Forecasting German Gross Domestic Product using DAX Iindex Values: MIDAS Analysis
Keywords:MIDAS, Forecasting, GDP, U-MIDAS, DAX
Stock markets are the driving force for the growth and development of national economies. Economic growth is directly proportional to the increase in production capacity. The increase in production capacity, and therefore, economic growth is measured by gross domestic product and gross national product. The frequencies of the variables used in financial time series differ. The frequency of the variables used can be daily, weekly, monthly or quarterly. Data such as gross domestic product is announced quarterly, while stock market data can be daily, monthly or weekly. This was a problem in studies on the relationships between variables with different frequency values. This problem has been solved with the Mixed Data Sampling (MIDAS) method. MIDAS is an effective method used in the empirical analysis of variables with different frequencies. In this study, financial time series with different frequencies were analyzed using the MIDAS method. The monthly data of the German DAX (Deutscher Aktienindex) index and the quarterly data of the German gross domestic product values are taken as variables. GDP values were tried to be estimated using DAX data. The period between 01.01.2000-01.01.2022 is included in the scope of the study. While the GDP series consists of 89 quarters of data, the DAX series consists of 265 data. In the analysis, STEP, U-MIDAS, PDL\ALMON and BETA models were created and their estimation performances were evaluated with RMSE, MAE, MAPE, SMAPE, THEIL U1, THEIL U2. As a result of the evaluation, Unrestricted MIDAS regressions (U-MIDAS) was the method with the best performance.
Aytekin, M. and Emirmahmutoğlu, F. (2021). GDP Flash Estimate with MIDAS and Mixed Frequency VAR Journal of Statistical Research, 11(1), 1-22.
Bai, J., Ghysels, E. and Wright, J. H. (2013). State Space Models and MIDAS Regressions. Econometric Reviews, 32(7), 779-813. doi: 10.1080/07474938.2012.690675
Bonino-Gayoso, N. and Garcia-Hiernaux, A. (2021). TF-MIDAS: a transfer function based mixed-frequency model. Journal of Statistical Computation and Simulation, 91(10), 1980-2017. doi: 10.1080/00949655.2021.1879082
Erdogdu, H. (2020). Forecasting Turkey’s GDP Growth with Mixed Data Sampling MIDAS Method. Igdir University Journal of Social Sciences(22), 519-542.
Foroni, C., Marcellino, M. and Schumacher, C. (2011). U-MIDAS: MIDAS regressions with unrestricted lag polynomials: Deutsche Bundesbank.
Ghysels, E., Hill, J. B. and Motegi, K. (2016). Testing for Granger causality with mixed frequency data. Journal of Econometrics, 192(1), 207-230.
Ghysels, E., Santa-Clara, P. and Valkanov, R. (2004). The MIDAS touch: Mixed data sampling regression models.
Guérin, P. and Marcellino, M. (2013). Markov-Switching MIDAS Models. Journal of Business & Economic Statistics, 31(1), 45-56. doi: 10.1080/07350015.2012.727721
Gul, S. and Kazdal, A. (2021). Nowcasting and Short-Term Forecasting Turkish GDP: Factor-MIDAS Approach. Central bank of the Turkish Republic (Vol. 21/11).
Kazaz, F. and Demireli, E. (2022). The Effect of Indicator Stock Exchange Index (Bist 100) on Macroeconomic Growth in Turkey: Midas Analysis. Gaziantep University Journal of Social Sciences, 21(2), 824-839.
Lamb, S. (2017). Investigation of the effect of oil prices and some macroeconomic variables on a number of indices in Borsa Istanbul. Management and Economics, 24(2), 579.
Marcellino, M. and Schumacher, C. (2010). Factor MIDAS for Nowcasting and Forecasting with Ragged‐Edge Data: A Model Comparison for German GDP. Oxford Bulletin of Economics and Statistics, 72(4), 518-550.
Uca, N., Civelek, M. E. and Çemberci, M. (2015). The Effect Of The Components Of Logıstıcs Performance Index On Gross Domestıc Product: Conceptual Model Proposal. Eurasian Business & Economics Journal, 1.
Yamak, N., Samut, S. and Koçak, S. (2018). Forecasting Economic Growth Rate With Mixedfrequency Data. Journal of Economic Sciences, 10(1), 34-49.