Periklis Gogas | Democritus University of Thrace (original) (raw)

Papers by Periklis Gogas

Research paper thumbnail of Testing Purchasing Power Parity Theory with the Hurst Exponent: The Case of 23 OECD Countries

International Finance eJournal, 2012

We test the validity of Purchasing Power Parity theory, examining the Real Exchange Rate of 23 OE... more We test the validity of Purchasing Power Parity theory, examining the Real Exchange Rate of 23 OECD countries for mean-reversion. In doing so, we estimate the Hurst exponent which is a well-established estimator of long memory in time series analysis. The innovation of our approach is that we employ the Detrended Fluctuation Analysis (DFA) for the estimation of Hurst on Real Exchange Rates bothin the full sample and also in rolling windows of three different sizes in an attempt to identify possible trends, breaks and the evolution of Hurst through time.

Research paper thumbnail of Income Inequality: A State-by-State Complex Network Analysis

SSRN Electronic Journal, 2015

Research paper thumbnail of Analyzing the Co-Movements of the US Gross State Product Growth with the Use of the Minimum Dominating Set

SSRN Electronic Journal, 2014

ABSTRACT Graph Theory and network analysis have received great attention lately in the process of... more ABSTRACT Graph Theory and network analysis have received great attention lately in the process of analyzing complex economic systems. Here we propose the use of the Minimum Dominating Set concept in order to identify a representing set of nodes which can describe the collective behavior of an entire network. We apply this technique in US Gross State Product (GSP) growth for the period 1997-2010. The correlation matrix is used to describe the co-movements of GSP growth. Our results show that the GSP of the whole 51 states network can be represented by the values of just 8 states.

Research paper thumbnail of The Slope and the Curvature of the Yield Curve in Recession Forecasting

SSRN Electronic Journal, 2014

ABSTRACT In this paper, we investigate the ability of two popular models to forecast the deviatio... more ABSTRACT In this paper, we investigate the ability of two popular models to forecast the deviation of GDP from its long-run trend, i.e. inflationary and output gaps. In doing so, we exploit the information provided by the yield curve that is documented in the literature as a good predictor of economic activity. We combine and train our forecasting model using interest rates from Treasury Bills and Government Bond rates for the period 1976Q3 to 2011Q4, in conjunction with the quarterly real seasonally adjusted GDP for the same period. Our results show that we can achieve an overall forecasting accuracy of 80% on out-of-sample data. However, our main focus in this paper is to construct a forecasting model for the recessions. Perfect accuracy in recession forecasting is achieved in more than one of the created models. The forecasting performance of our model strengthens the conviction that the yield curve can be a useful and accurate predictive tool.

Research paper thumbnail of Analysis of Network Topology Using the Threshold-Minimum Dominating Set

SSRN Electronic Journal, 2014

Graph theory is an essential tool for the modeling of complex networks. It can be used for networ... more Graph theory is an essential tool for the modeling of complex networks. It can be used for network control and monitoring. The latter is usually performed using a representative subset of the whole network identified with the Minimum Spanning Tree (MST) methodology. The MST, however, bears intrinsic limitations since it may lead to the formation of an inappropriate set of nodes that may withhold crucial information. In this study we unveil these limitations and propose the use of a novel methodology, the Threshold-Minimum Dominating Set (T-MDS), to derive a minimum subset of nodes that can adequately and efficiently describe the collective behavior of an entire network. We present two diverse empirical examples regarding a) a banking network of 200 U.S. banks and b) the GDP growth rates based network of 22 European countries to illustrate the advantages of the T-MDS methodology in the analysis of complex economic systems.

