https://doi.org/10.29312/remexca.v17i5.4141

elocation-id: elocation-id: e4141

Ponce-Gutiérrez, Ruiz-Pérez, González-Juárez, and Valenzuela-Reynaga: Interest rate and default risk in Mexican rural financing

Journal Metadata

Journal Identifier: remexca [journal-id-type=publisher-id]

Journal Title Group

Journal Title (Full): Revista mexicana de ciencias agrícolas

Abbreviated Journal Title: Rev. Mex. Cienc. Agríc [abbrev-type=publisher]

ISSN: 2007-0934 [pub-type=ppub]

ISSN: 2007-9230 [pub-type=epub]

Publisher

Publisher’s Name: Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias

Publisher’s Location: México

Article Metadata

Article Identifier: 10.29312/remexca.v17i5.4141 [pub-id-type=doi]

Article Grouping Data

Subject Group [subj-group-type=heading]

Subject Grouping Name: Article

Title Group

Article Title: Interest rate and default risk in Mexican rural financing

Contributor Group

Contributor [contrib-type=author]

Name of Person [name-style=western]

Surname: Ponce-Gutiérrez

Given (First) Names: Blanca Estela

X (cross) Reference [ref-type=aff; rid=aff1]

Superscript: 1

Contributor [contrib-type=author]

Name of Person [name-style=western]

Surname: Ruiz-Pérez

Given (First) Names: Roberto

X (cross) Reference [ref-type=aff; rid=aff1]

Superscript: 1

Contributor [contrib-type=author]

Name of Person [name-style=western]

Surname: González-Juárez

Given (First) Names: Araceli

X (cross) Reference [ref-type=aff; rid=aff2]

Superscript: 2

Contributor [contrib-type=author; corresp=yes]

Name of Person [name-style=western]

Surname: Valenzuela-Reynaga

Given (First) Names: Rodolfo

X (cross) Reference [ref-type=aff; rid=aff1]

Superscript: 1

X (cross) Reference [ref-type=corresp; rid=c1]

Superscript: §

Affiliation [id=aff1]

Label (of an Equation, Figure, Reference, etc.): 1

Institution Name: in an Address: Departamento de Contaduría y Finanzas-Instituto Tecnológico de Sonora. Avenida 5 de febrero 818 Sur, Col. Centro, Ciudad Obregón. Sonora [content-type=original]

Institution Name: in an Address: Departamento de Contaduría y Finanzas [content-type=orgdiv1]

Institution Name: in an Address: Instituto Tecnológico de Sonora [content-type=orgname]

Address Line: Avenida 5 de febrero 818 Sur, Col. Centro, Ciudad Obregón. Sonora.

Country: in an Address: México [country=MX]

Affiliation [id=aff2]

Label (of an Equation, Figure, Reference, etc.): 2

Institution Name: in an Address: División de Ciencias Económico Administrativas-Universidad Autónoma Chapingo. Carretera México-Texcoco km 38.5, Chapingo, Texcoco, Estado de México. [content-type=original]

Institution Name: in an Address: División de Ciencias Económico Administrativas [content-type=orgdiv1]

Institution Name: in an Address: Universidad Autónoma Chapingo [content-type=orgname]

Address Line: Carretera México-Texcoco km 38.5, Chapingo, Texcoco, Estado de México.

Country: in an Address: México [country=MX]

Author Note Group

Correspondence Information: [§] Autor para correspondencia: rodolfo.valenzuela@itson.edu.mx . [id=c1]

Publication Date [date-type=pub; publication-format=electronic]

Day: 01

Month: 08

Year: 2026

Electronic Location Identifier: e4173

History: Document History

Date [date-type=received]

Day: 01

Month: 06

Year: 2026

Date [date-type=accepted]

Day: 01

Month: 08

Year: 2026

Permissions

License Information [license-type=open-access; xlink:href=https://creativecommons.org/licenses/by-nc/4.0/; xml:lang=es]

Este es un artículo publicado en acceso abierto bajo una licencia Creative Commons

