This example shows us how annual real GDP growth related with annual inflation rate inversely.

İl Bazında Gayrisafi Yurt İçi Hasıla ($), 2004-2019

The literature shows that the nowcasting models generally use structured data such as real, financial and survey indicators. In recent years, it has been examined to find a way how to use the unstructured data in the nowcasting models. The search items, sentiments, or emotions gathered from internet platforms are used as unstructured data. In this study, it is analysed how ECB presidents\' speeches are included in the nowcasting model and what amount of their effects to the quarterly gross domestic product (GDP) of Germany. First, ECB presidents\' speeches are analysed to obtain the emotion indicators with the help of a newly harmonized complex dictionary. Afterward, these emotion indicators are added to the unbalanced and mixed frequency data, and a nowcasting model estimation for GDP is performed with the data by the expectation-maximization algorithm in the dynamic factor model representation. Moreover, the news analysis is performed to show how the revisions in the real-time data including emotion indicators affect the nowcasts for the current and next quarters of GDP. Finally, a forecast scenario is performed to see the effects of emotion indicators in the nowcasting model to realizations of GDP which shows a slowdown for the last two years. In conclusion, it is suggested that ECB presidents\' speeches may increase the performance of nowcasting models for German GDP.

The use of speech data in nowcasting models is a new topic while the use of sentiment and emotion indicators from microblogs and internet platforms in nowcasting models has been discussed in the literature. I investigate the effect of the speech data of European Central Bank’s (ECB) officials on nowcasting Euro Area GDP in this paper After performing a detailed descriptive analysis of the speech data, I obtain five emotion indicators as a result of the emotion analysis. I examine the contribution of these emotion indicators to a nowcasting model including indicators from the real sector and household/business surveys related to the Euro Area for the period of 1995:01-2019:12. I analyze the effects of emotion indicators on model root mean squared error (RMSE), impulse-response functions, variance decomposition analysis and revision analysis. Findings show that emotion indicators provide a decrease in RMSE of nowcasting model. It is found out that the shocks in the emotion indicators are significant on the GDP in the long term, and the emotion indicators are effective in explaining the variance of the forecast error variance of GDP. Revision analysis indicates that emotion indicators do not increase the revision of GDP nowcasts. As a result, it can be claimed that the emotion indicators obtained from the speeches of ECB officials have a noticeable effect on the nowcasting the Euro Area GDP.

The relationship between consumer prices and producer prices has been discussed in the literature. In this study, the size and existence of the relationship in question for Turkey are examined between the Consumer Price Index (CPI), Producer Price Index Domestic (D-PPI) and Agricultural Products Producer Price Index (AGRICULTURE-PPI). The stationarity characteristics of the price indices are examined using five different tests in parallel with the literature at first In the study in which the period between January 2011 and June 2020 is analyzed. The cointegration relationship between the indices is investigated with the Johansen,test Johansen test with level shift and Luthkepol test with level shift cointegration test which are based on the vector autoregressive (VAR) model, and the short-term dynamics of the relationship are analyzed with the Vector Error Correction (VEC) model. Unlike the literature, the long-term relationship between indices has been tried to be estimated by using two-stage least squares (2SLS), three-step least squares (3SLS) and seemingly unrelated regression (SUR) estimation methods. The findings indicate that the CPI, D-PPI and AGRICULTURE-PPI are integrated at the I(1) level and that there is statistically significant cointegration relationship between them, as well as short-term causality relationships. Considering the long-term relationship coefficients estimated by 2SLS, 3SLS and SUR, it is determined that the effect from AGRICULTURE-PPI and D-PPI to CPI is statistically significant in the long run but small in size. On the other hand, the effects from the CPI to AGRICULTURE-PPI and D-PPI are significant and bigger in size. In light of these findings , it can be claimed that a two-way relationship between producer prices and consumer prices are exist in Turkey for the years between 2011-2020, but consumer prices have stronger effect on producer prices.

The relationship between household and real sector expectations and other economic indicators has been frequently discussed in the literature. On the contrary, the existence and dimensions of the relationship between household expectations and real sector expectations are investigated between the sub-indices of Consumer Confidence Index (TGE) and the Real Sector Confidence Index (RKGE) in this study. Unlike the literature, the DCC-GARCH method, which estimates time-varying (dynamic) conditional correlations, is used in the study in which the period between January 2007-April 2020 is covered. Another difference of the study from the literature is that it tries to measure the effects of the USD/TL exchange rate and the COVID-19 outbreak on the dynamic relationship by panel data analysis. The findings revealed that the relationship between TGE and RKGE sub-indices do exists in general manner, and there are statistically significant relationship between the TGE sub-indices which represent the expectations of the household regarding the general economy and unemployment, and the RKGE sub-indices which represent the expectations of the real sector regarding the general course of the economy and order amounts. The panel data analysis findings indicate that USD/TL exchange rate adversely affects the relationship between TGE and RKGE sub-indices, whereas COVID-19 outbreak does not affect the relationship.

