Spelling suggestions: "subject:"[een] TIME-VARYING PARAMETER"" "subject:"[enn] TIME-VARYING PARAMETER""
1 |
International Housing Markets, Unconventional Monetary Policy and the Zero Lower BoundHuber, Florian, Punzi, Maria Teresa 25 January 2016 (has links) (PDF)
In this paper we propose a time-varying parameter VAR model for the housing market in the United States, the United Kingdom, Japan and the Euro Area. For these four economies, we answer the following research questions: (i) How can we evaluate the stance of monetary policy when the policy rate hits the zero lower bound? (ii) Can developments in the housing market still be explained by policy measures adopted by central banks? (iii) Did central banks succeed in mitigating the detrimental impact of the financial crisis on selected housing variables? We analyze the relationship between unconventional monetary policy and the housing markets by using the shadow interest rate estimated by Krippner (2013b). Our findings suggest that the monetary policy transmission mechanism to the housing market has not changed with the implementation of quantitative easing or forward guidance, and central banks can affect the composition of an investors portfolio through investment in housing. A counterfactual exercise provides some evidence that unconventional monetary policy has been particularly successful in dampening the consequences of the financial crisis on housing markets in the United States, while the effects are more muted in the other countries considered in this study. (authors' abstract) / Series: Department of Economics Working Paper Series
|
2 |
Assouplissement quantitatif : que tirer de l'experience japonaise ? / Quantitative easing : what can we learn from the japanese experience ?Moussa, Zakaria 06 December 2010 (has links)
La crise financière actuelle, en raison de sa similarité avec celle du Japon des années 1990, a poussé les autorités monétaires des plus grandes banques centrales à adopter l’assouplissement quantitatif. Seul le Japon, ayant connu une expérience d’assouplissement quantitatif récente mais depuis suffisamment d’années pour être étudiée, peut fournir des éléments de solution à cette crise.Cette thèse applique les techniques économétriques les plus appropriées et récentesà l’analyse de l’assouplissement quantitatif, appliqué par la Banque du Japon entre 2001 et 2006. En trois chapitres sont traitées les questions de savoir s’il était efficace ; sous quelles conditions ? Par quels canaux ?L’efficacité de cette stratégie de politique monétaire à stimuler l’activité et à stopperla spirale déflationniste a été montrée. Cette expérience met en avant le rôle important que la politique monétaire peut jouer pour sortir de la crise, même quand le taux directeur atteint zéro. Le canal des anticipations comme le canal de rééquilibrage des portefeuilles ont tous deux joué un rôle important dans la transmission de ces effets. Les principaux enseignements que l’on peut tirer de l’expérience japonaise sont, d’abord de remédier radicalement et immédiatement aux fragilités du secteur financier, deuxièmement, de mener une politique monétaire particulièrement agressive. Enfin, d’attendre le temps nécessaire pour que les fruits de cette politique viennent. L’expérience japonaise suggère que la Fed et la banque d’Angleterre doivent reporter leur sortie de cette stratégie, sortie qui doit être menée dansle cadre d’un programme et selon des objectifs numériques clairs. / The current financial crisis has now led most major central banks to rely on quantitative easing. The unique Japanese experience of quantitative easing is the only experience which enables us to judge this therapy’s effectiveness and the timing of the exit strategy. Is quantitative easing effective ? Under which conditions ? Through which canal ?This thesis, consisting of three essays, applies appropriate and recent econometrictechniques to examine the quantitative easing in Japan between 2001 and 2006. We show, for the first time, that quantitative easing was able not only to prevent further recession and deflation but also to provide considerable stimulation to both output and prices. Moreover, both expectation and portfolio-rebalancing channels play a crucial role in transmitting monetary policy effects. This experience shows that the monetary policy is still potent even when short-term interest rates reach a zero lower bound. The Japanese experience suggests that efforts to clean up the bank’s balance sheets significantly improved the effectiveness of quantitative easing. However, this effect, although considerable, was short-lived ; it became insignificant after one year. The short duration of this effect confirms the wisdom of the Fed’s decision to maintain quantitative easing longer, so that being short-lived, the positive effects could be exploited. In the light of the Japanese experience, we argue that, in addition to their fast reaction and the huge amount of CABs employed, which may have helped relieve short-term liquidity pressures in the financial system, the Fed was better off postponing its exit from quantitative easing.
|
3 |
Modelling of conditional variance and uncertainty using industrial process dataJuutilainen, I. (Ilmari) 14 November 2006 (has links)
Abstract
This thesis presents methods for modelling conditional variance and uncertainty of prediction at a query point on the basis of industrial process data. The introductory part of the thesis provides an extensive background of the examined methods and a summary of the results. The results are presented in detail in the original papers.
The application presented in the thesis is modelling of the mean and variance of the mechanical properties of steel plates. Both the mean and variance of the mechanical properties depend on many process variables. A method for predicting the probability of rejection in a quali?cation test is presented and implemented in a tool developed for the planning of strength margins. The developed tool has been successfully utilised in the planning of mechanical properties in a steel plate mill.
The methods for modelling the dependence of conditional variance on input variables are reviewed and their suitability for large industrial data sets are examined. In a comparative study, neural network modelling of the mean and dispersion narrowly performed the best.
A method is presented for evaluating the uncertainty of regression-type prediction at a query point on the basis of predicted conditional variance, model variance and the effect of uncertainty about explanatory variables at early process stages. A method for measuring the uncertainty of prediction on the basis of the density of the data around the query point is proposed. The proposed distance measure is utilised in comparing the generalisation ability of models. The generalisation properties of the most important regression learning methods are studied and the results indicate that local methods and quadratic regression have a poor interpolation capability compared with multi-layer perceptron and Gaussian kernel support vector regression.
