Applied Econometrics: A Practical Guide (Routledge Advanced Texts in Economics and Finance)

Description

Applied Econometrics: A Practical Guide is an extremely user-friendly and application-focused book on econometrics. Unlike many econometrics textbooks which are heavily theoretical on abstractions, this book is perfect for beginners and promises simplicity and practicality to the understanding of econometric models. Written in an easy-to-read manner, the book begins with hypothesis testing and moves forth to simple and multiple regression models. It also includes advanced topics: Endogeneity and Two-stage Least SquaresSimultaneous Equations ModelsPanel Data ModelsQualitative and Limited Dependent Variable ModelsVector Autoregressive (VAR) ModelsAutocorrelation and ARCH/GARCH ModelsUnit Root and Cointegration The book also illustrates the use of computer software (EViews, SAS and R) for economic estimating and modeling. Its practical applications make the book an instrumental, go-to guide for solid foundation in the fundamentals of econometrics. In addition, this book includes excerpts from relevant articles published in top-tier academic journals. This integration of published articles helps the readers to understand how econometric models are applied to real-world use cases. Cover Half Title Series Title Copyright Contents List of figures List of tables Preface Acknowledgments 1 Review of estimation and hypothesis tests 1.1 The problem 1.2 Population and sample 1.3 Hypotheses 1.4 Test statistic and its sampling distribution 1.5 Type I and Type II errors 1.6 Significance level 1.7 p-value 1.8 Powerful tests 1.9 Properties of estimators 1.10 Summary Review questions 2 Simple linear regression models 2.1 Introduction 2.1.1 A hypothetical example 2.1.2 Population regression line 2.1.3 Stochastic specification for individuals 2.2 Ordinary least squares estimation 2.3 Coefficient of determination (R2) 2.3.1 Definition and interpretation of R2 2.3.2 Application of R2: Morck, Yeung and Yu (2000) 2.3.3 Application of R2: Dechow (1994) 2.4 Hypothesis test 2.4.1 Testing H0 : β1 = 0 vs. H1 : β1 ≠ 0 2.4.2 Testing H0 : β1 = c vs. H1 : β1 ≠ c (c is a constant) 2.5 The model 2.5.1 Key assumptions 2.5.2 Gauss-Markov Theorem 2.5.3 Consistency of the OLS estimators 2.5.4 Remarks on model specification 2.6 Functional forms 2.6.1 Log-log linear models 2.6.2 Log-linear models 2.7 Effects of changing measurement units and levels 2.7.1 Changes of measurement units 2.7.2 Changes in the levels 2.8 Summary Review questions References Appendix 2 How to use EViews, SAS and R 3 Multiple linear regression models 3.1 The basic model 3.2 Ordinary least squares estimation 3.2.1 Obtaining the OLS estimates 3.2.2 Interpretation of regression coefficients 3.3 Estimation bias due to correlated-omitted variables 3.4 R2 and the adjusted R2 3.4.1 Definition and interpretation of R2 3.4.2 Adjusted R2 3.5 Hypothesis test 3.6 Model selection 3.6.1 General-to-simple approach 3.6.2 A comment on hypothesis testing 3.6.3 Guidelines for model selection 3.7 Applications 3.7.1 Mitton (2002) 3.7.2 McAlister, Srinivasan and Kim (2007) 3.7.3 Collins, Pincus and Xie (1999) 3.7.4 Angrist and Pixchke (2009, pp. 64–68) 3.8 Summary Review questions References Appendix 3A Hypothesis test using EViews and SAS Appendix 3B Geometric interpretation of the OLS regression equation 4 Dummy explanatory variables 4.1 Dummy variables for different intercepts 4.1.1 When there are two categories 4.1.2 When there are more than two categories 4.1.3 Interpretation when the dependent variable is in logarithm 4.1.4 Application: Mitton (2002) 4.1.5 Application: Hakes and Sauer (2006) 4.2 Dummy variables for different slopes 4.2.1 Use of a cross product with a dummy variable 4.2.2 Application: Basu (1997) 4.3 Structural stability of regression models 4.3.1 Test by splitting the sample (Chow test) 4.3.2 Test using dummy variables 4.4 Piecewise linear regression models 4.4.1 Using dummy variables 4.4.2 Using quantitative variables only 4.4.3 Morck, Shleifer and Vishny (1988) 4.5 Summary Review questions References Appendix 4 Dummy variables in EViews and SAS 5 More on multiple regression analysis 5.1 Multicollinearity 5.1.1 Consequences of multicollinearity 5.1.2 Solutions 5.2 Heteroscedasticity 5.2.1 Consequences of heteroscedasticity 5.2.2 Testing for heteroscedasticity 5.2.3 Application: Mitton (2002) 5.3 More on functional form 5.3.1 Quadratic function 5.3.2 Interaction terms 5.4 Applications 5.4.1 Bharadwaj, Tuli and Bonfrer (2011) 5.4.2 Ghosh and Moon (2005) 5.4.3 Arora and Vamvakidis (2005) 5.5 Summary Review questions