Credit-Delinquency Modelling and the Financial-Cooperative Gap

Credit-Delinquency Modelling and the Financial-Cooperative Gap: History

Please note this is an old version of this entry, which may differ significantly from the current revision.

Subjects: Business, Finance

Contributor: Henry Robert López-Núñez , Luis Marcelo Mantilla-Falcón , Oscar Patricio López-Solis , Alexander Fernando Haro-Sarango , Richard Oswaldo Manzano-Fernández

Credit delinquency is the central prudential risk of savings and credit cooperatives, yet the methods used to model it have been developed and tested mostly on banks, consumer credit and benchmark datasets. This review uses the broader credit-delinquency literature as a benchmark against which the position of financial cooperatives is measured. A curated corpus of 123 documents (1953–2026), assembled by expert selection and citation tracing and verified against bibliographic registries, was analysed with bibliometric laws, natural language processing (TF–IDF, non-negative matrix factorisation, co-word networks, semantic drift and burst detection), a binomial generalised linear model of paradigm diffusion based on manually verified coding, a citation-free contribution index and a contingency test of coverage deficit. Five phases are identified, with the largest semantic displacements at the 1997 and 2016 boundaries (cosine drift 0.642 and 0.567). Explainability diffuses fastest (odds ratio 1.354 per year, 95% CI 1.202–1.526), and the mean number of paradigms per document rises from 0.91 to 2.36. Cooperative research follows a different path: none of the 12 cooperative documents employs ensemble learning or explainability methods (Fisher exact p = 0.0001), and a reproducible search of 65,610 Semantic Scholar records indicates that cooperative credit-risk studies apply these methods at roughly one quarter of the rate observed elsewhere (0.8% versus 3.0%). The periodisation, the rapid diffusion of explainability and the cooperative coverage deficit hold when the incomplete year 2026 is excluded and under alternative filter thresholds, topic numbers, network and burst thresholds and coding procedures. A conceptual five-layer framework and a research agenda are proposed to guide empirical work on explainable delinquency modelling in cooperatives.

credit delinquency
financial cooperatives
credit unions
ensemble learning
explainable artificial intelligence
bibliometrics
natural language processing
non-performing loans
IFRS 9
EU AI Act

1. Introduction and History

Credit delinquency, the failure of a borrower to meet contractual repayment obligations, is simultaneously an accounting event, a prudential indicator and an economic signal. For savings and credit cooperatives it carries an additional meaning. Because a cooperative is owned by the members it lends to, a delinquent loan does not merely erode a shareholder’s return; it consumes the collective patrimony of the borrower’s peers. Delinquency management in cooperatives is therefore closely tied to governance, and predictive technologies deployed on member data raise questions of accountability towards the members who own the institution [1,2].

The methods available to model that risk have changed substantially over the last decade. Univariate ratio analysis [3] gave way to multivariate discriminant analysis [4], then to structural contingent-claim models [5] and probabilistic specifications [6]; credit rationing theory supplied informational microfoundations [7]; behavioural scoring industrialised the estimation of default probabilities [8,9]; ensemble learning challenged single classifiers [10,11,12,13]; and, most recently, explainable artificial intelligence (XAI) has become a central concern for the use of machine-learning models in regulated credit decisions [14,15,16].

Most of this methodological development has taken place in commercial banks, consumer-credit portfolios and public benchmark datasets rather than in financial cooperatives. This review therefore adopts a benchmark design. The broader credit-delinquency literature is treated as the reference population that defines the methodological frontier, and the cooperative literature is treated as the focal population whose position relative to that frontier is measured. The cooperative documents form a minority of the corpus by construction: the question is not what the cooperative literature contains in isolation, but whether it has taken part in the developments that the wider literature records. Cooperative finance has its own monitoring tradition, built around the PEARLS system [17,18], its own empirical literature on growth, diversification and failure [19,20,21,22,23] and its own claim to counter-cyclical resilience [2,24]. Whether that tradition has absorbed ensemble learning and explainability is an empirical question that a quantitative review can address.

1.1. Objectives and Research Questions

The review pursues five objectives, each operationalised as a measurable quantity:

RQ1: What is the growth regime of the credit-delinquency literature, and does its authorship and source structure follow the bibliometric regularities of Lotka, Bradford and Zipf?
RQ2: Can the methodological evolution of the field be partitioned into phases, and how large is the semantic displacement between them?
RQ3: At what rate has each analytical paradigm (structural–statistical, statistical learning, ensemble, explainability and regulatory–institutional) diffused through the literature, and how often are paradigms combined within the same document?
RQ4: Which documents display the strongest textual and structural signals of contribution to the field?
RQ5: Where do financial cooperatives stand relative to this benchmark, and is any coverage deficit robust to the way the corpus was constructed?

1.2. Contributions

The review makes five contributions.

First, it provides an explicit periodisation of the credit-delinquency literature supported by semantic drift between period centroids rather than asserted by inspection.
Second, it proposes a citation-free contribution index (the Textual Contribution Index, TCI) built from semantic novelty, diffusion-weighted influence, lexical generativity and structural bridging, and reports evidence on its stability and construct validity.
Third, it estimates paradigm diffusion with a binomial generalised linear model applied to manually verified paradigm codes, yielding adoption odds ratios, inflection years and transition times.
Fourth, it quantifies the cooperative coverage deficit with a contingency test and a trend test in the curated corpus, and replicates the comparison in a reproducible external search of the wider literature.
Fifth, it proposes a conceptual five-layer framework that links determinant modelling, imbalance-aware ensembles, explanation, governance translation and cooperative accountability, and derives a research agenda from the measured gaps.

This entry is adapted from the peer-reviewed paper https://doi.org/10.3390/encyclopedia6110243

© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.