Self-Normalized Dickey-Fuller Tests for a Unit Root
|
|
- Lisa Marshall
- 5 years ago
- Views:
Transcription
1 Self-Normalized Dickey-Fuller Tests for a Unit Root Gaowen Wang Department of Finance and Banking, Takming College Wei-Lin Mao Department of Economics, National Cheng-Chi University Keywords: domain of attraction, normal law, linear processes, long memory. JEL classification: C2, C3, C22. Tel.: GWWang@takming.edu.tw. Research supported by the National Science Council of the R.O.C. Grant NSC H Corresponding author. Address: No. 64, Sec. 2, Chih-nan Rd., Wenshan, Taipei 623, Taiwan, R.O.C. Tel.: Fax: maotai@nccu.edu.tw.
2 Abstract In this paper, we establish corresponding functionals of self-normalized partial sums for the Dickey-Fuller unit root tests. Under very weak conditions, we derive the asymptotic distributions of the tests with errors being iid random variables, short-memory linear processes, and long-memory linear processes. Under iid environment, the finite second moment condition used in previous research is weakened. For the short-memory linear processes with absolutely summable coefficients, we impose no other conditions on the coefficients. As to the long-memory linear processes, when the coefficients are assumed to be regularly varying with index in (, /2), the only extra condition on the coefficients is automatically satisfied.
3 Introduction Let {u k } k be a sequence of nondegenerate random variables. Set S n = n k= u k and Vn 2 = n k= u2 k. The asymptotic behavior of the self-normalized sums, {S n /V n } n, has been extensively studied in probability theory (Logan et al., 973; Egorov, 996; Giné et al., 997; Kruglov and Petrovskaya, 200; Csörgő et al., 2003; Chistyakov and Götze, 2004; Hall and Wang, 2004). It was proved that, as n, {S n /V n } converges weakly to a Gaussian distribution if and only if u k are independent and identically distributed (iid) random variables with E(u k ) = 0 and in the domain of attraction of the normal law, denoted by u k D N (Egorov, 996; Giné et al., 997). Particularly, S n and V n display synchronous behavior, implying that no moment conditions are needed for the limit distributions of the self-normalized sums (Chistyakov and Götze, 2004). In this paper, we employ the self-normalized sums to investigate asymptotic behavior of the Dickey-Fuller (DF) tests for a unit root (Dickey and Fuller, 979, 98). Testing unit root hypothesis is an important issue in econometrics and has been extensively studied in past years; see, for examples, Stock (994) and Phillips and Xiao (998) for recent surveys. The goal of this paper is threefold. The first one is to establish corresponding functionals of the self-normalized partial sums for the DF tests. The second goal is to demonstrate that the resulting functionals provide a convenient way to derive asymptotics for the tests by allowing a much wider class of
4 error processes. However, much research has been done in this direction. In this paper we restrict our attention to errors being assumed to be iid (a benchmark case), short-range dependent, and long-range dependent, respectively. Note that unit root tests in the presence of long-range dependence are recognized to be of increasing importance in econometrics. There has recently been a surging interest in such models. Investigations by Lubian (999), Wang et al. (2002, 2003), and Wu and Min (2004) have been at the forefront of research in this field. Nevertheless, from theoretical point of view, it is important to know what assumptions can be made for the models under consideration such that the asymptotics of the DF tests still remain asymptotically valid. The third major goal of this paper is to apply the resulting self-normalized DF test statistics to address this issue. Under iid environment, it is well-known that a finite second moment assumption on the underlying errors u k is often required for deriving asymptotics for the DF tests (White, 958; Dickey and Fuller, 979, 98; Phillips, 987; Phillips and Perron, 988; Chan and Wei, 988). In contrast, in this paper the required conditions are that u k D N and E(u k ) = 0. Clearly, our moment conditions are weaker since E(u 2 k ) = is allowable. It represents a significant improvement of the previously known results. On the other hand, the focus in this paper will be on linear processes which provide an excellent vehicle for modeling both short-range and longrange dependence. The process u k = j=0 θ jε k j, θ j R, where {ε i } i Z is a 2
5 sequence of iid(0,) innovations, is said to be short-memory if j=0 θ j < and long-memory if j=0 θ j = but j=0 θ2 j < (see Hall, 992). In the context of short-memory linear processes, the coefficients θ j are typically restricted to be one-summable (i.e. j=0 j θ j < ) or ½-summable (i.e. j=0 j/2 θ j < ) for validating the functional central limit theorem (FCLT) (Said and Dickey, 984; Phillips, 987; Stock, 994; Phillips and Xiao, 998; Chang and Park, 2002; Wang et al., 2003). For more results, see references cited in these papers and Phillips and Solo (992) for a demonstration of the asymptotics for linear processes. Recently, the studies by Wang et al. (2002) and by Wu and Min (2004) have obtained the invariance principles for short-memory linear processes under much weaker conditions than the ½-summability. However, our results for the asymptotic distributions of the DF tests only require that both j=0 θ j 0 and j=0 θ j <. Since we impose less restrictive assumptions on the coefficients θ j, it can be viewed as an improvement of the results in the latter two papers. As to the long-memory linear processes, the long-run variance of u k does not exist, so that the FCLT may not apply unless the norming constant has a faster rate than n and the innovations ε i have finite fourth moment, as is done in Lubian (999) who employed the invariance principle of Davydov (970) to derive asymptotics for the DF tests. As in Wang et al. (2002), in this paper only finite second moments for ε i are required, which abolishes the restriction of E(ε 4 i ) <, used in previous works (see, Wang et 3
