Econometric Modeling and Forecasting of Interest Rates in Montenegro

Abstract

In contemporary conditions, the econometric modeling and forecast have a growing importance in both the development of several theoretic models and approaches that may be used and in the necessity of forecasting to make proper decisions by authorities for their making. The paper tests the possibility of applying the Box-Jenkins approach and vector autoregressive models for modeling and forecasting interest rates in Montenegro. Box-Jenkins approach and vector autoregressive models are one of multiple, yet the most used approaches and models used for forecasting time series values. Thee comparison of forecasting models determines the more superior model. The time series of the interest rate to be modeled and forecasted is a monthly weighted average lending interest rate of banks on new loans in the period from December 2011 to January 2018.
 An example of the interest rate, which is extremely important in quite a bank-centric system in Montenegro, proved that the Box-Jenkins approach and VAR models may be used successfully for modeling and forecast. Moreover, the paper recommends the use of the Box-Jenkins approach and the assessed AR model for forecasting interest rate since it has better forecasting performances than the VAR model. Despite numerous limitations, primarily the inadequate statistical base, the AR model may find its application and help the decision-makers in the process of making economic decisions.
 
 Keywords: forecasting, autoregressive models, interest rate forecasting, Box-Jenkins, VAR, AR.