|
|
System Priors for Econometric Time Series
Andrle, Michal ; Plašil, Miroslav
This paper introduces “system priors” into Bayesian analysis of econometric time series and provides a simple and illustrative application. Unlike priors on individual parameters, system priors offer a simple and efficient way of formulating well-defined and economically meaningful priors about model properties that determine the overall behavior of the model. The generality of system priors is illustrated using an AR(2) process with a prior that its dynamics comes mostly from business-cycle frequencies.
Fulltext:  PDF
|
|
|
Gradient Boosting Machine and Artificial Neural Networks in R and H2O
Sabo, Juraj ; Bašta, Milan (advisor) ; Plašil, Miroslav (referee)
Artificial neural networks are fascinating machine learning algorithms. They used to be considered unreliable and computationally very expensive. Now it is known that modern neural networks can be quite useful, but their computational expensiveness unfortunately remains. Statistical boosting is considered to be one of the most important machine learning ideas. It is based on an ensemble of weak models that together create a powerful learning system. The goal of this thesis is the comparison of these machine learning models on three use cases. The first use case deals with modeling the probability of burglary in the city of Chicago. The second use case is the typical example of customer churn prediction in telecommunication industry and the last use case is related to the problematic of the computer vision. The second goal of this thesis is to introduce an open-source machine learning platform called H2O. It includes, among other things, an interface for R and it is designed to run in standalone mode or on Hadoop. The thesis also includes the introduction into an open-source software library Apache Hadoop that allows for distributed processing of big data. Concretely into its open-source distribution Hortonworks Data Platform.
|
|
|
In the Quest of Measuring the Financial Cycle
Plašil, Miroslav ; Konečný, Tomáš ; Seidler, Jakub ; Hlaváč, Petr
The recent financial crisis has demonstrated the importance of the linkages between the financial sector and the real economy. This paper sets out to develop two complementary methods for assessing the position of the economy in the financial cycle in order to identify emerging imbalances in timely manner. First, we construct a composite indicator using variables representing risk perceptions in the financial sector and calibrate this indicator to capture the credit losses the Czech banking sector experienced during the recent crisis. Second, we focus on the transitions of loans from one risk category to another, which allows us to capture the financial cycle from the perspective of the debt-paying ability of non-financial corporations. Both financial cycle measures can be used by policy makers for a wide range of policy decisions, including that on the setting of the countercyclical capital buffer.
Fulltext: PDF
|
|
|
Similarity and Clustering of Banks: Application to the Credit Exposures of the Czech Banking Sector
Brechler, Josef ; Hausenblas, Václav ; Komárková, Zlatuše ; Plašil, Miroslav
After the recent events in the global financial system there has been significant progress in the literature focusing on the sources of systemic importance of financial institutions. However, the concept of systemic importance is in practice often simplified to the problem of size and contagion due to interbank market interconnectedness. Against this backdrop, we explore additional features of systemic importance stemming from similarities between bank asset portfolios and investigate whether they can contribute to the build-up of systemic risks. We propose a set of descriptive methods to address this aspect empirically in the context of the Czech banking system. Our main findings suggest that the overall measure of the portfolio similarity of individual banks is relatively stable over time and is driven mainly by large and well-established banks. However, we identified several clusters of very similar banks whose market share is small individually but which could become systemically important when considered as a group. After taking into account the credit risk characteristics of portfolios we conclude that the importance of these clusters is even higher.
Fulltext: PDF
|
|
|
The Impact of Financial Variables on Czech Macroeconomic Developments: An Empirical Investigation
Adam, Tomáš ; Plašil, Miroslav
This paper investigates empirically to what extent financial variables can explain macroeconomic developments in the Czech Republic and how the results are sensitive to some (usually reasonable or routinely made) modeling choices. To this end, the dynamic model averaging/selection framework is applied to a universe of (potentially large) time-varying parameter VAR models, which allows one to assess the explanatory power of financial variables at each point in time. Based on a set of 27 competing models and an extensive ensemble of alternative specifications of those models, we find that financial variables were particularly relevant in explaining developments in the lead-up to and during economic downturns. By contrast, in tranquil times, models containing only traditional macroeconomic variables explained macroeconomic dynamics reasonably well. Within the broad set of financial variables considered, credit to the private sector, bank profitability, and leverage seem to be among the most relevant indicators.
