|
on Econometrics |
| By: | Nathaniel Beck, Jonathan N. Katz |
| Abstract: | This paper considers random coefficient models (RCMs) for time-series-cross-section data. These models allow for unit to unit variation in the model parameters. After laying out the various models, we assess several issues in specifying RCMs. We then consider the finite sample properties of some standard RCM estimators, and show that the most common one, associated with Hsiao, has very poor properties. These analyses also show that a somewhat awkward combination of estimators based on Swamy's work performs reasonably well; this awkward estimator and a Bayes estimator with an uninformative prior (due to Smith) seem to perform best. But we also see that estimators which assume full pooling perform well unless there is a large degree of unit to unit parameter heterogeneity. We also argue that the various data drive methods (whether classical or empirical Bayes or Bayes with gentle priors) tends to lead to much more heterogeneity than most political scientists would like. We speculate that fully Bayesian models, with a variety of informative priors, may be the best way to approach RCMs. |
| Date: | 2004–09 |
| URL: | http://d.repec.org/n?u=RePEc:clt:sswopa:1205&r=ecm |
| By: | Strikholm, Birgit (Dept. of Economic Statistics, Stockholm School of Economics); Teräsvirta, Timo (Dept. of Economic Statistics, Stockholm School of Economics) |
| Abstract: | <p> In this paper we propose a method for determining the number of regimes in threshold autoregressive models using smooth transition autoregression as a tool. As the smooth transition model is just an approximation to the threshold autoregressive one, no asymptotic properties are claimed for the proposed method. Tests available for testing the adequacy of a smooth transition autoregressive model are applied sequentially to determine the number of regimes. A simulation study is performed in order to find out the finite-sample properties of the procedure and to compare it with two other procedures available in the literature. We find that our method works reasonably well for both single and multiple threshold models. |
| Keywords: | Model specification; model selection criterion; nonlinear modelling; sequential testing; switching regression |
| JEL: | C22 C51 |
| Date: | 2005–01–11 |
| URL: | http://d.repec.org/n?u=RePEc:hhs:hastef:0578&r=ecm |
| By: | Brian P. McCall; John J. McCall |
| Abstract: | This paper is a selected overview of econometric methods for duration models and will appear in the forthcoming book The Economics of Search by the authors. The focus of the paper is on martingale methods for continuous time data and general methods for the analysis of discretetime data including multi-spell models and general life-history models. |
| URL: | http://d.repec.org/n?u=RePEc:hrr:papers:0205&r=ecm |
| By: | D van Dijk; D R Osborn; M Sensier |
| Abstract: | We examine the size properties of tests for causality in variance in the presence of structural breaks in volatility. Extensive Monte Carlo simulations demonstrate that these tests suffer from severe size distortions when such breaks are not taken into account. Pre-testing the series for structural changes in volatility is shown to largely remedy the problem. |
| Date: | 2004 |
| URL: | http://d.repec.org/n?u=RePEc:man:cgbcrp:45&r=ecm |
| By: | Roberto Reno' |
| Abstract: | In this paper a new fully nonparametric estimator of the diffusion coefficient is introduced, based on Fourier analysis of the observed trajectory. The proposed estimator is proved to be consistent and asymptotically normally distributed. After testing the estimator on Monte Carlo simulations, we use it to estimate an univariate model of the short rate with available interest rate data. Data analysis helps shedding new light on the functional form of the diffusion coefficient. |
| JEL: | C14 C6 E43 |
| Date: | 2004–11 |
| URL: | http://d.repec.org/n?u=RePEc:usi:wpaper:440&r=ecm |
| By: | Giulio Zanella |
| Abstract: | This paper tackles the issue of self-selection in social interactions models. I develop a theory of sorting and behavior, when the latter is subject to social influences, extending the model developed by Brock and Durlauf (2001a, 2003) to allow for equilibrium group formation. Individuals choose a group, and a behavior subject to an endogenous social effect. The latter turns out to be a segregating force, and stable equilibria are stratified. The sorting process may induce, inefficiently, multiple behavioral equilibria. Such a theory serves as a means to solve identification and selection problems that may undermine the empirical detection of social effects on individual behavior. I exploit the theoretical model to build a nonlinear (in the social effect) selection correction term. Such a term allows identification, and solves the selection problem that arises when individuals can choose the group whose effect the researcher is trying to disentangle. The resulting econometric model, although relying on strict parametric assumptions, indicates a viable alternative when reliable instrumental variables are not available, or randomized experiments not possible. |
| Keywords: | social interactions, neighborhood effects, sorting, self-selection, nested logit, identification of social effects |
| JEL: | C25 D85 E19 Z13 |
| Date: | 2004–11 |
| URL: | http://d.repec.org/n?u=RePEc:usi:wpaper:442&r=ecm |
| By: | Roberto Reno'; Antonio Roma; Stephen Schaefer |
| Abstract: | In this paper we discuss the estimation of the diffusion coefficient of one-factor models for the short rate via non-parametric methods. We test the estimators proposed by Ait Sahalia (1996a), Stanton (1997) and Bandi and Phillips (2003) on Monte Carlo simulation of the Vasicek and CIR model and show that all estimators, especially that proposed by Ait-Sahalia (1996a), are problematic for values of the mean reversion coefficient typically displayed by interest rate data. Moreover all estimators depend crucially on the choice of the bandwith parameter. Data analysis shows that the estimators lead to different estimates on the data set analyzed by Ait-Sahalia (1996a) and Stanton (1997); moreover we show that the two data set are inherently different. |
| JEL: | C14 E43 |
| Date: | 2004–12 |
| URL: | http://d.repec.org/n?u=RePEc:usi:wpaper:445&r=ecm |
| By: | Willa Chen (Texas A&M University); Rohit Deo (New york University) |
| Abstract: | We study the rate of convergence of moment conditions that have been commonly used in the literature for Generalised Method of Moments (GMM) estimation of short memory latent variable volatility models. We show that when the latent variable possesses long memory, these moment conditions have an n^{1/2-d} rate of convergence where 0<d<0.5 is the memory parameter. The resulting GMM estimators will thus not be ãn consistent. We then provide an alternative set of moment conditions that are ãn consistent and asymptotically normal under long memory in the latent variable, thus allowing for ãn consistent GMM estimation. |
| Keywords: | GMM, long memory, stochastic volatility and durations |
| JEL: | C1 C2 C3 C4 C5 C8 |
| Date: | 2005–01–14 |
| URL: | http://d.repec.org/n?u=RePEc:wpa:wuwpem:0501006&r=ecm |