Financial Literacy and Loan Repayment Performance in Informal Savings Groups: Evidence From Kericho Community Development Trust, Kenya

The global microfinance industry has witnessed exponential growth over the past three decades, with informal savings and credit associations emerging as critical vehicles for financial inclusion in sub-Saharan Africa (K​a​b​u​g​i​ ​&​ ​K​a​r​i​u​k​i​,​ ​2​0​2​0; M​a​g​a​n​g​a​,​ ​2​0​2​0). In Kenya, the accumulated savings and credit association (ASCA) model has gained particular traction among rural and peri-urban communities excluded from formal banking services. Unlike rotating savings and credit associations (ROSCAs), ASCAs accumulate pooled savings over defined cycles, extend loans to members at interest and distribute accumulated profits through share-out mechanisms at cycle end (C​h​i​n​e​k​a​ ​&​ ​M​t​e​t​w​a​,​ ​2​0​2​1; Z​h​a​o​ ​&​ ​L​i​,​ ​2​0​2​1).

Informal savings groups confront a fundamental sustainability challenge: loan delinquency notwithstanding their proliferation. The Kericho Community Development Trust (KCDT) financial report of 2022 documented a portfolio-at-risk (PAR) exceeding 30 days of 40.10%, with total outstanding arrears of Ksh. 4,355,349 (K​e​r​i​c​h​o​ ​C​o​m​m​u​n​i​t​y​ ​D​e​v​e​l​o​p​m​e​n​t​ ​T​r​u​s​t​,​ ​2​0​2​2). Such elevated default rates erode group capital, constrain future lending capacity and undermine member confidence, which ultimately threaten institutional survival.

Financial literacy is the combined knowledge, skills and attitudes enabling informed financial decision-making (G​a​i​t​a​n​-​A​n​g​u​l​o​ ​e​t​ ​a​l​.​,​ ​2​0​2​6). A growing body of literature identifies financial literacy as a critical determinant of loan repayment performance (A​l​i​ ​e​t​ ​a​l​.​,​ ​2​0​2​1; H​a​s​a​n​ ​e​t​ ​a​l​.​,​ ​2​0​2​1). Financially literate borrowers are better equipped to assess loan affordability, manage cash flows and prioritize debt obligations (D​e​z​s​o​,​ ​2​0​2​1; T​h​o​m​a​s​ ​&​ ​G​u​p​t​a​,​ ​2​0​2​1). Yet, empirical evidence on the financial literacy–repayment nexus remains sparse within the context of informal savings groups in East Africa.

While the ASCA model has expanded credit access for financially excluded populations in Kericho County, loan repayment performance remains distressingly poor. The KCDT recorded 45% of loans in arrears and a PAR of Ksh. 4,355,349 in 2022, signalling systemic repayment failure. Existing literature attributes this crisis to inadequate group security, weak enforcement mechanisms and economic shocks. However, the role of borrower-level financial literacy in explaining repayment behaviour within informal savings groups remains underexplored.

Specifically, it is unclear whether members’ understanding of interest calculations, loan terms and budgeting principles influences their propensity to default. If financial ignorance drives repayment failure, then literacy interventions may offer a cost-effective remedy. Conversely, if literacy has no bearing on repayment, then structural factors such as group governance or macroeconomic conditions demand priority attention. This ambiguity constitutes the research gap that this study addresses.

The study sought to examine the relationship between financial literacy and loan repayment performance among members of the KCDT, Kenya.

(1) To determine the level of financial literacy among ASCA members in KCDT.

(2) To assess the loan repayment performance of ASCA members in KCDT.

(3) To examine the relationship between financial literacy and loan repayment performance among ASCA members in KCDT.

Ho₁: There is no statistically significant relationship between financial literacy and loan repayment performance among ASCA members in KCDT, Kenya.

Ho₂: Financial literacy does not significantly predict loan repayment performance among ASCA members in KCDT, Kenya.

This study contributes to the financial inclusion discourse by shifting focus from credit supply to borrower demand-side capabilities. If financial literacy enhances repayment, then literacy investments become economically rational for group sustainability. The findings inform policymakers on whether to mandate financial education as a precondition for ASCA membership or loan eligibility. For practitioners, the study offers evidence-based guidance on designing literacy curricula tailored to informal group contexts. For scholars, the paper extends the financial literacy–debt management literature to an under-researched institutional setting.

The study holds implications for multiple stakeholders. KCDT administrators can use the findings to design pre-loan financial literacy screening and targeted training for high-risk borrowers. The Central Bank of Kenya and the Microfinance Act regulator may consider incorporating financial literacy requirements into ASCA accreditation standards. Development partners financing community savings initiatives can justify literacy programme budgets through demonstrated repayment impacts. Finally, the study enriches the academic literature on behavioural microfinance in informal settings.

