spent on essential goods and services that increase productivity and quality of life that would have been ignored otherwise like the purchase of additional fertilizers, better quality seeds, small agricultural machinery, higher veterinary expenses, increased children’s education quality, etc.

Credit improves savings, asset accumulation, health, food security, nutrition, education, women’s empowerment, housing, and job creation[1], reduces poverty incidence[2], and [3]promotes social cohesion. Promoting[4] access to inclusive rural financial services shows a positive impact at the microeconomic level, improving household welfare and local economic activities and at a macroeconomic level, the degree of financial intermediation is positively correlated with growth in the economy of a country.

Proper risk analysis of Ethiopian rural farmers will increase commerce in the rural community, creating jobs and businesses. Such economic activities have a way of compounding on each other with each business and trade compounding on the previous one helping create opportunities for entrepreneurs to tailor their services and products to farmers. This will create opportunities for private businesses, state-owned enterprises, and NGOs alike to impact the lives of rural farmers from a wide range of possible angles.

[1] Khandker, S.R. and Samad, H.A. (2013), “Are microcredit participants in Bangladesh trapped in poverty and debt ?”, Policy Research Working Paper No. 6404, World Bank: Washington, DC.

[2] Hossain, F. and Knight, T. (2008)  - “Financing the poor: can microcredit make a difference? Empirical observations from Bangladesh”, Brooks World Poverty Institute, University of Manchester, Manchester.

[3] Swaminathan, H., Du Bois, R.S. and Findeis, J.L. (2010), “Impact of access to credit on labor allocation patterns in Malawi”, World Development, Vol. 38 No. 4, pp. 555-566

[4] https://www.ifad.org/documents...

Our company was created with the main aim of introducing machine learning-reinforced credit score modeling and other quantitative risk analytics methods to bring about change in the lives of rural Ethiopians. The company’s mission is to help Rural farmers to access the technology and resources it needs while our vision is to be the biggest agricultural quantitative risk analyst company in Africa. The company employs highly technical and capable individuals educated in universities abroad and prestigious Ethiopian universities like the Addis Ababa Institute of Technology.

Our core competencies are quantitative risk analysis, risk modeling, implementation of Machine Learning Models, and installing and maintaining work on solar home systems. Our employees are highly educated technical individuals who have a specific set of skills and experience that has given them the capabilities of executing projects to the highest quality standard. The mixture of abroad and Ethiopian-educated staff has helped us bring about new technologies and way of doing business from abroad and implement it in such a way that takes the local poor rural farmer’s socio-economic and cultural experience into consideration.

Some of our employees have experience working as business consultants, accountants, and financial analysts for large consulting and auditing firms, while others have experience working in solar energy and construction companies. This has helped us design and implement various marketing, sales, and distribution strategies, conduct rigorous market research, etc. Since our employees have large Ethiopian companies as clients at different levels and positions, we can use their experience to help us develop solutions to various problems we have faced in the past and will face in the future. The strength of our company lies in leveraging our experience and connections to achieve our mission and realize our vision.

Our engineers who have experience working for solar companies have helped provide after-sales service to the first product we have offered by credit, which is a solar home system. As we start diversifying the products and services third-party companies offer using our credit model, we will add more technical engineers to offer our clients, sales service. Quality, efficient, and customer-oriented after-sales service is a core part of our marketing strategy.

Fixing malfunctioning solar systems for free has helped us build a necessary relationship with rural farmers and the farmer's union. Our Electromechanical Engineers have hands-on experience fixing systems that were previously purchased by other companies, which has served us well in building the trust and relationship we need to serve rural farmers effectively. For example, our lead sales agent is an engineering graduate with experience working as a sales agent and lead engineer for various farmer's unions. She has a deep connection to the leaders of the farmer's unions and knows how to get data, build a credit model, sell our product, and convince our customers (rural farmers) to continue sharing their experiences with us so we can measure the impact of our intervention.