6 al. (2003), Wu and Min (2004) and references therein). On the other hand, the condition j=n+ θ j = o( j=0 θ j) that we impose in this paper is sufficiently weak. In particular, when the coefficients have the form θ j = l(j)/j β for j, where β (/2, ) and l is a slowly varying function, we show that the imposed condition is automatically satisfied (see Remark 2 in Section 4). Therefore, our results are new and represent a significant improvement of the previously known results. The paper is organized as follows. The self-normalized DF test statistics are given in Section 2. These self-normalized statistics are applied in Section 3 to iid errors and in Section 4 to errors being, respectively, short-memory and long-memory linear processes with iid innovations. Section 5 concludes the paper. 2 Self-Normalized Unit Root Test Statistics In this section, we first review some known facts about the DF tests for a unit root and then present their self-normalized versions. Suppose {y t } t is generated according to y t = φy t + u t with φ =. () Given y 0 = 0 and n observations y,..., y n, to test φ = against φ <, two appropriate test statistics for the null are as follows: ( ˆρ n := n( ˆφ n ) = n 2 t= 4 y 2 t ) ( n ) y t u t, (2) t=
7 ˆτ n := ( ˆφ n ) ( n t= y2 t ˆσ 2 u ) /2, (3) where ˆφ n = ( n t= y2 t ) n t= y t y t is the least squares estimator (LSE) of φ in () by regressing y t on y t, and ˆσ 2 u = (n ) n t= (y t ˆφ n y t ) 2. Alternatively, by regressing y t on y t with a constant term c, the analog of the two statistics in (2) and (3) can be expressed, respectively, as [ ] [ ] ˆρ µn := n( ˆφ µn ) = n 2 (y t ȳ ) 2 (y t ȳ )u t (4), n t= t= [ n ˆτ µn := ( ˆφ t= µn ) (y t ȳ ) 2 ] /2, (5) ˆσ 2 µu where ˆσ 2 µu = (n 2) n t= (y t ĉ n ˆφ µn y t ) 2, ȳ = n n t= y t, and ˆφ µn and ĉ n are the LSEs of φ and c, respectively. These four test statistics have been well-studied and used as a standard for detecting the presence of a unit root in (). In the case that the error sequence {u t } is assumed to be iid with zero mean and finite positive variance, a number of authors (White, 958; Dickey and Fuller, 979, 98; Phillips, 987; Phillips and Perron, 988; Chan and Wei, 988) have reported the asymptotic distributions of the unit root statistics, which are given by: (see Chapter 7 of Hamilton (994) for full details) ˆρ n W ρ := W 2 () 2 0 W 2 (r)dr, (6) ˆτ n W τ := W 2 () 2 [ ] /2, (7) 0 W 2 (r)dr [ 2 W 2 () ] W () 0 ˆρ µn W ρµ := W (r)dr [ ] 0 W 2 2, (8) (r)dr 0 W (r)dr 5
8 ˆτ µn W τµ := [ 2 W 2 () ] W () 0 { W (r)dr [ ] }, (9) 0 W 2 2 /2 (r)dr 0 W (r)dr where {W (r) : 0 r } stands for a standard Brownian motion, and the symbol signifies the weak convergence of the space D[0, ] endowed with the Skorohod topology. Note that Dickey and Fuller (979, 98) in fact derived alternative but equivalent representations for the asymptotic distributions above (see Corollary 3..3 of Chan and Wei (988)). We now proceed to elaborate on the self-normalized DF tests. Again, let S n = u k, Vn 2 = k= u 2 k. (0) The quotient S n /V n is of the so-called self-normalized sum. We set S n /V n = 0 if V n = 0. Under () with y 0 = 0, we have the representations y t = S t = t k= u k and S 0 = 0. It then follows that k= n yt 2 = St 2, t= t= ȳ = n y t = n S t. () n t= t= Note further that y 2 t = (y t + u t ) 2 = y 2 t + 2y t u t + u 2 t, (2) implying that ( y t u t = yn 2 2 t= t= u 2 t ) = 2 ( S 2 n Vn 2 ). (3) Since, as n, n/(n d) tends to for finite d, it is easy to show that ˆσ 2 u = n u 2 t + o p () = n V n 2 + o p () and ˆσ µu 2 = n V n 2 + o p (), (4) t= 6
9 where the notation o p () denotes a term converging in probability to zero. Similarly, denote by O p () a sequence of random variables that is bounded in probability. Using (0) (4), the four test statistics in (2) (5) can be expressed in terms of functionals of the self-normalized partial sum processes {S t /V n, t n}, respectively, as follows: ˆρ n = ˆτ n = ˆρ µn = = ˆτ µn = = 2 (S2 n Vn 2 ) 2 n = [(S n/v n ) 2 ] n t= S2 n t n t= (S t/v n ), (5) 2 2 (S2 n Vn 2 ) n = 2[(S n/v n ) 2 ] t= S2 t [ n V n 2 + o p (), (6) + o p ()] n t= (S t/v n ) 2 n 2 (S2 n Vn 2 ) n S n [ n t= S2 t n n t= S t ( n t= S t ) 2 ] 2 [(S n/v n ) 2 ] n (S n/v n ) n t= S t/v n [ n ( n t= (S t/v n ) 2 n ) ] 2, (7) n t= S t/v n 2 (S2 n Vn 2 ) n S n n t= S t [ n ( n ) ] 2 [ t= S t n V n 2 + o p () ] t= S2 t n n 2 [(S n/v n ) 2 ] n (S n/v n ) n t= S t/v n [ n ( n t= (S t/v n ) 2 n ) ] + o p(). (8) 2 n t= S t/v n Clearly, if the sequence {S t /V n, t n} has asymptotic distribution, so do the DF test statistics ˆρ n, ˆτ n, ˆρ µn and ˆτ µn. Recently, it was proved that the sequence {S t /V n, t n} converges weakly to a standard Brownian motion {W (r), 0 r } if and only if u k are in the domain of attraction of the normal law and E(u k ) = 0 (Logan et al., 973; Egorov, 996; Giné et al., 997; Kruglov and Petrovskaya, 200; Csörgő et al., 2003). It should 7
