Testing for Regime Switching: A Comment

Size: px
Start display at page:

Download "Testing for Regime Switching: A Comment"

Transcription

1 Testing for Regime Switching: A Comment Andrew V. Carter Department of Statistics University of California, Santa Barbara Douglas G. Steigerwald Department of Economics University of California Santa Barbara May 5, 0 JEL Classi cation: C3 Key Words: consistent, Markov regime switching, quasi-maximum likelihood Introduction In Cho and White (007) Testing for Regime Switching the authors obtain the asymptotic null distribution of a quasi-likelihood ratio (QLR) statistic. The statistic is designed to test the null hypothesis of one regime against the alternative of Markov switching between two regimes. Likelihood ratio statistics are used because the test involves nuisance parameters that are not identi ed under the null hypothesis, together with other nonstandard features. Cho and White focus on a quasi-likelihood, which ignores certain serial correlation properties but allows for a tractable factorization of the likelihood. While the majority of their paper focuses on asymptotic behavior under the null hypothesis, Theorem (b) states that the quasi-maximum likelihood estimator (QMLE) is consistent under the alternative hypothesis. Consistency of the QMLE requires that the expected quasi-log-likelihood attain a global maximum at the population parameter values. This requirement holds for some Markov regimeswitching processes but, as we show below, not for an autoregressive process as analyzed in Cho and White. The quasi-likelihood approximates the Markov likelihood by replacing the probabilities of the state variable that indicates regimes. Speci cally, the probabilities from the distribution of state variables conditional on past values of the observed data (the conditional state probabilities) are replaced with the Corresponding author: doug@econ.ucsb.edu We thank Jin Cho and Hal White for helpful comments.

2 unconditional probabilities from the stationary distribution of state variables. In so doing the quasi-likelihood ignores serial correlation in the state variables. Ignoring this serial correlation can lead to inconsistency if the conditional state probabilities depend upon the regressors that enter the state-speci c conditional densities. To understand the source of inconsistency, we use the classic structure that Wald (949) proposed to demonstrate consistency of the MLE. The structure was adapted by Levine (983) to demonstrate a general property of consistency for a QMLE, where the quasi-log-likelihood is constructed from conditional density functions. As we show below, in applying the logic of Levine to Markov regime-switching processes, a key requirement is that the conditional state probabilities be independent of the regressors that enter the state-speci c conditional densities. For a process in which the conditional densities depend only on exogenous regressors, the conditional state probabilities are independent of the regressors and a QMLE is consistent. For the autoregressive process analyzed in Cho and White (Section 3, p. 697), however, the conditional state probabilities depend upon regressors that enter the state-speci c conditional densities and a QMLE is inconsistent. Lack of consistency of a QMLE extends generally to autoregressive processes as lagged values of the dependent variable, which are regressors in the conditional densities, contain information about lagged values of the state variable, which in turn contain information about the current value of the state variable. We organize the results as follows. The key assumptions from Cho and White together with the quasi-likelihood are contained in Section. In Section 3 we present the Markov-regime switching autoregressive process analyzed by Cho and White and show that the gradient of the quasi-log-likelihood does not equal zero when evaluated at the population parameter values. In Section 4 we establish a condition that ensures that the expected quasi-log-likelihood attains a global maximum (under the alternative) at the population parameter values. To understand the distinction between our su cient condition and the condition contained in Cho and White we rst relate the autoregressive process to each condition and then refer to the proof of Theorem (b) in Cho and White to determine the impact of each condition. We then discuss the implications of these results for the consistency of a QLR test statistic. Markov Regime-Switching Processes Rather than presenting all the assumptions from Cho and White, we focus only on the assumptions that pertain to establishing a global maximum of the quasilog-likelihood at the population parameter values. The rst assumption de nes the class of Markov regime-switching process, which have strictly stationary Other quasi-likelihood approximations for Markov regime-switching processes have been shown to lead to inconsistent estimators. Campbell (00) analyzes the case in which the underlying innovation density is misspeci ed. Kim, Piger and Startz (008) discuss the case in which the endogeneity of regime switching is ignored.

3 random variables. Assumption : (i) The observable random variables X t R d n, d N, are generated as t= a sequence of strictly stationary -mixing random variables such that for some c > 0 and [0; ) the -mixing coe cient,, is at most c. (ii) The sequence of unobserved state variables that indicate regimes, fs t f; gg n t=, is generated as a rst-order Markov process such that P (S t = js t = ) = p and P (S t = js t = ) = p with p ij [0; ] (i; j = ; ). (iii) The given fx t g is a Markov regime-switching process. That is, for some := ( 0; ; ) R r0+, F jx t X t jf t ; 0; if St = F jx t ; 0; if St = ; where F t := X t ; S t is the smallest -algebra generated by X t ; S t := Xt 0 ; : : : ; X; 0 S t ; : : : ; S ; r0 N; and the conditional cumulative distribution function of X t jf, F jx t ; 0; j has a probability density function f jx t ; 0; j (j = ; ). Further, for (p ; p ) (0; ](0; ] n f(; )g, is unique in R r0+. The point p = p = is excluded from the parameter space to rule out a deterministically periodic process for fs t g, which would imply that fx t g is not strictly stationary. Throughout the following discussion, E and P refer to the distribution of the process described in Assumption. To ensure the likelihood is well de ned, it is assumed that the data generating process speci es an F t -measurable probability density function. Assumption : (i) A model for f jx t ; 0; n j is f jx t ; j : j := ( 0 ; j ) ~ o, where ~ := 0 R r0+. Further, for each j ~, f jx t ; j is a measurable probability density function with cumulative distribution function F jx t ; j (j = ; ). The relevant hypotheses, for test of one regime against the alternative of two regimes, are H 0 : p = 0 ; p = 0 ; or = ; H : (p ; p ) (0; ] (0; ] n f(; )g and 6= : Because the log-likelihood cannot be reduced to a sum of individual loglikelihoods, if p = 0 (or p = 0) then the population variance of the associated rst derivative grows geometrically under the null. To avoid this di culty, Cho and White focus on the quasi-log-likelihood for a mixture model. The quasi-loglikelihood replaces the (likelihood) Markov conditional density of X t j X t, which is given by P S t = j X t f X t jx t ; 0; +P St = j X t f X t jx t ; 0; ; () 3