A solar home system is the first product we are offering for credit, and we have managed to sell more than 118 solar home systems to 126 families since we began operations. These systems have 4 lights, 1 torch, a radio, and a phone charging kit and can light one large (by Ethiopian standards) household. All our customers are rural families engaged in agriculture. None of them are members of any modern financial institution and have no history of taking loans from microfinance institutions. In all households, both men and women are economically active. All their children are currently enrolled in school, despite the highest-educated farmer being a 6th-grade dropout. About 65% of the farmers are seasonal agricultural producers. But the women, who are most active in the trading and service industries, ensure the family has a steady income throughout the year. So even if most of their income is dependent on agriculture, they still earn income throughout the year.

Our credit sales have affected not only the farmers but their families as well. In total, 648 people use our products (all people in the households). We have noticed a 65% increase in phone usage from those who have purchased SHS from us. We have also noticed their children are studying for longer hours a week (up to 7 hours). They have also saved an average of 34.5% of the money they would have paid for products by purchasing the SHS compared to purchases from MFIs. With the money they saved, we have also seen an investment of more than a 17-28% (on average) increase in their annual fertilizer and seed purchases. This, we believe, is further proof of our assumption that any money farmers save on their purchases will be used to increase their productivity and/or improve their quality of life.

We have also agreed with 10 farmers’ unions to use their credit histories to create credit models. We have used the data of their members to create the credit risk profiles of more than 10,000 farmers from these 10 unions. That means we also have a credit score of more than 10,000 farmers. The farmers who bought the product were not from these unions. So, we have used the data of those who had a credit history to predict the behavior of those who don’t have one. We currently sell our SHS system for 4800 (90 USD) ETB. Our revenue is currently more than 10,124 USD. 

We also have an agreement with a local third-party company that is currently, as of April 2023, importing farming machinery like pumps, with a market price of more than 95,000 USD, and has agreed to use our model to distribute its product to farmers.

Our order book also has more than 400 people looking to buy our products. Lack of access to credit and foreign currency has been a bottleneck in addressing this demand. As our products are cheaper than the alternative solutions, it’s not surprising there is a huge demand for them. With more than 55% of the population having no access to electricity, the demand for cheap, green electricity is significant.

The first reason we are applying is to gain access to MIT’s wide networks and access valuable technical knowledge and resource in risk management, quantitative analytics, and data relating to developing countries. Most developing countries, especially those in sub-Saharan Africa don’t have the necessary financial markets needed to finance the various activities they need to develop. Most of these financial resources come from western countries. The valuable insight we will gain from MIT’s network that have worked in financing developing projects and their experience relating to risk management and quantitative analytics will be crucial. The technical expertise will help us increase our capacity in creating systems that will allocate risks and resources at the right place.

The second reason is financial assistance to help us provide guarantees on loans issued to farmers by private companies. Our business model is simple. We will want to import agricultural machinery and sell them by credit to farmers ourselves and use the cash flow to guarantee the sales third-party companies make using our model. Third-party companies who will use our models will have confidence knowing we have “skin in the game” when we will sell products by credit to farmers. In due time, we will want to be a sort of insurer of transactions relating to farmers by leveraging model and quantitative data and stop selling our own products.

Third, we want to gain extremely valuable experience and insight from any MIT network who have experience working in bond or credit markets. As Ethiopia’s financial & capital markets are opening, we will have a considerable advantage if we have experience sharing with any MIT network member who is active in the industry. 

Machine learning utilizing and statistics-based credit risk modeling is unheard of in Ethiopia as Banks and MFIs use qualitative analysis to determine the risk profile of their customers. Our cutting-edge method will be the first-time data-driven credit risk modeling is implemented in Ethiopia potentially allowing millions who don’t have a credit history or the necessary type of assets that are accepted by MFIs and Banks to access capital goods and services by credit directly from Ethiopian companies and service providers. By using the data of those with credit history, we will be able to accurately predict the behavior of those with no history increasing the financial inclusion of the rural population.

