|
on Econometrics |
| By: | Milda Norkuté; Vasilis Sarafidis; Takashi Yamagata |
| Abstract: | This paper develops an instrumental variable (IV) estimator for consistent estimation of dynamic panel data models with a multifactor error structure when both N and T, the cross-sectional and time series dimensions respectively, are large. Our approach projects out the common factors from observed variables, the exogenous regressors of the model, using principal components analysis and then uses the defactored regressors as instruments to estimate the unknown parameters, as in a standard 2SLS procedure. The approach requires estimating solely the common factors contained in the regressors, leaving those that only influence the dependent variable into the errors. Hence our approach is computationally attractive. Since our estimator is based on instrumental variables, it is not subject to the Nickell bias that arises with least squares type estimators in dynamic panel data models. The finite sample performance of the proposed estimator is investigated using simulated data. The results show that the estimator performs well in terms of bias, RMSE and size. The performance of an overidentifying restrictions test is also explored and the evidence suggests that it has high power when the key assumption, strong exogeneity of (a subset of) the regressors, is violated. |
| Date: | 2018–02 |
| URL: | http://d.repec.org/n?u=RePEc:dpr:wpaper:1019&r=ecm |
| By: | Astill, Sam; Taylor, AM Robert |
| Abstract: | We develop tests for the presence of deterministic seasonal behaviour and seasonal mean shifts in a seasonally observed univariate time series. These tests are designed to be asymptotically robust to the order of integration of the series at both the zero and seasonal frequencies. Motivated by the approach of Hylleberg, Engle, Granger and Yoo [1990, Journal of Econometrics vol. 44, pp. 215-238], we base our approach on linear filters of the data which remove any potential unit roots at the frequencies not associated with the deterministic component(s) under test. Test statistics are constructed using the filtered data such that they have well defined limiting null distributions regardless of whether the data are either integrated or stationary at the frequency associated with the deterministic component(s) under test. In the same manner as Vogelsang [1998, Econometrica vol. 66, pp. 123-148], Bunzel and Vogelsang [2005, Journal of Business and Economic Statistics vol. 23, pp. 381-394] and Sayginsoy and Vogelsang [2011, Econometric Theory vol. 27, pp. 992-1025], we scale these statistics by a function of an auxiliary seasonal unit root statistic. This allows us to construct tests which are asymptotically robust to the order of integration of the data at both the zero and seasonal frequencies. Monte Carlo evidence suggests that our proposed tests have good finite sample size and power properties. An empirical application to U.K. GDP indicates the presence of seasonal mean shifts in the data. |
| Keywords: | Seasonality, Seasonal Level Breaks, Seasonal Unit Roots, Robust Tests |
| Date: | 2018–01 |
| URL: | http://d.repec.org/n?u=RePEc:esy:uefcwp:21470&r=ecm |
| By: | Luciano de Castro (University of Iowa); Antonio F. Galvao (University of Arizona); David M. Kaplan (University of Missouri); Xin Liu |
| Abstract: | This paper develops theory for feasible estimation and testing of finite-dimensional parameters identified by general conditional quantile restrictions, under much weaker assumptions than previously seen in the literature. This includes instrumental variables nonlinear quantile regression as a special case. More specifically, we consider a set of unconditional moments implied by the conditional quantile restrictions, providing conditions for local identification. Since estimators based on the sample moments are generally impossible to compute numerically in practice, we study feasible estimators based on smoothed sample moments. We propose a method of moments estimator for exactly identified models, as well as a generalized method of moments estimator for over-identified models. We establish consistency and asymptotic normality of both estimators under general conditions that allow for weakly dependent data and nonlinear structural models. Simulations with iid and dependent data illustrate the finite-sample properties. Our in-depth empirical application concerns the consumption Euler equation derived from quantile utility maximization. Advantages of the quantile Euler equation include robustness to fat tails, decoupling of risk attitude from the elasticity of intertemporal substitution, and log-linearization without any approximation error. For the four countries we examine, the quantile estimates of discount factor and elasticity of intertemporal substitution are economically reasonable for a range of quantiles above the median, even when two-stage least squares estimates are not reasonable. |
| Keywords: | instrumental variables, nonlinear quantile regression, quantile utility maximization |
| JEL: | C31 C32 C36 |
| Date: | 2017–07–10 |
| URL: | http://d.repec.org/n?u=RePEc:umc:wpaper:1803&r=ecm |