Abstract

Title: Abstract

The study aimed to analyze the structural factors that limit access to rural credit in Mexico to show how interest rates, climate risks, document formalization, and territorial location operate as institutional filters of financial exclusion. To this end, a binary logistic model was applied to a national database of 250 observations, compiled from administrative records and secondary data from 2023, including variables that identify exclusion patterns not attributable to the borrower’s productive performance. The model was statistically significant and had robust explanatory power for default on rural loans. The quantitative results show that the interest rate is the factor with the greatest impact: for each additional percentage point in the rate, the probability of default increases by 21.5 percentage points. A change in climate risk category, such as going from frost to pest raises this probability by 2.23%. In addition, each additional month in the term reduces risk by 2.01%, suggesting a differentiated allocation based on the solvency profile. These findings allow us to conclude that the interest rate does not reflect the producer’s productivity but rather the system’s structural fragility, acting as the main driver of exclusion and that, in the rural financing system, there is a rigid architecture that penalizes structural risk without offering adaptation mechanisms for vulnerable producers.

Keyword Group [xml:lang=en]

Title: Keywords:

Keyword: borrower segmentation

Keyword: credit exclusion

Keyword: institutional filters

Keyword: marginal effects

Keyword: predicted probability

Counts

Figure Count [count=0]

Table Count [count=6]

Equation Count [count=2]

Reference Count [count=25]

Abstract

The study aimed to analyze the structural factors that limit access to rural credit in Mexico to show how interest rates, climate risks, document formalization, and territorial location operate as institutional filters of financial exclusion. To this end, a binary logistic model was applied to a national database of 250 observations, compiled from administrative records and secondary data from 2023, including variables that identify exclusion patterns not attributable to the borrower’s productive performance. The model was statistically significant and had robust explanatory power for default on rural loans. The quantitative results show that the interest rate is the factor with the greatest impact: for each additional percentage point in the rate, the probability of default increases by 21.5 percentage points. A change in climate risk category, such as going from frost to pest raises this probability by 2.23%. In addition, each additional month in the term reduces risk by 2.01%, suggesting a differentiated allocation based on the solvency profile. These findings allow us to conclude that the interest rate does not reflect the producer’s productivity but rather the system’s structural fragility, acting as the main driver of exclusion and that, in the rural financing system, there is a rigid architecture that penalizes structural risk without offering adaptation mechanisms for vulnerable producers.

Keywords:

borrower segmentation, credit exclusion, institutional filters, marginal effects, predicted probability.

Introduction

In Mexico, financing for the rural sector has historically been channeled through development banking schemes such as the Trust Funds Instituted in Relation to Agriculture (FIRA), by its Spanish acronym and the now-defunct National Financial Institution for Agricultural, Rural, Forestry and Fisheries Development (FND), by its Spanish acronym. Nevertheless, international organizations such as the International Monetary Fund ( IMF, 2022 ) and the World Bank ( WB, 2021 ) have warned that the Mexican financial system has limited coverage in rural areas, conditioned by institutional, structural, and territorial barriers.

These restrictions do not respond to the borrower’s productive performance but to frictions such as the absence of collateral, credit history, and document formalization ( Beck et al ., 2009 ; De Olloqui and Fernández, 2017 ). Various studies indicate that the interest rate on rural credit does not reflect the producer’s technical efficiency but rather the fragility of the financial channel that serves producers ( De la Vega et al ., 2014 ; Cassimon, 2022 ; Yu et al ., 2023 ).

In this sense, the rate operates as an instrument of financial exclusion rather than as an indicator of economic risk ( Khan et al ., 2024 ). These authors argue that in Mexico the interest rate has been used as a mechanism of financial discipline to compensate for the informality, territorial dispersion, and productive volatility of the rural sector, replicating urban logics without adjustment to the agricultural context.

From a more structural perspective, FAO (2020) ; Assairh et al . (2024) state that, in economies with inelastic capital supply and segmented entrepreneurs, high rates can exclude productive agents if the financial system filters by institutional frictions instead of technical efficiency. These frictions include weak governance, information problems, contractual constraints and territorial segmentation ( Franjo et al ., 2022 ).

