30 research outputs found
Modern approaches for nonlinear data analysis of economic and financial time series
No full textThis thesis centers on introducing modern non-linear approaches for data analysis in economics and finance with special attention on business cycles and financial crisis. It is now well stated in the statistical and economic literature that major economic variables
display non-linear behaviour over the different phases of the business cycle. As such, nonlinear approaches/models are required to capture the features of the data generating mechanism of inherently asymmetric realizations, since linear models are incapable of
generating such behavior.
In this respect, the thesis provides an interdisciplinary and open-minded approach to analyzing economic and financial systems in a novel way. The thesis presents approaches that are robust to extreme values, non-stationarity, applicable to both short and long data length, transparent and adaptive to any financial/economic time series. The thesis provides step-by-step procedures in analyzing economic/financial indicators by incor-
porating concepts based on surrogate data method, wavelets, phase space embedding, ’delay vector variance’ (DVV) method and recurrence plots. The thesis also centers on transparent ways of identifying, dating turning points, evaluating impact of economic
and financial crisis. In particular, the thesis also provides a procedure on how to an-ticipate future crisis and the possible impact of such crisis. The thesis shows that the incorporation of these techniques in learning the structure and interactions within and
between economic and financial variables will be very useful in policy-making, since it facilitates the selection of appropriate processing methods, suggested by the data itself.
In addition, a novel procedure to test for linearity and unit root in a nonlinear framework is proposed by introducing a new model – the MT-STAR model – which has similar properties of the ESTAR model but reduces the effects of the identification problem and
can also account for asymmetry in the adjustment mechanism towards equilibrium. The asymptotic distributions of the proposed unit root test is non-standard and is derived.
The power of the test is evaluated through a simulation study and some empirical illustrations on real exchange rates show its accuracy. Finally, the thesis defines a
multivariate Self–Exciting Threshold Autoregressive with eXogenous input (MSETARX) models and present an estimation procedure for the parameters. The modeling procedure
for the MSETARX models and problems of estimation are briefly considered
Turning point chronology for the Euro-Zone: A Distance Plot Approach
No full textWe propose a transparent way of establishing a turning point chronology for the euro area business cycle. Our analysis is achieved by exploiting the concept of recurrence plots, in particular distance plots, to characterise and detect turning points of the business cycle. Firstly, we apply the concept of recurrence plots on the US Industrial Production Index (IPI) series; this serves as a benchmark for our analysis since it already contains a reference chronology for the US business cycle, as provided by the Dating Committee of the National Bureau of Economic Research (NBER). We then use this concept to construct a turning point chronology for the euro area business cycle. In particular, we show that this approach detects turning points and helps with the study of the business cycle without a priori assumptions on the statistical properties of the underlying economic indicator
A test for a new modelling: The Univariate MT-STAR Model
No full textIn ESTAR models it is usually quite dicult to obtain parameter estimates, as it is discussed
in the literature. The problem of properly distinguishing the transition function in relation to
extreme parameter combinations often leads to getting strongly biased estimators. This paper
proposes a new procedure to test for the unit root in a nonlinear framework, and contributes to
the existing literature in three separate directions. First, we propose a new alternative model – the
MT-STAR model – which has similar properties as the ESTAR model but reduces the eects of
the identication problem and can also account for cases where the adjustment mechanism towards
equilibrium is not symmetric. Second, we develop a testing procedure to detect the presence of
a nonlinear stationary process by establishing the limiting non-standard asymptotic distributions
of the proposed test-statistics. Finally, we perform Monte Carlo simulations to assess the small
sample performance of the test and then to highlight its power gain over existing tests for a unit
root
Understanding Exchange Rates Dynamics
