Konjonktür hareketleri ve reel ekonomi anlık tahmini ve öngörüsü üzerine makaleler

This dissertation consists of two essays about nowcasting and forecasting business cycles and real economy. First essay is joint work with Asst. Prof. Cem Çakmaklı and Prof. Sumru Güler Altuğ and it proposes a unified framework for joint estimation of the indexes that can broadly capture economic and financial conditions together with their cyclical regimes of recession and expansion. We aim to estimate the time when the real economy enters a recession or a expansion in Turkish economy by the proposed framework. Second essay is joint work with Asst. Prof. Cem Çakmaklı and it formulates a nowcasting model by incorporating Survey of Professional Forecasters (SPF) for improving the density nowcasting of gross domestic product growth at monthly frequency using Bayesian methods for U.S. economy. More specifically, first essay propose a method for real-time prediction of recessions using large sets of economic and financial variables with mixed frequencies. This method combines a dynamic factor model for the extraction of economic and financial conditions together with a tailored Markov regime switching specification for capturing their cyclical behavior. Departing from conventional methods estimating a single common cycle governing economic and financial conditions, or extracting economic and financial cycles in isolation of each other, the model allows for a common cycle which is reected with potential phase shifts to the financial conditions estimated alongside with other parameters. This in turn provides timely recession predictions by making efficient modeling of the financial cycle systematically leading the business cycle. We examine the model using a mixed frequency ragged-edge dataset for Turkey in real-time. The results show evidence for the superior predictive power of our specification by signaling oncoming recessions (expansions) as early as 3.6 (3.0) months ahead of the actual realization. In the second essay, we utilize a dynamic factor model with stochastic volatility specification for nowcasting U.S. gross domestic product by using the surveys conducted by the Federal Reserve Bank of Philadelphia. Specifically, our model produces now/forecasts of predictive densities that are aligned with survey expectations at difierent horizons, thereby integrating the predictive content of the survey expectations into the conventional dynamic factor model. We further incorporate a stochastic volatility structure into the baseline model to accommodate the changing volatility of the GDP growth, which also enables us to make use of the disagreement between individual forecasters as a proxy for uncertainty to provide more accurate density nowcasts. We provide results on the accuracy of nowcasts of U.S. GDP growth in a real-time exercise from 1977 through 2017. Comparison over diffierent specifications through predictive likelihoods and probability integral transforms (PIT) reveals the improvements on predictive power of proposed specifications. This is due to the fact that the model adapts to the rapidly changing conditions much faster than the conventional specification thanks to the stochastic volatility structure and exploitation of data provided by SPF.