4 with the mixture conditional density, ( ) f X t jx t ; 0; + f X t jx t ; 0; : () The mixture model () captures the serial correlation in X t, through f X t jx t ;, but ignores the serial correlation in fs t g as the random variable t := P S t = j X t is replaced with = E P S t = j X t = p = (p + p ). Note that because t = ( p ) P S t = j X t + p P S t = j X t ; if p + p = then t = = p and the quasi-log-likelihood is identical to the log-likelihood. The resultant quasi-log-likelihood is de ned for each (; ) [0; ] with := 0, as nx L n (; ) := l t (; ) ; where l t (; ) := log ( ) f X t jx t ; + f X t jx t ;. To ensure that expectations are well de ned, the following regularity condition is assumed. Assumption 3: For all (; ) [0; ], the quantity E [l t (; )] exists and is nite. t= 3 QMLE Inconsistency: An Example The Markov-regime switching autoregressive process analyzed by Cho and White (Section 3, p. 697) is X t = fst=g fst=g + :5X t + u t ; where u t i:i:d:n (0; ) and the transition probabilities satisfy P (S t = js t = ) = P (S t = js t = ) ; so that p = p. The quasi-log-likelihood employed by Cho and White for this process is constructed from the mixture model (p. 697, line 4) ( ) N + X t ; + N + X t ; : (3) To isolate the source of inconsistency in the QMLE from (3), we set the variance to and let = and =. The conditional density functions that enter (3) are N ( + X t ; ) = f X t jx t ; = p exp (X t X t ) N ( + + X t ; ) = f X t jx t ; = p exp (X t X t ) : 4

5 A necessary condition for consistency is that n E [ P n t= l t (; ; ; )] be maximized at the population parameter values. Because the process is stationary n We then have nx E [l t (; ; ; )] = E [l t (; ; ; )] := M (; ; ; ) : t= M (; ; ; ) = E log [ (X t ; X t ) + ( )] + E log f X t jx t ; ; where X t ; X t ; ; = f X tjx t ; f X t jx t ; = exp (X t X t ) Thus M (; ; ; ) = E log exp (X t X t ) h log + E (X t X t ) i : : + ( ) The key is to calculate the rst derivative of M (; ; ; ) with respect to the autoregressive coe cient evaluated at the population values of the parameters ( ; ; ; ). To begin, let t = X t X t. The derivative for is 0 M (; ; ; = X t exp h ( ) t i A +E (X ( ; ; ; ) exp h ( ) t t ) : t + ( ) To calculate these expectations we must account for the correlation between t and X t. The distribution of t conditional on X t depends on this previous observation through S t. The conditional density is f (zjx t ) = (z ) P (S t = jx t ) + (z) P (S t = jx t ) " " # # = t exp ( ) t + ( t ) (z) where (z) is the density of a standard Gaussian random variable. The rst expectation in (4) is 0 i X t exp h ( ) t i A exp h ( ) t + ( ) " # 0 h i = E 4X t exp ( ) 3 t exp ( z ) + ( t ) h i A (z) dz5 : exp z + ( ) ( ) (4) 5

6 h Because exp z = = i ( ) (z) = (z ), i ( ) " # 0 h exp ( ) t exp z + ( t ) h i A (z) dz exp ( z ) + ( ) 0 h t exp ( z ) + ( t ) h i A (z ) dz exp ( z ) + ( ) 0 t exp h v + ( ) + ( t ) i A (v) dv; exp h v + ( ) + ( ) and it follows that this integral is 0 t exp h v + ( ) + ( t ) i A (v) dv exp h v + ( ) + ( ) = i i t t exp h v + ( ) + ( t ) t exp h v + ( ) + t ( ) i i 5 (v) dv h exp h v + ( ) + ( ) = " # t + t exp " v + ( ) + ( )# (v) dv: Therefore, substituting this expression back into the above M (; ; ; = E (X t t )+ E [X t ( t )] C ; +E (X t t ) ; ( ; ; ; ) where C ; is the expectation " " # E exp + ( ) + ( ) for a standard Gaussian random variable. Clearly, C ; does not depend on X t or S t and is a bounded positive quantity: 0 < C ; ( ). Furthermore, we can use that E ( t jx t ) = P (S t = jx t ) = t to simplify the expression M (; ; ; = C ; E [X t ( t )] : ( ; ; ; ) The last factor in this expression is E [X t ( t )] = E (X t t ) EX t ES t = Cov (X t ; S t ) : 6

7 The last equality follows from E (X t S t ) = E (X t E (S t jx t )) and E (S t jx t ) = t. An Expression for the Covariance To obtain an expression for the covariance of X t and S t, we rst use the recursive expression X t = + X t X +S t + t = k + S t k + t k ; k=0 where t i:i:d:n (0; ). This implies that the covariance is X Cov (X t ; S t ) = k Cov (S t k ; S t ) : k=0 The covariance of the binary state variables is Cov (S t k ; S t ) = P (S t = js t k = ) ; where = P (S t = ) = P (S t k = ) is the stationary probability in the Markov chain. It can be shown that the conditional probability is k P (S t = js t k = ) = + ( ) : Thus and Cov (S t k ; S t ) = ( ) Cov (X t ; S t ) = ( ) p k p ; p X k k=0 = ( ) ( p ) ( p ) k p Therefore, the expression for the partial derivative of the M function with respect M (; ; ; ) = ( ) C ; ( p ) : ( p ) Therefore, if the Markov regime process includes a dependence between subsequent time points, the gradient along is not equal to 0 at the population parameter values and the expected value of the quasi-log-likelihood is maximized away from the population parameter values. Of course, this derivative vanishes under the null hypothesis where = 0, = 0 or = 0, as under the null hypothesis there is e ectively only one regime. This derivative also vanishes for = p because, if = p then P (S t = j) = P (S t = jjs t ) and the Markov regime process reduces to independent draws from the stationary distribution. In this case L n is the population log-likelihood rather than the quasi-log-likelihood. : 7

8 4 QMLE Consistency: A Su cient Condition To de ne a class of processes for which a QMLE is consistent, it is helpful to distinguish the elements of X that correspond to endogenous variables. Let a partition of X be given by X = (Y 0 ; W 0 ) 0, where Y denotes the sub-vector of endogenous variables and W denotes the sub-vector of exogenous variables. Let X t := Y t ; W t = Yt 0 ; : : : ; Y; 0 Wt; 0 : : : ; W 0. Exogeneity implies that P S t = j X t is independent of W t. To obtain a su cient condition we augment Assumption to remove the correlation between the conditional state probabilities, P S t = j X t, and the regressors that enter the conditional densities. Let Y t ; X t ; ; = f Y tjy t ; W t ; f Y t jy t ; W t ; : Assumption : (ii) The ratio Y t ; X t ; ; is a random variable that is independent of Y t for every value of. Assumption (ii) is parallel to the assumption used to de ne a state-space model in the statistical literature. Because Assumption (ii) implies that the statespeci c conditional density of Y t is independent of Y s for s < t, exogeneity of W t implies that f Y t jx t ; W t ; i = f Y t jw t ; i. While the assumption does not include the autoregressive model analyzed by Cho and White, it does include the second model they analyze, namely a simultaneous equations model of the type in Gri en et al. (009). We begin with the classic structure adapted by Levine (983) to establish that the expected quasi-log-likelihood attain a global maximum at the population parameter values. We employ the structure to prove that, under the su cient condition Assumption (ii), the unconditional population mean of the quasi-log-likelihood attains a global maximum at ( ; ). The object of analysis is E [l t (; ) l t ( ; )], where E denotes expectation with respect to the Markov conditional density (). Theorem : Under Assumptions, (i)-(ii) and 3, E [l t (; ) l t ( ; )] 0. Proof: We have E [l t (; ) l t ( ; )] = E " log log E " Y t ; X t ; ; + ( ) Y t ; X t ; ;!# Y t ; X t ; ; + ( ) Y t ; X t ; ; + ( ) Y t ; X t ; ;!# Y t ; X t ; ; ; + ( ) See, for example, Durbin and Koopman (00). 8

