
on Econometric Time Series 
By:  Kapetanios, George; Marcellino, Massimiliano 
Abstract:  The estimation of dynamic factor models for large sets of variables has attracted considerable attention recently, due to the increased availability of large datasets. In this paper we propose a new parametric methodology for estimating factors from large datasets based on state space models and discuss its theoretical properties. In particular, we show that it is possible to estimate consistently the factor space. We also develop a consistent information criterion for the determination of the number of factors to be included in the model. Finally, we conduct a set of simulation experiments that show that our approach compares well with existing alternatives. 
Keywords:  factor models; principal components; subspace algorithms 
JEL:  C32 C51 E52 
Date:  2006–04 
URL:  http://d.repec.org/n?u=RePEc:cpr:ceprdp:5620&r=ets 
By:  Kapetanios, George; Marcellino, Massimiliano 
Abstract:  The estimation of structural dynamic factor models (DFMs) for large sets of variables is attracting considerable attention. In this paper we briefly review the underlying theory and then compare the impulse response functions resulting from two alternative estimation methods for the DFM. Finally, as an example, we reconsider the issue of the identification of the driving forces of the US economy, using data for about 150 macroeconomic variables. 
Keywords:  factor models; principal components; structural identification; structural VAR; subspace algorithms 
JEL:  C32 C51 E52 
Date:  2006–04 
URL:  http://d.repec.org/n?u=RePEc:cpr:ceprdp:5621&r=ets 
By:  Roberto Patuelli (Department of Spatial Economics, Vrije Universiteit Amsterdam); Aura Reggiani (Department of Economics, University of Bologna, Italy); Peter Nijkamp (Department of Spatial Economics, Vrije Universiteit Amsterdam); Uwe Blien (Institut für Arbeitsmarkt und Berufsforschung (IAB), Nuremberg) 
Abstract:  In this paper, a set of neural network (NN) models is developed to compute shortterm forecasts of regional employment patterns in Germany. NNs are modern statistical tools based on learning algorithms that are able to process large amounts of data. NNs are enjoying increasing interest in several fields, because of their effectiveness in handling complex data sets when the functional relationship between dependent and independent variables is not explicitly specified. The present paper compares two NN methodologies. First, it uses NNs to forecast regional employment in both the former West and East Germany. Each model implemented computes single estimates of employment growth rates for each German district, with a 2year forecasting range. Next, additional forecasts are computed, by combining the NN methodology with ShiftShare Analysis (SSA). Since SSA aims to identify variations observed among the labour districts, its results are used as further explanatory variables in the NN models. The data set used in our experiments consists of a panel of 439 German districts. Because of differences in the size and time horizons of the data, the forecasts for West and East Germany are computed separately. The outofsample forecasting ability of the models is evaluated by means of several appropriate statistical indicators. 
Keywords:  networks; forecasts; regional employment; shiftshare analysis; shiftshare regression 
JEL:  C23 E27 R12 
Date:  2006–02–17 
URL:  http://d.repec.org/n?u=RePEc:dgr:uvatin:20060020&r=ets 
By:  Cars Hommes (Faculty of Economics and Econometrics, Universiteit van Amsterdam); Sebastiano Manzan (University of Leicester) 
Abstract:  This short paper is a comment on ``Testing for Nonlinear Structure and Chaos in Economic Time Series'' by Catherine Kyrtsou and Apostolos Serletis. We summarize their main results and discuss some of their conclusions concerning the role of outliers and noisy chaos. In particular, we include some new simulations to investigate whether economic time series may be characterized by low dimensional noisy chaos. 
Keywords:  nonlinearity; chaos; noise; selforganized criticality; MackeyGlassGARCH 
JEL:  C22 C45 C61 
Date:  2006–03–27 
URL:  http://d.repec.org/n?u=RePEc:dgr:uvatin:20060030&r=ets 
By:  Elena Pesavento; Barbara Rossi 
Abstract:  This paper is a comprehensive comparison of existing methods for constructing confidence bands for univariate impulse response functions in the presence of high persistence. Monte Carlo results show that Kilian (1998a), Wright (2000), Gospodinov (2004) and Pesavento and Rossi (2005) have favorable coverage properties, although they differ in terms of robustness at various horizons, median unbiasedness, and reliability in the possible presence of a unit or mildly explosive root. On the other hand, methods like Runkle's (1987) bootstrap, Andrews and Chen (1994), and regressions in levels or first differences (even when based on pretests) may not have accurate coverage properties. The paper makes recommendations as to the appropriateness of each method in empirical work. 