According to B​e​c​k​e​r​ ​(​1​9​6​4​), human capital theory posits that investments in knowledge and skills enhance individual productivity and economic outcomes. Applied to financial behaviour, financial literacy represents a form of human capital that improves decision-making quality (L​u​s​a​r​d​i​ ​&​ ​M​i​t​c​h​e​l​l​,​ ​2​0​1​4). Borrowers with superior financial human capital are hypothesized to make better loan choices, manage repayment obligations more effectively and avoid over-indebtedness; thereby reducing default probability.

According to B​a​n​d​u​r​a​ ​(​1​9​8​6​), social cognitive theory emphasizes self-efficacy; belief in one’s capability to execute behaviours as a determinant of action. A​l​i​ ​e​t​ ​a​l​.​ ​(​2​0​2​1​) demonstrated that financial self-efficacy, cultivated through literacy and socialization, directly influences financial coping behaviours and empowerment. In the ASCA context, members with high financial self-efficacy may exhibit greater confidence in managing loan repayments, seeking extensions when necessary and communicating proactively with group leadership during distress.

According to A​k​e​r​l​o​f​ ​(​1​9​7​0​), market-for-lemons framework highlights how information asymmetries between lenders and borrowers distort credit markets. Financially illiterate borrowers may misrepresent their repayment capacity or misunderstand loan terms, exacerbating adverse selection and moral hazard. Financial literacy mitigates these asymmetries by enabling borrowers to accurately assess their own creditworthiness and comprehend contractual obligations.

H​a​s​a​n​ ​e​t​ ​a​l​.​ ​(​2​0​2​1​) examined how financial literacy impacts inclusive finance in Bangladesh, finding that literacy significantly enhances participation in formal financial services and improves debt management practices. The study employed household surveys and structural equation modelling, concluding that literacy interventions should accompany credit expansion. However, the study focused on formal banking contexts, leaving informal group dynamics unexplored.

A​l​i​ ​e​t​ ​a​l​.​ ​(​2​0​2​1​) proposed a conceptual model examining financial literacy, financial socialization and financial self-efficacy as antecedents of financial empowerment among Saudi women. Using partial least squares path modelling on 1,368 respondents, the study found significantly positive associations between literacy, coping behaviour and well-being. The positive role of saving as a financial self-efficacy mechanism was highlighted. Nonetheless, the study’s exclusive focus on women limits generalizability to mixed-gender ASCA groups.

T​h​o​m​a​s​ ​&​ ​G​u​p​t​a​ ​(​2​0​2​1​) synthesized social capital theory, social exchange theory and social cognitive theory to develop a conceptual framework linking financial literacy to financial well-being through knowledge sharing as a moderator. Their bibliometric analysis revealed that financial literacy serves as a foundational enabler of informed decision-making across diverse cultural contexts. The framework suggests that literacy effects may be amplified in group settings where peer learning and social monitoring operate.

I​n​d​r​i​a​n​i​ ​e​t​ ​a​l​.​ ​(​2​0​2​3​) examined group lending management in Indonesia, finding that social capital strengthened group lending control dynamics to increase repayment capacity. While not explicitly measuring literacy, the study implied that informed group members—through training and experience—exhibited superior repayment outcomes. The research adopted moderating regression analysis on 97 group respondents in Sragen Regency.

P​r​a​s​t​y​a​n​i​n​g​t​y​a​s​ ​e​t​ ​a​l​.​ ​(​2​0​2​2​) investigated group loan phenomena in Nganjuk, East Java, Indonesia, where women borrowers frequently defaulted due to poor financial planning and multiple borrowing. The study recommended women’s empowerment activities such as financial literacy training as mechanisms to anticipate and minimize group loan risk. The findings suggest that literacy deficits directly contribute to repayment failure in informal group lending contexts.

M​e​n​g​s​t​i​e​ ​(​2​0​2​2​) examined the impact of microfinance on women’s economic empowerment in Ethiopia, using multiple regression on 346 women clients. The study found that education level, share-out amount and number of training sessions significantly affected economic empowerment, while previous business experience did not. Paired sample t-tests revealed significant mean differences in income, assets and savings before and after share-outs. The study’s reliance on paired t-tests rather than predictive modelling limits causal inference.

C​h​i​n​e​k​a​ ​&​ ​M​t​e​t​w​a​ ​(​2​0​2​1​) analysed savings and credit schemes in Zimbabwe, establishing that loaning schemes empowered members to survive but recommended obligatory rules to safeguard institutional stability. The study employed purposive sampling of 10 contributors; a methodological limitation that the current study addresses through probabilistic sampling.