The possibility of adaptively modelling a time-varying conditional variance function is disclosed. Two methods for adaptive modelling of the variance function are proposed. The background of the developed adaptive variance modelling methods is presented.
|
4 |
[en] GAS MODELS APPLIED TO TIME SERIES OF STREAMFLOW AND WIND / [pt] MODELOS GAS APLICADOS A SERIES TEMPORAIS DE VAZAO E VENTOGILSON GONCALVES DE MATOS 04 October 2013 (has links)
[pt] Os modelos GAS (generalized autoregressive score) são modelos de séries
temporais com parâmetros variantes no tempo, os quais possuem sua evolução
ditada pelo vetor score ponderado da função de verossimilhança. A avaliação da
verossimilhança nestes modelos é bastante simples, bem como incorporação de
efeitos como assimetria, memória longa e outras dinâmicas. Por serem baseados
no score da verossimilhança, exporta-se a estrutura completa da distribuição
preditiva para o mecanismo de atualização dos parâmetros, e não apenas
a média ou momentos de ordem superior. Estas características, somadas á
capacidade de lidar com processos multivariados e não estacionários, tornam
a classe em estudo uma nova alternativa para a construção de modelos com
parâmetros variantes, particularmente para séries temporais não gaussianas.
Nesta dissertação, foram desenvolvidos modelos GAS univariados para a
análise das séries mensais de vazão do Rio Paraibuna (MG) e de fator de
capacidade de uma usina é olica não divulgada do Nordeste, utilizando as
distribuições gama e beta, respectivamente. Além disso, foi derivado um novo
modelo GAS bivariado com marginais gama e beta para a modelagem conjunta
dos processos de vazão e vento, de modo a explorar a complementaridade
sazonal entre as séries. / [en] The GAS models (generalized autoregressive score) are time series models
with time-varying parameters, which have their update mechanism drived
by the scaled score of the likelihood function. The likelihood evaluation in
these models is quite simple, as well as the incorporation of effects like
asymmetry, long memory and other dynamics. Because they are based in the
scaled score of the likelihood, it exploits the full structure of the predictive
distribution to the update mechanism of the parameters, and not just mean
or higher order moments. These characteristics, coupled with the ability to
handle with multivariate and non-stationary processes, make the studied class
a new alternative to the construction of models with time-varying parameters,
particularly for non-Gaussian time series. In this dissertation, univariate GAS
models were developed to analyze monthly series of streamflow of Paraibuna
river (MG) and of capacity factor of a wind farm undisclosed in Northeast,
using the gamma and beta distributions, respectively. In addition, a new
bivariate GAS model with gamma and beta marginals was derived for the
joint modeling of the streamflow and wind processes, in order to explore the
seasonal complementarity between the series.
|
5 |
Preferências assimétricas variantes no tempo na função perda do Banco Central do Brasil.Lopes, Kennedy Carvalho 13 August 2012 (has links)
Made available in DSpace on 2015-05-08T14:44:50Z (GMT). No. of bitstreams: 1
Arquivototal.pdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5)
Previous issue date: 2012-08-13 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This paper estimates a reaction function with forward-looking time-varying parameters for changes in the Brazilian monetary policy under inflation targeting regime. As the policy rule has endogenous regressors, the conventional Kalman filter can t be applied. Thus, a two-step procedure of the type Heckman (1976) is used to estimate the hyperparameters consistent model. The results show that: i) there is strong empirical evidence of endogeneity of the regressors of monetary policy rule, ii) the expected interest rate was above 10% throughout the analysis period to an average of 11%; iii) response the Selic rate to inflation varies considerably throughout the period and has shown a declining trend, iv) the response of interest rates relative to inflation deviation from the target with the principle of Taylor; v) the coefficient of smoothing rate interest has been constant throughout the period; vi) that the BCB had in much of the period analyzed an aversion recession by allowing inflation above target. / Este trabalho estima uma função de reação forward-looking com parâmetros variando no tempo para verificar mudanças na condução da política monetária brasileira sob o regime de metas de inflação. Como a regra de política apresenta regressores endógenos, o filtro de Kalman convencional não pode ser aplicado. Diante disso, um procedimento em dois passos do tipo de Heckman (1976) é utilizado para estimação
consistente dos hiperparâmetros do modelo. Os resultados mostram que: i) há forte evidência empírica de endogeneidade dos regressores da regra de política monetária; ii) que a taxa de juros esperada esteve acima de 10% durante todo o período analisado, tendo uma média de 11%; iii) a resposta da taxa Selic à inflação varia consideravelmente ao longo do período e tem mostrado uma tendência decrescente; iv) a resposta da taxa de juros em relação ao desvio da inflação a meta respeita o princípio de Taylor; v) o coeficiente de suavização da taxa de juros foi constante durante todo o período analisado; vi) que o BCB teve em boa parte do período analisado uma aversão recessão, permitindo
uma inflação acima da meta.
