This volume contains an accessible discussion examining computationally-intensive techniques and bootstrap methods, providing ways to improve the finite-sample performance of well-known asymptotic tests for regression models. The book uses the linear regression model as a framework for introducing simulation-based tests to help perform econometric analyses.
Author(s): Leslie Godfrey
Year: 2009
Language: English
Pages: 224
Cover......Page 1
Contents......Page 7
Preface......Page 11
1.1. Introduction......Page 15
1.2. Tests for the classical linear regression model......Page 17
1.3. Tests for linear regression models under weaker assumptions: random regressors and non-Normal IID errors......Page 24
1.4. Tests for generalized linear regression models......Page 28
1.4.1. HCCME-based tests......Page 32
1.4.2. HAC-based tests......Page 35
1.5. Finite-sample properties of asymptotic tests......Page 39
1.5.1. Testing the significance of a subset of regressors......Page 41
1.5.2. Testing for non-Normality of the errors......Page 45
1.5.3. Using heteroskedasticity-robust tests of significance......Page 47
1.6. Non-standard tests for linear regression models......Page 49
1.7. Summary and concluding remarks......Page 56
2.1. Introduction......Page 58
2.2. Some key concepts and simple examples of tests for IID variables......Page 60
2.2.1. Monte Carlo tests......Page 61
2.2.2. Bootstrap tests......Page 64
2.3.1. The classical Normal model......Page 69
2.3.2. Models with IID errors from an unspecified distribution......Page 73
2.3.3. Dynamic regression models and bootstrap schemes......Page 78
2.3.4. The choice of the number of artificial samples......Page 81
2.4. Asymptotic properties of bootstrap tests......Page 83
2.5. The double bootstrap......Page 86
2.6. Summary and concluding remarks......Page 91
3.1. Introduction......Page 95
3.2. A Monte Carlo test of the assumption of Normality......Page 97
3.3. Simulation-based tests for heteroskedasticity......Page 102
3.3.1. Monte Carlo tests for heteroskedasticity......Page 105
3.3.2. Bootstrap tests for heteroskedasticity......Page 108
3.3.3. Simulation experiments and tests for heteroskedasticity......Page 109
3.4.1. Regression models with strictly exogenous regressors......Page 115
3.4.2. Stable dynamic regression models......Page 123
3.4.3. Some simulation evidence concerning asymptotic and bootstrap F tests......Page 124
3.5. Bootstrapping LM tests for serial correlation in dynamic regression models......Page 132
3.5.1. Restricted or unrestricted estimates as parameters of bootstrap worlds......Page 133
3.5.2. Some simulation evidence on the choice between restricted and unrestricted estimates......Page 137
3.6. Summary and concluding remarks......Page 146
4.1. Introduction......Page 148
4.2.1. Asymptotic analysis for predictive test statistics......Page 150
4.2.2. Single and double bootstraps for predictive tests......Page 153
4.2.3. Simulation experiments and results......Page 158
4.2.4. Dynamic regression models......Page 162
4.3. Using bootstrap methods with a battery of OLS diagnostic tests......Page 163
4.3.1. Regression models and diagnostic tests......Page 165
4.3.2. Bootstrapping the minimum p-value of several diagnostic test statistics......Page 166
4.3.3. Simulation experiments and results......Page 169
4.4. Bootstrapping tests for structural breaks......Page 174
4.4.1. Testing constant coefficients against an alternative with an unknown breakpoint......Page 176
4.4.2. Simulation evidence for asymptotic and bootstrap tests......Page 180
4.5. Summary and conclusions......Page 187
5.1. Introduction......Page 191
5.2. Bootstrap methods for independent heteroskedastic errors......Page 192
5.2.1. Model-based bootstraps......Page 195
5.2.2. Pairs bootstraps......Page 197
5.2.3. Wild bootstraps......Page 199
5.2.4. Estimating function bootstraps......Page 202
5.2.5. Bootstrapping dynamic regression models......Page 204
5.3. Bootstrap methods for homoskedastic autocorrelated errors......Page 207
5.3.1. Model-based bootstraps......Page 208
5.3.2. Block bootstraps......Page 212
5.3.3. Sieve bootstraps......Page 215
5.3.4. Other methods......Page 219
5.4.1. Asymptotic theory tests......Page 221
5.4.2. Block bootstraps......Page 224
5.4.3. Other methods......Page 227
5.5. Summary and concluding remarks......Page 228
6.1. Introduction......Page 232
6.2.1. The forms of test statistics......Page 235
6.2.2. Simulation experiments......Page 240
6.3. Bootstrapping heteroskedasticity-robust autocorrelation tests for dynamic models......Page 245
6.3.1. The forms of test statistics......Page 246
6.3.2. Simulation experiments......Page 249
6.4. Bootstrapping heteroskedasticity-robust structural break tests with an unknown breakpoint......Page 255
6.5.1. The forms of test statistics......Page 261
6.5.2. Simulation experiments......Page 268
6.6. Summary and conclusions......Page 276
7.1. Introduction......Page 280
7.2. Asymptotic tests for models with non-nested regressors......Page 282
7.2.1. Cox-type LLR tests......Page 283
7.2.2. Artificial regression tests......Page 287
7.2.4. Regularity conditions and orthogonal regressors......Page 288
7.2.5. Testing with multiple alternatives......Page 289
7.2.6. Tests for model selection......Page 291
7.2.7. Evidence from simulation experiments......Page 293
7.3.1. One non-nested alternative regression model: significance levels......Page 295
7.3.2. One non-nested alternative regression model: power......Page 303
7.3.3. One non-nested alternative regression model: extreme cases......Page 304
7.3.4. Two non-nested alternative regression models: significance levels......Page 307
7.3.5. Two non-nested alternative regression models: power......Page 309
7.4. Bootstrapping the LLR statistic with non-nested models......Page 311
7.5. Summary and concluding remarks......Page 314
8 Epilogue......Page 317
Bibliography......Page 319
E......Page 333
L......Page 334
U......Page 335
Z......Page 336
B......Page 337
D......Page 338
I......Page 339
O......Page 340
S......Page 341
W......Page 342
Y......Page 343