M​a​g​a​n​g​a​ ​(​2​0​2​0​) explored village savings and loan associations in Malawi, finding positive contributions to women’s financial empowerment but failing to address literacy as a mediating variable. The descriptive design and small sample (70 women) restrict generalizability.

M​u​s​t​o​n​e​n​ ​e​t​ ​a​l​.​ ​(​2​0​2​5​) investigated microfinance loan impacts on women entrepreneurs in Kenya, finding significant empowerment effects but neglecting literacy and repayment challenges. The study did not address high-interest rates or over-indebtedness risks, gaps the current study partially fills.

The study proposes that financial literacy, which is the independent variable, comprises four distinct dimensions of knowledge of interest calculations, budgeting, loan terms and risk assessment. The study proposes that financial literacy operates as a composite construct that influences loan repayment performance, which is the dependent variable. Loan repayment performance is measured through on-time repayment ratio, default frequency and PAR contribution. The relationship is moderated by group social capital (trust, peer monitoring) and mediated by financial self-efficacy.

The study adopted a correlational research design and cross-sectional descriptive survey. Correlational design was deemed appropriate for establishing the relationship between financial literacy and repayment performance (B​h​a​n​d​a​r​i​,​ ​2​0​2​2). The cross-sectional approach captured data at a single point, providing a snapshot of literacy–repayment dynamics without longitudinal tracking.

The study was conducted in Kericho County, Kenya, focusing on the KCDT and surrounding areas. Kericho’s agricultural economy—dominated by tea farming—creates seasonal income fluctuations that test borrowers’ repayment capacity, making it an ideal setting for examining literacy effects.

The target population comprised 380 members across 14 active KCDT groups. The formula proposed by K​r​e​j​c​i​e​ ​&​ ​M​o​r​g​a​n​ ​(​1​9​7​0​) yielded a sample of 191 members, distributed proportionally across groups using simple random sampling. Of 191 questionnaires distributed, 135 were returned valid (70.7% response rate).

Primary data were collected using a semi-structured questionnaire with four sections: (i) demographic and group characteristics; (ii) financial literacy assessment (15 items measuring interest calculation, budgeting, loan term comprehension and risk awareness); (iii) loan repayment behaviour (10 items measuring repayment timeliness, default history and communication with group leadership); and (iv) financial empowerment indicators. Secondary data on PAR and loan performance were extracted from KCDT financial records.

Content validity was established through expert review while a Construct Validity Index (CVI) of 0.75 adopted. Internal consistency was tested using Cronbach’s Alpha, with α ≥ 0.70 deemed satisfactory.

Descriptive statistics (frequencies, percentages, means, standard deviations) summarized literacy levels and repayment performance. Inferential statistics comprised Pearson correlation and simple linear regression. Prior to interpreting the regression coefficients, standard diagnostic procedures were conducted to verify ordinary least squares (OLS) assumptions. These included the Shapiro-Wilk test for residual normality, the Breusch-Pagan test for homoscedasticity and the Durbin-Watson statistic for autocorrelation. Residual-versus-fitted plots were also examined to assess linearity and independence of errors. Hypothesis testing used α = 0.05. The regression model was specified as:

Y = α + β₁X₁ + ε

where,

Y = Loan repayment performance;

X₁ = Financial literacy;

α = Constant;

β₁ = Coefficient of financial literacy;

ε = Error term.

The 135 respondents comprised 56.3% male and 43.7% female. The majority (34.8%) were aged 51–61 years, followed by 40–50 years and over 61 years (23.7% each). Group tenure varied: 20.7% had existed for 3 years, 17.8% for 5 years and 14.8% for 1 year. Service focus was predominantly loans and savings combined (43.0%) or loans alone (42.2%).

As shown in Table 1, members demonstrated the strongest budgeting skills (mean = 3.68) but the weakest risk awareness (mean = 3.21), suggesting that while members can manage routine finances, they struggle to anticipate and mitigate borrowing risks.

Descriptive analysis revealed moderate financial literacy among members. On a 5-point scale:

Table 2 shows that the secondary data confirmed poor portfolio health: 40.10% PAR > 30 days and total compulsory savings of Ksh. 15,822,927 with zero voluntary savings, indicating constrained financial flexibility.

Repayment performance was assessed through self-reported behaviour and secondary PAR data:

As shown in Table 3, financial literacy demonstrated a moderate positive and statistically significant correlation with repayment performance (r = 0.412, p < 0.001). This suggests that members with higher literacy scores exhibit better repayment discipline.