|
6 |
[en] FORECAST OF THE JOINT DENSITY OF WIND CAPACITY FACTOR THROUGH THE USE OF A MULTIVARIATE GAS MODEL / [pt] PREVISÃO DA DENSIDADE CONJUNTA DE FATOR DE CAPACIDADE EÓLICO VIA MODELO GAS MULTIVARIADOHENRIQUE HELFER HOELTGEBAUM 06 October 2015 (has links)
[pt] Neste trabalho usamos o arcabouço dos modelos GAS para gerar previsões conjuntas de fator de capacidade eólico, pertencentes a diferentes usinas localizadas em áreas geográficas distintas. Esses cenários são insumos para gerar uma distribuição de fluxo de caixa associada a um portfólio de contratos atrelados aos parques eólicos em questão. Inicialmente modelamos as densidades marginais via um modelo GAS, supondo densidade Beta. De maneira a capturar a estrutura de dependência entre esses fatores de capacidade, usamos uma cópula t-Student com a matriz de correlação também sendo atualizada via mecanismo GAS. Uma das contribuições importantes desse trabalho para o setor elétrico está na geração de cenários conjuntos apenas em um passo, evitando a necessidade de modelar variáveis transformadas e posteriormente transforma-las para retornar às suas respectivas escalas originais. Assim como é feito no caso supondo normalidade para as marginais. Como é sabido, exponenciar valores simulados a partir de uma densidade normal pode gerar resultados equivocados para fatores de capacidade eólico, e por propagação, isso pode afetar severamente as medidas de risco que são obtidas a partir da distribuição simulada de fluxo de caixa associada com o portfolio das usinas eólicas. Nossos resultados mostram que quando a dependência é levada em consideração, os fluxos de caixa tendem a ser maiores do que quando ignora-se a dependência. / [en] In this work we use the framework of GAS models to generate joint forecasts for capacity factors of several wind plants belonging to different geographical areas. Such scenarios are then used as input to raise the distribution of cash flows associated with a portfolio of contracts attached to these wind plants. We first model the marginal density of each capacity factor using a GAS model with Beta density. In order to capture the observed dependence among these capacity factors, we use a copula t- Student with correlation matrix evolving through a GAS mechanism. One of the important contributions of our framework is that generation of scenarios is accomplished in just one step, avoiding the need of transforming back variables to its original scale, as it is the case under a Gaussian assumption for the marginals. As it is known, exponentiation of simulated Gaussian values can result in unrealistic sampling paths for the wind capacity factor, and by propagation, this can badly a ect the risk measures obtained from the simulated distribution of the cash flows associated with a particular portfolio of wind plants. Our results shows that when taking into account dependence the cash flows are higher than when ignoring dependence.
|
7 |
Factor models, VARMA processes and parameter instability with applications in macroeconomicsStevanovic, Dalibor 05 1900 (has links)
Avec les avancements de la technologie de l'information, les données temporelles économiques et financières sont de plus en plus disponibles. Par contre, si les techniques standard de l'analyse des séries temporelles sont utilisées, une grande quantité d'information est accompagnée du problème de dimensionnalité. Puisque la majorité des séries d'intérêt sont hautement corrélées, leur dimension peut être réduite en utilisant l'analyse factorielle. Cette technique est de plus en plus populaire en sciences économiques depuis les années 90.
Étant donnée la disponibilité des données et des avancements computationnels, plusieurs nouvelles questions se posent. Quels sont les effets et la transmission des chocs structurels dans un environnement riche en données? Est-ce que l'information contenue dans un grand ensemble d'indicateurs économiques peut aider à mieux identifier les chocs de politique monétaire, à l'égard des problèmes rencontrés dans les applications utilisant des modèles standards? Peut-on identifier les chocs financiers et mesurer leurs effets sur l'économie réelle? Peut-on améliorer la méthode factorielle existante et y incorporer une autre technique de réduction de dimension comme l'analyse VARMA? Est-ce que cela produit de meilleures prévisions des grands agrégats macroéconomiques et aide au niveau de l'analyse par fonctions de réponse impulsionnelles? Finalement, est-ce qu'on peut appliquer l'analyse factorielle au niveau des paramètres aléatoires? Par exemple, est-ce qu'il existe seulement un petit nombre de sources de l'instabilité temporelle des coefficients dans les modèles macroéconomiques empiriques?
Ma thèse, en utilisant l'analyse factorielle structurelle et la modélisation VARMA, répond à ces questions à travers cinq articles. Les deux premiers chapitres étudient les effets des chocs monétaire et financier dans un environnement riche en données. Le troisième article propose une nouvelle méthode en combinant les modèles à facteurs et VARMA. Cette approche est appliquée dans le quatrième article pour mesurer les effets des chocs de crédit au Canada. La contribution du dernier chapitre est d'imposer la structure à facteurs sur les paramètres variant dans le temps et de montrer qu'il existe un petit nombre de sources de cette instabilité.
Le premier article analyse la transmission de la politique monétaire au Canada en utilisant le modèle vectoriel autorégressif augmenté par facteurs (FAVAR). Les études antérieures basées sur les modèles VAR ont trouvé plusieurs anomalies empiriques suite à un choc de la politique monétaire. Nous estimons le modèle FAVAR en utilisant un grand nombre de séries macroéconomiques mensuelles et trimestrielles. Nous trouvons que l'information contenue dans les facteurs est importante pour bien identifier la transmission de la politique monétaire et elle aide à corriger les anomalies empiriques standards. Finalement, le cadre d'analyse FAVAR permet d'obtenir les fonctions de réponse impulsionnelles pour tous les indicateurs dans l'ensemble de données, produisant ainsi l'analyse la plus complète à ce jour des effets de la politique monétaire au Canada.