10 be noticed that {S t /V n, t n} does not depend on the variance of u k such that E(u 2 k ) = is allowable if the distribution function of u k belongs to the domain of attraction of the normal law. In the following section we shall show that the asymptotic behavior of the DF unit root tests can be derived without the requirement of finite second moment on the iid errors, as commonly assumed in the literature (Dickey and Fuller, 979, 98; Phillips, 987; Chan and Wei, 988; Fuller, 996). 3 DF Tests with Iid Errors Let {u k } in () be iid symmetric random variables with zero mean and a common distribution function from the domain of attraction of a normal law, i.e., u k D N. To begin, we first introduce some notation and then discuss some results for self-normalized sums related to the central limit theorem (CLT) and the invariance principle. Denote by L(Y ) the probability distribution of the random variable Y, by (A) the indicator function of a set A, and by N (0, ) the standard normal distribution. Recent results show that self-normalized sums satisfy the CLT and the invariance principle provided that u k D N and E(u k ) = 0. Giné et al. (997) proved that lim L(S x 2 P( u k > x) n/v n ) = N (0, ) if and only if lim n x E(u 2 = 0. (9) k ( u k x)) The latter condition is equivalent to saying that u k are in the D N (Gnedenko 8
11 and Kolmogorov, 968, p.72; Araujo and Giné, 980, Theorem 6.7). This self-normalized CLT was conjectured by Logan et al. (973). Note that the normal law is the limit for a wide class of distributions, including every distribution with E(u 2 k ( u k x)) being a slowly varying function. This is the case whenever a second moment exists, but E(u 2 k ( u k x)) may also vary slowly when no variance exists (Feller, 966). Thus, the self-normalized CLT above holds under weaker moment conditions than those in the classical limit theorems. In addition, Csörgő et al. (2003, Theorem ) further proved that S [nr] V n W (r), 0 r, (20) where [x] denotes the largest integer less than or equal to x. Csörgő et al s (2003) theorem is a self-normalized version of the FCLT generalizing the theorem of Giné et al. (997). We particularly note that the identically distributed assumption on the sequence {u k } is not necessary for deriving the results shown in (9) (20), see Egorov (996) and Kruglov and Petrovskaya (200). Based on (9) and (20) together with the continuous mapping theorem (Billingsley, 968, p.3) and the fact that W () has a N (0, ) distribution, it is readily seen that if u k are iid symmetric errors and if their distribution functions belong to the D N, then the asymptotic distributions of the DF test statistics in (5) (8) are given by (ˆρ n, ˆτ n, ˆρ µn, ˆτ µn ) (W ρ, W τ, W ρµ, W τµ ). (2) 9
12 Result (2) is identical to those of the corresponding statistics reported in (6) (9), but to derive these asymptotic distributions the only requirement we use is that u k D N. This weakens the finite second moment condition assumed by Dickey and Fuller (979, 98) and by Fuller (996). To illustrate that the self-normalized approach is flexible enough to be applied in a wide class of error processes, in the following section {u k } is assumed to be short-memory and long-memory linear processes, respectively. 4 DF Tests with Long- and Short-Memory Errors Let u k be generated by the linear process u k = θ(l)ε k, θ(l) = θ j L j, (22) where L is the lag operator, {ε k } k Z is an iid random sequence with zero j=0 mean and unit variance, and {θ j } j 0 is a sequence of real numbers. In the context of linear processes, it is well-known that stringent restrictions typically have to be imposed on the rate of decay of θ j. As mentioned in Introduction, {u k } is said to be stationary with short-memory if j=0 θ j < and with long-memory if j=0 θ j = but j=0 θ2 j < (see Hall, 992). Recall S t = t k= u k, t n, and V 2 n = n k= u2 k. Let S εt = t k= ε k, t n, and V 2 εn = n k= ε2 k. Then from Section 3, S ε[nr]/v εn W (r), 0 r. In this section, we establish two results, one for the 0
13 long-memory case (Theorem below), the other for the short-memory case (Theorem 2 below). Theorem. Let {y t } and {u t } be generated according to () and (22), respectively, and let ε k, k Z, be iid random variables with E(ε 0 ) = 0 and E(ε 2 0 ) =. Let θ() = but j=0 θ2 j Then we have j=n+ <. Suppose that, as n, θ j = o(θ()). (23) λ ln S [nr] V n W (r), (24) where 0 r and λ 2 ln := V 2 n /[θ()v εn ] 2 0 in probability, and (a) ˆρ n 2 W 2 () ; 0 W 2 (r)dr (b) ˆτ n in probability; (c) ˆρ µn 2 W 2 () W () 0 W (r)dr 0 W 2 (r)dr [ ; 0 W (r)dr]2 (d) ˆτ µn in probability. Proof of Theorem. To prove (24), we star by writing S [nr] V n = S [nr]/[θ()v εn ] V n /[θ()v εn ]. (25) From this, we have λ ln S [nr] /V n = S [nr] /[θ()v εn ]. Put S j,ε[nr] = [nr] k= ε k j and V 2 j,εn = n k= ε2 k j. In particular, let S ε[nr] = S 0,ε[nr] and V 2 εn = V 2 0,εn. Noting S [nr] = θ()s ε[nr] + n j=0 θ j(s j,ε[nr] S ε[nr] ) + j=n+ θ j(s j,ε[nr]
14 S ε[nr] ) and Vj,εn 2 /V εn 2 p as n, by using (23) and the fact that for any j 0, S j,ε[nr] S ε[nr] = S j,ε[nr] V j,εn S ε[nr] V εn V j,εn V εn V εn = S j,ε[nr] V j,εn S ε[nr] V εn + o p () = O p () (26) is stochastically bounded self-normalized partial sums (Giné et al., 997; Chistyakov and Götze, 2004), we have that S [nr] θ()v εn = S ε[nr] V εn + = S ε[nr] V εn + j=0 j=0 θ j (S j,ε[nr] S ε[nr] ) θ()v εn + j=n+ θ j (S j,ε[nr] S ε[nr] ) θ()v εn θ j (S j,ε[nr] S ε[nr] ) θ()v εn + o p (). (27) Note that for 0 j < [nr], S j,ε[nr] and S ε[nr] are overlapped. Put Θ j = j k= θ n+ k, j n. Note that as n, Θ n and Θ n /θ(). After a straightforward calculation, the second term of the right hand side of (27) can be written as θ j (S j,ε[nr] S ε[nr] ) j=0 θ()v εn = Θ n θ() j= Θ j ε j n Θ n V εn Θ j ε j (n [nr]). (28) Θ n By Kolmogorov s Three Series Theorem, it is easy to show that as n, n j= ε j n/v εn and n j= ε j (n [nr])/v εn converge almost surely. This fact, together with Θ n, implies that j= Θ j Θ n ε j n V εn 0 a.s. and j= j= V εn Θ j Θ n ε j (n [nr]) V εn 0 a.s. (29) by Kronecker s lemma (see Petrov, 995, p.209), where notation a.s. stands for almost surely. Then we have S [nr] θ()v εn = S ε[nr] V εn + o p () W (r), 0 r, (30) 2