| By: | Mazur, Stepan (Örebro University School of Business); Otryakhin, Dmitry (Aarhus University); Podolskij, Mark (Aarhus University) |
| Abstract: | In this paper we investigate the parametric inference for the linear fractional stable motion in high and low frequency setting. The symmetric linear fractional stable motion is a three-parameter family, which constitutes a natural non-Gaussian analogue of the scaled fractional Brownian motion. It is fully characterised by the scaling parameter $\sigma>0$, the self-similarity parameter $H \in (0,1)$ and the stability index $\alpha \in (0,2)$ of the driving stable motion. The parametric estimation of the model is inspired by the limit theory for stationary increments L\'evy moving average processes that has been recently studied in \cite{BLP}. More specifically, we combine (negative) power variation statistics and empirical characteristic functions to obtain consistent estimates of $(\sigma, \alpha, H)$. We present the law of large numbers and some fully feasible weak limit theorems. |
| Keywords: | fractional processes; limit theorems; parametric estimation; stable motion |
| JEL: | C00 C13 |
| Date: | 2018–02–20 |
| URL: | http://d.repec.org/n?u=RePEc:hhs:oruesi:2018_003&r=ecm |
| By: | Mark Fisher (Federal Reserve Bank of Atlanta, USA); Mark J. Jensen (Federal Reserve Bank of Atlanta, USA; Rimini Centre for Economic Analysis) |
| Abstract: | Change point models using hierarchical priors share in the information of each regime when estimating the parameter values of a regime. Because of this sharing hierarchical priors have been very successful when estimating the parameter values of short-lived regimes and predicting the out-of-sample behavior of the regime parameters. However, the hierarchical priors have been parametric. Their parametric nature leads to global shrinkage that biases the estimates of the parameter coefficient of extraordinary regimes towards the value of the average regime. To overcome this shrinkage we model the hierarchical prior nonparametrically by letting the hyperparameter's prior, in other words, the hyperprior, be unknown and modeling it with a Dirichlet processes prior. To apply a nonparametric hierarchical prior to the probability of a break occurring we extend the change point model to a multiple-change-point panel model. The hierarchical prior then shares in the cross-sectional information of the break processes to estimate the transition probabilities. We apply our multiple-change-point panel model to a longitudinal data set of actively managed, US equity, mutual fund returns to measure fund performance and investigate what the chances are of a skilled fund being skilled in the future. |
| Keywords: | Bayesian nonparametric analysis, change points, Dirichlet process, hierarchical priors, mutual fund performance |
| Date: | 2018–02 |
| URL: | http://d.repec.org/n?u=RePEc:rim:rimwps:18-12&r=ecm |
| By: | Biørn, Erik (Dept. of Economics, University of Oslo) |
| Abstract: | Since identification, instrumental variables and variables exclusion, core concepts in econometrics, are entwined, several questions arise: How is identification related to the existence of IVs? How are identification criteria related to omitted variables? Is omission/inclusion of variables from a model’s equations part of the definition of IVs? Is exogeneity a critical claim to an IV? Is ‘omitted variables’ a meaningful term for a single equation when its ‘environment’ is incompletely described? Which are the borderlines between omitted, observable variables, omitted non-modeled variables, latent variables represented by proxies or measurement error mechanisms? These are among the questions addressed in this paper, partly with reference to the conflict between ‘experimentalists’ and ‘structuralists’, specifically relating to: (i) the contrast between ‘rudimentary models’ and models for ‘limited information inference’, (ii) the distinction between exogeneity of variables and the orthogonality claim for IVs and disturbances or errors, (iii) the role of predetermined variables in selecting IVs, and (iv) the ‘omitted variables’ concept and the role of IVs in ‘handling’ such variables, when considering the ‘origin’ of the omission. |
| Keywords: | Identification; Instrumental variables; Omitted variables; Limited information; Experimental approach |
| JEL: | B23 C21 C26 C31 C51 |
| Date: | 2017–12–18 |
| URL: | http://d.repec.org/n?u=RePEc:hhs:osloec:2017_013&r=ecm |
| By: | Francisco (F.) Blasques (VU Amsterdam, The Netherlands); Paolo Gorgi (VU Amsterdam, The Netherlands); Siem Jan (S.J.) Koopman (VU Amsterdam, The Netherlands) |
| Abstract: | We argue that existing methods for the treatment of missing observations in observation-driven models lead to inconsistent inference. We provide a formal proof of this inconsistency for a Gaussian model with time-varying mean. A Monte Carlo simulation study supports this theoretical result and illustrates how the inconsistency problem extends to score-driven and, more generally, to observation-driven models, which include well-known models for conditional volatility. To overcome the problem of inconsistent inference, we propose a novel estimation procedure based on indirect inference. This easy-to-implement method delivers consistent inference. The asymptotic properties are formally derived. Our proposed method shows a promising performance in both a Monte Carlo study and an empirical study concerning the measurement of conditional volatility from financial returns data. |