9 where the last line follows by Jensen s inequality. The expected value of the ratio is " Yt ; X t ; ; + ( ) Y t ; X t ; ; # E = E ( E " Y t ; X t ; ; + ( ) Y t ; X t ; ; + ( ) Y t ; X t ; ;! Y t ; X t ; ; + ( ) = Ef R (y;x t ; ; )+( )(y;x t ; ; ) (y;x t ; ; )+( ) Xt #) t y; X t ; ; + ( t ) f yjx t ; dyg = E y; X t ; ; + ( ) y; X t ; ; f yjx t ; dy R (y;x +E t ; ; )+( )(y;x t ; ; ) t (y;x t ; ; )+( ) y; X t ; ; f yjx t ; dy ; where t = t is Y t -measurable and so can be moved outside the conditional expectation. The rst expectation is because this is the expectation of a density. For the second term, let g X t = y; Xt ; ; + ( ) y; X t ; ;! y; X t ; ; + ( ) y; X t ; ; : Under Assumption (ii), g X t is independent of Y t, so the second term becomes E [ t g ( (W t ))]. Because t is independent of W t, which follows from the exogeneity of W t, the second term is E t g X t = E Y t [ t ] E W t g W t : The quantity t has expectation zero because E Y t implies E t =. Therefore [ t ] = E [ t ] and stationarity E [l t (; ) l t ( ; )] log () = 0: Given Theorem, establishing that the QMLE for the mixture model, ^ n ; ^ n, is consistent follows from standard application of a uniform law of large numbers as referenced in Cho and White (Proof of Theorem (b) p. 704). At this point, it is natural to ask how the su cient condition for consistency of a QMLE that we introduce, Assumption (ii), relates to the condition 9

10 that Cho and White refer to in their proof of Theorem (b) on consistency of a QMLE. To answer, we use the autoregressive process analyzed in Cho and White and presented in (3). Assumption (ii), which restricts the class of processes allowed by Assumptions and (i), explicitly rules out dependence of the state-speci c conditional densities on past values of the endogenous variables, and so rules out autoregressive processes. Cho and White consider all processes allowed under Assumptions and (i). Given that the autoregressive process that enters (3) is a leading example from their paper, it is perhaps not surprising that this process appears to be allowed under Assumptions and (i). To see this, note that under Assumption the density of X t given the entire past of X and the current and past values of the state variable is X t jf t = fst=gf X t jx t ; + fst=gf X t jx t ; : (5) The density of X t given only the entire past of X, which enters the likelihood, is obtained by replacing fst=ig with P S t = ijx t, so that the mixture model in (3) follows from (5) by replacing fst=g with and fst=g with. In consequence f X t jx t ; = N + X t ; and f X t jx t ; = N + X t ;. If the autoregressive process that enters (3) is allowed under Assumptions and (i), how can the proof of Theorem (b) in Cho and White be reconciled with the result in Section 3? It appears that a key step in the proof of Theorem (b) requires that, given S t =, the marginal density of X t equals the density of X t conditional on X t. In detail, the marginal density of X t given S t = is R f (x ; : : : x t js t = ) dx t. As Cho and White note (p. 704, line 8 from the bottom of the page) f (x ; : : : x t js t = ) = 0 F ( ) t = P F ( ) P F t ( ) [ 0] 0 ; which implies that the marginal density of X t given S t = is f (x ; : : : x t js t = ) dx t = 0 F ( ) t = P F ( ) P F t ( ) [ 0] 0 dx t : Because F is a function of x, the expression on the right side is = 0 F ( ) t = P F ( ) P F t ( ) [ 0] 0 dx t 0 F ( ) P F ( ) P F t ( ) P F t ( ) dx dx dx t [ 0] 0 : If F is not a function of x, then 0 F ( ) P F ( ) P F t ( ) P F t ( ) dx dx dx t [ 0] 0 = 0 F ( ) dx P F ( ) dx P F t ( ) dx t P F t ( ) [ 0] 0 : 0

11 In their proof of Theorem (b), Cho and White state (p. 704, line 5 from the bottom) f (x ; : : : x t js t = ) dx t = 0 F ( ) dx P F ( ) dx P F t ( ) dx t P F t ( ) [ 0] 0 ; which holds only if, given S t =, the distribution of X t does not depend on X s for s < t. Because Assumptions and (i) do not imply that, given the value of S t, the distribution of X t not depend on X s for s < t, proof of Theorem (b) requires an additional condition of the form of Assumption (ii). 5 Remarks The inconsistency of a QMLE for Markov regime-switching processes with autoregressive components extends to processes with moving-average components. Inconsistency of a QMLE, however, does not imply that a test based on the quasi-likelihood ratio is inconsistent. Consistency of a QLR test requires only that E l t (; ) be maximized at some point outside the null hypothesis space. For the autoregressive process in Section 3, the gradient of the M function is zero in every coordinate except, which indicates that M is maximized away from the null hypothesis and that the class of models for which the QLR test is consistent is larger than the class of models for which the QMLE is consistent. A de nitive treatment of consistency could be based on bounding the value of the likelihood under the null hypothesis and then demonstrating that under the alternative there is always a point in the alternative space for which the value of the likelihood exceeds the bound. Even for a consistent test, the power may be a ected by the consistency properties of the QMLE under the alternative.

12 References [] Campbell, S., Speci cation Testing and Semiparametric Estimation of Regime Switching Models: An Examination of the US Short Term Interest Rate, Economics Department Working Papers, Brown University, 00. [] Cho, J. and H. White, Testing for Regime-Switching, Econometrica, 007, 75, [3] Durbin, J. and S. Koopman, Time Series Analysis by State Space Methods, Oxford University: 00. [4] Gri n, J., C. Kolstad and D. Steigerwald, Estimating Market Power with Multiple Product Markets, Economics Department Working Papers, UC Santa Barbara, 009. [5] Kim, C., J. Piger and R. Startz, Estimation of Markov Regime-Switching Regression Models with Endogenous Switching, Journal of Econometrics, 008, 43, [6] Levine, D., A Remark on Serial Correlation in Maximum Likelihood, Journal of Econometrics 983, 3, [7] Wald, A., Note on the Consistency of the Maximum Likelihood Estimate, Annals of Mathematical Statistics, 949, 0,

SUPPLEMENT TO TESTING FOR REGIME SWITCHING: A COMMENT (Econometrica, Vol. 80, No. 4, July 2012, )