15 proving (24). To deal with the term λ ln = V n /[θ()v εn ] in (25) we note that V 2 n = [θ(l)ε k ] 2 = k= j=0 θ 2 j V 2 j,εn + 2 j=0 i=j+ θ j θ i ε k j ε k i =: A + A 2. (3) k= Noting θ 2 () = j=0 θ2 j + 2 j=0 i=j+ θ jθ i, by assumption this equality implies that 2 j=0 i=j+ θ j θ i /θ 2 (). (32) Note that, as n, V 2 εn/n p and n n k= ε k jε k i p 0 by the law of large numbers (LLN). Therefore, by (32) we have A 2 θ 2 ()V 2 εn Similar to (33), it is readily seen that = 2 j=0 i=j+ θ jθ n i k= ε k jε k i θ 2 () Vεn 2 p 0. (33) A θ 2 ()V 2 εn = j=0 θ2 j θ 2 () V 2 j,εn V 2 εn p 0, (34) since Vj,εn 2 /V εn 2 p and j=0 θ2 j /θ2 () 0 by assumption. So we conclude from (33) and (34) that λ 2 ln = V 2 n /[θ()v εn ] 2 p 0. The results of parts (a) (d) follow directly from (5) (8) and (24). The proof of Theorem is complete. Remark. If θ j are of the form l(j)/j β, j, β (/2, ), for some slowly varying function l, then θ j are regularly varying with index β and {u k } 3
16 has long-range dependence. By Karamata s theorem (see Theorem.5.3 and Proposition.5. in Bingham et al. (987)), sup {θ j } θ n+ 0 as n, (35) j n+ where a n b n denotes lim n a n /b n =. In this case the condition in (23) is automatically satisfied since (35) implies that j=n+ θ j 0 as n. Remark 2. Confining attention to ε k being iid random variables, Theorem asserts that even though the coefficients θ j decay slowly, the asymptotic distribution is the standard Brownian motion (as in Wang et al. (2002)) rather than the fractional Brownian motions (cf., Lubian, 999; Wang et al., 2003; Wu and Min, 2004). Under long-range dependence, since the only condition given in the theorem automatically holds (see Remark above), we believe that the theorem imposes the weakest conditions (or no condition) on the coefficients θ j than previous ones in the literature, for example in Wang et al. (2002). Specifically, as done in the latter paper, Theorem also abolishes the condition E(ε 4 k ) <, which represents a further improvement of the result of the latter paper in the respect that the θ j are allowed to decay sufficiently slowly. Theorem 2. Let {y t } and {u t } be generated according to () and (22), respectively, and let ε k, k Z, be iid random variables with E(ε 0 ) = 0 and E(ε 2 0 ) =. Let j=0 θ j < and θ() 0. Then we have S [nr] V n λ s W (r), 0 r, (36) 4
17 where λ 2 s := j=0 θ2 j /θ2 () <, and (a) ˆρ n 2 W 2 () λ 2 s ; 0 W 2 (r)dr (b) ˆτ n W 2 () λ 2 s 2λ s [ ; 0 W 2 (r)dr] /2 (c) ˆρ µn 2[W 2 () λ 2 s] W () 0 W (r)dr 0 W 2 (r)dr [ ; 0 W (r)dr]2 (d) ˆτ µn λ s 2[W 2 () λ 2 s] W () 0 { W (r)dr 0 W 2 (r)dr [ 0 W (r)dr]2} /2. Proof of Theorem 2. The proof of Theorem 2 is similar to the one of Theorem. First, since by assumption j=0 θ j <, it is necessary to have θ j 0 as j. It implies that as n, j=n+ θ j 0. By this fact and by (26), similar to (27) we have S [nr] θ()v εn = S ε[nr] V εn + j=0 θ j (S j,ε[nr] S ε[nr] ) θ()v εn + o p (). (37) Put Θ j = n k=j θ k, j n. Then the second term of the right hand side of (37) can be written as j=0 θ j (S j,ε[nr] S ε[nr] ) θ()v εn = j= Θ j ε j θ()v εn j= Θ j ε [nr]+ j θ()v εn =: M M 2. (38) Again, since the coefficients θ j are absolutely summable, it is sufficient to show that Θ j /θ() 0 as j. Noting that V 2 εn/n p as n, it thus follows from the Toeplitz lemma (Stout, 974, p.20) that E(M 2 ) = n n j= [ Θ j /θ()] 2 0. Then the weak LLN (Petrov, 995, p.34) implies that M p E(M ) = 0. By the same method used above, we can also show that M 2 p 0, and thus details are omitted. Now, we conclude from (37) and (38) that S [nr] /[θ()v εn ] = S ε[nr] /V εn + o p () W (r), 0 r. 5
18 Next, define λ 2 sn = V 2 n /[θ()v εn ] 2. Similar to (3), (33) and (34), we have that λ 2 sn p j=0 θ2 j /θ2 () =: λ 2 s <. Then (36) follows immediately from this and the result of the previous paragraph. Finally, the results of parts (a) (d) follows directly from (5) (8) and (36). The proof of Theorem 2 is complete. Remark 3. For short-memory linear processes with iid innovations, Theorem 2 holds under no condition on the coefficients θ j except that both j=0 θ j 0 and j=0 θ j <. Again, the required moment conditions are E(ε 2 k ) < rather than E(ε4 k ) < (cf. Wu and Min, 2004). It is easily seen that the theorem improves the previous similar results given by Phillips and Solo (992), Phillips and Xiao (998), Chang and Park (2002), Wang et al. (2002), and Wu and Min (2004). 5 Conclusion In contrast to the classical FCLT, the self-normalized version of the theorem possesses a major advantage that moment conditions are not the main prerequisite for convergence of the distribution of statistics based on the self-normalized partial sums (Giné et al., 997; Csörgő et al., 2003; Hall and Wang, 2004; Chistyakov and Götze, 2004). The paper demonstrates that the self-normalized limit theorems are particularly useful in characterizing the asymptotic distributions of the DF tests for a unit root. Thus, it seems 6
19 reasonable to expect that it would be helpful in deriving asymptotic distributions for many statistics arising from the study of nonstationary time series. 7