Research paper thumbnail of Asymmetric Effects of Monetary Policy in the U.S. and Brazil

SSRN Electronic Journal, 2013

Research paper thumbnail of Directional Forecasting in Financial Time Series Using Support Vector Machines: The USD/EURO Exchange Rate

SSRN Electronic Journal, 2011

Research paper thumbnail of Common Stochastic Trends and the Ricardian Equivalence in the OECD

SSRN Electronic Journal, 2014

The recent ceiling of U.S. federal debt and the European sovereign debt crises raised once again ... more The recent ceiling of U.S. federal debt and the European sovereign debt crises raised once again the interest upon balanced government budgets. The Ricardian Equivalence proposition appears as an attractive alternative for policy makers, since postponing taxes to be paid once growth is restored seems a very efficient scheme that relieves public discomfort. This paper attempts to investigate the long-run relationship between public debt and private consumption in order to test for the potential validity of the Ricardian equivalence proposition. We use a wide dataset of fifteen OECD countries using annual and quarterly data for the period 1980-2010. For the empirical estimation we employ both a univariate time series and a panel cointegration approach. Our empirical findings fail to provide empirical evidence in support of the Ricardian equivalence proposition for all countries of the sample, since the assumptions proposed by theory cannot be fulfilled.

Research paper thumbnail of A Novel Banking Supervision Method Using a Threshold-Minimum Dominating Set

SSRN Electronic Journal, 2014

ABSTRACT The magnitude of the recent financial crisis, which started from the U.S. and expanded i... more ABSTRACT The magnitude of the recent financial crisis, which started from the U.S. and expanded in Europe, change the perspective on banking supervision. The recent consensus is that to preserve a healthy and stable banking network, the monitoring of all financial institutions should be under a single regulator, the Central Bank. In this paper we study the interrelations of banking institutions under the framework of Complex Networks. Specifically, our goal is to provide an auxiliary early warning system for the banking system's supervisor that would be used in addition to the existing schemes of control. We employ the Minimum Dominating Set (MDS) methodology to reveal the most strategically important banks of the banking network and use them as alarm triggers. By monitoring the MDS subset the regulators can have an overview of the whole network. Our dataset is formed from the 200 largest American banks and we examine their interconnection through their total deposits. The MDS concept is applied for the first time in this setting and the results show that it may be an essential supplementary tool to the arsenal of a Central Bank.

Research paper thumbnail of Convergence of European Business Cycles: Evidence from a Graph Theory-Based Model

SSRN Electronic Journal, 2014

ABSTRACT This paper examines the co-movement patterns of European business cycles during the peri... more ABSTRACT This paper examines the co-movement patterns of European business cycles during the period 1986-2011, having as a focal point the year of the euro coin introduction, in 1999. We work within a Graph Theory context and apply a rolling window to study the evolution of the network that corresponds to the GDP growth cross-correlations of 22 European economies. The network is analyzed using the metrics of node degree and network density as well as the Minimum Dominating Set, providing us not only with a quantitative but also with a qualitative insight of the studied data set. Our main empirical results indicate that despite some distinct signs of divergence, the business cycles of the European countries display overall increased synchronization throughout the selected time sample.

Research paper thumbnail of A Novel Banking Supervision Method Using the Minimum Dominating Set

SSRN Electronic Journal, 2014

Research paper thumbnail of Forecasting the NOK/USD Exchange Rate with Machine Learning Techniques

SSRN Electronic Journal, 2014

Research paper thumbnail of Asynchronous Business Cycles in the E.U.: The Effect of the Common Currency

SSRN Electronic Journal, 2012

Research paper thumbnail of Forecasting the U.S. Real House Price Index

SSRN Electronic Journal, 2014

Research paper thumbnail of Fiscal shocks and asymmetric effects: a comparative analysis

We empirically test the effects of unanticipated fiscal policy shocks on the growth rate and the ... more We empirically test the effects of unanticipated fiscal policy shocks on the growth rate and the cyclical component of real private output and reveal different types of asymmetries in fiscal policy implementation. The data used are quarterly U.S. observati ons over the period 1967:1 to 2011:4. In doing so, we use both a vector autoregressive and the novel support vector machines systems in order to extract the fiscal policy shocks series. The latter has never been used before in a similar macroeconomic setting. Within our research framework, in order to test the robustness of our results to alternative aggregate money supply definitions we use two alternative moentary aggregates. These are the commonly reported by central banks and policy makers simple sum monetary aggregates at the MZM level of aggregation and the alternative CFS Divisia MZM aggregate. From each of these four systems we extracted four types of shocks: a negative and a positive government spending shock and a negati...