Pearson correlation analysis revealed:

Table 4 shows that the model yielded an R² of 0.170, indicating that financial literacy accounted for 17.0% of the variance in loan repayment performance.

The analysis of variance as shown in Table 5 shows that the calculated F(1, 133) = 26.647 at p < 0.001; was greater than the F-critical. This implies that the model was statistically significant.

As shown in Table 6, the unstandardized coefficient (B = 0.471) indicates that for every one-unit increase in financial literacy, repayment performance improves by 0.471 units. The standardized coefficient (β = 0.412) confirms literacy as a meaningful predictor.

It is essential to verify that the fitted model satisfies the standard assumptions underlying ordinary least squares inference before presenting the regression estimates. Table 7 summarizes the diagnostic checks conducted on the residuals.

As seen in Table 7, the Shapiro-Wilk test confirmed that residuals were approximately normally distributed (W = 0.987, p = 0.156). The Breusch-Pagan test indicated no significant heteroscedasticity (χ² = 2.134, p = 0.144), supporting the assumption of constant error variance. The Durbin-Watson statistic (d = 1.902) fell within the acceptable range (1.5–2.5), suggesting no serious autocorrelation. Visual inspection of residual-versus-fitted plots displayed a random scatter around zero with no discernible funnel pattern, confirming linearity and homoscedasticity assumptions. Collectively, these diagnostics support the validity of the regression inference.

Ho₁: There is no statistically significant relationship between financial literacy and loan repayment performance.

Decision: r = 0.412, p < 0.001. Since p < 0.05, reject Ho₁. There is a statistically significant positive relationship between financial literacy and loan repayment performance.

Ho₂: Financial literacy does not significantly predict loan repayment performance.

Decision: β = 0.412, p = 0.000. Since p < 0.05, reject Ho₂. Financial literacy significantly predicts loan repayment performance.

The findings corroborate human capital theory: financial literacy constitutes a productive investment that enhances borrower decision-making and repayment capacity. The moderate correlation (r = 0.412) suggests that while literacy matters, it is not the sole determinant of repayment since group dynamics, economic conditions and loan terms also exert influence.

While the study modelled financial literacy as a composite construct, the descriptive statistics suggest that the four dimensions may exert differential effects on repayment behaviour. Members demonstrated strongest budgeting skills (mean = 3.68) but weakest risk awareness (mean = 3.21), with loan term understanding falling in between (mean = 3.55). This pattern implies that risk awareness may be a more critical deficit than budgeting competence in explaining default propensity, a conjecture consistent with A​d​e​k​u​n​l​e​ ​(​2​0​2​6​), who found that borrowers’ comprehension of loan terms and risk implications was a stronger predictor of repayment behaviour than basic budgeting knowledge in Nigerian microfinance banks. The current aggregate model does not permit direct comparison of dimension-specific effects; however, the descriptive gaps suggest that future research should employ multiple regression with disaggregated literacy dimensions to test whether interest calculation comprehension, budgeting proficiency, loan term understanding and risk awareness exert unique or differential effects on repayment outcomes. Such disaggregation would inform targeted curriculum design that prioritizes the weakest-performing dimensions rather than delivering uniform literacy content.

The findings from Kericho County converge with emerging evidence from comparable informal savings group contexts across sub-Saharan Africa and South Asia. In Nigeria, A​d​e​k​u​n​l​e​ ​(​2​0​2​6​) documented that financial literacy explained 36.9% of variance in loan repayment behaviour among microfinance bank clients, a substantially higher explanatory power than the 17.0% observed in the present study, possibly reflecting the more formalized regulatory environment and structured literacy programmes in Nigerian Microfinance institutions (MFIs). In South Africa, M​s​o​m​i​ ​&​ ​N​z​a​m​a​ ​(​2​0​2​2​) similarly found that financial literacy significantly influenced small and medium enterprises (SME) loan repayments during the COVID-19 era, underscoring the cross-contextual stability of the literacy–repayment nexus. Recent work in Cambodia by S​a​m​r​e​t​h​ ​e​t​ ​a​l​.​ ​(​2​0​2​5​) further corroborates that financial literacy among microfinance borrowers enhances repayment discipline, while O​s​u​m​a​ ​e​t​ ​a​l​.​ ​(​2​0​2​5​) demonstrated that microfinance and financial inclusion interventions reduce vulnerability across sub-Saharan Africa. These cross-national parallels indicate that the literacy–repayment relationship observed in KCDT is not idiosyncratic to Kericho but reflects broader challenges facing informal and semi-formal financial intermediaries in agrarian and transition economies. Consequently, the policy recommendations advanced here may have transferable relevance for ASCA-type institutions operating in Tanzania, Uganda and Rwanda, where similar seasonal income volatility and low financial literacy prevail.