SUPPLEMENT TO TESTING FOR REGIME SWITCHING: A COMMENT (Econometrica, Vol. 80, No. 4, July 2012, ) Econometrica Supplementary Material SUPPLEMENT TO TESTING FOR REGIME SWITCHING: A COMMENT Econometrica, Vol. 80, No. 4, July 01, 1809 181 BY ANDREW V. CARTER ANDDOUGLAS G. STEIGERWALD We present proof

More information

On the Power of Tests for Regime Switching

On the Power of Tests for Regime Switching On the Power of Tests for Regime Switching joint work with Drew Carter and Ben Hansen Douglas G. Steigerwald UC Santa Barbara May 2015 D. Steigerwald (UCSB) Regime Switching May 2015 1 / 42 Motivating

More information

ECONOMETRICS FIELD EXAM Michigan State University May 9, 2008

ECONOMETRICS FIELD EXAM Michigan State University May 9, 2008 ECONOMETRICS FIELD EXAM Michigan State University May 9, 2008 Instructions: Answer all four (4) questions. Point totals for each question are given in parenthesis; there are 00 points possible. Within

More information

Obtaining Critical Values for Test of Markov Regime Switching

Obtaining Critical Values for Test of Markov Regime Switching University of California, Santa Barbara From the SelectedWorks of Douglas G. Steigerwald November 1, 01 Obtaining Critical Values for Test of Markov Regime Switching Douglas G Steigerwald, University of

More information

ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Spring 2013 Instructor: Victor Aguirregabiria

ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Spring 2013 Instructor: Victor Aguirregabiria ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Spring 2013 Instructor: Victor Aguirregabiria SOLUTION TO FINAL EXAM Friday, April 12, 2013. From 9:00-12:00 (3 hours) INSTRUCTIONS:

More information

Markov-Switching Models with Endogenous Explanatory Variables. Chang-Jin Kim 1

Markov-Switching Models with Endogenous Explanatory Variables. Chang-Jin Kim 1 Markov-Switching Models with Endogenous Explanatory Variables by Chang-Jin Kim 1 Dept. of Economics, Korea University and Dept. of Economics, University of Washington First draft: August, 2002 This version:

More information

Speci cation of Conditional Expectation Functions

Speci cation of Conditional Expectation Functions Speci cation of Conditional Expectation Functions Econometrics Douglas G. Steigerwald UC Santa Barbara D. Steigerwald (UCSB) Specifying Expectation Functions 1 / 24 Overview Reference: B. Hansen Econometrics

More information

Economics 241B Estimation with Instruments

Economics 241B Estimation with Instruments Economics 241B Estimation with Instruments Measurement Error Measurement error is de ned as the error resulting from the measurement of a variable. At some level, every variable is measured with error.

More information

GMM-based inference in the AR(1) panel data model for parameter values where local identi cation fails

GMM-based inference in the AR(1) panel data model for parameter values where local identi cation fails GMM-based inference in the AR() panel data model for parameter values where local identi cation fails Edith Madsen entre for Applied Microeconometrics (AM) Department of Economics, University of openhagen,

More information

ECONOMET RICS P RELIM EXAM August 24, 2010 Department of Economics, Michigan State University

ECONOMET RICS P RELIM EXAM August 24, 2010 Department of Economics, Michigan State University ECONOMET RICS P RELIM EXAM August 24, 2010 Department of Economics, Michigan State University Instructions: Answer all four (4) questions. Be sure to show your work or provide su cient justi cation for

More information

Chapter 2. GMM: Estimating Rational Expectations Models

Chapter 2. GMM: Estimating Rational Expectations Models Chapter 2. GMM: Estimating Rational Expectations Models Contents 1 Introduction 1 2 Step 1: Solve the model and obtain Euler equations 2 3 Step 2: Formulate moment restrictions 3 4 Step 3: Estimation and

More information

Parametric Inference on Strong Dependence

Parametric Inference on Strong Dependence Parametric Inference on Strong Dependence Peter M. Robinson London School of Economics Based on joint work with Javier Hualde: Javier Hualde and Peter M. Robinson: Gaussian Pseudo-Maximum Likelihood Estimation

More information

MC3: Econometric Theory and Methods. Course Notes 4

MC3: Econometric Theory and Methods. Course Notes 4 University College London Department of Economics M.Sc. in Economics MC3: Econometric Theory and Methods Course Notes 4 Notes on maximum likelihood methods Andrew Chesher 25/0/2005 Course Notes 4, Andrew

More information

13 Endogeneity and Nonparametric IV

13 Endogeneity and Nonparametric IV 13 Endogeneity and Nonparametric IV 13.1 Nonparametric Endogeneity A nonparametric IV equation is Y i = g (X i ) + e i (1) E (e i j i ) = 0 In this model, some elements of X i are potentially endogenous,

More information

ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Winter 2014 Instructor: Victor Aguirregabiria

ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Winter 2014 Instructor: Victor Aguirregabiria ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Winter 2014 Instructor: Victor guirregabiria SOLUTION TO FINL EXM Monday, pril 14, 2014. From 9:00am-12:00pm (3 hours) INSTRUCTIONS:

More information

1. The Multivariate Classical Linear Regression Model

1. The Multivariate Classical Linear Regression Model Business School, Brunel University MSc. EC550/5509 Modelling Financial Decisions and Markets/Introduction to Quantitative Methods Prof. Menelaos Karanasos (Room SS69, Tel. 08956584) Lecture Notes 5. The

More information

Time Series Models and Inference. James L. Powell Department of Economics University of California, Berkeley

Time Series Models and Inference. James L. Powell Department of Economics University of California, Berkeley Time Series Models and Inference James L. Powell Department of Economics University of California, Berkeley Overview In contrast to the classical linear regression model, in which the components of the

More information

(Y jz) t (XjZ) 0 t = S yx S yz S 1. S yx:z = T 1. etc. 2. Next solve the eigenvalue problem. js xx:z S xy:z S 1

(Y jz) t (XjZ) 0 t = S yx S yz S 1. S yx:z = T 1. etc. 2. Next solve the eigenvalue problem. js xx:z S xy:z S 1 Abstract Reduced Rank Regression The reduced rank regression model is a multivariate regression model with a coe cient matrix with reduced rank. The reduced rank regression algorithm is an estimation procedure,

More information

Simple Estimators for Semiparametric Multinomial Choice Models

Simple Estimators for Semiparametric Multinomial Choice Models Simple Estimators for Semiparametric Multinomial Choice Models James L. Powell and Paul A. Ruud University of California, Berkeley March 2008 Preliminary and Incomplete Comments Welcome Abstract This paper

More information

A Course on Advanced Econometrics

A Course on Advanced Econometrics A Course on Advanced Econometrics Yongmiao Hong The Ernest S. Liu Professor of Economics & International Studies Cornell University Course Introduction: Modern economies are full of uncertainties and risk.

More information

Testing for a Trend with Persistent Errors

Testing for a Trend with Persistent Errors Testing for a Trend with Persistent Errors Graham Elliott UCSD August 2017 Abstract We develop new tests for the coe cient on a time trend in a regression of a variable on a constant and time trend where

More information

We begin by thinking about population relationships.