Research paper thumbnail of M P RA The North American natural gas liquids markets are chaotic

Research paper thumbnail of Emerging Methodologies in Economics and Finance

THE TEAM. Our research team is operates within the Department of Economics, Democritus University... more THE TEAM. Our research team is operates within the Department of Economics, Democritus University of Thrace, Greece. Our research efforts are funded by a Research Grant from the European Union (Research Funding Program THALES) under the title “Study and Forecasting of Economic Data Using Machine Learning”, MIS 380292. MEMBERS. The research team is led by associate professors Periklis Gogas an economist (B.A., M.A., Ph.D.) and Theophilos Papadimitriou a mathematician (B.A.) and electrical engineer (M.Sc., Ph.D). Seven Ph.D. candidates and nine Master’s students are actively working for the team. RESEARCH INTERESTS. Our team’s interests include both classic and emerging methodologies of Econometrics as they are applied to Economics and Finance. We currently work with: a) Machine Learning: Support Vector Machines for Classification and Regression and Deep Learning Architectures and b) Complex Networks: Threshold – Minimum Dominating Set, Weighted Dominating Set, and Multivariate Networks.

Research paper thumbnail of Data for: Forecasting energy markets using support vector machines

Abstract of associated article: In this paper we investigate the efficiency of a support vector m... more Abstract of associated article: In this paper we investigate the efficiency of a support vector machine (SVM)-based forecasting model for the next-day directional change of electricity prices. We first adjust the best autoregressive SVM model and then we enhance it with various related variables. The system is tested on the daily Phelix index of the German and Austrian control area of the European Energy Exchange (ΕΕΧ) wholesale electricity market. The forecast accuracy we achieved is 76.12% over a 200day period.

Research paper thumbnail of Oil Market Efficiency under a Machine Learning Perspective

Forecasting

Forecasting commodities and especially oil prices have attracted significant research interest, o... more Forecasting commodities and especially oil prices have attracted significant research interest, often concluding that oil prices are not easy to forecast and implying an efficient market. In this paper, we revisit the efficient market hypothesis of the oil market, attempting to forecast the West Texas Intermediate oil prices under a machine learning framework. In doing so, we compile a dataset of 38 potential explanatory variables that are often used in the relevant literature. Next, through a selection process, we build forecasting models that use past oil prices, refined oil products and exchange rates as independent variables. Our empirical findings suggest that the Support Vector Machines (SVM) model coupled with the non-linear Radial Basis Function kernel outperforms the linear SVM and the traditional logistic regression (LOGIT) models. Moreover, we provide evidence that points to the rejection of even the weak form of efficiency in the oil market.

Research paper thumbnail of Money Neutrality, Monetary Aggregates and Machine Learning

Algorithms

The issue of whether or not money affects real economic activity (money neutrality) has attracted... more The issue of whether or not money affects real economic activity (money neutrality) has attracted significant empirical attention over the last five decades. If money is neutral even in the short-run, then monetary policy is ineffective and its role limited. If money matters, it will be able to forecast real economic activity. In this study, we test the traditional simple sum monetary aggregates that are commonly used by central banks all over the world and also the theoretically correct Divisia monetary aggregates proposed by the Barnett Critique (Chrystal and MacDonald, 1994; Belongia and Ireland, 2014), both in three levels of aggregation: M1, M2, and M3. We use them to directionally forecast the Eurocoin index: A monthly index that measures the growth rate of the euro area GDP. The data span from January 2001 to June 2018. The forecasting methodology we employ is support vector machines (SVM) from the area of machine learning. The empirical results show that: (a) The Divisia mon...