Findings such as those by Madero-Salmerón and Núñez-Mora (2025) highlight that the behavior of the interest rate responds to macroeconomic factors such as expected inflation, the exchange rate, and central bank decisions, which generate distortions in the local rate curve. In rural contexts, this implies that the rate reflects not only the financial cost but also the institutional burden that producers face when trying to access formal financing schemes.

Despite the interest rate playing a central role in rural credit, empirical studies analyzing its effect as a structural variable of exclusion remain limited. Most studies focus on coverage or the operational sustainability of intermediaries, without addressing the specific impact of the rate on the default risk ( Del Ángel, 2003 ). This omission is relevant given that in Mexico formal credit is unevenly distributed, and the rate penalizes not technical performance but rather the producer’s location, formalization and documentary profile ( WB, 2021 ).

In this framework, the present work aims to empirically analyze the impact of interest rates on the default risk of Mexican agricultural units in order to show how this variable reflects the rural financial system’s structural fragility rather than the borrower’s productivity.

Materials and methods

The present work was observational and employed data from secondary sources and administrative records. Because it is not possible to access the individual microdata of FIRA borrowers, we aggregated and categorized the available aggregate data into a dichotomous variable, using several (polytomous) attributes available on the microsites of the National Banking and Securities Commission ( CNBV, 2025 ) by its Spanish initialism, the minimum risk search engine ( FIRA, 2025a ), and the microsite called in figures ( FIRA, 2025b ). The dataset used comprised 250 cross-sectional observations for 2023. The variables used in the estimated binary logistic model are shown in Table 1 .

Table 1

Table 1. Definition of the empirical model’s variables.

Variable Description Scale of measurement
REPRIS Loan outcome (repayment risk) Dichotomous: 1= default, 0= repayment (dependent variable)
CLIRIS Type of climate risk associated with the loan granted (events: hail, frost, heat wave, pests, drought, and rainy period) Nominal, categorical
LOAM Total value of the loan granted Continuous, quantitative in pesos ($)
LOTER Term to repay the loan granted Discrete, quantitative in months
INRAT Annual interest rate applied to the loan granted Continuous, quantitative in percentage (%)
COL Collateral granted by the borrower Ordinal, dichotomous: 0= A collateral was provided, 1= no collateral was provided

The methodological procedure used in the research was a binary logistic model, which estimates the probability of an event (default) occurring given a set of predictor variables. The empirical model to be estimated has the following mathematical expression:

ln ( P 1 - P ) = β 0 + β 1 CLIRIS + β 2 LOAM + β 3 LOTER + β 4 INRAT

Where: ln(•) is the logarithm of the odds ratio (odds logarithm); β0 is the constant term (intercept); β1, β2, β3, and β4, are the coefficients that weight the respective explanatory variables, which are those defined above; and e is the base of the natural logarithm (whose value is approximately 2.718). Regarding the expected signs for the model’s coefficients and based on the financial logic of the default risk, it is expected that the parameter β1 will be positive, since higher climate risk tends to increase the probability of default.

As for the amount of the loan granted, an inverse relationship with default risk (β2< 0) is expected, assuming that larger amounts are assigned to more solvent profiles. Similarly, the repayment term is also expected to have a negative effect (β3< 0), as longer terms offer greater flexibility to the borrower and reduce risk. Finally, the interest rate is expected to have a positive coefficient (β4> 0), since higher rates increase the financial burden and, therefore, the default risk ( Armendáriz and Morduch, 2010 ; Meki and Quinn, 2024 ; Khan et al ., 2024 ).

Although the logistic model estimates coefficients that indicate the direction and significance of each variable, these are expressed in terms of the logarithm of the odds ratio, which makes them difficult to interpret directly; therefore, to facilitate their interpretation, the so-called marginal effects are calculated, which show how much the probability of the event (default) changes in response to a unit variation in each explanatory variable, keeping the other variables constant.

The marginal effect is estimated using the following analytic expression:

P ( Y = 1 ) X j = β j P ( 1 - P )

Where the right side of the equation is the marginal effect of the variable on the probability of the event; it is the estimated coefficient for the variable in the logistic model; the estimated probability that the event (default) will occur and the complementary probability (that the event will not occur).

