|
on Econometrics |
| By: | Tong Li (Institute for Fiscal Studies and Vanderbilt University) |
| Abstract: | This paper proposes a formal model selection test for choosing between two competing structural econometric models. The procedure is based on a novel lack-of-fit criterion, namely, the simulated mean squared error of predictions (SMSEP), taking into account the complexity of structural econometric models. It is asymptotically valid for any fixed number of simulations, and allows for any estimator which has a vn asymptotic normality or is superconsistent with a rate at n. The test is bi-directional and applicable to non-nested models which are both possibly misspecified. The asymptotic distribution of the test statistic is derived. The proposed test is general regardless of whether the optimization criteria for estimation of competing models are the same as the SMSEP criterion used for model selection. An empirical application using timber auction data from Oregon is used to illustrate the usefulness and generality of the proposed testing procedure. |
| Keywords: | Lack-of-fit, Model selection tests, Non-nested models, Simulated mean squared error of predictions |
| JEL: | C12 C15 C52 |
| Date: | 2006–08 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:16/06&r=ecm |
| By: | Yingyao Hu; Susanne Schennach (Institute for Fiscal Studies and University of Chicago) |
| Abstract: | While the literature on nonclassical measurement error traditionally relies on the availability of an auxiliary dataset containing correctly measured observations, this paper establishes that the availability of instruments enables the identification of a large class of nonclassical nonlinear errors-in-variables models with continuously distributed variables. The main identifying assumption is that, conditional on the value of the true regressors, some "measure of location" of the distribution of the measurement error (e.g. its mean, mode or median) is equal to zero. The proposed approach relies on the eigenvalue-eigenfunction decomposition of an integral operator associated with specific joint probability densities. The main identifying assumption is used to order the eigenfunctions so that the decomposition is unique. The authors propose a convenient sieve-based estimator, derive its asymptotic properties and investigate its finite-sample behavior through Monte Carlo simulations. An example of application to the relationship between earnings and divorce rates is also provided. |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:17/06&r=ecm |
| By: | Russell Davidson; James MacKinnon |
| Abstract: | We study several tests for the coefficient of the single right-hand-side endogenous variable in a linear equation estimated by instrumental variables. We show that all the test statistics -- Student's t, Anderson-Rubin, Kleibergen's K, and likelihood ratio (LR) -- can be written as functions of six random quantities. This leads to a number of interesting results about the properties of the tests under weak-instrument asymptotics. We then propose several new procedures for bootstrapping the three non-exact test statistics and a conditional version of the LR test. These use more efficient estimates of the parameters of the reduced-form equation than existing procedures. When the best of these new procedures is used, $K$ and conditional LR have excellent performance under the null, and LR also performs very well. However, power considerations suggest that the conditional LR test, bootstrapped using this new procedure when the sample size is not large, is probably the method of choice. |
| JEL: | C10 C12 C15 C30 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:mcl:mclwop:2006-21&r=ecm |
| By: | Hidehiko Ichimura (Institute for Fiscal Studies and University of Tokyo); Sokbae 'Simon' Lee (Institute for Fiscal Studies and University College London) |
| Abstract: | This paper develops a concrete formula for the asymptotic distribution of two-step, possibly non-smooth semiparametric M-estimators under general misspecification. Our regularity conditions are relatively straightforward to verify and also weaker than those available in the literature. The first-stage nonparametric estimation may depend on finite dimensional parameters. We characterize: (1) conditions under which the first-stage estimation of nonparametric components do not affect the asymptotic distribution, (2) conditions under which the asymptotic distribution is affected by the derivatives of the first-stage nonparametric estimator with respect to the finite-dimensional parameters, and (3) conditions under which one can allow non-smooth objective functions. Our framework is illustrated by applying it to three examples: (1) profiled estimation of a single index quantile regression model, (2) semiparametric least squares estimation under model misspecification, and (3) a smoothed matching estimator. |
