Konstrukce nestochastických předpovědí ekonomických časových řad

This thesis is devoted to the construction of non-stochastic economic time series forecasts. The secondary objective of the thesis is to analyze the relationship between energy indicators and macroeconomic aggregates. Statistical and econometric models such as moving average processes, autoregressive processes, or a combination of both mentioned processes, exponential smoothing, etc. are used in the thesis in order to create accurate models for predicting the evolution of indicators. This paper examines the cointegration of individual time series and aims to quantify the long- and short-term relationships between indicators using an error correction model. Statistical methods such as bootstrap are used. The thesis thus provides an alternative perspective on the issue of economic forecasting.