In this study, it is attempted to test for Turkish economy the validity of Ramsey-Cass-Koopmans (RCK) economic growth model which analyse the balance between production, consumption and capital, and which is the starting point of growth theory. The panel dataset prepared for the institutional sectors (financial and non-financial companies, household and government) and covering the period of 2009-2018 is analysed in the study. The existence of production, consumption and capital optimization is investigated using the ARDL error correction model specification. New critical values are calculated with the bootstrap approach and statistical significance is evaluated according to these new critical values in order to ensure the statistical significance of short- and long-term elasticity coefficients obtained in panel ARDL estimation. The short- and long-term elasticity coefficients estimates obtained from OLS, fixed effects and random effects estimators are also included in the study in order to compare the estimation results. Although the standard OLS, fixed and random effects approaches all show the opposite direction, it is concluded that the RCK economic growth model is not valid for the analysis period in Turkey according to the results of the Panel ARDL approach evaluated with bootstrapped critical values. This result implies that the institutional sectors do not have an economic structure that harmoniously balances their production, consumption and capital growth in the analysis period.

The synchronization of the GDP growth cycles between countries have been discussed in the literature. The existence and dimensions of synchronization between the GDP growth cycles of the OECD countries are examined in this study. The existence of synchronization is revealed by estimating the time-varying (dynamic) conditional correlations used in measuring the level of synchronization in two stages by VAR-DCC-MGARCH methods in the study in which the period between 1961-2019 is analyzed. Unlike the literature, K-Means cluster analysis is applied to examine the differentiation of dynamic conditional correlations with respect to member countries. In addition, the effects of economic crises on dynamic conditional correlations are measured by panel data analysis by clusters. The findings showed that there is a synchronization among the member countries and the synchronization levels differ according to the countries. According to the synchronization levels, the member countries are determined to be divided into two clusters and it is observed that the countries in Europe and the Americas concentrated in one cluster. Finally, it is determined that the 1974 oil and 2009 global economic crises reduced the synchronization between the member countries while the 1963 European and 1997 Asian crises increased synchronization in the analysis period. It is observed that the effects of economic crises on synchronization are similar in direction but different in size by clusters.

Nowadays, the socio-economic effects of the COVID-19 outbreak have been discussing in the literature. In this study, the effects of the COVID-19 outbreak on the card payments in Turkey is examined in the sectoral detail. The effects of COVID-19 outbreak on synchronization between sectors are analyzed for the period between 07-03-2014 and 24-04-2020 using weekly data. Firstly, the time-varying (dynamic) conditional correlations used in measuring the level of synchronization are estimated by AR-DCC-GARCH method and the existence of sectoral synchronization among card payments is revealed. The effects of COVID-19 outbreak on this synchronization are tried to be measured with the help of regression model estimation with the assumption of four different types of effects. The findings are that there is a synchronization in terms of card payments across many sectors in Turkey and levels of synchronization are differs by sector. Regarding the effect of COVID-19 outbreak by sectors, it is determined that the most common effect is “permanent,” and “first panic then deepening” and “first panic then permanent” effects are also dominant. It is observed that the market and shopping sector differentiated from other sectors during the COVID-19 outbreak period, whereas the level of synchronization of all other sectors, particularly the casino / distillery, food and airline sectors, increased, in other words, shared the downward trend.

This paper aims to examine whether CBRT’s statements are affected by ECB and FED statements. The sentiment analysis is performed to build the balance sentiment indicators (BSI) for CBRT, ECB, and FED from 2008 January to 2020 August using by their statements. Individual and group properties of BSI’s are examined by unit root tests and the Bounds cointegration tests. The short and long-run effects of ECB and FED sentiments on CBRT’s sentiments are investigated by the ARDL and UECM models. The analysis applied over the full sample period and two sub-samples which they represent the periods before and after 2013 June. The results imply that the sentiment indicators of CBRT, ECB, and FED have statistically significant cointegration relationship over the full sample period and the period after 2013 June. Particularly after June 2013, the CBRT's statements are positively related with the ECB's and FED's statements in the long run however CBRT's statements are negatively affected by FED's statements in the short run.

The transmission between consumer prices and producer prices have been discussed in the literature. In this study, the existence and dimensions of the transmission in question are examined between the Consumer Price Index (CPI) and the Domestic Producer Price Index (D-PPI) sub-indices. The analysis period is determined between 2005 February - 2020 April. Unlike the literature, the VAR-DCC-GARCH method, which predicts the time-varying (dynamic) conditional correlations, has been used in determining the presence of the transmission and measuring its level. Another difference of the study from the literature is that K-Means cluster analysis is applied to the average transmission coefficients in order to examine the differentiation in the transmission coefficients among sub-indices. Finally, the effects of USD /TL exchange rate on the dynamic transmission coefficients among the sub-indices are measured with the dynamic panel ARDL error correction model. The findings points out that there is a valid transmission between the CPI and D-PPI sub-indices, and the most of sub-indices shows statistically significant transmission. The cluster analysis shows that the CPI and D-PPI sub-indices are divided into two clusters, and the CPI sub-indices which are related to service prices and all D-PPI sub-indices are concentrated in a cluster. USD/TL exchange rate is found to have a statistically significant and positive effect on the transmission between the CPI and D-PPI sub-indices. The dynamic panel model estimation results indicate that USD/TL exchange rate has an impact on pass-through in both short and long term for both clusters.