| Date: | 2006–08 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:15/06&r=ecm |
| By: | Russell Davidson; James MacKinnon |
| Abstract: | We develop a new method, based on the use of polar coordinates, to investigate the existence of moments for instrumental variables and related estimators in the linear regression model. For generalized IV estimators, we obtain familiar results. For JIVE, we obtain the new result that this estimator has no moments at all. Simulation results illustrate the consequences of its lack of moments. |
| JEL: | C10 C12 C30 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:mcl:mclwop:2006-22&r=ecm |
| By: | James J. Heckman (University of Chicago, University College Dublin, American Bar Foundation and IZA Bonn); Sergio Urzua (University of Chicago); Edward Vytlacil (Columbia University) |
| Abstract: | This paper examines the properties of instrumental variables (IV) applied to models with essential heterogeneity, that is, models where responses to interventions are heterogeneous and agents adopt treatments (participate in programs) with at least partial knowledge of their idiosyncratic response. We analyze two-outcome and multiple-outcome models including ordered and unordered choice models. We allow for transition-specific and general instruments. We generalize previous analyses by developing weights for treatment effects for general instruments. We develop a simple test for the presence of essential heterogeneity. We note the asymmetry of the model of essential heterogeneity: outcomes of choices are heterogeneous in a general way; choices are not. When both choices and outcomes are permitted to be symmetrically heterogeneous, the method of IV breaks down for estimating treatment parameters. |
| Keywords: | unobserved heterogeneity, instrumental variables, treatment parameters, policy evaluation |
| JEL: | C31 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp2320&r=ecm |
| By: | Lennard van Gelder; Ad Stokman |
| Abstract: | Formal testing and estimation of nonlinear relations require a substantial number of observations which are typically lacking in annual models. In this paper, a novel two-step procedure is introduced to model nonlinearities in yearly asset-price based leading indicator models for growth. In the first step, quarterly data are explored to test for the presence of regime switches, the identif ication of transition variables and estimation of the accompanying thresholds. In the second step, we implement the quarterly thresholds in the annual indicator models. Results for the US and the Netherlands show that the annual forecasts improve compared to the linear model, despite the poor out-of-sample performance of the quarterly regime switching models. |
| Keywords: | leading indicators; gdp growth; non-linear models. |
| JEL: | C53 E37 |
| Date: | 2006–08 |
| URL: | http://d.repec.org/n?u=RePEc:dnb:dnbwpp:106&r=ecm |
| By: | George Kapetanios; Vincent Labhard; Simon Price |
| Abstract: | In recent years there has been increasing interest in forecasting methods that utilise large data sets, driven partly by the recognition that policymaking institutions need to process large quantities of information. Factor analysis is a popular way of doing this. Forecast combination is another, and it is on this that we concentrate. Bayesian model averaging methods have been widely employed in this area, but a neglected alternative approach employed in this paper uses information theoretic based weights. We consider the use of model averaging in forecasting UK inflation with a large data set from this perspective. We find that an information theoretic model averaging scheme can be a powerful alternative both to the more widely used Bayesian model averaging scheme and to factor models. |
| URL: | http://d.repec.org/n?u=RePEc:boe:boeewp:268&r=ecm |
| By: | Xiujian Chen; Shu Lin; W. Robert Reed (University of Canterbury) |
| Abstract: | Our study revisits Beck and Katz' (1995) comparison of the Parks and PCSE estimators using time-series, cross-sectional data (TSCS). Our innovation is that we construct simulated statistical environments that are designed to approximate actual TSCS data. We pattern our statistical environments after income and tax data on U.S. states from 1960-1999. While PCSE generally does a better job than Parks in estimating standard errors/confidence intervals, it too can be unreliable, sometimes producing standard errors/confidence intervals that are substantially off the mark. Further, we find that the benefits of PCSE can come at a large cost in estimator efficiency. |
| Keywords: | Panel data, Parks model; PCSE estimator; Monte Carlo methods |
| JEL: | C23 C15 |
| Date: | 2006–03–31 |
| URL: | http://d.repec.org/n?u=RePEc:cbt:econwp:06/04&r=ecm |
| By: | Russell Davidson; James MacKinnon |