Research paper thumbnail of Testing Purchasing Power Parity Theory with the Hurst Exponent: The Case of 23 OECD Countries

International Finance eJournal, 2012

We test the validity of Purchasing Power Parity theory, examining the Real Exchange Rate of 23 OE... more We test the validity of Purchasing Power Parity theory, examining the Real Exchange Rate of 23 OECD countries for mean-reversion. In doing so, we estimate the Hurst exponent which is a well-established estimator of long memory in time series analysis. The innovation of our approach is that we employ the Detrended Fluctuation Analysis (DFA) for the estimation of Hurst on Real Exchange Rates bothin the full sample and also in rolling windows of three different sizes in an attempt to identify possible trends, breaks and the evolution of Hurst through time.

Research paper thumbnail of Income Inequality: A State-by-State Complex Network Analysis

SSRN Electronic Journal, 2015

Research paper thumbnail of Analyzing the Co-Movements of the US Gross State Product Growth with the Use of the Minimum Dominating Set

SSRN Electronic Journal, 2014

ABSTRACT Graph Theory and network analysis have received great attention lately in the process of... more ABSTRACT Graph Theory and network analysis have received great attention lately in the process of analyzing complex economic systems. Here we propose the use of the Minimum Dominating Set concept in order to identify a representing set of nodes which can describe the collective behavior of an entire network. We apply this technique in US Gross State Product (GSP) growth for the period 1997-2010. The correlation matrix is used to describe the co-movements of GSP growth. Our results show that the GSP of the whole 51 states network can be represented by the values of just 8 states.

Research paper thumbnail of The Slope and the Curvature of the Yield Curve in Recession Forecasting

SSRN Electronic Journal, 2014

ABSTRACT In this paper, we investigate the ability of two popular models to forecast the deviatio... more ABSTRACT In this paper, we investigate the ability of two popular models to forecast the deviation of GDP from its long-run trend, i.e. inflationary and output gaps. In doing so, we exploit the information provided by the yield curve that is documented in the literature as a good predictor of economic activity. We combine and train our forecasting model using interest rates from Treasury Bills and Government Bond rates for the period 1976Q3 to 2011Q4, in conjunction with the quarterly real seasonally adjusted GDP for the same period. Our results show that we can achieve an overall forecasting accuracy of 80% on out-of-sample data. However, our main focus in this paper is to construct a forecasting model for the recessions. Perfect accuracy in recession forecasting is achieved in more than one of the created models. The forecasting performance of our model strengthens the conviction that the yield curve can be a useful and accurate predictive tool.

Research paper thumbnail of Analysis of Network Topology Using the Threshold-Minimum Dominating Set

SSRN Electronic Journal, 2014

Graph theory is an essential tool for the modeling of complex networks. It can be used for networ... more Graph theory is an essential tool for the modeling of complex networks. It can be used for network control and monitoring. The latter is usually performed using a representative subset of the whole network identified with the Minimum Spanning Tree (MST) methodology. The MST, however, bears intrinsic limitations since it may lead to the formation of an inappropriate set of nodes that may withhold crucial information. In this study we unveil these limitations and propose the use of a novel methodology, the Threshold-Minimum Dominating Set (T-MDS), to derive a minimum subset of nodes that can adequately and efficiently describe the collective behavior of an entire network. We present two diverse empirical examples regarding a) a banking network of 200 U.S. banks and b) the GDP growth rates based network of 22 European countries to illustrate the advantages of the T-MDS methodology in the analysis of complex economic systems.

