Statistical, Econometric and State Space Models for Time and Savings Deposits of Households
Statistical, Econometric and State Space Models for Time and Savings Deposits of Households

Summary/Abstract: This article gives an applied and simple presentation of the time series models of the amount of the time and savings deposits of households useful in forecasting its data. The amount of the time and savings deposits of households is a very important variable which can be used by commercial banks for long-terms credits, which has a great importance for the restructuring of the enterprise sphere in our conditions. The article contains three sections. The first section presents the classical addi-tive statistical model in which the various components in time series data are de-fined: trend – the downward movement of the amount of the time and savings deposits of households data over the time of four years, seasonal variations, a pat-tern of change in the data that completes itself within a calendar year and then is repeated on a yearly basis, and the error, the erratic movements in the data that have no definable pattern. In the second section we consider the concept of mul-tiple regression causal (econometric) model in which time, seasonal factor and other time series variables (the independent variables) are included to explain the behaviour of the time and savings deposits of households time series. In the third section the „state-space“ (or Markovian [9]) representation of the relationship between input variables is used. Especially, we consider the Kalman solution. We derive the Kalman filter as a structural model based on so called structural time series models where the trend and seasonal components are explicitly designed. A simple interpretation of the Kalman one-step recursion procedures (recursive filter, smoothing and prediction) is presented....

• Details
• Contents