Banks and MFIs use qualitative risk analysis that doesn’t take a lot of Ethiopia’s farmer situation into consideration leading both to cater to big businesses in the capital city and regional areas respectively. Due to a lack of technical know-how, access to credit history, and the average rural farmers’ lack of assets that qualify for collateral, farmers are perceived to have high credit risk. This couldn’t be further from the truth as Ethiopian MFIs[1] are experiencing less than 2% of default on their loans to Rural farmers. Such inaccurate risk measurement is given as the reason for the high-interest rate. We have validated this showing the risk analysis of banks & MFIs is false, and farmers aren’t as risky as they are made out to be.

Information and data of farmers, when properly analyzed can be used to predict the probability of default, loss given default, and the exposure at default. By using the socioeconomic variables that describe these farmers, we can accurately predict the behavior and risk of other farmers who don’t have this history. As third-party companies sell more products and services using our model, we will be able to collect more data and better optimize improving the services we offer. So, the accuracy of our model, its differentiation power, and its ability to accurately predict behavior are only going to grow and get better.

Our models will also allow the financial inclusion of those who don’t have any loan history, or the necessary asset required by banks for collateral. By using the loan history of others, we have singled out the variables that have high statistical significance in determining the probability of loan repayment to a high degree and use it to predict the behavior and default probability of those who don’t have a history. This will allow us to give out loans to those farmers that don’t have a loan history and help them access capital goods that could increase their production and productivity.

[1] MFI Characteristics & Loan Preferences of farmers household level evidence from Rural Ethiopia – Fekadu Gelaw 

• Partner with 100 farmers’ unions across the country

We intend to collaborate with at least 100 farmers' unions over the next five years to develop a particular credit model for their members, and with each farmers union having an average of 5,000 members, we expect to have modeled the behavior of at least 500,000 farmers. As our ability to accurately predict and model the credit and counterparty risk behavior of farmers grows exponentially with the amount of data we have, so will our capacity and ability to accurately price and quantify risk.

Farmers will be able to capitalize on their numbers and risk analysis to buy goods and services directly from Ethiopian companies at lower prices. Farmers will be able to buy the products they require at a fair price because it will be market-driven. Ethiopian companies will be able to access previously inaccessible customer demographics for the first time in our country's history. Such a linkage is bound to increase productivity through the supply of machinery, agrochemicals, and fertilizers; promote gender equality as women have a lower level of risk according to our model; promote the use of green energy; and indirectly impact children's health, environment, and education.

• Distribute Agri-products & services of third-party companies using our models so farmers can have access to the technology and machinery they need.

By the end of the 5 years, we plan for our partnering farmers union to buy, by credit, a minimum of 200 million ETB (4 million USD) worth of goods and services from third-party companies. The additional services and goods farmers will be able to get by leveraging their collective data and our credit model will help increase their productivity and economic output.

We also expect to be providing different types of goods and products, both ours (imported by us) and those of third-party companies, to rural farmers. We expect these products and services to include, green energy equipment, small and heavy agricultural machinery, agrochemicals, other agricultural capital goods, veterinary services, etc. 

• Ensure the credit transactions between companies (those who use our model) & farmers for a small premium.

In 5 years, we expect to have directly (our own imported products) sold a total of 50,000 SHS products on credit, mining invaluable data about the rural population, and further enhancing our credit model and analysis. This means additional 50,000 Households will be able to access cheap green electricity because of our interventions.

We want to use the cash flow from this to ensure the credit sales of third-party companies make to farmers. We can leverage our data and knowledge about farmers to guarantee credit sales encouraging businesses to expand their business. At first, we expect most of our revenue to come from our sales but after the third year, we expect the premium companies will pay will constitute more than 50% of our total revenue and grow exponentially till we stop selling our own products. 