Research paper thumbnail of Asymmetric Effects of Monetary Policy in the U.S. and Brazil

SSRN Electronic Journal, 2013

Research paper thumbnail of Directional Forecasting in Financial Time Series Using Support Vector Machines: The USD/EURO Exchange Rate

SSRN Electronic Journal, 2011

Research paper thumbnail of Common Stochastic Trends and the Ricardian Equivalence in the OECD

SSRN Electronic Journal, 2014

The recent ceiling of U.S. federal debt and the European sovereign debt crises raised once again ... more The recent ceiling of U.S. federal debt and the European sovereign debt crises raised once again the interest upon balanced government budgets. The Ricardian Equivalence proposition appears as an attractive alternative for policy makers, since postponing taxes to be paid once growth is restored seems a very efficient scheme that relieves public discomfort. This paper attempts to investigate the long-run relationship between public debt and private consumption in order to test for the potential validity of the Ricardian equivalence proposition. We use a wide dataset of fifteen OECD countries using annual and quarterly data for the period 1980-2010. For the empirical estimation we employ both a univariate time series and a panel cointegration approach. Our empirical findings fail to provide empirical evidence in support of the Ricardian equivalence proposition for all countries of the sample, since the assumptions proposed by theory cannot be fulfilled.

Research paper thumbnail of A Novel Banking Supervision Method Using a Threshold-Minimum Dominating Set

SSRN Electronic Journal, 2014

ABSTRACT The magnitude of the recent financial crisis, which started from the U.S. and expanded i... more ABSTRACT The magnitude of the recent financial crisis, which started from the U.S. and expanded in Europe, change the perspective on banking supervision. The recent consensus is that to preserve a healthy and stable banking network, the monitoring of all financial institutions should be under a single regulator, the Central Bank. In this paper we study the interrelations of banking institutions under the framework of Complex Networks. Specifically, our goal is to provide an auxiliary early warning system for the banking system's supervisor that would be used in addition to the existing schemes of control. We employ the Minimum Dominating Set (MDS) methodology to reveal the most strategically important banks of the banking network and use them as alarm triggers. By monitoring the MDS subset the regulators can have an overview of the whole network. Our dataset is formed from the 200 largest American banks and we examine their interconnection through their total deposits. The MDS concept is applied for the first time in this setting and the results show that it may be an essential supplementary tool to the arsenal of a Central Bank.

Research paper thumbnail of Convergence of European Business Cycles: Evidence from a Graph Theory-Based Model

SSRN Electronic Journal, 2014

ABSTRACT This paper examines the co-movement patterns of European business cycles during the peri... more ABSTRACT This paper examines the co-movement patterns of European business cycles during the period 1986-2011, having as a focal point the year of the euro coin introduction, in 1999. We work within a Graph Theory context and apply a rolling window to study the evolution of the network that corresponds to the GDP growth cross-correlations of 22 European economies. The network is analyzed using the metrics of node degree and network density as well as the Minimum Dominating Set, providing us not only with a quantitative but also with a qualitative insight of the studied data set. Our main empirical results indicate that despite some distinct signs of divergence, the business cycles of the European countries display overall increased synchronization throughout the selected time sample.

Research paper thumbnail of A Novel Banking Supervision Method Using the Minimum Dominating Set

SSRN Electronic Journal, 2014

Research paper thumbnail of Forecasting the NOK/USD Exchange Rate with Machine Learning Techniques

SSRN Electronic Journal, 2014

Research paper thumbnail of Asynchronous Business Cycles in the E.U.: The Effect of the Common Currency

SSRN Electronic Journal, 2012

Research paper thumbnail of Forecasting the U.S. Real House Price Index

SSRN Electronic Journal, 2014

Research paper thumbnail of Fiscal shocks and asymmetric effects: a comparative analysis

We empirically test the effects of unanticipated fiscal policy shocks on the growth rate and the ... more We empirically test the effects of unanticipated fiscal policy shocks on the growth rate and the cyclical component of real private output and reveal different types of asymmetries in fiscal policy implementation. The data used are quarterly U.S. observati ons over the period 1967:1 to 2011:4. In doing so, we use both a vector autoregressive and the novel support vector machines systems in order to extract the fiscal policy shocks series. The latter has never been used before in a similar macroeconomic setting. Within our research framework, in order to test the robustness of our results to alternative aggregate money supply definitions we use two alternative moentary aggregates. These are the commonly reported by central banks and policy makers simple sum monetary aggregates at the MZM level of aggregation and the alternative CFS Divisia MZM aggregate. From each of these four systems we extracted four types of shocks: a negative and a positive government spending shock and a negati...

