Loan Evaluation Research Articles

Application of artificial neural network to loan recovery prediction

PurposeThis study assesses the classification accuracy of an artificial neural network (ANN) model. It examines the application of loan recovery probability rather than odds of default as the case with traditional credit evaluation models.Design/methodology/approachData on 2,300 loans granted over the period 2001-2012 was obtained from the databases of Nigerian commercial banks and primary mortgage institutions. A multilayer feed-forward ANN model with back-propagation learning algorithm was developed having classified the sample into training (38 per cent), testing (41 per cent) and validation (21 per cent) sub-samples.FindingsThe model exhibits a high overall percentage classification accuracy of 92.6 per cent. It also achieves relatively low misclassification Type I and Type II errors at 6.5 per cent and 8.2 per cent, respectively. Macroeconomic variables such as gross domestic product, inflation and interest rates have the strongest influence on the ANN model classification power. The result of the analysis shows that adopting odds of recovery in ANN classification models can lead to improved loan evaluation.Originality/valueThe paper is distinct from extant studies in that it presents a new dimension to loan evaluation in Nigerian lending market. To the best knowledge of the authors, the paper is among the first to explore probability of loan recovery as the basis for credit evaluation in the country.

Fuzzy Logic Based Loan Evaluation System

Retail loans play a key role in the banking of many countries. At the same time loans to individuals are regarded as more risky than business loans. For these reasons, the efficiency of retail credit granting is important for the welfare of both households and of banking system.In this paper a fuzzy logic model for retail loan evaluation is proposed. The fuzzy model consists of five input variables such as “income”, “credit history”, “employment”, “character”, and “collateral condition” and single output variable which indicates credit standing. Whether applicant's credit standing shall be regarded as “low”, “medium” or “high” depends on the degree of membership for the linguistic terms of fuzzy output.

Role of Financial Banks in Promoting the Entrepreneurship: A Mixed Methodology Approach from Kingdom of Saudi Arabia

The issue of getting finances for the small businesses and entrepreneurs is always been in debate and remain unresolved in many countries due to unavailability of qualified venture capitalists. The developing and emerging economies set the micro finance banks for this purpose, however, it is argued that the owner and entrepreneur faces many problems like collaterals, documentation, etc. This research focuses on the role of financial banks in promoting the small business and entrepreneurial culture in the Saudi Arabia in providing credit. The research applied a mixed methodology and at the first stage, qualitative data is collected and then the results of these structured interviews were used to construct a survey questionnaire for the quantitative analysis. The result of study shows that the levels of business cooperation and information sharing and quality of business have an important significance on the success of loan application. Furthermore, the results also support that the bureaucracy of bank in terms of loan documents requirement and loan evaluation procedure can make small business hesitate when applying for loans.

The computation of optimised credit transition matrices

Transition probability matrices (TPMs) indicate the likelihood of obligor credit migration over a specific time horizon. TPMs are used in various risk management applications ranging from loan evaluation, portfolio risk analysis, economic capital assessment and the pricing of financial instruments. The standard methodology is to compute single period TPMs from empirical data using the discrete cohort approach and estimate t period TPMs as the t year product of single period empirical TPMs (exponentiation). This methodology assumes credit migration is Markovian and time homogeneous. When these assumptions do not hold, this methodology results in significant forecasting errors for multi-period transition horizons. In this paper, a framework for generating empirically consistent TPMs using an optimisation methodology is presented. Optimised TPMs produce default term structures that are substantially more accurate in terms of their ability to match empirical observations over multiple time periods than those produced by exponentiating single period empirical TPMs. Additionally, it is found that optimised TPMs show smoother surfaces with consistent probability mass distributions (monotonicity), reduce the impact of the Markov assumption, and reduce discrepancies of credit migration over multiple time horizons.

Revisiting the BOT Structure in Direct Foreign Investment Transactions

Revisiting the BOT Structure in Direct Foreign Investment Transactions

A hybrid algorithm of improved case-based reasoning and multi-attribute decision making in fuzzy environment for investment loan evaluation

This paper presents a model to support the banking managerial decisions in the evaluation of investment plans, especially on rejecting inappropriate plans that can be done in short time (less than hour) and with minimal cost. Because there are some uncertainties in the evaluation process, our proposed model utilises fuzzy set theory to define the problem space in which an acceptance or rejection decision for a submitted investment plan is made. The model is based on lessons-learned concept and developed through the combination of case-based reasoning (CBR) and multiple attribute decision making in fuzzy environment. The model uses an enhanced version of CBR in which a novel concept as solution's truth value is implemented. A set of investment plans is evaluated to show the applicability and efficiency of the model. Different scenarios in terms of sensitivity analysis are also mentioned to capture managerial insights. Comparing the obtained results of the model with those of other algorithms shows its better proximity to human reasoning and decision making.