Results and discussion

The Table 2 presents the descriptive statistics for the discrete and continuous quantitative variables used in the research.

Table 2

Table 2. Descriptive statistics of quantitative variables.

Variable Mean Standard deviation Value
Minimum Maximum
LOAM 266 000 67 282.3 180 000 420 000
LOTER 18.3 6.79 12 36
INRAT 14.98 1.10 13 17

The variables are described in Table 1. Based on statistical runs from Stata 15.

In the case of categorical variables, since it is not possible to obtain the above descriptive indicators, it is usual to show the frequencies and proportions of each category. The Table 3 shows these statistics for the study’s categorical variables.

Table 3

Table 3. Frequency and proportions of categorical variables.

Variable-categories Frequency Distribution
Percentual Cumulative
CLIRIS
Hail 7 2.8 2.8
Frost 18 7.2 10
Heat wave 34 13.6 23.6
Pests 76 30.4 54
Drought 86 34.4 88.4
Rainy period 29 11.6 100
Total 250 100
REPRIS
Default (1) 110 44 44
Repayment (0) 140 56 100
Total 250 100
COL
Not provided (0) 110 44 44
Provided (1) 140 56 100
Total 250 100

Based on Stata 15 outputs.

The results of the empirical logistic model, in which the response variable is the repayment risk (REPRIS), are shown in Table 4 .

Table 4

Table 4. Estimated parameters of the empirical logistic model.

Variable Parameter Standard error z P>|z|
Intercept -156.9097 28.89789 -5.43 0
CLIRIS 1.241496 0.4827 2.57 0.01
LOAM -0.0000372 0.0001 -3.31 0.001
LOTERM -1.124856 0.2552 -4.41 0
INRAT 11.98603 2.2032 5.44 0

Based on Stata 15 outputs.

There is an inverse relationship between the amount of loan granted (LOAM) and the repayment risk (REPRIS), which is consistent with the logic of credit selection, in which more favorable contractual conditions tend to be assigned to borrowers with better solvency profiles and lower default risk ( Meki and Quinn, 2024 ; Khan et al ., 2024 ). Table 5 shows the goodness-of-fit statistics of the empirical model.

Table 5

Table 5. Goodness of fit statistics of the empirical model.

Observations Log-likelihood Chi-square likelihood ratio Probability associated with chi-square Pseudo R2
n L(θ) LRX2(4) Prob> X2 R2McFadden
250 -17.3579 308.25 0 0.8988

Based on Stata 15 outputs.

The chi-square likelihood ratio statistic (308.25 with four degrees of freedom) and its associated probability (0) indicate that the model with explanatory variables is highly significant and substantially improves the fit compared to a model without predictors.

This result implies that at least one of the independent variables has a relevant effect on the probability of default on rural loans. In addition, the pseudo-R2value (0.8988) indicates that the model achieves a notable improvement in the fit to the observed data, reinforcing its explanatory power ( Menar, 2000 ). The average marginal effects of the empirical model are shown in Table 6 .

Table 6

Table 6. Average marginal effects of the empirical logistic model.

Delta method z P> |z|
∂Y∂X Standard error
CLIRIS 0.0222695 0.0085895 2.59 0.01
LOAM -6.68e-07 1.62e-07 -4.12 0
LOTER -0.0201772 0.0039901 -5.06 0
INRAT 0.2150008 0.0349705 6.1 0

Based on Stata 15 outputs.

The interpretation of the average marginal effects is the key point in the logistic model, as it allows the estimated coefficients to be expressed in terms of probability. In the case of climate risk (CLIRIS), it can be stated that a change of one category in the type of climate risk; for example, from ‘frost’ to ‘pests’, is associated with a 2.23% increase in the probability of default, keeping the other variables constant.

Additionally, for each 1-unit increase in the loan amount (LOAM), the probability of default decreases by 0.000067%. For the loan term (LOTER), the probability of default increases by 2.02% per additional month in the loan term. According to the results, for each 1-percentage point increase in the interest rate (INRAT), the probability of default would increase by 21.5 percentage points.