No full textURL des Documents de travail : http://ces.univ-paris1.fr/cesdp/cesdp2013.htmlDocuments de travail du Centre d'Economie de la Sorbonne 2013.23 - ISSN : 1955-611XWith the emergence of the chaos theory and the method of surrogates data, nonlinear approaches employed in analysing time series typically suffer from high computational complexity and lack of straightforward explanation. Therefore, the need for methods capable of characterizing time series in terms of their linear, nonlinear, deterministic and stochastic nature are preferable. In this paper, we provide a signal modality analysis on a variety of exchange rates. The analysis is achieved by using the recently proposed "delay vector variance" (DVV) method, which examines local predictability of a signal in the phase space to detect the presence of determinism and nonlinearity in a time series. Optimal embedding parameters used in the DVV analysis are obtain via differential entropy based method using wavelet-based surrogates. A comprehensive analysis of the feasibility of this approach is provided. The empirical results show that the DVV method can be opted as an alternative way to understanding exchange rates dynamics.Avec l'émergence de la théorie du chaos et de la méthode des données surrogate, les approches non linéaires utilisées dans l'analyse des séries temporelles sont complexes et manquent d'explication simple. Néanmoins il est intéressant d'utiliser des techniques permettant de prendre en compte les caractères linéaire, non linéaire, déterministe et stochastique des données observées dans la pratique. Dans cet article, nous analysons une grande variété de données de taux de change. L'analyse est obtenue en utilisant la méthode "delay vector variance" (DVV), qui étudie la prévisibilité locale d'un signal dans l'espace de phase afin de détecter la présence de déterminisme et de non linéarité dans les séries chronologique
The Univariate MT-STAR Model and a new linearity and unit root test procedure
No full textInternational audienceA novel procedure to test for linearity and unit root in a nonlinear framework is proposed by introducing a new model–the MT-STAR model–which has similar properties of the ESTAR model but reduces the effects of the identification problem and can also account for asymmetry in the adjustment mechanism towards equilibrium. The asymptotic distribution of the proposed unit root test is non standard and is derived. The power of the test is evaluated through a simulation study and some empirical illustrations on real exchange rates show its accuracy
Insights to the European debt crisis using recurrence quantification and network analysis
No full textURL des Documents de travail : http://centredeconomiesorbonne.univ-paris1.fr/documents-de-travail/Documents de travail du Centre d'Economie de la Sorbonne 2015.35 - ISSN : 1955-611XThe turmoil in the sovereign debt markets in Europe has raised concerns on the usefulness of sovereign credit default swaps and government bond yields in periods of distress. In addressing this issue, we introduce a novel nonlinear approach for the analysis of non-stationary multivariate data based on complex networks and recurrence analysis. We show the relevance of the approach in studying joint risk connections, extracting hidden spatial information, time dependence, detection of regime changes and providing early warning indicators. The feasibility and relevance of the approach in studying systemic risk is discussed. Finally, we share more light on possible extensions and applications of the approach to systemic risk.Les turbulences sur les marchés de la dette souveraine en Europe ont suscité les intérêts sur l'utilité des contrats d'échange sur risque de crédits (CDS) et des rendements sur les obligations d'État dans les périodes de crise. En examinant cette question, nous introduisons une nouvelle approche non linéaire pour l'analyse des données à variables multiples non-stationnaires, basée sur les réseaux complexes et l'analyse de récurrence. Nous montrons la pertinence de l'approche en étudiant les risques communs, en extrayant les informations spatiales cachées et la dépendance par rapport au temps, en détectant des changements de régime et en fournissant des indicateurs d'alerte précoce. La faisabilité et la pertinence de l'approche dans l'étude du risque systémique est discutée. Enfin, la lumière est faite sur d'éventuelles extensions et applications de l'approche du risque systémique
Coupling direction of the European Banking and Insurance sectors using inter-system recurrence networks
No full textURL des Documents de travail : http://centredeconomiesorbonne.univ-paris1.fr/documents-de-travail/Documents de travail du Centre d'Economie de la Sorbonne 2015.51 - ISSN : 1955-611XModern financial systems exhibit a high degree of interdependence making it difficult in predicting. This has raise concerns on the correct identification of coupling direction in financial sectors of the economy. This study explores a “two-way” risk connection between the European banking and insurance sector based on geometrical closeness of observations. Specifically, the study looks at the inter-system recurrence networks in tracing dynamical transitions and detecting coupling direction between these sectors. The overall results shows that the banking sector is central in risk transmission compared to the insurance sector. A comprehensive discussion of the feasibility and relevance of the approach in studying systemic risk is provided.Les systèmes financiers modernes présentent un degré élevé d'interdépendance rendant difficile la prédiction. Cela a soulevé des questions concernant l'identification correcte d'une direction de couplage dans les secteurs financiers de l'économie. Cette étude explore "en deux sens" la connexion des risques entre le système bancaire européen et le secteur de l'assurance, basée sur la proximité géométrique des observations. Plus précisément, l'étude se penche sur les réseaux de récurrence inter-système en traçant des transitions dynamiques et en détectant la direction de couplage entre ces secteurs. Les résultats globaux montrent que le secteur bancaire est un élément central dans la transmission de risque par rapport au secteur de l'assurance. Une discussion complète de la faisabilité et la pertinence de l'approche dans l'étude du risque systémique est fournie