20 References Araujo, A. and E. Giné (980), The Central Limit Theorem for Real and Banach Valued Random Variables, New York: Wiley. Billingsley, P. (968), Convergence of Probability Measure, New York: Wiley. Bingham, N.H., C.M. Goldie, and J.L. Teugels (987), Regular Variation, Cambridge, UK: Cambridge University Press. Chan, N.H. and C.Z. Wei (988), Limiting Distributions of Least Squares Estimates of Unstable Autoregressive Processes, Annals of Statistics, 6, Chang, Y. and J.Y. Park (2002), On the Asymptotics of ADF Tests for Unit Roots, Econometric Reviews, 2, Chistyakov, G.P. and F. Götze (2004), Limit Distributions of Studentized Means, Annals of Probability, 32, Csörgő, M., B. Szyszkowicz, and Q. Wang (2003), Donsker s Theorem for Self-Normalized Partial Sums Processes, Annals of Probability, 3, Davydov, Yu.A. (970), The Invariance Principle for Stationary Processes, Theory of Probability and Its Applications, 5, Dickey, D.A. and W.A. Fuller (979), Distribution of the Estimators for 8
21 Autoregressive Time Series with a Unit Root, Journal of the American Statistical Association, 74, Dickey, D.A. and W.A. Fuller (98), Likelihood Ratio Statistics for Autoregressive Time Series with a Unit Root, Econometrica, 49, Egorov, V.A. (996), On the Asymptotic Behavior of Self-Normed Sums of Random Variables, Theory of Probability and Its Applications, 4, Feller, W. (966), An Introduction to Probability Theory and Its Applications 2, New York: Wiley. Fuller, W.A. (996), Introduction to Statistical Time Series, 2nd ed, New York: Wiley. Giné, E., F. Götze, and D.M. Mason (997), When is the Student t-statistic Asymptotically Standard Normal? Annals of Probability, 25, Gnedenko, B.V. and A.N. Kolmogorov (968), Limit Distributions for Sums of Independent Random Variables, transl. by K.L. Chung, New York: Addison-Wesley. Hall, P. (992), Convergence Rates in the Central Limit Theorem for Means of Autoregressive and Moving Average Sequences, Stochastic Processes and Their Applications, 43,
22 Hall, P. and Q. Wang (2004), Exact Convergence Rate and Leading Term in Central Limit Theorem for Student s t Statistic, Annals of Probability, 32, Hamilton, J.D. (994), Time Series Analysis, Princeton: Princeton University Press. Kruglov, V.M. and G.N. Petrovskaya (200), Weak Convergence of a Certain Functional, Theory of Probability and Its Applications, 46, Logan, B.F., C.L. Mallows, S.O. Rice, and L.A. Shepp (973), Limit Distributions of Self-Normalized Sums, Annals of Probability,, Lubian, D. (999), Long-Memory Errors in Time Series Regressions with a Unit Root, Journal of Time Series Analysis, 20, Petrov, V.V. (995), Limit Theorems of Probability Theory, Oxford: Clarendon Press. Phillips, P.C.B. (987), Time Series Regression with a Unit Root, Econometrica, 55, Phillips, P.C.B. and P. Perron (988), Testing for a Unit Root in Time Series Regression, Biometrika, 75, Phillips, P.C.B. and V. Solo (992), Asymptotics for Linear Processes, Annals of Statistics, 20,
23 Phillips, P.C.B. and Z. Xiao (998), A Primer on Unit Root Testing, Journal of Economic Surveys, 2, Said, S.E. and D.A. Dickey (984), Testing for Unit Roots in Autoregressive-Moving Average Models of Unknown Order, Biometrika, 7, Stock, J.H. (994), Unit Root, Structural Breaks and Trends, in R.F. Engle and D.L. McFadden (ed), Handbook of Econometrics, vol. IV, , Elsevier, Amsterdam. Stout, W.F. (974), Almost Sure Convergence, New York: Academic Press. Wang, Q., Y.-X. Lin, and C.M. Gulati (2002), The Invariance Principle for Linear Processes with Applications, Econometric Theory, 8, Wang, Q., Y.-X. Lin, and C.M. Gulati (2003), Asymptotics for General Fractionally Integrated Processes with Applications to Unit Root Tests, Econometric Theory, 9, White, J.S. (958), The Limiting Distribution of the Serial Correlation Coefficient in the Explosive Case, Annals of Mathematical Statistics, 29, Wu, W.B. and W. Min (2004), On Linear Processes with Dependent Innovations, Stochastic Processes and Their Applications, forthcoming. 2
Taking a New Contour: A Novel View on Unit Root Test 1
Taking a New Contour: A Novel View on Unit Root Test 1 Yoosoon Chang Department of Economics Rice University and Joon Y. Park Department of Economics Rice University and Sungkyunkwan University Abstract
More informationQED. Queen s Economics Department Working Paper No. 1244
QED Queen s Economics Department Working Paper No. 1244 A necessary moment condition for the fractional functional central limit theorem Søren Johansen University of Copenhagen and CREATES Morten Ørregaard
More informationDarmstadt Discussion Papers in Economics
Darmstadt Discussion Papers in Economics The Effect of Linear Time Trends on Cointegration Testing in Single Equations Uwe Hassler Nr. 111 Arbeitspapiere des Instituts für Volkswirtschaftslehre Technische
More informationCREATES Research Paper A necessary moment condition for the fractional functional central limit theorem
CREATES Research Paper 2010-70 A necessary moment condition for the fractional functional central limit theorem Søren Johansen and Morten Ørregaard Nielsen School of Economics and Management Aarhus University
More informationUNIT ROOT TESTING FOR FUNCTIONALS OF LINEAR PROCESSES
Econometric Theory, 22, 2006, 1 14+ Printed in the United States of America+ DOI: 10+10170S0266466606060014 UNIT ROOT TESTING FOR FUNCTIONALS OF LINEAR PROCESSES WEI BIAO WU University of Chicago We consider
More informationMoreover, the second term is derived from: 1 T ) 2 1
170 Moreover, the second term is derived from: 1 T T ɛt 2 σ 2 ɛ. Therefore, 1 σ 2 ɛt T y t 1 ɛ t = 1 2 ( yt σ T ) 2 1 2σ 2 ɛ 1 T T ɛt 2 1 2 (χ2 (1) 1). (b) Next, consider y 2 t 1. T E y 2 t 1 T T = E(y