Motivée par la dernière crise économique, la recherche sur le rôle du secteur financier a repris de l'importance. Dans le deuxième article nous examinons les effets et la propagation des chocs de crédit sur l'économie réelle en utilisant un grand ensemble d'indicateurs économiques et financiers dans le cadre d'un modèle à facteurs structurel. Nous trouvons qu'un choc de crédit augmente immédiatement les diffusions de crédit (credit spreads), diminue la valeur des bons de Trésor et cause une récession. Ces chocs ont un effet important sur des mesures d'activité réelle, indices de prix, indicateurs avancés et financiers. Contrairement aux autres études, notre procédure d'identification du choc structurel ne requiert pas de restrictions temporelles entre facteurs financiers et macroéconomiques. De plus, elle donne une interprétation des facteurs sans restreindre l'estimation de ceux-ci.
Dans le troisième article nous étudions la relation entre les représentations VARMA et factorielle des processus vectoriels stochastiques, et proposons une nouvelle classe de modèles VARMA augmentés par facteurs (FAVARMA). Notre point de départ est de constater qu'en général les séries multivariées et facteurs associés ne peuvent simultanément suivre un processus VAR d'ordre fini. Nous montrons que le processus dynamique des facteurs, extraits comme combinaison linéaire des variables observées, est en général un VARMA et non pas un VAR comme c'est supposé ailleurs dans la littérature. Deuxièmement, nous montrons que même si les facteurs suivent un VAR d'ordre fini, cela implique une représentation VARMA pour les séries observées. Alors, nous proposons le cadre d'analyse FAVARMA combinant ces deux méthodes de réduction du nombre de paramètres. Le modèle est appliqué dans deux exercices de prévision en utilisant des données américaines et canadiennes de Boivin, Giannoni et Stevanovic (2010, 2009) respectivement. Les résultats montrent que la partie VARMA aide à mieux prévoir les importants agrégats macroéconomiques relativement aux modèles standards. Finalement, nous estimons les effets de choc monétaire en utilisant les données et le schéma d'identification de Bernanke, Boivin et Eliasz (2005). Notre modèle FAVARMA(2,1) avec six facteurs donne les résultats cohérents et précis des effets et de la transmission monétaire aux États-Unis. Contrairement au modèle FAVAR employé dans l'étude ultérieure où 510 coefficients VAR devaient être estimés, nous produisons les résultats semblables avec seulement 84 paramètres du processus dynamique des facteurs.
L'objectif du quatrième article est d'identifier et mesurer les effets des chocs de crédit au Canada dans un environnement riche en données et en utilisant le modèle FAVARMA structurel. Dans le cadre théorique de l'accélérateur financier développé par Bernanke, Gertler et Gilchrist (1999), nous approximons la prime de financement extérieur par les credit spreads. D'un côté, nous trouvons qu'une augmentation non-anticipée de la prime de financement extérieur aux États-Unis génère une récession significative et persistante au Canada, accompagnée d'une hausse immédiate des credit spreads et taux d'intérêt canadiens. La composante commune semble capturer les dimensions importantes des fluctuations cycliques de l'économie canadienne. L'analyse par décomposition de la variance révèle que ce choc de crédit a un effet important sur différents secteurs d'activité réelle, indices de prix, indicateurs avancés et credit spreads. De l'autre côté, une hausse inattendue de la prime canadienne de financement extérieur ne cause pas d'effet significatif au Canada. Nous montrons que les effets des chocs de crédit au Canada sont essentiellement causés par les conditions globales, approximées ici par le marché américain. Finalement, étant donnée la procédure d'identification des chocs structurels, nous trouvons des facteurs interprétables économiquement.
Le comportement des agents et de l'environnement économiques peut varier à travers le temps (ex. changements de stratégies de la politique monétaire, volatilité de chocs) induisant de l'instabilité des paramètres dans les modèles en forme réduite. Les modèles à paramètres variant dans le temps (TVP) standards supposent traditionnellement les processus stochastiques indépendants pour tous les TVPs. Dans cet article nous montrons que le nombre de sources de variabilité temporelle des coefficients est probablement très petit, et nous produisons la première évidence empirique connue dans les modèles macroéconomiques empiriques. L'approche Factor-TVP, proposée dans Stevanovic (2010), est appliquée dans le cadre d'un modèle VAR standard avec coefficients aléatoires (TVP-VAR). Nous trouvons qu'un seul facteur explique la majorité de la variabilité des coefficients VAR, tandis que les paramètres de la volatilité des chocs varient d'une façon indépendante. Le facteur commun est positivement corrélé avec le taux de chômage. La même analyse est faite avec les données incluant la récente crise financière. La procédure suggère maintenant deux facteurs et le comportement des coefficients présente un changement important depuis 2007. Finalement, la méthode est appliquée à un modèle TVP-FAVAR. Nous trouvons que seulement 5 facteurs dynamiques gouvernent l'instabilité temporelle dans presque 700 coefficients. / As information technology improves, the availability of economic and finance time series grows in terms of both time and cross-section sizes. However, a large amount of information can lead to the curse of dimensionality problem when standard time series tools are used. Since most of these series are highly correlated, at least within some categories, their co-variability pattern and informational content can be approximated by a smaller number of (constructed) variables. A popular way to address this issue is the factor analysis. This framework has received a lot of attention since late 90's and is known today as the large dimensional approximate factor analysis.
Given the availability of data and computational improvements, a number of empirical and theoretical questions arises. What are the effects and transmission of structural shocks in a data-rich environment? Does the information from a large number of economic indicators help in properly identifying the monetary policy shocks with respect to a number of empirical puzzles found using traditional small-scale models? Motivated by the recent financial turmoil, can we identify the financial market shocks and measure their effect on real economy? Can we improve the existing method and incorporate another reduction dimension approach such as the VARMA modeling? Does it help in forecasting macroeconomic aggregates and impulse response analysis? Finally, can we apply the same factor analysis reasoning to the time varying parameters? Is there only a small number of common sources of time instability in the coefficients of empirical macroeconomic models?