References Appendix 5 Testing and correcting for heteroscedasticity 6 Endogeneity and two-stage least squares estimation 6.1 Measurement errors 6.1.1 Measurement errors in the dependent variable 6.1.2 Measurement errors in an explanatory variable 6.2 Specification errors 6.2.1 Omitted variables 6.2.2 Inclusion of irrelevant variables 6.2.3 A guideline for model selection 6.3 Two-stage least squares estimation 6.4 Generalized method of moments (GMM) 6.4.1 GMM vs. 2SLS 6.5 Tests for endogeneity 6.5.1 Ramsey (1969) test 6.5.2 Hausman (1978) test 6.6 Applications 6.6.1 Dechow, Sloan and Sweeney (1995) 6.6.2 Beaver, Lambert and Ryan (1987) 6.6.3 Himmelberg and Petersen (1994) 6.7 Summary Review questions References Appendix 6A Estimation of 2SLS and GMM using EViews and SAS Appendix 6B Hausman test for endogeneity using EViews and SAS 7 Models for panel data 7.1 One big regression 7.2 Fixed effects model 7.2.1 Using time dummies (for bt) 7.2.2 Using cross-section dummies (for a1) 7.2.3 Applying transformations 7.3 Applications 7.3.1 Cornwell and Trumbull (1994) 7.3.2 Blackburn and Neumark (1992) 7.3.3 Garin-Munoz (2006) 7.3.4 Tuli, Bharadwaj and Kohli (2010) 7.4 Random effects 7.5 Fixed vs. random effects models 7.6 Summary Review questions References Appendix 7A Controlling for fixed effects using EViews and SAS Appendix 7B Is it always possible to control for unit-specific effects? 8 Simultaneous equations models 8.1 Model description 8.2 Estimation methods 8.2.1 Two-stage least squares (2SLS) 8.2.2 Three-stage least squares (3SLS) 8.2.3 Generalized method of moments (GMM) 8.2.4 Full-information maximum likelihood (FIML) 8.3 Identification problem 8.4 Applications 8.4.1 Cornwell and Trumbull (1994) 8.4.2 Beaver, McAnally and Stinson (1997) 8.4.3 Barton (2001) 8.4.4 Datta and Agarwal (2004) 8.5 Summary Review questions References Appendix 8 Estimation of simultaneous equations models using EViews and SAS 9 Vector autoregressive (VAR) models 9.1 VAR models 9.2 Estimation of VAR models 9.3 Granger-causality test 9.4 Forecasting 9.5 Impulse-response analysis 9.6 Variance decomposition analysis 9.7 Applications 9.7.1 Stock and Watson (2001) 9.7.2 Zhang, Fan, Tsai and Wei (2008) 9.7.3 Trusov, Bucklin and Pausels (2009) 9.8 Summary Review questions References Appendix 9 Estimation and analysis of VAR models using SAS 10 Autocorrelation and ARCH/GARCH 10.1 Autocorrelation 10.1.1 Consequences of autocorrelation 10.1.2 Test for autocorrelation 10.1.3 Estimation of autocorrelation 10.2 ARCH-type models 10.2.1 ARCH model 10.2.2 GARCH (Generalized ARCH) model 10.2.3 TGARCH (Threshold GARCH) model 10.2.4 EGARCH (Exponential GARCH) model 10.2.5 GARCH-M model 10.3 Applications 10.3.1 Wang, Salin and Leatham (2002) 10.3.2 Zhang, Fan, Tsai and Wei (2008) 10.3.3 Value at Risk (VaR) 10.4 Summary Review questions References Appendix 10A Test and estimation of autocorrelation using EViews and SAS Appendix 10B Test and estimation of ARCH/GARCH models using SAS 11 Unit root, cointegration and error correction model 11.1 Spurious regression 11.2 Stationary and nonstationary time series 11.3 Deterministic and stochastic trends 11.4 Unit root tests 11.4.1 Dickey-Fuller (DF) test 11.4.2 Augmented Dickey-Fuller (ADF) test 11.4.3 Example: unit root test using EViews 11.5 Cointegration 11.5.1 Tests for cointegration 11.5.2 Vector error correction models (VECMs) 11.5.3 Example: test and estimation of cointegration using EViews 11.6 Applications 11.6.1 Stock and Watson (1988) 11.6.2 Baillie and Selover (1987) 11.6.3 Granger (1988) 11.6.4 Dritsakis (2004) 11.6.5 Ghosh (1993) 11.7 Summary Review questions References Appendix 11A Unit root test using SAS Appendix 11B Johansen test for cointegration Appendix 11C Vector error correction modeling (VECM): test and estimation using SAS 12 Qualitative and limited dependent variable models 12.1 Linear probability model 12.2 Probit model 12.2.1 Interpretation of the coefficients 12.2.2 Measuring the goodness-of-fit 12.3 Logit model 12.3.1 Interpretation of the coefficients 12.3.2 Logit vs. probit 12.3.3 Adjustment for unequal sampling rates: Maddala (1991), Palepu (1986) 12.4 Tobit model 12.4.1 The Tobit model 12.4.2 Applications of the Tobit model 12.4.3 Estimation using EViews and SAS 12.5 Choice-based models 12.5.1 Self-selection model 12.5.2 Choice-based Tobit model 12.5.3 Estimation using SAS 12.6. Applications 12.6.1 Bushee (1998) 12.6.2 Leung, Daouk and Chen (2000) 12.6.3 Shumway (2001) 12.6.4 Robinson and Min (2002) 12.6.5 Leuz and Verrecchia (2000) 12.7 Summary Review questions References Appendix 12 Maximum likelihood estimation (MLE) Index