Our main aim is to make sure farmers will access the necessary goods and services needed for their agricultural activities at an expensive interest rate that holds farmers in a perpetual cycle of poverty. To this end, we will track KPIs that measure the performance of the model, the businesses, and the impact on farmers.

Model performance KPIs will include ROC, Cumulative Accuracy Profile, Accuracy Ratio, Gini coefficient, and Kolmogorov-Smirnov among others. These KPIs will help measure the performance of our model in predicting the credit behavior of farmers. We will also track the expected default rate per score range and see the results (actual defaults). Such quantitative measurements will help us see trends that may be described by socio-economic, environmental, or geographical factors and will be included in our model KPIs. For example, a group of rural farmers from the same area might all default due to unforeseen events like floods, etc. Model performance is essential in helping us attract and retain partner third-party companies.

To track the growth of the business so it can reach a level of self-sufficiency, we will use KPIs that will track the total number of products we sell directly by credit, how many different types of products are sold, how much money is generated from sales, and the impact of default rate effect on the financial health of the company, the number of third-party companies we have as a partner. For geographical diversity and for our model to accurately predict a diverse farmer profile, we will be partnering with different farmer unions from different parts of the country. Thus, we will be tracking the number of farmers’ unions we have partnered with as metrics to track the growth of our business.

Farmers will be at the center of our operations, and extensive data and key performance indicators (KPIs) will be gathered to monitor interest rates in real-time, as well as purchase (investment) effects on quality of life and productivity. Data pertaining to agricultural production are collected as part of the farmers union's normal business operations, and we have an arrangement with the farmers union to track the financial data of willing participants, both those who buy agricultural, green energy machinery and services on credit (experimental group) and those who do not purchase products and services on credit for at least a year and a half using our model.

With this strategy, we will be able to compare the performance of our customers who purchase our products on credit versus those who do not. We will be able to assess the effect of our credit sales on productivity by tracking increases in investment in fertilizers, machinery, and other equipment, as well as outputs such as yields, total sales, and cattle numbers. Changes in quality of life will be evaluated by metrics such as changes in health expenditure, changes in housing status, changes in electricity type, increased use of modern financial infrastructure, children's education status, and so on.  

Access to machinery and goods on credit means that farmers will be able to purchase equipment that improves overall productivity and quality of life. Traditional means of production are one of the major reasons for the low production yield per hectare in Ethiopia. The high capital required for the purchase of technological equipment means unless they access these products by credit, most will be stuck in the perpetual cycle of low production, income, and investment.

Using our models, companies can collect interest payments on the products they sell directly to farmers without going through the unnecessary regulatory requirements of MFIs. By selling directly, companies save time by avoiding waiting for 3–4 months before getting the necessary documents to distribute their products through MFIs and save money by avoiding the 10% savings requirement of the MFI. The time saved is important as inflation is currently in the late 30s and the time saved is expected to cut costs by 8–10%. Service fees that the rural farmer is expected to pay an MFI for equipment through them is an average of 5–7.5%, which is also avoided. In total, prices are predicted to decrease by 18–26% just by avoiding MFIs as a distribution channel.

Our model categorizes customers into 5 levels of risk and is expected to pay interest ranging from 6% to 10%, they will be saving an additional 15-20% on interest paid to MFIs. In total, rural farmers are expected to save 32-38.5%.

Ethiopian farmers are poor, and any money saved from purchases of goods and services will be spent on improving their quality of life or productivity by investing in additional capital or farming equipment, etc. Savings will enable farmers to make investments by helping them allocate and store money to buy additional inputs, such as fertilizer or pesticides, or other technologies. In[1] households of farmers, the objective of savings was to ensure provisions for running consumption expenditure, purchasing durable goods, and expanding their economic activity.