Date:  2006–03 
URL:  http://d.repec.org/n?u=RePEc:emo:wp2003:0603&r=ets 
By:  Brännäs, Kurt (Department of Economics, Umeå University); Lönnbark, Carl (Department of Economics, Umeå University) 
Abstract:  This note gives dynamic effects of discrete and continuous explanatory variables for count data or integervalued moving average models. An illustration based on a model for the number of transactions in a stock is included. 
Keywords:  INMA model; Marginal effect; Intraday; Financial data 
JEL:  C22 C25 G12 
Date:  2006–04–05 
URL:  http://d.repec.org/n?u=RePEc:hhs:umnees:0679&r=ets 
By:  Quoreshi, Shahiduzzaman (Department of Economics, Umeå University) 
Abstract:  A model to account for the long memory property in a count data framework <p> is proposed and applied to high frequency stock transactions data. <p> The unconditional and conditional first and second order moments are <p> given. The CLS and FGLS estimators are discussed. In its empirical <p> application to two stock series for AstraZeneca and Ericsson B, we find <p> that both series have a fractional integration property. 
Keywords:  Intraday; High frequency; Estimation; Fractional integration; Reaction time 
JEL:  C13 C22 C25 C51 G12 G14 
Date:  2006–04–11 
URL:  http://d.repec.org/n?u=RePEc:hhs:umnees:0673&r=ets 
By:  Quoreshi, Shahiduzzaman (Department of Economics, Umeå University) 
Abstract:  A vector integervalued moving average (VINMA) model is introduced. <p> The VINMA model allows for both positive and negative correlations <p> between the counts. The conditional and unconditional first and second <p> order moments are obtained. The CLS and FGLS estimators are discussed. <p> The model is capable of capturing the covariance between and <p> within intraday time series of transaction frequency data due to macroeconomic <p> news and news related to a specific stock. Empirically, it is <p> found that the spillover effect from Ericsson B to AstraZeneca is larger <p> than that from AstraZeneca to Ericsson B 
Keywords:  Count data; Intraday; Time series; Estimation; Reaction 
JEL:  C13 C22 C25 C51 G12 G14 
Date:  2006–04–11 
URL:  http://d.repec.org/n?u=RePEc:hhs:umnees:0674&r=ets 
By:  Quoreshi, Shahiduzzaman (Department of Economics, Umeå University) 
Abstract:  This thesis comprises four papers concerning modelling of financial count data. Paper [1], [2] <p> and [3] advance the integervalued moving average model (INMA), a special case of integervalued <p> autoregressive moving average (INARMA) model class, and apply the models to the number of <p> stock transactions in intraday data. Paper [4] focuses on modelling the long memory property of <p> time series of count data and on applying the model in a financial setting. <p> Paper [1] advances the INMA model to model the number of transactions in stocks in intraday <p> data. The conditional mean and variance properties are discussed and model extensions to <p> include, e.g., explanatory variables are offered. Least squares and generalized method of moment <p> estimators are presented. In a small Monte Carlo study a feasible least squares estimator comes out <p> as the best choice. Empirically we find support for the use of longlag moving average models in a <p> Swedish stock series. There is evidence of asymmetric effects of news about prices on the number <p> of transactions. <p> Paper [2] introduces a bivariate integervalued moving average (BINMA) model and applies the <p> BINMA model to the number of stock transactions in intraday data. The BINMA model allows <p> for both positive and negative correlations between the count data series. The study shows that <p> the correlation between series in the BINMA model is always smaller than one in an absolute sense. <p> The conditional mean, variance and covariance are given. Model extensions to include explanatory <p> variables are suggested. Using the BINMA model for AstraZeneca and Ericsson B it is found that <p> there is positive correlation between the stock transactions series. Empirically, we find support for <p> the use of longlag bivariate moving average models for the two series. <p> Paper [3] introduces a vector integervalued moving average (VINMA) model. The VINMA <p> model allows for both positive and negative correlations between the counts. The conditional and <p> unconditional first and second order moments are obtained. The CLS and FGLS estimators are <p> discussed. The model is capable of capturing the covariance between and within intraday time <p> series of transaction frequency data due to macroeconomic news and news related to a specific <p> stock. Empirically, it is found that the spillover effect from Ericsson B to AstraZeneca is larger <p> than that from AstraZeneca to Ericsson B. <p> Paper [4] develops models to account for the long memory property in a count data framework <p> and applies the models to high frequency stock transactions data. The unconditional and conditional <p> first and second order moments are given. The CLS and FGLS estimators are discussed. <p> In its empirical application to two stock series for AstraZeneca and Ericsson B, we find that both <p> series have a fractional integration property. 
