Importance The article reviews the process of forecasting the solvency of the credit institution's customers. To outline an appropriate strategy for handling various loan balances, it is reasonable to devise more accurate methods for forecasting the solvency of borrowers at the loan repayment stage. Objectives The research pursues setting up methods, which would adequately reflect each borrower's tendency of repaying the loan depending on the loan terms, borrower's profile and credit history, and ensuring sufficient accuracy of the borrower's solvency forecast for the following period. Methods We proposed methods for predicting the borrower's solvency through Markov chains of the 1st and 2nd order that allowed considering the previous status of the credit history, and machine learning methods to evaluate probabilities of the credit account changes, including factors that could possibly influence the ultimate solvency. The approach implies a preliminary analysis of the credit history data and recovers missing indicators. Results Based on the methods, we propose our own algorithm for forecasting the borrowers' solvency. The efficiency of the approach is illustrated with an example. As for data of credit accounts, we implemented prediction methods and identified the best ones per each change in the status of the credit account. When evaluating the quality of estimates, we found matrices of differences between the real and projected indicators. Conclusions and Relevance The proposed methods will improve the solvency forecasts for each change in the credit account's status and enable the credit institution to take relevant measures to reduce the anomaly risk
Thomas L.C. A Survey of Credit and Behavioural Scoring: Forecasting Financial Risk of Lending to Consumers. International Journal of Forecasting, 2000, vol. 16, iss. 2, pp. 149–172.
Thomas L.C., Ho J., Scherer W.T. Time will Tell: Behavioral Scoring and the Dynamics of Consumer Credit Assessment. IMA Journal of Management Mathematics, 2001, vol. 12, iss. 1, pp. 89–103.
Grimshaw S., Alexander W. Markov Chain Models for Delinquency: Transition Matrix Estimation and Forecasting. Applied Stochastic Models in Business and Industry, 2011, vol. 27, iss. 3, pp. 267–279.
Timofeeva G.A., Timofeev N.A. Prognozirovanie sostavlyayushchikh kreditnogo portfelya na osnove modeli markovskoi tsepi [Forecasting credit portfolio components with a Markov chain model]. Avtomatika i telemehanika = Automation and Remote Control, 2012, no. 4, pp. 47–65.
Diggle P.J., Liang K.Y., Zeger S.L. Analysis of Longitudinal Data. Oxford, Oxford University Press, 2002.
Klyachkin V.N., Dontsova Yu.S. Sravnitel'nyi analiz tochnosti nelineinykh modelei pri prognozirovanii sostoyaniya sistemy na osnove markovskoi tsepi [A comparative analysis of the accuracy of nonlinear models in forecasting the condition of system on the basis of the Markov chain]. Izvestiya Samarskogo nauchnogo tsentra Rossiiskoi akademii nauk = Proceedings of Samara Scientific Center of RAS, 2013, vol. 15, no. 4, pp. 924–927.
Chizhova A.S. Ekonometricheskaya model' otsenki matrits veroyatnostei perekhoda kreditnykh reitingov [An econometric model for evaluating the probability matrices of changes in credit ratings]. Prikladnaya ekonometrika = Applied Econometrics, 2007, no. 3, pp. 11–26.
Merkov A.B. Raspoznavanie obrazov: vvedenie v metody statisticheskogo obucheniya [Pattern recognition: an introduction into statistical learning methods]. Moscow, URSS Publ., 2010, 254 p.
Aivazyan S.A., Enyukov I.S., Meshalkin L.D. Prikladnaya statistika. Osnovy modelirovaniya i pervichnaya obrabotka dannykh [Applied statistics. Fundamentals of modeling and primary processing of data]. Moscow, Finansy i statistika Publ., 1983, 471 p.
Little R.J.A., Rubin D.B. Statisticheskii analiz dannykh s propuskami [Statistical Analysis with Missing Data]. Moscow, Finansy i statistika Publ., 1990, 336 p.
Grin' N.V. Metodologicheskie aspekty postroeniya skoringovykh modelei. V kn.: Ekonomika, modelirovanie, prognozirovanie. Vypusk 6 [Methodological aspects of setting scoring models. In: Economics, modeling, forecasting. Issue 6]. Minsk, R&D Institute of Ministry of Economy of Republic of Belarus Publ., 2012, pp. 174–180.
Pfeifer P.E., Carraway R.L. Modeling Customer Relationships as Markov Chains. Journal of Interactive Marketing, 2000, vol. 14, iss. 2, pp. 43–55.
Sokolov G.A., Chistyakova N.A. Teoriya veroyatnostei. Upravlyaemye tsepi Markova v ekonomike [Probability theory. Controlled Markov chains in economy]. Moscow, FIZMATLIT Publ., 2005, 248 p.
Wasserman P. Neirokomp'yuternaya tekhnika: teoriya i praktika [Neural Computing: Theory and Practice]. Moscow, Mir Publ., 1992, 184 p.
Vapnik V.N. Vosstanovlenie zavisimostei po empiricheskim dannym [Restoring correlations through empirical data]. Moscow, Nauka Publ., 1979, 448 p.
Vagin V.N., Golovina E.Yu., Zagoryanskaya A.A., Fomina M.V. Dostovernyi i pravdopodobnyi vyvod v intellektual'nykh sistemakh [Credible and plausible inference in intelligent systems]. Moscow, FIZMATLIT Publ., 2004, 704 p.
Thomas L.C., Edelman D.B., Crook J.N. Credit Scoring and its Applications: SIAM Monographs on Mathematical Modeling and Computation. SIAM, Philadelphia, USA, 2002, 248 p.
Alekseeva V.A., Dontsova Yu.S., Klyachkin V.N. Vosstanovlenie propushchennykh nablyudenii pri klassifikatsii ob"ektov [Missing observations recovery during objects classification]. Izvestiya Samarskogo nauchnogo tsentra Rossiiskoi akademii nauk = Proceedings of Samara Scientific Center of Russian Academy of Sciences, 2014, vol. 16, no. 6, pp. 357–359.