This thesis concentrates on the structural factor analysis and VARMA modeling and answers these questions through five articles. The first two articles study the effects of monetary policy and credit shocks in a data-rich environment. The third article proposes a new framework that combines the factor analysis and VARMA modeling, while the fourth article applies this method to measure the effects of credit shocks in Canada. The contribution of the final chapter is to impose the factor structure on the time varying parameters in popular macroeconomic models, and show that there are few sources of this time instability.
The first article analyzes the monetary transmission mechanism in Canada using a
factor-augmented vector autoregression (FAVAR) model. For small open economies
like Canada, uncovering the transmission mechanism of monetary policy using VARs
has proven to be an especially challenging task. Such studies on Canadian data have
often documented the presence of anomalies such as a price, exchange rate, delayed overshooting and uncovered interest rate parity puzzles. We estimate a FAVAR model using large sets of monthly and quarterly macroeconomic time series. We find that the information summarized by the factors is important to properly identify the monetary transmission mechanism and contributes to mitigate the puzzles mentioned above, suggesting that more information does help. Finally, the FAVAR framework allows us to check impulse responses for all series in the informational data set, and thus provides the most comprehensive picture to date of the effect of Canadian monetary policy.
As the recent financial crisis and the ensuing global economic have illustrated, the financial sector plays an important role in generating and propagating shocks to the real economy. Financial variables thus contain information that can predict future economic conditions. In this paper we examine the dynamic effects and the propagation of credit shocks using a large data set of U.S. economic and financial indicators in a structural factor model. Identified credit shocks, interpreted as unexpected deteriorations of the credit market conditions, immediately increase credit spreads, decrease rates on Treasury securities and cause large and persistent downturns in the activity of many economic sectors. Such shocks are found to have important effects on real activity measures, aggregate prices, leading indicators and credit spreads. In contrast to other recent papers, our structural shock identification procedure does not require any timing restrictions between the financial and macroeconomic factors, and yields an interpretation of the estimated factors without relying on a constructed measure of credit market conditions from a large set of individual bond prices and financial series.
In third article, we study the relationship between VARMA and factor representations of a vector stochastic process, and propose a new class of factor-augmented VARMA (FAVARMA) models. We start by observing that in general multivariate series and associated factors do not both follow a finite order VAR process. Indeed, we show that when the factors are obtained as linear combinations of observable series, their dynamic process is generally a VARMA and not a finite-order VAR as usually assumed in the literature. Second, we show that even if the factors follow a finite-order VAR process, this implies a VARMA representation for the observable series. As result, we propose the FAVARMA framework that combines two parsimonious methods to represent the dynamic interactions between a large number of time series: factor analysis and VARMA modeling. We apply our approach in two pseudo-out-of-sample forecasting exercises using large U.S. and Canadian monthly panels taken from Boivin, Giannoni and Stevanovic (2010, 2009) respectively. The results show that VARMA factors help in predicting several key macroeconomic aggregates relative to standard factor forecasting models. Finally, we estimate the effect of monetary policy using the data and the identification scheme as in Bernanke, Boivin and Eliasz (2005). We find that impulse responses from a parsimonious 6-factor FAVARMA(2,1) model give an accurate and comprehensive picture of the effect and the transmission of monetary policy in U.S.. To get similar responses from a standard FAVAR model, Akaike information criterion estimates the lag order of 14. Hence, only 84 coefficients governing the factors dynamics need to be estimated in the FAVARMA framework, compared to FAVAR model with 510 VAR parameters.
In fourth article we are interested in identifying and measuring the effects of credit shocks in Canada in a data-rich environment. In order to incorporate information from a large number of economic and financial indicators, we use the structural factor-augmented VARMA model. In the theoretical framework of the financial accelerator, we approximate the external finance premium by credit spreads. On one hand, we find that an unanticipated increase in US external finance premium generates a significant and persistent economic slowdown in Canada; the Canadian external finance premium rises immediately while interest rates and credit measures decline. From the variance decomposition analysis, we observe that the credit shock has an important effect on several real activity measures, price indicators, leading indicators, and credit spreads. On the other hand, an unexpected increase in Canadian external finance premium shows no significant effect in Canada. Indeed, our results suggest that the effects of credit shocks in Canada are essentially caused by the unexpected changes in foreign credit market conditions. Finally, given the identification procedure, we find that our structural factors do have an economic interpretation.
The behavior of economic agents and environment may vary over time (monetary policy strategy shifts, stochastic volatility) implying parameters' instability in reduced-form models. Standard time varying parameter (TVP) models usually assume independent stochastic processes for all TVPs. In the final article, I show that the number of underlying sources of parameters' time variation is likely to be small, and provide empirical evidence on factor structure among TVPs of popular macroeconomic models. To test for the presence of, and estimate low dimension sources of time variation in parameters, I apply the factor time varying parameter (Factor-TVP) model, proposed by Stevanovic (2010), to a standard monetary TVP-VAR model. I find that one factor explains most of the variability in VAR coefficients, while the stochastic volatility parameters vary in the idiosyncratic way. The common factor is highly and positively correlated to the unemployment rate. To incorporate the recent financial crisis, the same exercise is conducted with data updated to 2010Q3. The VAR parameters present an important change after 2007, and the procedure suggests two factors. When applied to a large-dimensional structural factor model, I find that four dynamic factors govern the time instability in almost 700 coefficients.
|
8 |
Factor models, VARMA processes and parameter instability with applications in macroeconomicsStevanovic, Dalibor 05 1900 (has links)
Avec les avancements de la technologie de l'information, les données temporelles économiques et financières sont de plus en plus disponibles. Par contre, si les techniques standard de l'analyse des séries temporelles sont utilisées, une grande quantité d'information est accompagnée du problème de dimensionnalité. Puisque la majorité des séries d'intérêt sont hautement corrélées, leur dimension peut être réduite en utilisant l'analyse factorielle. Cette technique est de plus en plus populaire en sciences économiques depuis les années 90.