Keywords:  Count data; Intraday; High frequency; Time series; Estimation; Long memory; Finance 
JEL:  C13 C22 C25 C51 G12 G14 
Date:  2006–04–11 
URL:  http://d.repec.org/n?u=RePEc:hhs:umnees:0675&r=ets 
By:  Dirk Baur; Brian M. Lucey 
Abstract:  This paper analyzes the existence of flighttoquality from stocks to bonds and contagion between the two asset classes. Flighttoquality is present if correlations between stocks and bonds strongly decrease in falling stock markets since this constitutes a movement of the asset classes in opposite directions. A movement in the same direction characterized by strongly increasing correlations in falling stock markets implies contagion across asset classes. We estimate dynamic conditional correlations and analyze normal and extreme changes of these correlations through time without an a priori specification of any crisis period. Daily MSCI stock and government bond returns are analyzed for a selection of European countries and the US. Our findings show that the correlation between the asset classes is characterized by large fluctuations and negative on average for the whole sample period. Extreme negative and positive correlation changes explained with flighttoquality and contagion are relatively frequent phenomena. Examples of flighttoquality are in the Asian and Russian crisis 1997 and 1998 and contagion is found after September 11. Controlling for the regime of correlations further shows that stock market volatility contributes to flighttoquality and bond volatility to contagion. 
Keywords:  flighttoquality, contagion, multivariate GARCH 
JEL:  F36 G11 G14 G15 
Date:  2006–04–05 
URL:  http://d.repec.org/n?u=RePEc:iis:dispap:iiisdp122&r=ets 
By:  Raj Aggarwal; Brian M. Lucey; Sunil K. Mohanty 
Abstract:  An important puzzle in international finance is the failure of the forward exchange rate to be a rational forecast of the future spot rate. It has often been suggested that this puzzle may be resolved by using better statistical procedures that correct for both nonstationarity and nonnormality in the data. We document that even after accounting for nonstationarity, nonnormality, and heteroscedasticity using parametric and nonparametric tests on data for over a quarter century, US dollar forward rates for horizons ranging from one to twelve months for the major currencies, the British pound, Japanese yen, Swiss franc, and the German mark, are generally not rational forecasts of future spot rates. These findings of nonrationality in forward exchange rates for the major currencies continue to be puzzling especially as these foreign exchange markets are some of the most liquid asset markets with very low trading costs. 
Keywords:  flighttoquality, contagion, multivariate GARCH 
JEL:  F31 G14 F47 G15 
Date:  2006–04–05 
URL:  http://d.repec.org/n?u=RePEc:iis:dispap:iiisdp123&r=ets 
By:  Hugo Kruiniger (Queen Mary, University of London) 
Abstract:  In this paper we consider GMM based estimation and inference for the panel AR(1) model when the data are persistent and the time dimension of the panel is fixed. We find that the nature of the weak instruments problem of the ArellanoBond estimator depends on the distributional properties of the initial observations. Subsequently, we derive local asymptotic approximations to the finite sample distributions of the ArellanoBond estimator and the System estimator, respectively, under a variety of distributional assumptions about the initial observations and discuss the implications of the results we obtain for doing inference. We also propose two LM type panel unit root tests. 