Before discussing the results, it is necessary to note that the application of logistic models or other approaches, such as generalized linear models and multivariate methods, to obtain the probability of default risk among borrowers is scarce, and those that could be consulted, such as Meky and Quinn (2024) , do not use average marginal effects to interpret the parameters estimated with the logistic model.

The problem with interpreting these coefficients is that it is done in terms of the odds ratio; for example, as stated in Assairh et al . (2021) , if the logistic model parameter for the gender variable (woman= 0, man= 1) is -0.29, then its odds ratio is 0.75.

This result can be interpreted as meaning that men have a 24.2% lower default odds ratio than women, holding the other variables in the model constant. It is obvious that this way of analyzing logistic regression results is not intuitive and straightforward as in the classical linear regression model. However, the use of average marginal effects derived from the estimated parameters does allow a direct interpretation, as is done in the classic linear regression model, but in terms of probabilities when the explanatory variable in question changes by one unit.

On the other hand, the work by Korkulu and Akan (2024) aimed to identify the factors influencing the demand for subsidized (low-interest) agricultural loans among agricultural producers in the province of Erzurum, Turkey, using the binary logistic regression approach.

The said research determined the predicted probabilities of the variables that define the demand for credit in that country; although the present research estimated the probability of default on the loan received by rural borrowers, in both works, the use of the average marginal effects of the parameters of the logistic model directly provided the predicted probability of the demand for loan and the probability of default on the loan, respectively.

For example, Korkulu and Akan (2024) found that producers located near the center of one of the ten districts in which they conducted their survey had an 80% probability of applying for credit, whereas those farther away only reached a 60% probability. This 20% difference in predicted probability, derived from a single change in geographic location, illustrates how marginal effects can offer a more direct and useful reading for decision-making.

For example, the present research found that for each additional percentage point in the interest rate, the probability of default increases by 21.5%. The apparently unrelated results for Turkey and Mexico in the previous paragraph provide empirical support for this research proposal: access to rural credit is determined not by the producer’s technical efficiency but by an institutional architecture that filters applicants based on their geographical location, level of formalization, and documentary traceability. Using logit regression for agricultural loans in China, Yin et al . (2020) also reported a significant effect of the base interest rate on default, a result consistent with the findings of the present research.

In this way, it can be said that exclusion in the granting of rural credit manifests not only in explicit rejection but also in invisible filters that condition access through institutional design.

Additionally, in Mexico, climate risk acts as a structural factor that directly affects default, since according to the results of the logistic model, a change in the category of the type of risk, such as moving from drought to pests, increases the probability of default by 2.23%. This effect reveals that the financial system does not absorb natural risk but instead transfers it to the producer, also making it a criterion of institutional exclusion beyond technical performance.

Korkulu and Akan (2024) also found that producers without formal collateral have an 18.7% lower probability of applying for loans than those with documentary support, even though both groups express the real need for financing.

This finding provides empirical support for the fact that Turkey’s financial system filters access not by productive viability but by the level of formalization. Finally, it is necessary to highlight that the vast majority of studies that address the analysis of financial exclusion, the structural penalization of climate risk, territorial segmentation in access to credit, and documentary informality as institutional barriers, all articulated by a financial architecture that reproduces inequalities under criteria unrelated to technical efficiency, do so through qualitative or descriptive analytical analysis, as is done in Muñoz-Rodríguez et al . (2002) ; FAO (2020) .

In Amanullah and Chana (2020) ; Nyebar et al . (2023) following a similar approach, it is noted that in rural economies with imperfect markets, access to credit will be conditioned by structural factors such as the lack of collateral, productive informality, and institutional segmentation. These studies point out that the poorest households face severe restrictions in accessing financing, which will perpetuate cycles of low investment and productive vulnerability, and conclude that exclusion does not respond to technical efficiency but to systemic failures in the rural financial architecture.