Productivity & quality of life[2] depends on farmers’ ability to accumulate income and their way of spending. Saving is a strategic variable in rural economic growth, improving quality of life, and promoting sustainable farming development. Purchases of capital goods machinery, agrochemicals, and fertilizer purchases will increase yields, and diversify their income streams, etc. Rural development will also contribute to reducing poverty in urban areas by reducing excessive population influxes from rural areas. There[3] is increasingly strong evidence that promoting access to inclusive rural financial services shows a positive impact at the microeconomic level, improving household welfare and local economic activities. Also at the macroeconomic level, the degree of financial intermediation is positively correlated with growth in the economy of a country.

[1] SAVINGS BEHAVIOUR IN HOUSEHOLDS OF FARMERS AS COMPARED TO OTHER SOCIO-ECONOMIC GROUPS IN POLAND Agnieszka Kozera*, Romana Głowicka-Wołoszyn, Joanna Stanisławska

[2] Savings of Small Farms: Their Magnitude, Determinants and Role in Sustainable Development. Example of Poland - Barbara Wieliczko  , Agnieszka Kurdy´s-Kujawska , and Agnieszka Sompolska-Rzechuła

[3] https://www.ifad.org/documents...

In machine learning, credit risk scoring is basically a classification problem. Using a credit risk evaluation approach, farmers can be classified as "good" or "bad" customers. Based on the credit scoring findings, financial institutions can make loan approval decisions and risk pricing decisions.

We started by collaborating with five farmer unions as they offer government-supplied animal feeds, fertilizers, and seeds on credit to their members.  Their data on their members can be comprehensive and include socioeconomic variables such as family size, land size, yield rate, yield type, cattle no, cattle vaccination status, and so on. Developing a credit risk model that considers the Ethiopian farmer's context is critical, as farmers or consumers behave differently in various parts of the globe, and thus we build that focuses on the Ethiopian farmer. Because all the material was on paper, we started by digitizing and cleaning it. After removing outliers and cleaning the data we removed variables that were showing multicollinearity. [

We divided our data into two, train and test datasets after defining the dependent variables (good and bad determined by presence of delinquency). Then, we used a variety of statistical analyses and feature engineering to identify our variables that are statistically significant and have a high degree of predicting loan status. We discovered numerous statistically significant factors that could predict the likelihood of loan repayment to a higher degree of accuracy that are unique to the Ethiopian Farmer by using machine learning and additional data gathering. As our selection parameter, we use the usual P-value of 0.01.

We used the logit function to calculate the likelihood of an individual defaulting based on the factors we gathered. The logit function is commonly used for linking functions to isolate those with high statistical relevance in predicting the dependent variable. To assist our supervised learning model, we used the sci-kit learn machine learning module. This module is a powerful instrument that businesses and scientists use to analyses predictive data. 

Supervised learning is a type of machine learning algorithm in which the goal is to classify variables and predict outcomes, which in our case are either default or complete payback. In predictive analysis, supervised algorithms are used because they provide a deep look into various independent variables, which in our case were socioeconomic variables of farmers. Therefore, we can predict certain outcomes based on our dependent variables. (Default or loan payback). We were able to create a model that could predict the likelihood of default through proper data analysis using modern statistical analysis. This was used to quantify Probability of Default, and simulated Loss given default and Exposure at default for different products and services that can be provided to farmers.

Our model has helped us quantify the risk and price of selling agricultural machineries to farmers with no history of credit purchases from any financial institutions using the data of those who have history of procuring agricultural inputs from their unions.

Our work necessitates that we completely comprehend the culture of the societal segment that we seek to serve. Our success is dependent on how well we fully understand the various cultural and sociological idiosyncrasies of the community we serve. As Ethiopia is a diverse nation where people from different ethnic and religious backgrounds call home, it is not a surprise that our team reflects this, with individuals of various ethnic and religious backgrounds all working together to achieve a common objective. Diversity is essential for our company's success.