Keywords:  Dynamic panel data, GMM, Weak instruments, Weak identification, Local asymptotics, Multiindex asymptotics, Diagonal path asymptotics, LM test, Panel unit root test 
JEL:  C12 C13 C23 
Date:  2006–04 
URL:  http://d.repec.org/n?u=RePEc:qmw:qmwecw:wp560&r=ets 
By:  Gang Liu, Terje Skjerpen, Anders Rygh Swensen and Kjetil Telle (Statistics Norway) 
Abstract:  Timeseries regressions including nonlinear transformations of an integrated variable are not uncommon in various fields of economics. In particular, within the Environmental Kuznets Curve (EKC) literature, where the effect on the environment of income levels is investigated, it is standard procedure to include a third order polynomial in the income variable. When the income variable is an I(1)variable and this variable is also included nonlinearly in the regression relation, the properties of the estimators and standard inferential procedures are unknown. Surprisingly, such problems have received rather limited attention in applied work, and appear disregarded in the EKC literature. We investigate the properties of the estimators of longrun parameters using MonteCarlo simulations. We find that the mean of the ordinary least squares estimates are very similar to the true values and that standard testing procedures based on normality behave rather well. 
Keywords:  Emissions; Environmental Kuznets Curve; Unit Roots; Monte Carlo Simulations 
JEL:  C15 C16 C22 C32 O13 
Date:  2006–01 
URL:  http://d.repec.org/n?u=RePEc:ssb:dispap:443&r=ets 
By:  Juan J. Dolado; Jesús Gonzalo; Laura Mayoral 
Abstract:  This paper proposes a new timedomain test of a process being I(d), 0 < d = 1, under the null, against the alternative of being I(0) with deterministic components subject to structural breaks at known or unknown dates, with the goal of disentangling the existing identification issue between longmemory and structural breaks. Denoting by AB(t) the di?erent types of structural breaks in the deterministic components of a time series considered by Perron (1989), the test statistic proposed here is based on the tratio (or the infimum of a sequence of tratios) of the estimated coefficient on yt1 in an OLS regression of ?dyt on a simple transformation of the abovementioned deterministic components and yt1, possibly augmented by a suitable number of lags of ?dyt to account for serial correlation in the error terms. The case where d = 1 coincides with the Perron (1989) or the Zivot and Andrews (1992) approaches if the break date is known or unknown, respectively. The statistic is labelled as the SBFDF (Structural BreakFractional Dickey Fuller) test, since it is based on the same principles as the wellknown DickeyFuller unit root test. Both its asymptotic behavior and finite sample properties are analyzed, and two empirical applications are provided. 
Date:  2005–09 
URL:  http://d.repec.org/n?u=RePEc:upf:upfgen:954&r=ets 
By:  Laura Mayoral 
Abstract:  Although it is commonly accepted that most macroeconomic variables are nonstationary, it is often difficult to identify the source of the nonstationarity. In particular, it is wellknown that integrated and short memory models containing trending components that may display sudden changes in their parameters share some statistical properties that make their identification a hard task. The goal of this paper is to extend the classical testing framework for I(1) versus I(0)+ breaks by considering a a more general class of models under the null hypothesis: nonstationary fractionally integrated (FI) processes. A similar identification problem holds in this broader setting which is shown to be a relevant issue from both a statistical and an economic perspective. The proposed test is developed in the time domain and is very simple to compute. The asymptotic properties of the new technique are derived and it is shown by simulation that it is very wellbehaved in finite samples. To illustrate the usefulness of the proposed technique, an application using inflation data is also provided. 
Date:  2005–10 
URL:  http://d.repec.org/n?u=RePEc:upf:upfgen:956&r=ets 
By:  Juan J. Dolado; Jesús Gonzalo; Laura Mayoral 
Abstract:  This paper discusses the role of deterministic components in the DGP and in the auxiliary regression model which underlies the implementation of the Fractional DickeyFuller (FDF) test for I(1) against I(d) processes with d € [0, 1). This is an important test in many economic applications because I(d) processess with d < 1 are meanreverting although, when 0.5 = d < 1, like I(1) processes, they are nonstationary. We show how simple is the implementation of the FDF in these situations, and argue that it has better properties than LM tests. A simple testing strategy entailing only asymptotically normally distributed tests is also proposed. Finally, an empirical application is provided where the FDF test allowing for deterministic components is used to test for longmemory in the per capita GDP of several OECD countries, an issue that has important consequences to discriminate between growth theories, and on which there is some controversy. 