Research paper thumbnail of M P RA The North American natural gas liquids markets are chaotic

Research paper thumbnail of Emerging Methodologies in Economics and Finance

THE TEAM. Our research team is operates within the Department of Economics, Democritus University... more THE TEAM. Our research team is operates within the Department of Economics, Democritus University of Thrace, Greece. Our research efforts are funded by a Research Grant from the European Union (Research Funding Program THALES) under the title “Study and Forecasting of Economic Data Using Machine Learning”, MIS 380292. MEMBERS. The research team is led by associate professors Periklis Gogas an economist (B.A., M.A., Ph.D.) and Theophilos Papadimitriou a mathematician (B.A.) and electrical engineer (M.Sc., Ph.D). Seven Ph.D. candidates and nine Master’s students are actively working for the team. RESEARCH INTERESTS. Our team’s interests include both classic and emerging methodologies of Econometrics as they are applied to Economics and Finance. We currently work with: a) Machine Learning: Support Vector Machines for Classification and Regression and Deep Learning Architectures and b) Complex Networks: Threshold – Minimum Dominating Set, Weighted Dominating Set, and Multivariate Networks.

Research paper thumbnail of Data for: Forecasting energy markets using support vector machines

Abstract of associated article: In this paper we investigate the efficiency of a support vector m... more Abstract of associated article: In this paper we investigate the efficiency of a support vector machine (SVM)-based forecasting model for the next-day directional change of electricity prices. We first adjust the best autoregressive SVM model and then we enhance it with various related variables. The system is tested on the daily Phelix index of the German and Austrian control area of the European Energy Exchange (ΕΕΧ) wholesale electricity market. The forecast accuracy we achieved is 76.12% over a 200day period.

Research paper thumbnail of Oil Market Efficiency under a Machine Learning Perspective

Forecasting

Forecasting commodities and especially oil prices have attracted significant research interest, o... more Forecasting commodities and especially oil prices have attracted significant research interest, often concluding that oil prices are not easy to forecast and implying an efficient market. In this paper, we revisit the efficient market hypothesis of the oil market, attempting to forecast the West Texas Intermediate oil prices under a machine learning framework. In doing so, we compile a dataset of 38 potential explanatory variables that are often used in the relevant literature. Next, through a selection process, we build forecasting models that use past oil prices, refined oil products and exchange rates as independent variables. Our empirical findings suggest that the Support Vector Machines (SVM) model coupled with the non-linear Radial Basis Function kernel outperforms the linear SVM and the traditional logistic regression (LOGIT) models. Moreover, we provide evidence that points to the rejection of even the weak form of efficiency in the oil market.

Research paper thumbnail of Money Neutrality, Monetary Aggregates and Machine Learning

Algorithms

The issue of whether or not money affects real economic activity (money neutrality) has attracted... more The issue of whether or not money affects real economic activity (money neutrality) has attracted significant empirical attention over the last five decades. If money is neutral even in the short-run, then monetary policy is ineffective and its role limited. If money matters, it will be able to forecast real economic activity. In this study, we test the traditional simple sum monetary aggregates that are commonly used by central banks all over the world and also the theoretically correct Divisia monetary aggregates proposed by the Barnett Critique (Chrystal and MacDonald, 1994; Belongia and Ireland, 2014), both in three levels of aggregation: M1, M2, and M3. We use them to directionally forecast the Eurocoin index: A monthly index that measures the growth rate of the euro area GDP. The data span from January 2001 to June 2018. The forecasting methodology we employ is support vector machines (SVM) from the area of machine learning. The empirical results show that: (a) The Divisia mon...