We begin by thinking about population relationships. Conditional Expectation Function (CEF) We begin by thinking about population relationships. CEF Decomposition Theorem: Given some outcome Y i and some covariates X i there is always a decomposition where

More information

ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Winter 2016 Instructor: Victor Aguirregabiria

ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Winter 2016 Instructor: Victor Aguirregabiria ECOOMETRICS II (ECO 24S) University of Toronto. Department of Economics. Winter 26 Instructor: Victor Aguirregabiria FIAL EAM. Thursday, April 4, 26. From 9:am-2:pm (3 hours) ISTRUCTIOS: - This is a closed-book

More information

A New Approach to Robust Inference in Cointegration

A New Approach to Robust Inference in Cointegration A New Approach to Robust Inference in Cointegration Sainan Jin Guanghua School of Management, Peking University Peter C. B. Phillips Cowles Foundation, Yale University, University of Auckland & University

More information

Time Series Analysis. James D. Hamilton PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY

Time Series Analysis. James D. Hamilton PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY Time Series Analysis James D. Hamilton PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY PREFACE xiii 1 Difference Equations 1.1. First-Order Difference Equations 1 1.2. pth-order Difference Equations 7

More information

Tests for Cointegration, Cobreaking and Cotrending in a System of Trending Variables

Tests for Cointegration, Cobreaking and Cotrending in a System of Trending Variables Tests for Cointegration, Cobreaking and Cotrending in a System of Trending Variables Josep Lluís Carrion-i-Silvestre University of Barcelona Dukpa Kim y Korea University May 4, 28 Abstract We consider

More information

Cointegration Tests Using Instrumental Variables Estimation and the Demand for Money in England

Cointegration Tests Using Instrumental Variables Estimation and the Demand for Money in England Cointegration Tests Using Instrumental Variables Estimation and the Demand for Money in England Kyung So Im Junsoo Lee Walter Enders June 12, 2005 Abstract In this paper, we propose new cointegration tests

More information

ECONOMICS 7200 MODERN TIME SERIES ANALYSIS Econometric Theory and Applications

ECONOMICS 7200 MODERN TIME SERIES ANALYSIS Econometric Theory and Applications ECONOMICS 7200 MODERN TIME SERIES ANALYSIS Econometric Theory and Applications Yongmiao Hong Department of Economics & Department of Statistical Sciences Cornell University Spring 2019 Time and uncertainty

More information

Chapter 6. Maximum Likelihood Analysis of Dynamic Stochastic General Equilibrium (DSGE) Models

Chapter 6. Maximum Likelihood Analysis of Dynamic Stochastic General Equilibrium (DSGE) Models Chapter 6. Maximum Likelihood Analysis of Dynamic Stochastic General Equilibrium (DSGE) Models Fall 22 Contents Introduction 2. An illustrative example........................... 2.2 Discussion...................................

More information

Estimation and Inference with Weak Identi cation

Estimation and Inference with Weak Identi cation Estimation and Inference with Weak Identi cation Donald W. K. Andrews Cowles Foundation Yale University Xu Cheng Department of Economics University of Pennsylvania First Draft: August, 2007 Revised: March

More information

Regression. Econometrics II. Douglas G. Steigerwald. UC Santa Barbara. D. Steigerwald (UCSB) Regression 1 / 17

Regression. Econometrics II. Douglas G. Steigerwald. UC Santa Barbara. D. Steigerwald (UCSB) Regression 1 / 17 Regression Econometrics Douglas G. Steigerwald UC Santa Barbara D. Steigerwald (UCSB) Regression 1 / 17 Overview Reference: B. Hansen Econometrics Chapter 2.20-2.23, 2.25-2.29 Best Linear Projection is

More information

Estimation of Dynamic Regression Models

Estimation of Dynamic Regression Models University of Pavia 2007 Estimation of Dynamic Regression Models Eduardo Rossi University of Pavia Factorization of the density DGP: D t (x t χ t 1, d t ; Ψ) x t represent all the variables in the economy.

More information

Chapter 2. Dynamic panel data models

Chapter 2. Dynamic panel data models Chapter 2. Dynamic panel data models School of Economics and Management - University of Geneva Christophe Hurlin, Université of Orléans University of Orléans April 2018 C. Hurlin (University of Orléans)

More information

Notes on Time Series Modeling

Notes on Time Series Modeling Notes on Time Series Modeling Garey Ramey University of California, San Diego January 17 1 Stationary processes De nition A stochastic process is any set of random variables y t indexed by t T : fy t g

More information

Introduction: structural econometrics. Jean-Marc Robin

Introduction: structural econometrics. Jean-Marc Robin Introduction: structural econometrics Jean-Marc Robin Abstract 1. Descriptive vs structural models 2. Correlation is not causality a. Simultaneity b. Heterogeneity c. Selectivity Descriptive models Consider

More information

Likelihood Ratio Based Test for the Exogeneity and the Relevance of Instrumental Variables

Likelihood Ratio Based Test for the Exogeneity and the Relevance of Instrumental Variables Likelihood Ratio Based est for the Exogeneity and the Relevance of Instrumental Variables Dukpa Kim y Yoonseok Lee z September [under revision] Abstract his paper develops a test for the exogeneity and

More information

Instrumental Variables. Ethan Kaplan

Instrumental Variables. Ethan Kaplan Instrumental Variables Ethan Kaplan 1 Instrumental Variables: Intro. Bias in OLS: Consider a linear model: Y = X + Suppose that then OLS yields: cov (X; ) = ^ OLS = X 0 X 1 X 0 Y = X 0 X 1 X 0 (X + ) =)

More information

Obtaining Critical Values for Test of Markov. Regime Switching

Obtaining Critical Values for Test of Markov. Regime Switching Obtaining Critical Values for Test of Markov Regime Switching Valerie K. Bostwick and Douglas G. Steigerwald Department of Economics University of California, Santa Barbara February 25, 2014 Abstract For

More information

Simple Estimators for Monotone Index Models

Simple Estimators for Monotone Index Models Simple Estimators for Monotone Index Models Hyungtaik Ahn Dongguk University, Hidehiko Ichimura University College London, James L. Powell University of California, Berkeley (powell@econ.berkeley.edu)

More information

Approximately Most Powerful Tests for Moment Inequalities

Approximately Most Powerful Tests for Moment Inequalities Approximately Most Powerful Tests for Moment Inequalities Richard C. Chiburis Department of Economics, Princeton University September 26, 2008 Abstract The existing literature on testing moment inequalities

More information

GMM based inference for panel data models

GMM based inference for panel data models GMM based inference for panel data models Maurice J.G. Bun and Frank Kleibergen y this version: 24 February 2010 JEL-code: C13; C23 Keywords: dynamic panel data model, Generalized Method of Moments, weak