| Keywords: | missing data; observation-driven models; consistency; indirect inference; volatility |
| JEL: | C22 C58 |
| Date: | 2018–02–09 |
| URL: | http://d.repec.org/n?u=RePEc:tin:wpaper:20180013&r=ecm |
| By: | Aurelio F. Bariviera; Angelo Plastino; George Judge |
| Abstract: | This paper offers a general and comprehensive definition of the day-of-the-week effect. Using symbolic dynamics, we develop a unique test based on ordinal patterns in order to detect it. This test uncovers the fact that the so-called "day-of-the-week" effect is partly an artifact of the hidden correlation structure of the data. We present simulations based on artificial time series as well. Whereas time series generated with long memory are prone to exhibit daily seasonality, pure white noise signals exhibit no pattern preference. Since ours is a non parametric test, it requires no assumptions about the distribution of returns so that it could be a practical alternative to conventional econometric tests. We made also an exhaustive application of the here proposed technique to 83 stock indices around the world. Finally, the paper highlights the relevance of symbolic analysis in economic time series studies. |
| Date: | 2018–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1801.07941&r=ecm |
| By: | Licht, Adrian; Escribano Sáez, Álvaro; Blazsek, Szabolcs Istvan |
| Abstract: | We introduce the Seasonal-QVAR (quasi-vector autoregressive) model for world crude oil production and global real economic activity that identifies the hidden seasonality not found in linear VAR and VARMA models. World crude oil production has an annual seasonality component, and global real economic activity as measured by ocean freight rates has a six-month seasonality component.Seasonal-QVAR is a dynamic conditional score (DCS) model for the multivariate t distribution.Seasonal-VARMA and Seasonal-VAR are special cases of Seasonal-QVAR, this latter being superior to the two former models and also superior to the basic structural model with local level and stochastic seasonality components |
| Keywords: | Crude oil production; Vector autoregressive moving average (VARMA) model; Vector autoregressive (VAR) model; Basic structural model; Nonlinear multivariate dynamic location models; Score-driven stochastic seasonality; Dynamic conditional score (DCS) models |
| JEL: | C52 C32 |
| Date: | 2018–02 |
| URL: | http://d.repec.org/n?u=RePEc:cte:werepe:26316&r=ecm |
| By: | Rumman Khan |
| Abstract: | Pseudo-panels allow estimation of panel models when only repeated cross-sections are available. This involves grouping individuals into cohorts and using the cohort means as if they are observations in a genuine panel. Their practical use is constrained by a lack of consensus on how the pseudo-panels should be formed, particularly to address potential sampling error bias. We show that grouping can also create substantial aggregation bias, calling into question how well pseudo-panels can mimic panel estimates. We create two metrics for assessing the grouping process, one for each potential source of bias. If both metrics are above certain recommended values, the biases from aggregation and sampling error are minimised, meaning results can be interpreted as if they were from genuine panels. |
| Keywords: | pseudo-panel, estimation bias, sampling error; aggregation bias, repeated cross-section; household surveys |
| Date: | 2018 |
| URL: | http://d.repec.org/n?u=RePEc:not:notcre:18/01&r=ecm |
| By: | Yusuke Matsuki |
| Abstract: | Abstract This study develops a simple distribution-free test of monotonicity of conditional expectations. The test is based solely on ordinary least squares (OLS) and exploits the property between conditional expectation and projection; we prove that the monotonicity of a conditional expectation function restricts the sign of a corresponding projection coefficient. The estimated projection coefficient is used for a one-tailed t-test. The test -- which is notably simpler than other monotonicity tests -- is applied to bidding data from Japanese construction procurement auctions to empirically test first-price sealed bid auction models with independent private values (IPV), assuming the data are generated from a symmetric Bayesian Nash equilibrium. We regress the bid level on the number of bidders and use the estimated projection coefficient for testing. We find that the test results depend on public work categories.Length: 25 pages |
| URL: | http://d.repec.org/n?u=RePEc:tcr:wpaper:e110&r=ecm |
| By: | Lorenzo Camponovo (University of Surrey) |
| Abstract: | We study the validity of bootstrap methods in approximating the sampling distribution of penalized GMM estimators with oracle properties. More precisely, we focus on bridge estimators with L_q penalty for 0 |
| JEL: | C12 C13 C52 |
| Date: | 2018–02 |
| URL: | http://d.repec.org/n?u=RePEc:sur:surrec:0618&r=ecm |