Generating Transition Probability Matrices through an Optimization Framework

Transition probability matrices (TPMs), which indicate the likelihood of obligor credit migration over a given time horizon, have been used in various credit applications ranging from pricing of financial instruments, loan evaluation, portfolio risk analysis, and economic capital assessment. The standard methodology for calculating TPMs is the discrete, cohort approach. We present an alternative method for calculating TPMs using an optimization methodology. The optimization framework incorporates multiple business requirements, such as: ensuring smooth surfaces with consistent probability mass distributions, reduction of impact from time homogeneity and Markov assumptions, and reduction of forecast errors for multiple time steps. The resultant optimized TPMs are compared with traditional discrete TPMs developed using the cohort method. The proposed optimization method results in TPMs with significantly better predictive power and properties that better suit many business applications.

The evaluation of consumer loans using support vector machines

The commencement of the Basel II requirement, popularization of consumer loans and the intense competition in financial market has increased the awareness of the critical delinquency issue for financial institutions in granting loans to potential applicants. In the past few decades, the scheme of artificial neural networks has been successfully applied to the financial field. Recently, the Support Vector Machine (SVM) has emerged as the better neural network in dealing with classification and forecasting problems due to its superior features of generalization performance and global optimum. This study develops a loan evaluation model using SVM to identify potential applicants for consumer loans. In addition to conducting experiments on performance comparison via cross-validation and paired t test, we analyze misclassification errors in terms of Type I and Type II and their effect on selecting network parameters of SVM. The analysis findings facilitate the development of a useful visual decision-support tool. The experimental results using a real-world data set reveal that SVM surpasses traditional neural network models in generalization performance and visualization via the visual tool, which helps decision makers determine appropriate loan evaluation strategies.

ALEES: an agricultural loan evaluation expert system

Financial institutions such as banks have been active in the development of a variety of expert systems. Several banks have developed expert systems that capture the experience of their top loan officers in evaluating loan applications. The research literature reveals a number of criteria and measures that have been employed by researchers and lenders. While some consensus is apparent for several measures, the research clearly indicates that both quantitative and qualitative information must be employed in evaluating a loan proposal. Further, the research on rural lending has shown that subjective assessments play an important role in the evaluation of agricultural loans, but they have rarely been included as part of an expert system to appraise agricultural loan applications. Several agricultural loan evaluation systems are operational, but not all are ‘true’ expert systems. While they have all focused on analysing the borrower's financial position given the current market and economic conditions, most have not taken into account qualitative factors such as the character of the borrower. The purpose of this paper is to present an expert system that has been specifically designed and developed to incorporate both qualitative and quantitative assessments in agricultural loan evaluation. The development of the expert system is outlined; the knowledge acquisition process is described; system validation is explained, as is the loan officers’ perception of the expert system. An evaluation of several agricultural loan evaluation expert systems is also included in this paper.

Inductive modeling of expert decision making in loan evaluation: a decision strategy perspective

There have been two dominant paradigms in understanding and modeling an expert's decision-making behavior: output analysis and process-tracing. While the two paradigms are complementary, they have not been used yet in a combined manner. This study extends the previous research work in the two paradigms to expert system research by (1) analyzing individual experts' decision strategies, (2) comparing performance of four popular inductive modeling methods, and (3) matching their performance against decision strategy type.

Empirical Bayes analysis of the commercial loan evaluation process

In this paper, logistic regression models containing random effects are explored as a means of simulating the decision process of commercial loan officers. Empirical Bayes estimates, determined with the EM Algorithm, are calculated for the random parameters. Analysis of data from a major Canadian bank is presented.

Information Asymmetries and the Provision of Finance to Small Firms

Dr M. R. Binks and G.V. Reed are with the department of economics, nottingham university, England, and C.T. Ennew with the Department of industrial economics of the same university. Ascertaining the risk associated with the provision of external finance to small firms involves the collection of considerable amounts of external fiance to small firms involves the collection of considrable amounts of inforation. The cost incurred in this process are sufficient to cause both an equity and a debt gap. Mainstream econmic therorists are divided as to whetehr the inforatmion asymmetries concerened cause an under or an over-supply of credit. This paper outlines the main elements of the eimpact of asymmetries concrened cause an under of over supply of credit.idicates the tendency of a cpaital as opposed to an incoming-generting approach to loan evaluation to emerge as a natural default process by United Kingdom banks. Although identifying the anture of asymmetries in practice is very difficult. it is important to make attempts to do so in order to guide and monitor progress in theri removal .the authors explain how a measure of demand, supply and attainment quality, with respect.The result illustrate tehrrelative experience of firms by sector, size and growth rate and reveal the possiblity tha those firms most in need of a prospectus based approach may be least well placed to receive it.