Conclusions

Based on the results of the logistics model estimated in this research, it is possible to assert that they confirm that rural financing in Mexico presents structural conditions that go beyond the producer’s mere technical performance and efficiency. Variables such as climate risk, document formalization, and territorial location operate as institutional filters that penalize conditions external to the productive subject.

In particular, the marginal effect of climate risk (2.23%) and the exclusion due to lack of collateral (up to 18.7% in comparative studies) show that inequalities are reproduced in the rural financial system under exclusionary financial architecture. It is possible to point out that territorial segmentation and the absence of compensatory mechanisms (for example, complementary guarantee funds, subsidized agricultural insurance, income stabilization funds, among others) reinforce this logic, turning credit into a privilege conditioned by criteria of formalization and location rather than by productive viability.

Based on these findings, it is possible to generalize that, in rural contexts with incomplete markets, the institutional design of financing functions as a filtering mechanism rather than an inclusion mechanism, and that the prevailing financial architecture neither assesses productive capacities nor cushions structural vulnerabilities; rather, it reproduces patterns of exclusion that penalize producers for their environment rather than for their performance.

In this way, it is concluded that any rural financial inclusion policy must be based on recognizing these structural distortions, and that any regulatory and operational reconfiguration of the system must incorporate criteria of institutional co-responsibility, mechanisms for hedging against climate risk, and evaluation models that recognize the territorial and productive diversity of the Mexican rural sector. Only in this way can credit become an instrument of inclusion and not a mechanism for systematic exclusion from access to financing for rural productive activity.

Bibliography

1 

Amanullah, L. G. and Channa, S. 2020. Credit constraints and rural farmers’ welfare in an agrarian economy. Heliyon. 6(10):1-10. https://doi.org/10.1016/j.heliyon.2020.e05252.

2 

Armendáriz, B. and Morduch, J. 2010. The economics of microfinance. 2nd. Edition. MIT Press. 67-1356 pp.

3 

Assairh, L.; Kaicer, M. and Jerry, M. 2021. Credit risk assessment in microfinance institutions through scoring: Moroccan case. International Journal of Accounting, Finance, Auditing, Management and Economics. 2(5):667-679. https://www.ijafame.org/index.php/home/article/view/191.

4 

Beck, T.; Demirgüç, K., A. and Martínez, P. M. S. 2009. Banking services for everyone? Barriers to bank access and use around the world. The World Bank Economic Review. 22(3):397-430. https://doi.org/10.1093/wber/lhn020.

5 

Cassimon, D.; Engelen, P. J.; Reyntjens, S. and Van Campenhout, B. 2022. Determinants of and barriers to people’s financial inclusion in Mexico. OECD Economics Department Working Papers. OECD Publishing, Paris. No. 1728. 5-30 pp. https://doi.org/10.1787/73e9341b-en.

6 

CNBV. 2025. Comisión Nacional Bancaria y de Valores. Portafolio de información. Gobierno de México. https://portafolioinfo.cnbv.gob.mx/Paginas/Inicio.aspx.

7 

De la Vega-Mena, M.; Santoyo-Cortés, V. H.; Muñoz-Rodríguez, M. y Altamirano-Cárdenas, J. R. 2014. Cobertura financiera de la banca de desarrollo para el sector rural de México. FIRA-Financiera Rural. Estudios Sociales. 22(44):137-164. https://www.scielo.org.mx/pdf/estsoc/v22n44/v22n44a9.pdf.

8 

Del Ángel-Mobarak, G. 2003. El Estado y los servicios financieros al medio rural. Análisis a través del caso Fideicomisos Instituidos en Relación con la Agricultura. Revista Estudios Agrarios. 17(46):75-109. https://biblat.unam.mx/es/revista/estudios-agrarios/6.

9 

De-Olloqui, F. y Fernández-Díez, M. C. 2017. Financiamiento del sector agroalimentario y desarrollo rural. Documento para discusión No. IDB-DP-512. Banco Interamericano de Desarrollo. https://publications.iadb.org/publications/spanish/document/Financiamiento-del-sector-agroalimentario-y-desarrollo-rural.pdf.