Our model will not operate at the intended level unless we include social and cultural norms. Credit, for example, is strongly forbidden and considered a sin in some cultures, particularly in Ethiopia's Muslim population. With the Muslim population expected to account for 40% of the overall population, we must work hard to convince them of the benefits of financing the purchase of capital goods and services. One option is to raise the processing and operating fees for Muslim rural farmers and provide zero-interest products, like Islamic banking. It is considered a serious social offence in different cultures to be unable to repay someone to whom you owe money.

This will have a detrimental impact on our capacity to estimate default risk since we need to know the conditions under which a typical borrower will fail.  Thus, we must fully understand the effect of social and cultural norms on farmers’ behaviors. 

To understand cultural norms like those given above, we actively recruit locals from the region where the farmer union is located so that we have a deeper knowledge of the population we hope to serve. This will allow us to better grasp the culture and quirks of the farmers we aim to serve. Working with folks that are well-versed in the culture will allow us to learn a great deal. Thus, diversity is embedded well into our business models.

Our environmental policy prohibits us from working with any company whose policy isn’t aligned to the well-being and health of farmers & communities, especially considering climate change. And our choice of first product is proof with our company offering rural farmers the option of purchasing solar technologies for by credit for a cheaper price.

Thus, ESG policies are engrained to the DNA of our company and crucial to our success. 

As a risk analytics company, we exist between private companies and farmers, aiming to create a vibrant credit-risk rural market that is not expensive for farmers and risky for private companies. We help farmers buy the products and services they need by credit with a low-interest rate and help companies minimize their risk exposure. This is helpful for both parties as they need each other. Currently, as we are selling our products, the farmers get products like solar home systems, which they need as they aren’t connected to the national grid yet, by credit at a cheaper rate than MFIs. Third-party companies who want to sell products to farmers by credit get to sell their products using our models at a cheaper rate than MFIs without the various fees and bureaucratic costs.

While the retail sales of products like solar home systems that we make are not part of our long-term strategy, in the short-term it plays a crucial role by helping us increase our cash flows as well as collect valuable data from farmers who purchase these products. The data is used to increase the predictive power of our model as machine learning models are highly dependent on high-quality data. The cash flow helps us insure the credit purchases farmers make using our model from third-party companies. The additional premium companies pay for ensuring their credit transactions can further help us expand our business, onboard more third-party companies, increase the number of products and services offered to credit through our model, and fuel further expansion.  

By being in the middle, we make money by selling our products to farmers, selling our credit risk analysis to companies, and leveraging our data and knowledge by selling insurance on the credit purchases of farmers from third-party companies. The Impact of full-scale implementation of our business model is far-reaching but can be summarized by the fact that we will be able to create a vibrant, cheaper, more efficient rural credit market. When companies pay premiums to insure their credit sales, they will share the risk of default with each other, bringing the overall risk they take by selling to one of the poorest farmers in the world. And ss the overall risk decreases, so does the cost of doing business, enabling us to hit our target of creating a vibrant and cheap rural credit market. 

We aim to achieve financial stability with our retail sales, the fees companies pay for using our model, and the premium companies pay us to insure their credit sales after initial investment and/or grant. It is critical for the business to generate revenue and not rely on outside investment or contributions, since doing so might be dangerous and unsustainable in the long run. Investments might come with conditions that compel a corporation to relinquish control over critical choices. Furthermore, donations and grants cannot be guaranteed and may fluctuate due to economic conditions or changes in donor preferences. To this end, it is paramount that the company is able to significantly increase its cash flow in order to expand and offer its businesses to farmers and businesses alike. 

Currently, our main source of income is the retail credit sales of products. Our cash balance is more than 10,000 USD, and a little more than 4 thousand USD is expected to come from credit sales (receivables). We have also partnered with a local company that is looking to distribute agricultural products to farmers. They are in the process of importing their product, and they will be among the first to use our model. We expect to make a little more than 1,500 USD, as we distribute the clients’ product, which has a market value of more than 95 thousand USD. As we grow the number of business partners we have, we aim to use the model fees and retail fees as a reserve for our insurance of credit transactions, with the premium paid to us fueling our growth in the long term.