More informationPh.D. Seminar Series in Advanced Mathematical Methods in Economics and Finance UNIT ROOT DISTRIBUTION THEORY I. Roderick McCrorie
Ph.D. Seminar Series in Advanced Mathematical Methods in Economics and Finance UNIT ROOT DISTRIBUTION THEORY I Roderick McCrorie School of Economics and Finance University of St Andrews 23 April 2009 One
More informationSelf-normalized Cramér-Type Large Deviations for Independent Random Variables
Self-normalized Cramér-Type Large Deviations for Independent Random Variables Qi-Man Shao National University of Singapore and University of Oregon qmshao@darkwing.uoregon.edu 1. Introduction Let X, X
More informationTesting for a unit root in an ar(1) model using three and four moment approximations: symmetric distributions
Hong Kong Baptist University HKBU Institutional Repository Department of Economics Journal Articles Department of Economics 1998 Testing for a unit root in an ar(1) model using three and four moment approximations:
More informationA TIME SERIES PARADOX: UNIT ROOT TESTS PERFORM POORLY WHEN DATA ARE COINTEGRATED
A TIME SERIES PARADOX: UNIT ROOT TESTS PERFORM POORLY WHEN DATA ARE COINTEGRATED by W. Robert Reed Department of Economics and Finance University of Canterbury, New Zealand Email: bob.reed@canterbury.ac.nz
More informationTesting for non-stationarity
20 November, 2009 Overview The tests for investigating the non-stationary of a time series falls into four types: 1 Check the null that there is a unit root against stationarity. Within these, there are
More informationUnivariate Unit Root Process (May 14, 2018)
Ch. Univariate Unit Root Process (May 4, 8) Introduction Much conventional asymptotic theory for least-squares estimation (e.g. the standard proofs of consistency and asymptotic normality of OLS estimators)
More informationNOTES AND PROBLEMS IMPULSE RESPONSES OF FRACTIONALLY INTEGRATED PROCESSES WITH LONG MEMORY
Econometric Theory, 26, 2010, 1855 1861. doi:10.1017/s0266466610000216 NOTES AND PROBLEMS IMPULSE RESPONSES OF FRACTIONALLY INTEGRATED PROCESSES WITH LONG MEMORY UWE HASSLER Goethe-Universität Frankfurt
More informationOn detection of unit roots generalizing the classic Dickey-Fuller approach
On detection of unit roots generalizing the classic Dickey-Fuller approach A. Steland Ruhr-Universität Bochum Fakultät für Mathematik Building NA 3/71 D-4478 Bochum, Germany February 18, 25 1 Abstract
More informationLM threshold unit root tests
Lee, J., Strazicich, M.C., & Chul Yu, B. (2011). LM Threshold Unit Root Tests. Economics Letters, 110(2): 113-116 (Feb 2011). Published by Elsevier (ISSN: 0165-1765). http://0- dx.doi.org.wncln.wncln.org/10.1016/j.econlet.2010.10.014
More informationThe Number of Bootstrap Replicates in Bootstrap Dickey-Fuller Unit Root Tests
Working Paper 2013:8 Department of Statistics The Number of Bootstrap Replicates in Bootstrap Dickey-Fuller Unit Root Tests Jianxin Wei Working Paper 2013:8 June 2013 Department of Statistics Uppsala
More informationA Test of Cointegration Rank Based Title Component Analysis.
A Test of Cointegration Rank Based Title Component Analysis Author(s) Chigira, Hiroaki Citation Issue 2006-01 Date Type Technical Report Text Version publisher URL http://hdl.handle.net/10086/13683 Right
More informationConsider the trend-cycle decomposition of a time series y t
1 Unit Root Tests Consider the trend-cycle decomposition of a time series y t y t = TD t + TS t + C t = TD t + Z t The basic issue in unit root testing is to determine if TS t = 0. Two classes of tests,
More informationTHE UNIVERSITY OF CHICAGO Booth School of Business Business 41914, Spring Quarter 2013, Mr. Ruey S. Tsay
THE UNIVERSITY OF CHICAGO Booth School of Business Business 494, Spring Quarter 03, Mr. Ruey S. Tsay Unit-Root Nonstationary VARMA Models Unit root plays an important role both in theory and applications
More informationGLS detrending and unit root testing
Economics Letters 97 (2007) 222 229 www.elsevier.com/locate/econbase GLS detrending and unit root testing Dimitrios V. Vougas School of Business and Economics, Department of Economics, Richard Price Building,
More informationNonsense Regressions due to Neglected Time-varying Means
Nonsense Regressions due to Neglected Time-varying Means Uwe Hassler Free University of Berlin Institute of Statistics and Econometrics Boltzmannstr. 20 D-14195 Berlin Germany email: uwe@wiwiss.fu-berlin.de
More informationOn Bootstrap Implementation of Likelihood Ratio Test for a Unit Root
On Bootstrap Implementation of Likelihood Ratio Test for a Unit Root ANTON SKROBOTOV The Russian Presidential Academy of National Economy and Public Administration February 25, 2018 Abstract In this paper
More informationCh. 14 Stationary ARMA Process
Ch. 14 Stationary ARMA Process A general linear stochastic model is described that suppose a time series to be generated by a linear aggregation of random shock. For practical representation it is desirable
More informationProf. Dr. Roland Füss Lecture Series in Applied Econometrics Summer Term Introduction to Time Series Analysis
Introduction to Time Series Analysis 1 Contents: I. Basics of Time Series Analysis... 4 I.1 Stationarity... 5 I.2 Autocorrelation Function... 9 I.3 Partial Autocorrelation Function (PACF)... 14 I.4 Transformation
More informationarxiv: v1 [math.pr] 7 Aug 2009
A CONTINUOUS ANALOGUE OF THE INVARIANCE PRINCIPLE AND ITS ALMOST SURE VERSION By ELENA PERMIAKOVA (Kazan) Chebotarev inst. of Mathematics and Mechanics, Kazan State University Universitetskaya 7, 420008