Approches modernes pour l'analyse non linéaire de données de séries chronologiques économiques et financières
No full textL’axe principal de la thèse est centré sur des approches non-linéaires modernes d’analyse des données économiques et financières, avec une attention particulière sur les cycles économiques et les crises financières. Un consensus dans la littérature statistique et financière s’est établie autour du fait que les variables économiques ont un comportement non-linéaire au cours des différentes phases du cycle économique. En tant que tel, les approches/modèles non-linéaires sont requis pour saisir les caractéristiques du mécanisme de génération des données intrinsèquement asymétriques, que les modèles linéaires sont incapables de reproduire.À cet égard, la thèse propose une nouvelle approche interdisciplinaire et ouverte à l’analyse des systèmes économiques et financiers. La thèse présente des approches robustes aux valeurs extrêmes et à la non-stationnarité, applicables à la fois pour des petits et de grands échantillons, aussi bien pour des séries temporelles économiques que financières. La thèse fournit des procédures dites étape par étape dans l’analyse des indicateurs économiques et financiers en intégrant des concepts basés sur la méthode de substitution de données, des ondelettes, espace incorporation de phase, la m´méthode retard vecteur variance (DVV) et des récurrences parcelles. La thèse met aussi en avant des méthodes transparentes d’identification, de datation des points de retournement et de l´évaluation des impacts des crises économiques et financières. En particulier, la thèse fournit également une procédure pour anticiper les crises futures et ses conséquences.L’étude montre que l’intégration de ces techniques dans l’apprentissage de la structure et des interactions au sein et entre les variables économiques et financières sera très utile dans l’élaboration de politiques de crises, car elle facilite le choix des méthodes de traitement appropriées, suggérées par les données.En outre, une nouvelle procédure pour tester la linéarité et la racine unitaire dans un cadre non-linéaire est proposé par l’introduction d’un nouveau modèle – le modèle MT-STAR – qui a des propriétés similaires au modèle ESTAR mais réduit les effets des problèmes d’identification et peut aussi représenter l’asymétrie dans le mécanisme d’ajustement vers l’équilibre. Les distributions asymptotiques du test de racine unitaire proposées sont non-standards et sont calculées. La puissance du test est évaluée par simulation et quelques illustrations empiriques sur les taux de change réel montrent son efficacité. Enfin, la thèse développe des modèles multi-variés Self-Exciting Threshold Autoregressive avec des variables exogènes (MSETARX) et présente une méthode d’estimation paramétrique. La modélisation des modèles MSETARX et des problèmes engendrés par son estimation sont brièvement examinés.This thesis centers on introducing modern non-linear approaches for data analysis in economics and finance with special attention on business cycles and financial crisis. It is now well stated in the statistical and economic literature that major economic variables display non-linear behaviour over the different phases of the business cycle. As such, nonlinear approaches/models are required to capture the features of the data generating mechanism of inherently asymmetric realizations, since linear models are incapable of generating such behavior.In this respect, the thesis provides an interdisciplinary and open-minded approach to analyzing economic and financial systems in a novel way. The thesis presents approaches that are robust to extreme values, non-stationarity, applicable to both short and long data length, transparent and adaptive to any financial/economic time series. The thesis provides step-by-step procedures in analyzing economic/financial indicators by incorporating concepts based on surrogate data method, wavelets, phase space embedding, ’delay vector variance’ (DVV) method and recurrence plots. The thesis also centers on transparent ways of identifying, dating turning points, evaluating impact of economic and financial crisis. In particular, the thesis also provides a procedure on how to anticipate future crisis and the possible impact of such crisis. The thesis shows that the incorporation of these techniques in learning the structure and interactions within and between economic and financial variables will be very useful in policy-making, since it facilitates the selection of appropriate processing methods, suggested by the data itself.In addition, a novel procedure to test for linearity and unit root in a nonlinear framework is proposed by introducing a new model – the MT-STAR model – which has similar properties of the ESTAR model but reduces the effects of the identification problem and can also account for asymmetry in the adjustment mechanism towards equilibrium. The asymptotic distributions of the proposed unit root test is non-standard and is derived.The power of the test is evaluated through a simulation study and some empirical illustrations on real exchange rates show its accuracy. Finally, the thesis defines a multivariate Self–Exciting Threshold Autoregressive with eXogenous input (MSETARX) models and present an estimation procedure for the parameters. The modeling procedure for the MSETARX models and problems of estimation are briefly considered