More information

Chapter 6: Endogeneity and Instrumental Variables (IV) estimator

Chapter 6: Endogeneity and Instrumental Variables (IV) estimator Chapter 6: Endogeneity and Instrumental Variables (IV) estimator Advanced Econometrics - HEC Lausanne Christophe Hurlin University of Orléans December 15, 2013 Christophe Hurlin (University of Orléans)

More information

Control Functions in Nonseparable Simultaneous Equations Models 1

Control Functions in Nonseparable Simultaneous Equations Models 1 Control Functions in Nonseparable Simultaneous Equations Models 1 Richard Blundell 2 UCL & IFS and Rosa L. Matzkin 3 UCLA June 2013 Abstract The control function approach (Heckman and Robb (1985)) in a

More information

On Econometric Analysis of Structural Systems with Permanent and Transitory Shocks and Exogenous Variables

On Econometric Analysis of Structural Systems with Permanent and Transitory Shocks and Exogenous Variables On Econometric Analysis of Structural Systems with Permanent and Transitory Shocks and Exogenous Variables Adrian Pagan School of Economics and Finance, Queensland University of Technology M. Hashem Pesaran

More information

LECTURE 13: TIME SERIES I

LECTURE 13: TIME SERIES I 1 LECTURE 13: TIME SERIES I AUTOCORRELATION: Consider y = X + u where y is T 1, X is T K, is K 1 and u is T 1. We are using T and not N for sample size to emphasize that this is a time series. The natural

More information

Supplemental Material 1 for On Optimal Inference in the Linear IV Model

Supplemental Material 1 for On Optimal Inference in the Linear IV Model Supplemental Material 1 for On Optimal Inference in the Linear IV Model Donald W. K. Andrews Cowles Foundation for Research in Economics Yale University Vadim Marmer Vancouver School of Economics University

More information

Nonparametric Identi cation and Estimation of Truncated Regression Models with Heteroskedasticity

Nonparametric Identi cation and Estimation of Truncated Regression Models with Heteroskedasticity Nonparametric Identi cation and Estimation of Truncated Regression Models with Heteroskedasticity Songnian Chen a, Xun Lu a, Xianbo Zhou b and Yahong Zhou c a Department of Economics, Hong Kong University

More information

MAXIMUM LIKELIHOOD ESTIMATION AND UNIFORM INFERENCE WITH SPORADIC IDENTIFICATION FAILURE. Donald W. K. Andrews and Xu Cheng.

MAXIMUM LIKELIHOOD ESTIMATION AND UNIFORM INFERENCE WITH SPORADIC IDENTIFICATION FAILURE. Donald W. K. Andrews and Xu Cheng. MAXIMUM LIKELIHOOD ESTIMATION AND UNIFORM INFERENCE WITH SPORADIC IDENTIFICATION FAILURE By Donald W. K. Andrews and Xu Cheng October COWLES FOUNDATION DISCUSSION PAPER NO. 8 COWLES FOUNDATION FOR RESEARCH

More information

Online Appendix to: Marijuana on Main Street? Estimating Demand in Markets with Limited Access

Online Appendix to: Marijuana on Main Street? Estimating Demand in Markets with Limited Access Online Appendix to: Marijuana on Main Street? Estating Demand in Markets with Lited Access By Liana Jacobi and Michelle Sovinsky This appendix provides details on the estation methodology for various speci

More information

Contents. University of York Department of Economics PhD Course 2006 VAR ANALYSIS IN MACROECONOMICS. Lecturer: Professor Mike Wickens.

Contents. University of York Department of Economics PhD Course 2006 VAR ANALYSIS IN MACROECONOMICS. Lecturer: Professor Mike Wickens. University of York Department of Economics PhD Course 00 VAR ANALYSIS IN MACROECONOMICS Lecturer: Professor Mike Wickens Lecture VAR Models Contents 1. Statistical v. econometric models. Statistical models

More information

Equivalence of several methods for decomposing time series into permananent and transitory components

Equivalence of several methods for decomposing time series into permananent and transitory components Equivalence of several methods for decomposing time series into permananent and transitory components Don Harding Department of Economics and Finance LaTrobe University, Bundoora Victoria 3086 and Centre

More information

Lecture Notes on Measurement Error

Lecture Notes on Measurement Error Steve Pischke Spring 2000 Lecture Notes on Measurement Error These notes summarize a variety of simple results on measurement error which I nd useful. They also provide some references where more complete

More information

Estimating the Number of Common Factors in Serially Dependent Approximate Factor Models

Estimating the Number of Common Factors in Serially Dependent Approximate Factor Models Estimating the Number of Common Factors in Serially Dependent Approximate Factor Models Ryan Greenaway-McGrevy y Bureau of Economic Analysis Chirok Han Korea University February 7, 202 Donggyu Sul University

More information

ECON 594: Lecture #6

ECON 594: Lecture #6 ECON 594: Lecture #6 Thomas Lemieux Vancouver School of Economics, UBC May 2018 1 Limited dependent variables: introduction Up to now, we have been implicitly assuming that the dependent variable, y, was

More information

Economics 620, Lecture 9: Asymptotics III: Maximum Likelihood Estimation

Economics 620, Lecture 9: Asymptotics III: Maximum Likelihood Estimation Economics 620, Lecture 9: Asymptotics III: Maximum Likelihood Estimation Nicholas M. Kiefer Cornell University Professor N. M. Kiefer (Cornell University) Lecture 9: Asymptotics III(MLE) 1 / 20 Jensen

More information

Models, Testing, and Correction of Heteroskedasticity. James L. Powell Department of Economics University of California, Berkeley

Models, Testing, and Correction of Heteroskedasticity. James L. Powell Department of Economics University of California, Berkeley Models, Testing, and Correction of Heteroskedasticity James L. Powell Department of Economics University of California, Berkeley Aitken s GLS and Weighted LS The Generalized Classical Regression Model

More information

Multinomial Discrete Choice Models

Multinomial Discrete Choice Models hapter 2 Multinomial Discrete hoice Models 2.1 Introduction We present some discrete choice models that are applied to estimate parameters of demand for products that are purchased in discrete quantities.