More informationEconomics Division University of Southampton Southampton SO17 1BJ, UK. Title Overlapping Sub-sampling and invariance to initial conditions
Economics Division University of Southampton Southampton SO17 1BJ, UK Discussion Papers in Economics and Econometrics Title Overlapping Sub-sampling and invariance to initial conditions By Maria Kyriacou
More informationUnderstanding Regressions with Observations Collected at High Frequency over Long Span
Understanding Regressions with Observations Collected at High Frequency over Long Span Yoosoon Chang Department of Economics, Indiana University Joon Y. Park Department of Economics, Indiana University
More informationNon-Stationary Time Series and Unit Root Testing
Econometrics II Non-Stationary Time Series and Unit Root Testing Morten Nyboe Tabor Course Outline: Non-Stationary Time Series and Unit Root Testing 1 Stationarity and Deviation from Stationarity Trend-Stationarity
More informationOn the Breiman conjecture
Péter Kevei 1 David Mason 2 1 TU Munich 2 University of Delaware 12th GPSD Bochum Outline Introduction Motivation Earlier results Partial solution Sketch of the proof Subsequential limits Further remarks
More informationTime Series Analysis. James D. Hamilton PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY
Time Series Analysis James D. Hamilton PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY & Contents PREFACE xiii 1 1.1. 1.2. Difference Equations First-Order Difference Equations 1 /?th-order Difference
More informationThe Role of "Leads" in the Dynamic Title of Cointegrating Regression Models. Author(s) Hayakawa, Kazuhiko; Kurozumi, Eiji
he Role of "Leads" in the Dynamic itle of Cointegrating Regression Models Author(s) Hayakawa, Kazuhiko; Kurozumi, Eiji Citation Issue 2006-12 Date ype echnical Report ext Version publisher URL http://hdl.handle.net/10086/13599
More informationE 4101/5101 Lecture 9: Non-stationarity
E 4101/5101 Lecture 9: Non-stationarity Ragnar Nymoen 30 March 2011 Introduction I Main references: Hamilton Ch 15,16 and 17. Davidson and MacKinnon Ch 14.3 and 14.4 Also read Ch 2.4 and Ch 2.5 in Davidson
More informationSome functional (Hölderian) limit theorems and their applications (II)
Some functional (Hölderian) limit theorems and their applications (II) Alfredas Račkauskas Vilnius University Outils Statistiques et Probabilistes pour la Finance Université de Rouen June 1 5, Rouen (Rouen
More informationTesting for Unit Roots with Cointegrated Data
Discussion Paper No. 2015-57 August 19, 2015 http://www.economics-ejournal.org/economics/discussionpapers/2015-57 Testing for Unit Roots with Cointegrated Data W. Robert Reed Abstract This paper demonstrates
More informationA BOOTSTRAP VIEW ON DICKEY-FULLER CONTROL CHARTS FOR AR(1) SERIES
A BOOTSTRAP VIEW ON DICKEY-FULLER CONTROL CHARTS FOR AR(1) SERIES Ansgar Steland Ruhr-Universität Bochum, Germany Abstract: Dickey-Fuller control charts aim at monitoring a random walk until a given time
More informationConstruction of Stationarity Tests Title Distortions.
Construction of Stationarity Tests Title Distortions Author(s) Kurozumi, Eiji Citation Issue 2005-11 Date Type Technical Report Text Version URL http://hdl.handle.net/10086/16931 Right Hitotsubashi University
More informationTime 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 informationDEPARTMENT OF ECONOMICS AND FINANCE COLLEGE OF BUSINESS AND ECONOMICS UNIVERSITY OF CANTERBURY CHRISTCHURCH, NEW ZEALAND
DEPARTMENT OF ECONOMICS AND FINANCE COLLEGE OF BUSINESS AND ECONOMICS UNIVERSITY OF CANTERBURY CHRISTCHURCH, NEW ZEALAND Testing For Unit Roots With Cointegrated Data NOTE: This paper is a revision of
More informationECON 616: Lecture Two: Deterministic Trends, Nonstationary Processes
ECON 616: Lecture Two: Deterministic Trends, Nonstationary Processes ED HERBST September 11, 2017 Background Hamilton, chapters 15-16 Trends vs Cycles A commond decomposition of macroeconomic time series
More informationE 4160 Autumn term Lecture 9: Deterministic trends vs integrated series; Spurious regression; Dickey-Fuller distribution and test
E 4160 Autumn term 2016. Lecture 9: Deterministic trends vs integrated series; Spurious regression; Dickey-Fuller distribution and test Ragnar Nymoen Department of Economics, University of Oslo 24 October
More informationLECTURE 12 UNIT ROOT, WEAK CONVERGENCE, FUNCTIONAL CLT
MARCH 29, 26 LECTURE 2 UNIT ROOT, WEAK CONVERGENCE, FUNCTIONAL CLT (Davidson (2), Chapter 4; Phillips Lectures on Unit Roots, Cointegration and Nonstationarity; White (999), Chapter 7) Unit root processes
More informationOn the Set of Limit Points of Normed Sums of Geometrically Weighted I.I.D. Bounded Random Variables
On the Set of Limit Points of Normed Sums of Geometrically Weighted I.I.D. Bounded Random Variables Deli Li 1, Yongcheng Qi, and Andrew Rosalsky 3 1 Department of Mathematical Sciences, Lakehead University,
More informationUnit 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 informationA New Test in Parametric Linear Models with Nonparametric Autoregressive Errors
A New Test in Parametric Linear Models with Nonparametric Autoregressive Errors By Jiti Gao 1 and Maxwell King The University of Western Australia and Monash University Abstract: This paper considers a
More informationRESIDUAL-BASED BLOCK BOOTSTRAP FOR UNIT ROOT TESTING. By Efstathios Paparoditis and Dimitris N. Politis 1
Econometrica, Vol. 71, No. 3 May, 2003, 813 855 RESIDUAL-BASED BLOCK BOOTSTRAP FOR UNIT ROOT TESTING By Efstathios Paparoditis and Dimitris N. Politis 1 A nonparametric, residual-based block bootstrap
More informationBCT Lecture 3. Lukas Vacha.