Neural Model Based Approach for Loan Evaluation

In this paper, the authors analyze the loan evaluation problem in India. They give a hierarchical decomposition of this problem which facilitates a functional fit of suitable architectures. After motivating the need for a neural approach, they propose a mixed-network architecture that consists partly of expert-system and partly of hybrid neural net independently and justify the need for such a mixed approach at appropriate places.

Home Ownership for Low Income Families: Key Findings from the Five Year Victorian Capital Indexed Loan Evaluation Study

This paper presents key findings from a recently completed five year evaluation study of the impact on low income families of a Victorian State Housing Authority home loan scheme. The research was funded by the Department of Planning and Housing, Victoria, and carried out by the Australian Institute of Family Studies. The study was designed to test the feasibility of providing families who were either in public housing or on the waiting list with low start, inflation-adjusted, home loans. This paper examines three outcomes: changes in employment patterns, home repairs and improvements, and estimated capital gain. The findings contribute to larger debates within sociology over the meaning and impact of home ownership and the privatisation of housing.

Low income home ownership: The privatisation of squalor or the great Australian dream?

Abstract Owning a home brings with it responsibility for meeting repair and renovation costs. Under a special low‐start loan scheme of the Victorian Ministry of Housing and Construction, many public and private renters are joining the ranks of Australian home owners. Using data gathered from the Australian Institute of Family Studies five year Capital Indexed Loan (CAPIL) Evaluation Study, this paper examines the home repair and renovation activity of a group of Victorian low income families and specifically attempts to distinguish between repairs and home improvements. The findings suggest that while 87 per cent of families involved themselves in some form of home activity in the two year period since moving into their home, home improvements occurred three times more frequently than basic repairs. Only about 17 per cent of the families reported structural, plumbing or heating work, defined as necessary repairs in the analysis. A combination of choice and necessity appeared to determine average expenditu...

Restructuring Home Loans for Low Income Families: Results from the Capital Indexed Loan Scheme

In the 1980s, rising house prices in Australia's major capital cities, coupled with the high cost of housing finance, have effectively excluded many families from purchasing a home. In response, the Victorian Ministry of Housing and Construction, in late 1984, introduced a home loan scheme that provided inflation-adjusted mortgages with repayments set at 25 per cent of family income. Using data from the first two rounds of the Australian Institute of Family Studies five year Capital Indexed Loan (CAPIL) Evaluation Study, this paper examines some of the outcomes of this scheme by comparing families in the Capital Indexed Loan Scheme with a matched group of families in the public and private rental sectors. The findings show the impact of both tenure and family type on economic gains, housing costs and housing satisfaction.

Inductive learning for risk classification

The authors discuss the application of inductive learning to credit risk analysis. Three risk classification problems are addressed: business loan evaluation, bond rating, and bankruptcy prediction. The use of a system called Marble, a knowledge-based decision-support system that uses approximately 80 decision rules to evaluate commercial loans, is discussed. In particular, an aspect of Marble that uses inductive learning to classify financial risks is described, and the effectiveness of the technique is discussed.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Inductive Learning Methods for Knowledge-Based Decision Support: A Comparative Analysis

This paper describes the inductive learning methods for generating decision rules in decision support systems. Three similarity-based learning systems are studied based on: (1) the AQ-Star method, (2) the Tree-Induction method, and (3) the Probabilistic Learning method. Loan evaluation examples and empirical data are used as a basis for comparing these inductive learning methods on their algorithmic characteristics and decision support performance.

LP: AN EXPERT SYSTEM FOR LEASE VERSUS PURCHASE DECISION

Only recently have the applications of expert systems in finance received tremendous attention from not only the practitioners but the academicians as well. Widespread applications of this technology in diverse financial problems include credit management loan evaluation portfolio analysis capital budgeting cash flow analysis, and so forth. Another problem area that can benefit from this technology is the lease versus purchase decision, the joint process of corporate capital budgeting decision and financing decision. Because of the complex nature in the evaluation of the lease versus purchase decision, expert systems are thus appropriate for this type of application.

Applying inductive learning to enhance knowledge-based expert systems

This paper describes the use of inductive learning in MARBLE, a knowledge-based expert system I have developed for assisting business loan evaluation. Inductive learning is the process of inferring classification concepts from raw data; I use this technique to generate loan-granting decision rules based on historical and proforma financial information. A learning method is presented in this paper that can induce decision rules from training examples.