Keywords:  Deterministic components, DickeyFuller test, fractionally DickeyFuller test, fractional processes, long memory, trends, unit roots 
JEL:  C12 C22 C40 
Date:  2005–02 
URL:  http://d.repec.org/n?u=RePEc:upf:upfgen:957&r=ets 
By:  Laura Mayoral 
Abstract:  The statistical properties of inflation and, in particular, its degree of persistence and stability over time is a subject of intense debate and no consensus has been achieved yet. The goal of this paper is to analyze this controversy using a general approach, with the aim of providing a plausible explanation for the existing contradictory results. We consider the inflation rates of 21 OECD countries which are modelled as fractionally integrated (FI) processes. First, we show analytically that FI can appear in inflation rates after aggregating individual prices from firms that face different costs of adjusting their prices. Then, we provide robust empirical evidence supporting the FI hypothesis using both classical and Bayesian techniques. Next, we estimate impulse response functions and other scalar measures of persistence, achieving an accurate picture of this property and its variation across countries. It is shown that the application of some popular tools for measuring persistence, such as the sum of the AR coefficients, could lead to erroneous conclusions if fractional integration is present. Finally, we explore the existence of changes in inflation inertia using a novel approach. We conclude that the persistence of inflation is very high (although nonpermanent) in most postindustrial countries and that it has remained basically unchanged over the last four decades. 
Keywords:  Inflation persistence, persistence stability, ARFIMA models, long memory, structural breaks, bayesian estimations 
JEL:  C22 E31 
Date:  2005–02 
URL:  http://d.repec.org/n?u=RePEc:upf:upfgen:958&r=ets 
By:  Laura Mayoral 
Abstract:  A new parametric minimum distance timedomain estimator for ARFIMA processes is introduced in this paper. The proposed estimator minimizes the sum of squared correlations of residuals obtained after filtering a series through ARFIMA parameters. The estimator is easy to compute and is consistent and asymptotically normally distributed for fractionally integrated (FI) processes with an integration order d strictly greater than 0.75. Therefore, it can be applied to both stationary and nonstationary processes. Deterministic components are also allowed in the DGP. Furthermore, as a byproduct, the estimation procedure provides an immediate check on the adequacy of the specified model. This is so because the criterion function, when evaluated at the estimated values, coincides with the BoxPierce goodness of fit statistic. Empirical applications and MonteCarlo simulations supporting the analytical results and showing the good performance of the estimator in finite samples are also provided. 
Keywords:  Fractional integration, nonstationary longmemory time series, minimum distance estimation 
JEL:  C13 C22 
Date:  2006–01 
URL:  http://d.repec.org/n?u=RePEc:upf:upfgen:959&r=ets 
By:  Ilias Lekkos (EFG Eurobank); Costas Milas (Keele University); Theodore Panagiotidis (Loughborough University) 
Abstract:  This paper explores the ability of factor models to predict the dynamics of US and UK interest rate swap spreads within a linear and a nonlinear framework. We reject linearity for the US and UK swap spreads in favour of a regimeswitching smooth transition vector autoregressive (STVAR) model, where the switching between regimes is controlled by the slope of the US term structure of interest rates. We compare the ability of the STVAR model to predict swap spreads with that of a nonlinear nearestneighbours model as well as that of linear AR and VAR models. We find some evidence that the nonlinear models predict better than the linear ones. At short horizons, the nearestneighbours (NN) model predicts better than the STVAR model US swap spreads in periods of increasing risk conditions and UK swap spreads in periods of decreasing risk conditions. At long horizons, the STVAR model increases its forecasting ability over the linear models, whereas the NN model does not outperform the rest of the models. 
Keywords:  Interest rate swap spreads, term structure of interest rates, factor models, regime switching, smooth transition models, nearestneighbours, forecasting. 
JEL:  C51 C52 C53 E43 
Date:  2006–03 
URL:  http://d.repec.org/n?u=RePEc:lbo:lbowps:2006_6&r=ets 