| Abstract: | We perform an extensive series of Monte Carlo experiments to compare the performance of the "Jacknife Instrumental Variables Estimator", or JIVE, with that of the more familiar 2SLS and LIML estimators. We find no evidence to suggest that JIVE should ever be used. It is always more dispersed than 2SLS, often very much so, and it is almost always inferior to LIML in all respects. Interestingly, JIVE seems to perform particularly badly when the instruments are weak. |
| JEL: | C10 C15 C20 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:mcl:mclwop:2004-02&r=ecm |
| By: | Russell Davidson; Jean-Yves Duclos |
| Abstract: | Asymptotic and bootstrap tests are studied for testing whether there is a relation of stochastic dominance between two distributions. These tests have a null hypothesis of nondominance, with the advantage that, if this null is rejected, then all that is left is dominance. This also leads us to define and focus on {\it restricted} stochastic dominance, the only empirically useful form of dominance relation that we can seek to infer in many settings. One testing procedure that we consider is based on an empirical likelihood ratio. The computations necessary for obtaining a test statistic also provide estimates of the distributions under study that satisfy the null hypothesis, on the frontier between dominance and nondominance. These estimates can be used to perform dominance tests that can turn out to provide much improved reliability of inference compared with the asymptotic tests so far proposed in the literature. |
| JEL: | C10 C12 C15 I32 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:mcl:mclwop:2006-20&r=ecm |
| By: | Annabelle Mourougane |
| Abstract: | The objective of this paper is to develop a short-term indicator-based model to predict quarterly GDP in Canada by efficiently exploiting all available monthly information. To this aim, monthly forecasting equations are estimated using the GDP series published every month by Statistics Canada as well as other monthly indicators. The procedures are automated and the model can be run whenever major monthly data are released, allowing the appropriate choice of the model according to the information set available. The most important gain from this procedure is for the current-quarter forecast when one or two months of GDP data are available, with all monthly models estimated in the paper outperforming a standard quarterly autoregressive model in terms of size of errors. The use of indicators also appears to improve forecasting performance, especially when an average of indicator-based models is used. Real-time forecasting performance of the average model appear to be good, with an apparent stability of the estimates from one update to the next, despite the extensive use of monthly data. The latter result should nonetheless be interpreted with caution and will need to be re-assessed when more data become available. <P>Prévoir le PIB mensuel au Canada <BR>L’objectif de cet article est de développer un modèle d’indicateurs conjoncturels pour prédire le PIB trimestriel au Canada en utilisant de manière efficace toute l’information mensuelle disponible. À cette fin, des équations mensuelles de prévisions de court terme sont estimées en utilisant la série de PIB publiée chaque mois par Statistique Canada et d’autres indicateurs conjoncturels. Les procédures ont été automatisées et le modèle peut être mis à jour chaque fois qu’une donnée importante est publiée, la spécification du modèle variant ainsi en fonctions de l’ensemble des données disponibles. Le gain le plus important de la procédure développée est obtenue pour les prévisions du trimestre courant quand un ou deux mois de données du PIB mensuel sont disponibles. Dans ce cas, tous les modèles mensuels estimés dans cet article ont des erreurs de prévisions inférieures à celle d’un modèle trimestriel autorégressif standard. L’utilisation d’indicateurs conjoncturels améliore les performances en termes de prévisions, en particulier lorsqu’une moyenne de tous les modèles d’indicateurs conjoncturels est utilisée. Les prévisions réalisées en temps réel en faisant la moyenne des différents modèles d’indicateurs conjoncturels se sont avérées de qualité satisfaisante, avec une stabilité apparente des estimations successives, malgré l’utilisation extensive de données mensuelles. Ces résultats doivent toutefois être interprétés avec prudence et devront être vérifiés quand plus de données seront disponibles. |
| Keywords: | Canada, Canada, indicator models, modèle d'indicateurs conjoncturels, monthly GDP, short-term forecasts, real-time estimations, PIB mensuel, prévisions de court terme, estimations en temps réel |
| JEL: | C52 C53 E37 |
| Date: | 2006–09–13 |
| URL: | http://d.repec.org/n?u=RePEc:oec:ecoaaa:515-en&r=ecm |
| By: | Strijbosch,L.W.G.; Heuts,R.M.J.; Moors,J.J.A. (Tilburg University, Center for Economic Research) |