10 

FAO. 2020. Food and Agriculture Organization. Mejores prácticas del financiamiento agrícola. Rural finance and investment learning centre. https://www.rfilc.org/wp-content/uploads/2020/08/1126195018429-better-practices-in-agricultural-lending--es.pd.

11 

FIRA. 2025a. Fideicomisos instituidos en relación con la agricultura. FIRA en cifras. https://www.fira.gob.mx/nd/firaencifras.jsp.

12 

FIRA. 2025b. Fideicomisos instituidos en relación con la agricultura buscador de riesgos mínimos. https://www.fira.gob.mx/Nd/BuscadorRiesgoMinimo.jsp.

13 

Franjo, L.; Pouokam, N. and Turino, F. 2022. Financial frictions and firm informality: A general equilibrium perspective. The Economic Journal 132(645):1790-1823. https://doi.org/10.1093/ej/ueac010.

14 

IMF. 2022. International Monetary Fund Mexico: financial sector assessment program- financial system stability assessment. IMF Country Report No. 2022/335. https://doi.org/10.5089/9798400223686.002.

15 

Khan, F. U.; Nouman, M.; Negrut, L.; Abban, J.; Cismas, L. M. and Siddiqi, M. F. 2024. Constraints to agricultural finance in underdeveloped and developing countries: a systematic literature review. International Journal of Agricultural Sustainability. 22(1):1-24. https://doi.org/10.1080/14735903.2024.2329388.

16 

Korkulu, A. and Akan, Y. 2024. Determining factors affecting agricultural credit demand: A research in Erzurum province, Türkiye. Journal of Agricultural Sciences. 30(4):712-724. https://doi.org/10.15832/ankutbd.1439139.

17 

Ledgerwood, J.; Earne, J. and Nelson, C. 2013. The new micro-finance handbook: A financial market system perspective. Washington, DC. World Bank Publications. https://doi.org/10.1596/978-0-8213-9750-5.

18 

Madero-Salmerón, B. L. y Núñez-Mora, J. A. 2025. Estimación de la tasa objetivo a través de un modelo econométrico empírico y su aplicación para estimar la curva de la tasa de interés. Panorama Económico. 17(35):45-68. https://doi.org/10.29201/peipn.v17i35.92.

19 

Meki, M. and Quinn, S. 2024. Microfinance: an overview. Oxford Review of Economic Policy. 40(1):1-35. https://doi.org/10.1093/oxrep/grad030.

20 

Menard, S. 2000. Coefficients of determination for multiple logistic regression analysis. The American Statistician, 54(1):17-24. https://doi.org/10.1080/00031305.2000.10474502.

21 

Muñoz-Rodríguez, R. M.; Santoyo-Cortés, V. H. y Altamirano-Cárdenas, J. R. 2002. Mercados e instituciones financieras rurales: una nueva arquitectura financiera rural para México. Universidad Autónoma Chapingo (UACH). CIESTAAM. https://repositorio.chapingo.edu.mx/bitstreams/99ec5360-1331-4f89-9104-677687725df7/download.

22 

Nyebar, A.; Obalade, A. A. and Muzindutsi, P. F. 2023. Effectiveness of credit risks management policies used by Ghanaian commercial banks in agricultural financing. In financial sector development in Ghana: exploring bank stability, financing models, and development challenges for sustainable financial markets Cham. Springer International Publishing. 231-264 pp.

23 

WB. 2021. World Bank. Expanding financial access for Mexico’s poor and supporting economic sustainability. Results Brief. https://www.worldbank.org/en/results/2021/04/09/expanding-financial-access-for-mexico-s-poor-and-supporting-economic-sustainability.

24 

Yin, Z.; Meng, L. and Sha, Y. 2020. Determinants of agriculture-related loan default: Evidence from China. Bulletin of Monetary Economics and Banking, 23(0):5-27. https://doi.org/10.21098/bemp.v23i0.1160.

25 

Yu, L.; Song, Y.; Wu, H. and Shi, H. 2023. Credit constraint, interlinked insurance and credit contract and farmers’ adoption of innovative seeds-field experiment of the loess plateau. Land MDPI. 12(2):1-28. https://doi.org/10.3390/land12020357.