More information

Comment on HAC Corrections for Strongly Autocorrelated Time Series by Ulrich K. Müller

Comment on HAC Corrections for Strongly Autocorrelated Time Series by Ulrich K. Müller Comment on HAC Corrections for Strongly Autocorrelated ime Series by Ulrich K. Müller Yixiao Sun Department of Economics, UC San Diego May 2, 24 On the Nearly-optimal est Müller applies the theory of optimal

More information

Nonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments

Nonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments Nonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments Xiaohong Chen Yale University Yingyao Hu y Johns Hopkins University Arthur Lewbel z

More information

Economics 620, Lecture 13: Time Series I

Economics 620, Lecture 13: Time Series I Economics 620, Lecture 13: Time Series I Nicholas M. Kiefer Cornell University Professor N. M. Kiefer (Cornell University) Lecture 13: Time Series I 1 / 19 AUTOCORRELATION Consider y = X + u where y is

More information

Simultaneous Equation Models Learning Objectives Introduction Introduction (2) Introduction (3) Solving the Model structural equations

Simultaneous Equation Models Learning Objectives Introduction Introduction (2) Introduction (3) Solving the Model structural equations Simultaneous Equation Models. Introduction: basic definitions 2. Consequences of ignoring simultaneity 3. The identification problem 4. Estimation of simultaneous equation models 5. Example: IS LM model

More information

Bootstrapping the Grainger Causality Test With Integrated Data

Bootstrapping the Grainger Causality Test With Integrated Data Bootstrapping the Grainger Causality Test With Integrated Data Richard Ti n University of Reading July 26, 2006 Abstract A Monte-carlo experiment is conducted to investigate the small sample performance

More information

Chapter 1. GMM: Basic Concepts

Chapter 1. GMM: Basic Concepts Chapter 1. GMM: Basic Concepts Contents 1 Motivating Examples 1 1.1 Instrumental variable estimator....................... 1 1.2 Estimating parameters in monetary policy rules.............. 2 1.3 Estimating

More information

Vector Auto-Regressive Models

Vector Auto-Regressive Models Vector Auto-Regressive Models Laurent Ferrara 1 1 University of Paris Nanterre M2 Oct. 2018 Overview of the presentation 1. Vector Auto-Regressions Definition Estimation Testing 2. Impulse responses functions

More information

Unit roots in vector time series. Scalar autoregression True model: y t 1 y t1 2 y t2 p y tp t Estimated model: y t c y t1 1 y t1 2 y t2

Unit roots in vector time series. Scalar autoregression True model: y t 1 y t1 2 y t2 p y tp t Estimated model: y t c y t1 1 y t1 2 y t2 Unit roots in vector time series A. Vector autoregressions with unit roots Scalar autoregression True model: y t y t y t p y tp t Estimated model: y t c y t y t y t p y tp t Results: T j j is asymptotically

More information

Empirical Market Microstructure Analysis (EMMA)

Empirical Market Microstructure Analysis (EMMA) Empirical Market Microstructure Analysis (EMMA) Lecture 3: Statistical Building Blocks and Econometric Basics Prof. Dr. Michael Stein michael.stein@vwl.uni-freiburg.de Albert-Ludwigs-University of Freiburg

More information

GENERIC RESULTS FOR ESTABLISHING THE ASYMPTOTIC SIZE OF CONFIDENCE SETS AND TESTS. Donald W.K. Andrews, Xu Cheng and Patrik Guggenberger.

GENERIC RESULTS FOR ESTABLISHING THE ASYMPTOTIC SIZE OF CONFIDENCE SETS AND TESTS. Donald W.K. Andrews, Xu Cheng and Patrik Guggenberger. GENERIC RESULTS FOR ESTABLISHING THE ASYMPTOTIC SIZE OF CONFIDENCE SETS AND TESTS By Donald W.K. Andrews, Xu Cheng and Patrik Guggenberger August 2011 COWLES FOUNDATION DISCUSSION PAPER NO. 1813 COWLES

More information

SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER. Donald W. K. Andrews. August 2011

SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER. Donald W. K. Andrews. August 2011 SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER By Donald W. K. Andrews August 2011 COWLES FOUNDATION DISCUSSION PAPER NO. 1815 COWLES FOUNDATION FOR RESEARCH IN ECONOMICS

More information

MA 8101 Stokastiske metoder i systemteori

MA 8101 Stokastiske metoder i systemteori MA 811 Stokastiske metoder i systemteori AUTUMN TRM 3 Suggested solution with some extra comments The exam had a list of useful formulae attached. This list has been added here as well. 1 Problem In this

More information

VAR Models and Applications

VAR Models and Applications VAR Models and Applications Laurent Ferrara 1 1 University of Paris West M2 EIPMC Oct. 2016 Overview of the presentation 1. Vector Auto-Regressions Definition Estimation Testing 2. Impulse responses functions

More information

Nonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments

Nonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments Nonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments Xiaohong Chen Yale University Yingyao Hu y Johns Hopkins University Arthur Lewbel z

More information

Some Recent Developments in Spatial Panel Data Models

Some Recent Developments in Spatial Panel Data Models Some Recent Developments in Spatial Panel Data Models Lung-fei Lee Department of Economics Ohio State University l ee@econ.ohio-state.edu Jihai Yu Department of Economics University of Kentucky jihai.yu@uky.edu

More information

Spatial panels: random components vs. xed e ects

Spatial panels: random components vs. xed e ects Spatial panels: random components vs. xed e ects Lung-fei Lee Department of Economics Ohio State University l eeecon.ohio-state.edu Jihai Yu Department of Economics University of Kentucky jihai.yuuky.edu

More information

Non-Markovian Regime Switching with Endogenous States and Time-Varying State Strengths

Non-Markovian Regime Switching with Endogenous States and Time-Varying State Strengths Non-Markovian Regime Switching with Endogenous States and Time-Varying State Strengths January 2004 Siddhartha Chib Olin School of Business Washington University chib@olin.wustl.edu Michael Dueker Federal

More information

1 Quantitative Techniques in Practice

1 Quantitative Techniques in Practice 1 Quantitative Techniques in Practice 1.1 Lecture 2: Stationarity, spurious regression, etc. 1.1.1 Overview In the rst part we shall look at some issues in time series economics. In the second part we

More information

Single-Equation GMM: Endogeneity Bias

Single-Equation GMM: Endogeneity Bias Single-Equation GMM: Lecture for Economics 241B Douglas G. Steigerwald UC Santa Barbara January 2012 Initial Question Initial Question How valuable is investment in college education? economics - measure

More information

Environmental Econometrics

Environmental Econometrics Environmental Econometrics Syngjoo Choi Fall 2008 Environmental Econometrics (GR03) Fall 2008 1 / 37 Syllabus I This is an introductory econometrics course which assumes no prior knowledge on econometrics;

More information

Bias in the Mean Reversion Estimator in the Continuous Time Gaussian and Lévy Processes

Bias in the Mean Reversion Estimator in the Continuous Time Gaussian and Lévy Processes Bias in the Mean Reversion Estimator in the Continuous Time Gaussian and Lévy Processes Aman Ullah a, Yun Wang a, Jun Yu b a Department of Economics, University of California, Riverside CA 95-47 b School

More information

Switching Regime Estimation

Switching Regime Estimation Switching Regime Estimation Series de Tiempo BIrkbeck March 2013 Martin Sola (FE) Markov Switching models 01/13 1 / 52 The economy (the time series) often behaves very different in periods such as booms

More information

Discussion Paper Series

Discussion Paper Series INSTITUTO TECNOLÓGICO AUTÓNOMO DE MÉXICO CENTRO DE INVESTIGACIÓN ECONÓMICA Discussion Paper Series Size Corrected Power for Bootstrap Tests Manuel A. Domínguez and Ignacio N. Lobato Instituto Tecnológico