Étant donnée la disponibilité des données et des avancements computationnels, plusieurs nouvelles questions se posent. Quels sont les effets et la transmission des chocs structurels dans un environnement riche en données? Est-ce que l'information contenue dans un grand ensemble d'indicateurs économiques peut aider à mieux identifier les chocs de politique monétaire, à l'égard des problèmes rencontrés dans les applications utilisant des modèles standards? Peut-on identifier les chocs financiers et mesurer leurs effets sur l'économie réelle? Peut-on améliorer la méthode factorielle existante et y incorporer une autre technique de réduction de dimension comme l'analyse VARMA? Est-ce que cela produit de meilleures prévisions des grands agrégats macroéconomiques et aide au niveau de l'analyse par fonctions de réponse impulsionnelles? Finalement, est-ce qu'on peut appliquer l'analyse factorielle au niveau des paramètres aléatoires? Par exemple, est-ce qu'il existe seulement un petit nombre de sources de l'instabilité temporelle des coefficients dans les modèles macroéconomiques empiriques?
Ma thèse, en utilisant l'analyse factorielle structurelle et la modélisation VARMA, répond à ces questions à travers cinq articles. Les deux premiers chapitres étudient les effets des chocs monétaire et financier dans un environnement riche en données. Le troisième article propose une nouvelle méthode en combinant les modèles à facteurs et VARMA. Cette approche est appliquée dans le quatrième article pour mesurer les effets des chocs de crédit au Canada. La contribution du dernier chapitre est d'imposer la structure à facteurs sur les paramètres variant dans le temps et de montrer qu'il existe un petit nombre de sources de cette instabilité.
Le premier article analyse la transmission de la politique monétaire au Canada en utilisant le modèle vectoriel autorégressif augmenté par facteurs (FAVAR). Les études antérieures basées sur les modèles VAR ont trouvé plusieurs anomalies empiriques suite à un choc de la politique monétaire. Nous estimons le modèle FAVAR en utilisant un grand nombre de séries macroéconomiques mensuelles et trimestrielles. Nous trouvons que l'information contenue dans les facteurs est importante pour bien identifier la transmission de la politique monétaire et elle aide à corriger les anomalies empiriques standards. Finalement, le cadre d'analyse FAVAR permet d'obtenir les fonctions de réponse impulsionnelles pour tous les indicateurs dans l'ensemble de données, produisant ainsi l'analyse la plus complète à ce jour des effets de la politique monétaire au Canada.
Motivée par la dernière crise économique, la recherche sur le rôle du secteur financier a repris de l'importance. Dans le deuxième article nous examinons les effets et la propagation des chocs de crédit sur l'économie réelle en utilisant un grand ensemble d'indicateurs économiques et financiers dans le cadre d'un modèle à facteurs structurel. Nous trouvons qu'un choc de crédit augmente immédiatement les diffusions de crédit (credit spreads), diminue la valeur des bons de Trésor et cause une récession. Ces chocs ont un effet important sur des mesures d'activité réelle, indices de prix, indicateurs avancés et financiers. Contrairement aux autres études, notre procédure d'identification du choc structurel ne requiert pas de restrictions temporelles entre facteurs financiers et macroéconomiques. De plus, elle donne une interprétation des facteurs sans restreindre l'estimation de ceux-ci.
Dans le troisième article nous étudions la relation entre les représentations VARMA et factorielle des processus vectoriels stochastiques, et proposons une nouvelle classe de modèles VARMA augmentés par facteurs (FAVARMA). Notre point de départ est de constater qu'en général les séries multivariées et facteurs associés ne peuvent simultanément suivre un processus VAR d'ordre fini. Nous montrons que le processus dynamique des facteurs, extraits comme combinaison linéaire des variables observées, est en général un VARMA et non pas un VAR comme c'est supposé ailleurs dans la littérature. Deuxièmement, nous montrons que même si les facteurs suivent un VAR d'ordre fini, cela implique une représentation VARMA pour les séries observées. Alors, nous proposons le cadre d'analyse FAVARMA combinant ces deux méthodes de réduction du nombre de paramètres. Le modèle est appliqué dans deux exercices de prévision en utilisant des données américaines et canadiennes de Boivin, Giannoni et Stevanovic (2010, 2009) respectivement. Les résultats montrent que la partie VARMA aide à mieux prévoir les importants agrégats macroéconomiques relativement aux modèles standards. Finalement, nous estimons les effets de choc monétaire en utilisant les données et le schéma d'identification de Bernanke, Boivin et Eliasz (2005). Notre modèle FAVARMA(2,1) avec six facteurs donne les résultats cohérents et précis des effets et de la transmission monétaire aux États-Unis. Contrairement au modèle FAVAR employé dans l'étude ultérieure où 510 coefficients VAR devaient être estimés, nous produisons les résultats semblables avec seulement 84 paramètres du processus dynamique des facteurs.