Fulltext: PDF
|
|
|
Inflation and the Steeplechase Between Economic Activity Variables
Baxa, Jaromír ; Plašil, Miroslav ; Vašíček, Bořek
A sharp increase in unemployment accompanied by a relatively muted response of inflation during the Great Recession cast further doubts on the validity of the Phillips curve. With the aid of dynamic model averaging (Raftery et al., 2010), this paper aims to highlight that the existence of a systemic relation between real activity and inflation is blurred due to (i) the failure to capture inflationary pressures by means of a single measure of economic activity, and (ii) the existence of a non-linear response of inflation to the driving variable. Based on data for the U.S. and other G7 countries, our results show that the relation between economic activity and inflation is quite sturdy when one allows for more complex assessment of the former. We find that inflation responds to different measures of economic activity across time and space, and no measure of economic activity clearly dominates. The output gap is often outperformed by unemployment-related variables such as the short-term unemployment rate, the unemployment expansion gap, and the unemployment recession gap. Finally, our results confirm a weakening of the inflation–activity relationship (i.e., a flattening of the Phillips curve) in the last decades, which might be attributed to structural changes in the economy and monetary policy, that is robust both across activity measures and across countries.
Fulltext: PDF
|
|
| |
|
|
Empirical Testing of the New Keynesian Phillips Curve in the Czech Republic
Plašil, Miroslav ; Arlt, Josef (advisor) ; Pánková, Václava (referee) ; Komárek, Luboš (referee)
New keynesian Phillips curve (NKPC) has become a central model to study the relation between inflation and real economic activity, notably in the framework of optimal monetary policy design. However, some recent evidence suggests that empirical data are usually at odds with the underlying theory. The model due to its inherent structure represents a statistical challenge in its own right. Since Galí and Gertler (1999) published their seminal paper introducing estimation via GMM techniques, they have triggered a heated debate on its empirical relevance. Their approach has been heavily criticised by later authors, mainly on the grounds of questionable behaviour of GMM estimator in the NKPC context and/or its small sample properties. The common criticism includes sensitivity to the choice of instrument set, weak identification and small sample bias. In this thesis I propose a new estimation strategy that provides a remedy to above mentioned shortcomings and allows to obtain reliable estimates. The procedure exploits recent advances in GMM theory as well as in other fields of statistics, in particular in the area of time series factor analysis and bootstrap. The proposed estimation strategy consists of several consecutive steps: first, to reduce a small sample bias resulting from excessive use of instruments I summarize all available information by employing factor analysis and include estimated factors into information set. In the second step I use statistical information criteria to select optimal instruments and eventually I obtain confidence intervals on parameters using bootstrap method. In NKPC context all these methods were used for the first time and can also be used independently. Their combination however provides synergistic effect that helps to improve the properties of estimates and to check the efficiency of given steps. Obtained results suggest that NKPC model can explain Czech inflation dynamics fairly well and provide some support for underlying theory. Among other things the results imply that the policy of disinflation may not be as costly with respect to a loss in aggregate product as earlier versions of Phillips curve would indicate. However, finding a good proxy for real economic activity has proved to be a difficult task. In particular we demonstrated that results are conditional on how the measure is calculated, some measures even showed countercyclical behaviour. This issue -- in the thesis discussed only in passing -- is a subject of future research. In addition to the proposed strategy and provided parameter estimates the thesis brings some partial simulation-based findings. Simulations elaborate on earlier literature on naive bootstrap in GMM context and study performance of bootstrap modifications of unit root and KPSS test.
|
|
|
Correspondence analysis
Konrádová, Lucie ; Hebák, Petr (advisor) ; Plašil, Miroslav (referee)
The aim of this thesis is to introduce statistical method called Correspondence analysis as a strong instrument for exploratory data analysis. The main purpose is to understand how to interpret the correspondence map, the graphical output of this method, correctly. The method is presented both in its simple version, and its extension to multivariate data. Usage of method is demonstrated on data of non-financial subjects of Czech republic, which are entered in the register of economic subjects.
|
guest ::
