Publication: Bankalarda takipteki krediler: Türk bankacılık sektöründe takipteki kredilerin tahminine yönelik bir model uygulaması
Abstract
Tasarruf açığı bulunan kişilerin çeşitli ihtiyaçlarını karşılamak üzere kredi kuruluşlarından belirli bir maliyetle geri ödenmek üzere aldıkları bir borç olan kredi, vadesi 90 günü geçmesine rağmen kısmen veya tamamen ödenmeyerek takibe düşebilir. Ekonominin genel durumu açısından öncü gösterge niteliği taşıyan takipteki krediler, ekonomide bireylerin ve kurumların ödeme kabiliyetini, bankalarda da aktif kalitesini ve risk düzeyini gösterir. Oranın sağlıklı bir şekilde tahmin edilebilmesi ekonomik birimlerin politikalarını, bankaların da bilançolarını etkin bir şekilde yönetmelerine imkân tanır. Literatürde takipteki kredilerin tahminine yönelik gözleme ve ekonometrik modellere dayalı çeşitli çalışmalar yapılmıştır. Etkin bir risk yönetimi için, her iki yaklaşıma dayalı sistemlerin birlikte kullanılmasının, kontrol mekanizmasının kurulması açısından daha faydalı sonuçlar doğuracağı düşünülebilir. Hâlihazırda Türkiye’ye özgü çalışmalar oldukça sınırlıdır Bu çalışmada Türk bankacılık sektöründe takipteki kredi oranının aylık bazda tahminine yönelik bir model uygulaması sunulmuştur. İstatistikî testler, modelin iyi bir tahmin edici olduğunu göstermiştir. Model, takipteki kredilerin önemli ölçüde stok sorunu olduğuna işaret etmektedir. Diğer bir deyişle belirli bir dönemde takipteki krediler iyi yönetilirse, sonraki dönemlerde ekonomik koşullar bozulsa bile, takipteki krediler artısı görece sınırlı kalacaktır.
Loan, which is a debt borrowed from loan institutions to be paid back with a certain cost by people who have saving gaps to meet their various needs; can fall into monitoring by not being paid partially or completely despite the fact that the maturity of the debt has passed 90 days. Non performing loans that carry a leading indicator quality in terms of the general state of economics, shows the solvency of the individuals and enterprise in economy, and also shows the active quality and risk level in banks. To predict the rate healthily enables the economic units to manage their policies and the banks to manage their financial statements effectively. Various studies are performed in literature, based on observing the non performing loan predictions and econometric models. For an effective risk management, it can be thought that using systems based on both approaches together can lead to more useful results in terms of establishing the control mechanism. Studies unique to Turkey are quite limited at present. In this study, a model implementation intended for monthly based predictions of non performing loans in the sector of Turkish banking, is presented. The statistical tests shows that the model is a good predictor. The model points at the fact that non performing loans have serios stock problems. In other words, if non performing loans in a certain period are managed well, even if economic conditions deteriorate in following periods, increase in non performing loans will be relatively limited.
Loan, which is a debt borrowed from loan institutions to be paid back with a certain cost by people who have saving gaps to meet their various needs; can fall into monitoring by not being paid partially or completely despite the fact that the maturity of the debt has passed 90 days. Non performing loans that carry a leading indicator quality in terms of the general state of economics, shows the solvency of the individuals and enterprise in economy, and also shows the active quality and risk level in banks. To predict the rate healthily enables the economic units to manage their policies and the banks to manage their financial statements effectively. Various studies are performed in literature, based on observing the non performing loan predictions and econometric models. For an effective risk management, it can be thought that using systems based on both approaches together can lead to more useful results in terms of establishing the control mechanism. Studies unique to Turkey are quite limited at present. In this study, a model implementation intended for monthly based predictions of non performing loans in the sector of Turkish banking, is presented. The statistical tests shows that the model is a good predictor. The model points at the fact that non performing loans have serios stock problems. In other words, if non performing loans in a certain period are managed well, even if economic conditions deteriorate in following periods, increase in non performing loans will be relatively limited.