L'objectif du quatrième article est d'identifier et mesurer les effets des chocs de crédit au Canada dans un environnement riche en données et en utilisant le modèle FAVARMA structurel. Dans le cadre théorique de l'accélérateur financier développé par Bernanke, Gertler et Gilchrist (1999), nous approximons la prime de financement extérieur par les credit spreads. D'un côté, nous trouvons qu'une augmentation non-anticipée de la prime de financement extérieur aux États-Unis génère une récession significative et persistante au Canada, accompagnée d'une hausse immédiate des credit spreads et taux d'intérêt canadiens. La composante commune semble capturer les dimensions importantes des fluctuations cycliques de l'économie canadienne. L'analyse par décomposition de la variance révèle que ce choc de crédit a un effet important sur différents secteurs d'activité réelle, indices de prix, indicateurs avancés et credit spreads. De l'autre côté, une hausse inattendue de la prime canadienne de financement extérieur ne cause pas d'effet significatif au Canada. Nous montrons que les effets des chocs de crédit au Canada sont essentiellement causés par les conditions globales, approximées ici par le marché américain. Finalement, étant donnée la procédure d'identification des chocs structurels, nous trouvons des facteurs interprétables économiquement.
Le comportement des agents et de l'environnement économiques peut varier à travers le temps (ex. changements de stratégies de la politique monétaire, volatilité de chocs) induisant de l'instabilité des paramètres dans les modèles en forme réduite. Les modèles à paramètres variant dans le temps (TVP) standards supposent traditionnellement les processus stochastiques indépendants pour tous les TVPs. Dans cet article nous montrons que le nombre de sources de variabilité temporelle des coefficients est probablement très petit, et nous produisons la première évidence empirique connue dans les modèles macroéconomiques empiriques. L'approche Factor-TVP, proposée dans Stevanovic (2010), est appliquée dans le cadre d'un modèle VAR standard avec coefficients aléatoires (TVP-VAR). Nous trouvons qu'un seul facteur explique la majorité de la variabilité des coefficients VAR, tandis que les paramètres de la volatilité des chocs varient d'une façon indépendante. Le facteur commun est positivement corrélé avec le taux de chômage. La même analyse est faite avec les données incluant la récente crise financière. La procédure suggère maintenant deux facteurs et le comportement des coefficients présente un changement important depuis 2007. Finalement, la méthode est appliquée à un modèle TVP-FAVAR. Nous trouvons que seulement 5 facteurs dynamiques gouvernent l'instabilité temporelle dans presque 700 coefficients. / As information technology improves, the availability of economic and finance time series grows in terms of both time and cross-section sizes. However, a large amount of information can lead to the curse of dimensionality problem when standard time series tools are used. Since most of these series are highly correlated, at least within some categories, their co-variability pattern and informational content can be approximated by a smaller number of (constructed) variables. A popular way to address this issue is the factor analysis. This framework has received a lot of attention since late 90's and is known today as the large dimensional approximate factor analysis.
Given the availability of data and computational improvements, a number of empirical and theoretical questions arises. What are the effects and transmission of structural shocks in a data-rich environment? Does the information from a large number of economic indicators help in properly identifying the monetary policy shocks with respect to a number of empirical puzzles found using traditional small-scale models? Motivated by the recent financial turmoil, can we identify the financial market shocks and measure their effect on real economy? Can we improve the existing method and incorporate another reduction dimension approach such as the VARMA modeling? Does it help in forecasting macroeconomic aggregates and impulse response analysis? Finally, can we apply the same factor analysis reasoning to the time varying parameters? Is there only a small number of common sources of time instability in the coefficients of empirical macroeconomic models?
This thesis concentrates on the structural factor analysis and VARMA modeling and answers these questions through five articles. The first two articles study the effects of monetary policy and credit shocks in a data-rich environment. The third article proposes a new framework that combines the factor analysis and VARMA modeling, while the fourth article applies this method to measure the effects of credit shocks in Canada. The contribution of the final chapter is to impose the factor structure on the time varying parameters in popular macroeconomic models, and show that there are few sources of this time instability.
The first article analyzes the monetary transmission mechanism in Canada using a
factor-augmented vector autoregression (FAVAR) model. For small open economies
like Canada, uncovering the transmission mechanism of monetary policy using VARs
has proven to be an especially challenging task. Such studies on Canadian data have
often documented the presence of anomalies such as a price, exchange rate, delayed overshooting and uncovered interest rate parity puzzles. We estimate a FAVAR model using large sets of monthly and quarterly macroeconomic time series. We find that the information summarized by the factors is important to properly identify the monetary transmission mechanism and contributes to mitigate the puzzles mentioned above, suggesting that more information does help. Finally, the FAVAR framework allows us to check impulse responses for all series in the informational data set, and thus provides the most comprehensive picture to date of the effect of Canadian monetary policy.
As the recent financial crisis and the ensuing global economic have illustrated, the financial sector plays an important role in generating and propagating shocks to the real economy. Financial variables thus contain information that can predict future economic conditions. In this paper we examine the dynamic effects and the propagation of credit shocks using a large data set of U.S. economic and financial indicators in a structural factor model. Identified credit shocks, interpreted as unexpected deteriorations of the credit market conditions, immediately increase credit spreads, decrease rates on Treasury securities and cause large and persistent downturns in the activity of many economic sectors. Such shocks are found to have important effects on real activity measures, aggregate prices, leading indicators and credit spreads. In contrast to other recent papers, our structural shock identification procedure does not require any timing restrictions between the financial and macroeconomic factors, and yields an interpretation of the estimated factors without relying on a constructed measure of credit market conditions from a large set of individual bond prices and financial series.