More information

Bayesian IV: the normal case with multiple endogenous variables

Bayesian IV: the normal case with multiple endogenous variables Baesian IV: the normal case with multiple endogenous variables Timoth Cogle and Richard Startz revised Januar 05 Abstract We set out a Gibbs sampler for the linear instrumental-variable model with normal

More information

Estimation and Inference with Weak, Semi-strong, and Strong Identi cation

Estimation and Inference with Weak, Semi-strong, and Strong Identi cation Estimation and Inference with Weak, Semi-strong, and Strong Identi cation Donald W. K. Andrews Cowles Foundation Yale University Xu Cheng Department of Economics University of Pennsylvania This Version:

More information

(c) i) In ation (INFL) is regressed on the unemployment rate (UNR):

(c) i) In ation (INFL) is regressed on the unemployment rate (UNR): BRUNEL UNIVERSITY Master of Science Degree examination Test Exam Paper 005-006 EC500: Modelling Financial Decisions and Markets EC5030: Introduction to Quantitative methods Model Answers. COMPULSORY (a)

More information

A time series plot: a variable Y t on the vertical axis is plotted against time on the horizontal axis

A time series plot: a variable Y t on the vertical axis is plotted against time on the horizontal axis TIME AS A REGRESSOR A time series plot: a variable Y t on the vertical axis is plotted against time on the horizontal axis Many economic variables increase or decrease with time A linear trend relationship

More information

Limited Dependent Variables and Panel Data

Limited Dependent Variables and Panel Data and Panel Data June 24 th, 2009 Structure 1 2 Many economic questions involve the explanation of binary variables, e.g.: explaining the participation of women in the labor market explaining retirement

More information

Modelling Panel Count Data With Excess Zeros: A Hurdle Approach

Modelling Panel Count Data With Excess Zeros: A Hurdle Approach Modelling Panel Count Data With Excess Zeros: A Hurdle Approach José M. R. Murteira Faculdade de Economia, Universidade de Coimbra, and CEMAPRE September, 2007 Abstract This paper presents a hurdle-type

More information

Cointegration and the joint con rmation hypothesis

Cointegration and the joint con rmation hypothesis Cointegration and the joint con rmation hypothesis VASCO J. GABRIEL Department of Economics, Birkbeck College, UK University of Minho, Portugal October 2001 Abstract Recent papers by Charemza and Syczewska

More information

x i = 1 yi 2 = 55 with N = 30. Use the above sample information to answer all the following questions. Show explicitly all formulas and calculations.

x i = 1 yi 2 = 55 with N = 30. Use the above sample information to answer all the following questions. Show explicitly all formulas and calculations. Exercises for the course of Econometrics Introduction 1. () A researcher is using data for a sample of 30 observations to investigate the relationship between some dependent variable y i and independent

More information

Modelling Panel Count Data With Excess Zeros: A Hurdle Approach

Modelling Panel Count Data With Excess Zeros: A Hurdle Approach Modelling Panel Count Data With Excess Zeros: A Hurdle Approach José M. R. Murteira Faculdade de Economia, Universidade de Coimbra, and CEMAPRE March, 2007 Abstract This paper presents a hurdle-type regression

More information

Large Panel Data Models with Cross-Sectional Dependence: A Surevey

Large Panel Data Models with Cross-Sectional Dependence: A Surevey Large Panel Data Models with Cross-Sectional Dependence: A Surevey Hashem Pesaran University of Southern California, CAFE, USA, and Trinity College, Cambridge, UK A Course on Panel Data Models, University

More information

Econometric Analysis of Cross Section and Panel Data

Econometric Analysis of Cross Section and Panel Data Econometric Analysis of Cross Section and Panel Data Jeffrey M. Wooldridge / The MIT Press Cambridge, Massachusetts London, England Contents Preface Acknowledgments xvii xxiii I INTRODUCTION AND BACKGROUND

More information

Regression with time series

Regression with time series Regression with time series Class Notes Manuel Arellano February 22, 2018 1 Classical regression model with time series Model and assumptions The basic assumption is E y t x 1,, x T = E y t x t = x tβ

More information

Testing Weak Convergence Based on HAR Covariance Matrix Estimators

Testing Weak Convergence Based on HAR Covariance Matrix Estimators Testing Weak Convergence Based on HAR Covariance atrix Estimators Jianning Kong y, Peter C. B. Phillips z, Donggyu Sul x August 4, 207 Abstract The weak convergence tests based on heteroskedasticity autocorrelation

More information

Binomial Autoregressive Moving Average Models with. an Application to U.S. Recessions

Binomial Autoregressive Moving Average Models with. an Application to U.S. Recessions Binomial Autoregressive Moving Average Models with an Application to U.S. Recessions Richard Startz Department of Economics University of Washington Working Paper No. 56 Center for Statistics in the Social

More information

Linear models. Linear models are computationally convenient and remain widely used in. applied econometric research

Linear models. Linear models are computationally convenient and remain widely used in. applied econometric research Linear models Linear models are computationally convenient and remain widely used in applied econometric research Our main focus in these lectures will be on single equation linear models of the form y

More information

Finite-sample quantiles of the Jarque-Bera test

Finite-sample quantiles of the Jarque-Bera test Finite-sample quantiles of the Jarque-Bera test Steve Lawford Department of Economics and Finance, Brunel University First draft: February 2004. Abstract The nite-sample null distribution of the Jarque-Bera

More information

When is it really justifiable to ignore explanatory variable endogeneity in a regression model?

When is it really justifiable to ignore explanatory variable endogeneity in a regression model? Discussion Paper: 2015/05 When is it really justifiable to ignore explanatory variable endogeneity in a regression model? Jan F. Kiviet www.ase.uva.nl/uva-econometrics Amsterdam School of Economics Roetersstraat

More information

DEPARTMENT OF ECONOMICS DISCUSSION PAPER SERIES

DEPARTMENT OF ECONOMICS DISCUSSION PAPER SERIES ISSN 1471-0498 DEPARTMENT OF ECONOMICS DISCUSSION PAPER SERIES DYNAMIC BINARY OUTCOME MODELS WITH MAXIMAL HETEROGENEITY Martin Browning and Jesus M. Carro Number 426 April 2009 Manor Road Building, Oxford

More information

SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER. Donald W. K. Andrews. August 2011 Revised March 2012

SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER. Donald W. K. Andrews. August 2011 Revised March 2012 SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER By Donald W. K. Andrews August 2011 Revised March 2012 COWLES FOUNDATION DISCUSSION PAPER NO. 1815R COWLES FOUNDATION FOR

More information

E cient Estimation and Inference for Di erence-in-di erence Regressions with Persistent Errors

E cient Estimation and Inference for Di erence-in-di erence Regressions with Persistent Errors E cient Estimation and Inference for Di erence-in-di erence Regressions with Persistent Errors Ryan Greenaway-McGrevy Bureau of Economic Analysis Chirok Han Korea University January 2014 Donggyu Sul Univeristy

More information