| Abstract: | In inventory management, hierarchical forecasting (HF) is a hot issue : families of items are formed for which total demand is forecasted; total forecast then is broken up to produce forecasts for the individual items. Since HF is a complicated procedure, analytical results are hard to obtain; consequently, most literature is based on simulations and case studies. This paper succeeds in following a more theoretical approach by simplifying the problem : we consider estimation instead of forecasting. So, from a random sample we estimate both total demand and the fraction of this total that individual items take; multiplying these two quantities gives a new estimate of individual demand. Then our research question is: can aggregation of items, followed by fractioning, lead to more accurate estimates of individual demand? Thirdly, a more practical situation is investigated by means of simulation. |
| Keywords: | hierarchical forecasting;aggregation;top-down approach |
| JEL: | C53 |
| Date: | 2006 |
| URL: | http://d.repec.org/n?u=RePEc:dgr:kubcen:200686&r=ecm |
| By: | Richard Blundell (Institute for Fiscal Studies and University College London); Martin Browning (Institute for Fiscal Studies and Nuffield College, Oxford); Ian Crawford (Institute for Fiscal Studies and University of Surrey) |
| Abstract: | This paper uses revealed preference inequalities to provide tight nonparametric bounds on consumer responses to price changes. Price responses are allowed to vary nonparametrically across the income distribution by exploiting microdata on consumer expenditures and incomes over a finite set of discrete relative price changes. This is achieved by combining the theory of revealed preference with the semiparametric estimation of consumer expansion paths (Engel curves). We label these expansion path based bounds as E-bounds. Deviations from revealed preference restrictions aremeasured by preference perturbations which are shown to usefully characterise taste change. |
| Keywords: | Demand responses, relative prices, revealed preference, semiparametric regression, changing tastes |
| JEL: | D12 C14 C43 |
| Date: | 2005–10 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:ifsewp:05/20&r=ecm |
| By: | Laurent Davezies (DEPP and CREST-INSEE); Xavier d’Haultfoeuille (ENSAE, CREST-INSEE and Université Paris I-Panthéon-Sorbonne); Denis Fougère (CNRS, CREST-INSEE, CEPR and IZA Bonn) |
| Abstract: | This paper considers the semiparametric identification of endogenous and exogenous peer effects based on group size variation. We show that Lee (2006)’s linear-in-means model is generically identified, even when all members of the group are not observed. While unnecessary in general, homoskedasticity may be required in special cases to recover all parameters. Extensions to asymmetric responses to peers and binary outcomes are also considered. Once more, most parameters are semiparametrically identified under weak conditions. However, recovering all of them requires more stringent assumptions. Eventually, we bring theoretical evidence that the model is more adapted to small groups. |
| Keywords: | social interactions, linear-in-means model, semiparametric identification |
| JEL: | C14 C21 C25 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp2324&r=ecm |
| By: | Russell Davidson; Emmanuel Flachaire |
| Abstract: | A random sample drawn from a population would appear to offer an ideal opportunity to use the bootstrap in order to perform accurate inference, since the observations of the sample are IID. In this paper, Monte Carlo results suggest that bootstrapping a commonly used index of inequality leads to inference that is not accurate even in very large samples,although inference with poverty indices is satisfactory. We find that the major cause is the extreme sensitivity of many inequality indices to the exact nature of the upper tail of the income distribution. This leads us to study two non-standard bootstraps, the m out of n bootstrap, which is valid in some situations where the standard bootstrap fails, and a bootstrap in which the upper tail is modelled parametrically. Monte Carlo results suggest that accurate inference can be achieved with this last method in moderately large samples. |
| JEL: | C14 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:mcl:mclwop:2005-06&r=ecm |
| By: | Zdenek Hlavka; Michal Pesta |
| Abstract: | State price density (SPD) contains important information concerning market expectations. In existing literature, a constrained estimator of the SPD is found by nonlinear least squares in a suitable Sobolev space. We improve the behavior of this estimator by implementing a covariance structure taking into account the time of the trade and by considering simultaneously both the observed Put and Call option prices. |
