This study designs and develops an interactive credit scoring model for Iran’s leasing industry to rigorously measure the joint effect of contract duration and repayment rate on customers’ credit risk. The motivation stems from a critical gap in mainstream credit risk literature: prevailing approaches—ranging from statistical frameworks such as logistic regression and classical linear models to scorecards and machine-learning methods—typically emphasize the individual effects of predictors while overlooking their structural interactions. In practice, however, leasing credit decisions are shaped by the simultaneous configuration of contract length and repayment intensity. Adopting a localized, comparative perspective vis-à-vis established global models, this research addresses that gap by explicitly modeling the multiplicative interaction between duration and repayment rate.
The study is applied and quantitative. It analyzes a dataset of 1,136 active leasing contracts drawn from Iran Khodro Leasing, Parsian Leasing, Mellat Leasing, and Karafarin Leasing over 2019–2023 (1398–1402). After rigorous data cleaning, validation, and variable encoding, credit risk was operationalized via default incidence and repayment behavior metrics. The empirical strategy relies on a modified multiple regression in which, besides the main effects of contract duration and repayment rate, the interaction term Duration×RepaymentRate captures the joint influence of these variables. Appropriate controls for contract features and customer characteristics were included to mitigate conventional estimation biases and enhance model interpretability
The findings show a positive relationship between contract duration and credit risk, indicating that longer terms significantly amplify the likelihood of risky repayment behavior. Conversely, the repayment rate exhibits a negative relationship with credit risk, implying that more intensive payment schedules reduce the odds of default. Crucially, the interaction between duration and repayment rate is significantly negative (β₄ = −0.192, Sig = 0.001). The model explains approximately 61% of the variance in credit risk, underscoring strong explanatory power. This pattern clearly indicates that in longer-term contracts, setting a high repayment rate not only offsets the risk-increasing effect of extended duration but also delivers a substantial risk reduction. In other words, the combined “length × intensity” effect is non-linear and policy-sensitive, and interactive specifications materially outperform models that evaluate contractual variables in isolation, improving default prediction accuracy.
From a managerial standpoint, the results advise leasing companies to co-optimize contract duration and repayment rate when designing credit products, pricing risk, and allocating financial resources. Particularly for long-term contracts, implementing robust repayment schedules (complemented by behavioral incentives and penalty structures) can markedly improve the risk profile. We further recommend embedding interaction-aware components into internal scorecards and revising credit assessment policies to reflect these joint effects. In terms of contribution, the proposed interactive model fills a notable gap in the Iranian literature, offering a transferable, data-driven framework that elevates predictive accuracy and strengthens credit decision-making in real operational settings. Finally, future research should extend the model by incorporating behavioral covariates, conducting stability tests across sub-markets, and benchmarking against advanced machine-learning algorithms to assess generalizability and performance at scale.