Yazar "Altinbas, Hazar" seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim Yok
    Öğe
    Forecasting and Evaluation of Non-Performing Loans in the Turkish Banking Sector
    (Istanbul Univ, Sch Business, 2023) Altinbas, Hazar; Hanisoglu, Gulay Selvi
    In recent years, there is an increasing trend in non-performing loan levels in Turkey which causes stress both on the real and financial sectors. Increasing non-performing loan volumes are an indication of problems in sectors or the general economy. It is also closely related with the stability of the banking system. It is therefore important for regulatory/ supervisory institutions and banks to be able to predict problematic loan levels successfully, for better policy making and management. For this purpose, non-performing loans to credit ratio in Turkey for the dates between the first quarter of 2015 and fourth quarter of 2019 were forecasted with two machine learning methods, namely random forests and boosted trees, by using data starting from the first quarter of 2003. Lagged values of several macroeconomic, bank-specific and uncertainty factors are included as determinant variables in the analyses. Methods provide insight about the relationship of included variables with non-performing loans. Our results indicate partial dependencies and positive relationship between non-performing loans and inflation, interest rate and capital adequacy ratios, and negative relationship with credit to gross domestic product ratio.
  • Küçük Resim Yok
    Öğe
    Public debt forecasts and machine learning: the Italian case
    (Emerald Group Publishing Ltd, 2023) Sica, Edgardo; Altinbas, Hazar; Marini, Gaetano Gabriele
    PurposePublic debt forecasts represent a key policy issue. Many methodologies have been employed to predict debt sustainability, including dynamic stochastic general equilibrium models, the stock flow consistent method, the structural vector autoregressive model and, more recently, the neuro-fuzzy method. Despite their widespread application in the empirical literature, all of these approaches exhibit shortcomings that limit their utility. The present research adopts a different approach to public debt forecasts, that is, the random forest, an ensemble of machine learning.Design/methodology/approachUsing quarterly observations over the period 2000-2021, the present research tests the reliability of the random forest technique for forecasting the Italian public debt.FindingsThe results show the large predictive power of this method to forecast debt-to-GDP fluctuations, with no need to model the underlying structure of the economy.Originality/valueCompared to other methodologies, the random forest method has a predictive capacity that is granted by the algorithm itself. The use of repeated learning, training and validation stages provides well-defined parameters that are not conditional to strong theoretical restrictions This allows to overcome the shortcomings arising from the traditional techniques which are generally adopted in the empirical literature to forecast public debt.