| Keywords: | isotonic regression, Sobolev spaces, monotonicity, multiple observations, covariance structure, option price |
| JEL: | C10 C13 C14 C20 C88 G13 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:hum:wpaper:sfb649dp2006-069&r=ecm |
| By: | Jumah, Adusei (Department of Economics and Finance, Institute for Advanced Studies, Vienna, Austria and Department of Economics, University of Vienna, Austria); Kunst, Robert M. (Department of Economics and Finance, Institute for Advanced Studies, Vienna, Austria and Department of Economics, University of Vienna, Austria) |
| Abstract: | This paper explores the seasonal cycles of European agricultural commodity prices. We focus on three food crops (barley, soft and durum wheat) and on beef. We investigate whether seasonality is deterministic or unit-root stochastic and whether seasonal cycle for specific agricultural commodities have converged over time. Finally, we develop time-series models that are capable of forecasting agricultural prices on a quarterly basis. Firstly, we find that seasonal cycles in agricultural commodity prices are mainly deterministic and that evidence on common cycles across countries varies over agricultural commodities. The prediction experiments, however, yield a ranking with respect to accuracy that does not always match the statistical in-sample evidence. |
| Keywords: | Seasonal cycles, Seasonal unit roots, Forecasting, Agricultural commodities |
| JEL: | C32 C53 Q11 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:ihs:ihsesp:192&r=ecm |
| By: | Westerlund, Joakim (Department of Economics, Lund University); Costantini, Mauro (Department of Public Economics) |
| Abstract: | Most econometric methods for testing the proposition of long-run monetary neutrality rely on the assumption that money and real output do not cointegrate, a result that is usually supported by the data. This paper argues that these results can be attributed in part to the low power of univariate tests, and that a violation of the noncointegration assumption is likely to result in a nonrejection of the neutrality proposition. To alleviate this problem, two new and more powerful panel cointegration tests are proposed that can be used under very general conditions. The tests are then applied to a panel covering 10 countries between 1870 and 1986. The results suggest money and real output are cointegrated, and that the neutrality proposition therefore must be rejected. |
| Keywords: | Monetary Neutrality; Panel Cointegration Testing |
| JEL: | C12 C22 C23 E30 E50 |
| Date: | 2006–08–09 |
| URL: | http://d.repec.org/n?u=RePEc:hhs:lunewp:2006_018&r=ecm |
| By: | Peter F. Christoffersen (McGill University and CIRANO); Francis X. Diebold (University of Pennsylvania and NBER); Roberto S. Mariano (School of Economics and Social Sciences, Singapore Management University); Anthony S. Tay (School of Economics and Social Sciences, Singapore Management University); Yiu Kuen Tse (School of Economics and Social Sciences, Singapore Management University) |
| Abstract: | Recent theoretical work has revealed a direct connection between asset return volatility forecastability and asset return sign forecastability. This suggests that the pervasive volatility forecastability in equity returns could, via induced sign forecastability, be used to produce direction-ofchange forecasts useful for market timing. We attempt to do so in the context of two key Asian equity markets, with some success, as assessed by formal probability forecast scoring rules such as the Brier score. An important ingredient is our conditioning not only on conditional variance information, but also conditional skewness and kurtosis information, when forming direction-of-change forecasts. |
| Keywords: | Volatility, variance, skewness, kurtosis, market timing, asset management, asset allocation, portfolio management. |
| JEL: | G10 G12 |
| Date: | 2004–07 |
| URL: | http://d.repec.org/n?u=RePEc:siu:wpaper:02-2005&r=ecm |
| By: | Jesus Crespo Cuaresma; Gernot Doppelhofer |
| Abstract: | We propose a framework for assessing the existence and quantifying the effect of threshold effects in cross-country growth regressions in the presence of model uncertainty. The method is based on Bayesian model averaging tech- niques and generalizes the Bayesian Averaging of Classical Estimates (BACE) method put forward by Sala-i-Martin, Doppelhofer, and Miller (2004). We ap- ply the method presented in this paper to a set of 21 variables that have been found to be robustly related to economic growth in a cross-section of 88 coun- tries. We find no evidence of robust threshold effects generated by the initial level of GDP per capita. However, we find that the proportion of years a country has been open to trade is an important source of nonlinear effects on economic growth. |
| JEL: | C11 C15 O20 O50 |
| Date: | 2006–09 |
| URL: | http://d.repec.org/n?u=RePEc:vie:viennp:0608&r=ecm |