The 17.0% explained variance aligns with prior studies acknowledging that repayment behaviour is multifactorial. I​n​d​r​i​a​n​i​ ​e​t​ ​a​l​.​ ​(​2​0​2​3​) similarly found social capital moderates lending outcomes, while P​r​a​s​t​y​a​n​i​n​g​t​y​a​s​ ​e​t​ ​a​l​.​ ​(​2​0​2​2​) identified multiple borrowing and poor planning as literacy-related default drivers. Notably, members scored lowest on risk awareness (mean = 3.21), suggesting that literacy programmes should prioritize risk education to help members assess affordability, anticipate income shocks and understand compound interest implications. The absence of voluntary savings (Ksh. 0 reported) further signals constrained financial buffers, exacerbating vulnerability to default.

The finding that 40.10% of loans are at risk in spite of moderate literacy levels implies that literacy alone is insufficient without complementary interventions: flexible repayment schedules, emergency credit lines and group-level insurance mechanisms.

Financial literacy among KCDT members is moderate (mean = 3.47), with strongest budgeting skills but weakest risk awareness. Loan repayment performance is suboptimal, with 40.10% PAR > 30 days and high self-reported default frequency. Financial literacy has a statistically significant positive relationship with loan repayment performance (r = 0.412, p < 0.001). Financial literacy significantly predicts repayment performance, explaining 17.0% of variance (β = 0.412, p < 0.001).

The study concludes that financial literacy is a significant but partial determinant of loan repayment performance in informal savings groups. While literacy enhances borrowers’ capacity to understand obligations and manage repayments, structural factors, including group governance, economic volatility, and savings flexibility, also demand attention. The ASCA model in Kericho County cannot achieve sustainable lending without embedding financial education into group operations. Literacy is necessary but not sufficient; it must be complemented by institutional reforms that address the 40.10% PAR crisis.

The study advances two categories of recommendations: (a) empirically-derived recommendations grounded directly in the statistical findings reported in Sections 4.4 and 4.5 and (b) broader management recommendations (reflecting theoretical frameworks and institutional best practice beyond direct regression findings) informed by the theoretical framework and institutional best-practice literature, extending beyond the direct regression results.

(1) Mandatory Pre-Loan Financial Literacy Screening. KCDT should administer brief literacy assessments before loan approval, with remedial training for low-scoring applicants. The regression model demonstrated that financial literacy significantly predicts repayment performance (β = 0.412, p < 0.001); screening operationalizes this predictive relationship by identifying high-risk borrowers prior to disbursement.

(2) Risk Awareness Prioritization in Training Curricula. Group training should emphasize risk anticipation, income-shock planning, over-indebtedness avoidance, and emergency-fund creation. Risk awareness was the weakest literacy dimension (mean = 3.21, SD = 1.18) and the aggregate literacy score was significantly correlated with repayment (r = 0.412, p < 0.001), suggesting that strengthening the weakest dimension should yield disproportionate repayment gains.

(3) Voluntary Savings Incentivization. The trust should introduce incentives for voluntary savings beyond compulsory contributions. Secondary data revealed zero voluntary savings alongside a 40.10% PAR > 30 days, indicating that borrowers lack financial buffers to absorb repayment shocks.

(1) Integrated Financial Education Modules. Group meetings should incorporate 15-minute literacy sessions covering interest calculations, budgeting and risk management. This recommendation draws on Human Capital Theory (B​e​c​k​e​r​,​ ​1​9​6​4) and Social Cognitive Theory (B​a​n​d​u​r​a​,​ ​1​9​8​6), which posit that sustained knowledge investment and self-efficacy cultivation improve financial behaviour, rather than on a specific direct regression coefficient.

(2) Post-Default Literacy Counselling. Members who default should receive targeted financial counselling rather than punitive exclusion. Grounded in the conceptual framework linking financial self-efficacy to coping behaviour (A​l​i​ ​e​t​ ​a​l​.​,​ ​2​0​2​1), this proposal reflects rehabilitative management practice rather than a direct empirical result from the present regression model.

(3) Policy Advocacy for ASCA Accreditation Standards. The Central Bank of Kenya and Ministry of Cooperatives should mandate financial literacy requirements in ASCA accreditation standards. Rationale: This reflects broader governance implications suggested by Asymmetric Information Theory (A​k​e​r​l​o​f​,​ ​1​9​7​0) and the institutional sustainability literature, extending beyond the sample-specific regression findings.