Analysis of the accuracy of macroeconomic forecasts of selected institutions in the Czech Republic in 2015-2023

The thesis compares macroeconomic forecasts made by selected institutions and analyzes their accuracy, utilizing three statistical metrics: mean absolute error, average forecast error, and Theil's inequality coefficient. This is followed by a point-based evaluation of individual subjects based on the quality of their estimates, depending on the observed macroeconomic variable and prediction horizon. The theoretical part provides basic information on the creation and evaluation of macroeconomic forecasts, which are subsequently utilized in the thesis. Data for the research in the practical part were obtained from the Czech Statistical Office and macroeconomic forecasts of the Ministry of Finance of the Czech Republic, the Czech National Bank, Komerční banka, UniCredit Bank, OECD, and the International Monetary Fund. The results of the thesis confirm hypothesis 2 for some macroeconomic indicators: „In the years of the COVID-19 pandemic and the war in Ukraine, the deviations of forecasts from reality were greater than in the remaining observed years“, and refute hypotheses 1 and 3: „The estimates of the Ministry of Finance and the Czech National Bank as public institutions were on average more accurate than the estimates of other subjects“ and „The Czech National Bank, as an institution targeting inflation, was the most accurate in its prediction“. The thesis addresses a significant research topic and contributes to understanding the differences in forecasts among various institutions, identifying reliable forecasts, and may also serve as a stimulus for improving forecasts and analyses in the future.