In third article, we study the relationship between VARMA and factor representations of a vector stochastic process, and propose a new class of factor-augmented VARMA (FAVARMA) models. We start by observing that in general multivariate series and associated factors do not both follow a finite order VAR process. Indeed, we show that when the factors are obtained as linear combinations of observable series, their dynamic process is generally a VARMA and not a finite-order VAR as usually assumed in the literature. Second, we show that even if the factors follow a finite-order VAR process, this implies a VARMA representation for the observable series. As result, we propose the FAVARMA framework that combines two parsimonious methods to represent the dynamic interactions between a large number of time series: factor analysis and VARMA modeling. We apply our approach in two pseudo-out-of-sample forecasting exercises using large U.S. and Canadian monthly panels taken from Boivin, Giannoni and Stevanovic (2010, 2009) respectively. The results show that VARMA factors help in predicting several key macroeconomic aggregates relative to standard factor forecasting models. Finally, we estimate the effect of monetary policy using the data and the identification scheme as in Bernanke, Boivin and Eliasz (2005). We find that impulse responses from a parsimonious 6-factor FAVARMA(2,1) model give an accurate and comprehensive picture of the effect and the transmission of monetary policy in U.S.. To get similar responses from a standard FAVAR model, Akaike information criterion estimates the lag order of 14. Hence, only 84 coefficients governing the factors dynamics need to be estimated in the FAVARMA framework, compared to FAVAR model with 510 VAR parameters.
In fourth article we are interested in identifying and measuring the effects of credit shocks in Canada in a data-rich environment. In order to incorporate information from a large number of economic and financial indicators, we use the structural factor-augmented VARMA model. In the theoretical framework of the financial accelerator, we approximate the external finance premium by credit spreads. On one hand, we find that an unanticipated increase in US external finance premium generates a significant and persistent economic slowdown in Canada; the Canadian external finance premium rises immediately while interest rates and credit measures decline. From the variance decomposition analysis, we observe that the credit shock has an important effect on several real activity measures, price indicators, leading indicators, and credit spreads. On the other hand, an unexpected increase in Canadian external finance premium shows no significant effect in Canada. Indeed, our results suggest that the effects of credit shocks in Canada are essentially caused by the unexpected changes in foreign credit market conditions. Finally, given the identification procedure, we find that our structural factors do have an economic interpretation.
The behavior of economic agents and environment may vary over time (monetary policy strategy shifts, stochastic volatility) implying parameters' instability in reduced-form models. Standard time varying parameter (TVP) models usually assume independent stochastic processes for all TVPs. In the final article, I show that the number of underlying sources of parameters' time variation is likely to be small, and provide empirical evidence on factor structure among TVPs of popular macroeconomic models. To test for the presence of, and estimate low dimension sources of time variation in parameters, I apply the factor time varying parameter (Factor-TVP) model, proposed by Stevanovic (2010), to a standard monetary TVP-VAR model. I find that one factor explains most of the variability in VAR coefficients, while the stochastic volatility parameters vary in the idiosyncratic way. The common factor is highly and positively correlated to the unemployment rate. To incorporate the recent financial crisis, the same exercise is conducted with data updated to 2010Q3. The VAR parameters present an important change after 2007, and the procedure suggests two factors. When applied to a large-dimensional structural factor model, I find that four dynamic factors govern the time instability in almost 700 coefficients.
|
9 |
[en] SCOREDRIVENMODELS.JL: A JULIA PACKAGE FOR GENERALIZED AUTOREGRESSIVE SCORE MODELS / [pt] SCOREDRIVENMODELS.JL: PACOTE EM JULIA PARA MODELOS GENERALIZADOS AUTORREGRESSIVOS COM SCOREGUILHERME MEIRELLES BODIN DE MORAES 03 February 2022 (has links)
[pt] Os modelos orientados por score, também conhecidos como modelos generalizados de score autorregressivo (GAS), representam uma classe de modelos
de séries temporais orientados por observação. Eles possuem propriedades
desejáveis para modelagem de séries temporais, como a capacidade de modelar diferentes distribuições condicionais e considerar parâmetros variantes
no tempo dentro de uma estrutura flexível. Neste trabalho, apresentamos
ScoreDrivenModels.jl, um pacote Julia de código aberto para modelagem,
previsão e simulação de séries temporais usando a estrutura de modelos
baseados em score. O pacote é flexível no que diz respeito à definição do
modelo, permitindo ao usuário especificar a estrutura de atraso e quais parâmetros são variantes no tempo ou constantes. Também é possível considerar várias distribuições, incluindo Beta, Exponencial, Gama, Lognormal,
Normal, Poisson, Student s t e Weibull. A interface fornecida é flexível,
permitindo aos usuários interessados implementar qualquer distribuição e
parametrização desejada. / [en] Score-driven models, also known as generalized autoregressive score (GAS)
models, represent a class of observation-driven time series models. They
possess desirable properties for time series modeling, such as the ability
to model different conditional distributions and to consider time-varying
parameters within a flexible framework. In this dissertation, we present
ScoreDrivenModels.jl, an open-source Julia package for modeling, forecasting, and simulating time series using the framework of score-driven models.
The package is flexible with respect to model definition, allowing the user to
specify the lag structure and which parameters are time-varying or constant.
It is also possible to consider several distributions, including Beta, Exponential, Gamma, Lognormal, Normal, Poisson, Student s t, and Weibull.
The provided interface is flexible, allowing interested users to implement
any desired distribution and parametrization.
|
Page generated in 0.0668 seconds