BCT Lecture 3 Lukas Vacha vachal@utia.cas.cz Stationarity and Unit Root Testing Why do we need to test for Non-Stationarity? The stationarity or otherwise of a series can strongly influence its behaviour
More informationOn the Long-Run Variance Ratio Test for a Unit Root
On the Long-Run Variance Ratio Test for a Unit Root Ye Cai and Mototsugu Shintani Vanderbilt University May 2004 Abstract This paper investigates the effects of consistent and inconsistent long-run variance
More informationAsymptotic Statistics-III. Changliang Zou
Asymptotic Statistics-III Changliang Zou The multivariate central limit theorem Theorem (Multivariate CLT for iid case) Let X i be iid random p-vectors with mean µ and and covariance matrix Σ. Then n (
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
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 informationSingle Equation Linear GMM with Serially Correlated Moment Conditions
Single Equation Linear GMM with Serially Correlated Moment Conditions Eric Zivot October 28, 2009 Univariate Time Series Let {y t } be an ergodic-stationary time series with E[y t ]=μ and var(y t )
More informationOn Perron s Unit Root Tests in the Presence. of an Innovation Variance Break
Applied Mathematical Sciences, Vol. 3, 2009, no. 27, 1341-1360 On Perron s Unit Root ests in the Presence of an Innovation Variance Break Amit Sen Department of Economics, 3800 Victory Parkway Xavier University,
More informationGARCH( 1, 1) processes are near epoch dependent
Economics Letters 36 (1991) 181-186 North-Holland 181 GARCH( 1, 1) processes are near epoch dependent Bruce E. Hansen 1Jnrrwsity of Rochester, Rochester, NY 14627, USA Keceived 22 October 1990 Accepted
More informationInvariance principles for fractionally integrated nonlinear processes
IMS Lecture Notes Monograph Series Invariance principles for fractionally integrated nonlinear processes Xiaofeng Shao, Wei Biao Wu University of Chicago Abstract: We obtain invariance principles for a
More informationSpatial autoregression model:strong consistency
Statistics & Probability Letters 65 (2003 71 77 Spatial autoregression model:strong consistency B.B. Bhattacharyya a, J.-J. Ren b, G.D. Richardson b;, J. Zhang b a Department of Statistics, North Carolina
More informationThis chapter reviews properties of regression estimators and test statistics based on
Chapter 12 COINTEGRATING AND SPURIOUS REGRESSIONS This chapter reviews properties of regression estimators and test statistics based on the estimators when the regressors and regressant are difference
More information11/18/2008. So run regression in first differences to examine association. 18 November November November 2008
Time Series Econometrics 7 Vijayamohanan Pillai N Unit Root Tests Vijayamohan: CDS M Phil: Time Series 7 1 Vijayamohan: CDS M Phil: Time Series 7 2 R 2 > DW Spurious/Nonsense Regression. Integrated but
More informationUnit Root and Cointegration
Unit Root and Cointegration Carlos Hurtado Department of Economics University of Illinois at Urbana-Champaign hrtdmrt@illinois.edu Oct 7th, 016 C. Hurtado (UIUC - Economics) Applied Econometrics On the
More informationNon-Stationary Time Series and Unit Root Testing
Econometrics II Non-Stationary Time Series and Unit Root Testing Morten Nyboe Tabor Course Outline: Non-Stationary Time Series and Unit Root Testing 1 Stationarity and Deviation from Stationarity Trend-Stationarity
More informationTest for Parameter Change in ARIMA Models
Test for Parameter Change in ARIMA Models Sangyeol Lee 1 Siyun Park 2 Koichi Maekawa 3 and Ken-ichi Kawai 4 Abstract In this paper we consider the problem of testing for parameter changes in ARIMA models
More informationECONOMETRICS II, FALL Testing for Unit Roots.
ECONOMETRICS II, FALL 216 Testing for Unit Roots. In the statistical literature it has long been known that unit root processes behave differently from stable processes. For example in the scalar AR(1)
More informationB y t = γ 0 + Γ 1 y t + ε t B(L) y t = γ 0 + ε t ε t iid (0, D) D is diagonal
Structural VAR Modeling for I(1) Data that is Not Cointegrated Assume y t =(y 1t,y 2t ) 0 be I(1) and not cointegrated. That is, y 1t and y 2t are both I(1) and there is no linear combination of y 1t and
More informationA Bootstrap View on Dickey-Fuller Control Charts for AR(1) Series
AUSTRIAN JOURNAL OF STATISTICS Volume 35 (26), Number 2&3, 339 346 A Bootstrap View on Dickey-Fuller Control Charts for AR(1) Series Ansgar Steland RWTH Aachen University, Germany Abstract: Dickey-Fuller
More informationSimulating Properties of the Likelihood Ratio Test for a Unit Root in an Explosive Second Order Autoregression
Simulating Properties of the Likelihood Ratio est for a Unit Root in an Explosive Second Order Autoregression Bent Nielsen Nuffield College, University of Oxford J James Reade St Cross College, University
More informationOn the Error Correction Model for Functional Time Series with Unit Roots
On the Error Correction Model for Functional Time Series with Unit Roots Yoosoon Chang Department of Economics Indiana University Bo Hu Department of Economics Indiana University Joon Y. Park Department
More informationModelling of Economic Time Series and the Method of Cointegration
AUSTRIAN JOURNAL OF STATISTICS Volume 35 (2006), Number 2&3, 307 313 Modelling of Economic Time Series and the Method of Cointegration Jiri Neubauer University of Defence, Brno, Czech Republic Abstract:
More informationTime series: Cointegration
Time series: Cointegration May 29, 2018 1 Unit Roots and Integration Univariate time series unit roots, trends, and stationarity Have so far glossed over the question of stationarity, except for my stating
More informationSteven Cook University of Wales Swansea. Abstract
On the finite sample power of modified Dickey Fuller tests: The role of the initial condition Steven Cook University of Wales Swansea Abstract The relationship between the initial condition of time series
More informationFinal Exam November 24, Problem-1: Consider random walk with drift plus a linear time trend: ( t
Problem-1: Consider random walk with drift plus a linear time trend: y t = c + y t 1 + δ t + ϵ t, (1) where {ϵ t } is white noise with E[ϵ 2 t ] = σ 2 >, and y is a non-stochastic initial value. (a) Show
More informationAdditive functionals of infinite-variance moving averages. Wei Biao Wu The University of Chicago TECHNICAL REPORT NO. 535
Additive functionals of infinite-variance moving averages Wei Biao Wu The University of Chicago TECHNICAL REPORT NO. 535 Departments of Statistics The University of Chicago Chicago, Illinois 60637 June
More informationEcon 423 Lecture Notes: Additional Topics in Time Series 1
Econ 423 Lecture Notes: Additional Topics in Time Series 1 John C. Chao April 25, 2017 1 These notes are based in large part on Chapter 16 of Stock and Watson (2011). They are for instructional purposes
More informationFinancial Econometrics
Financial Econometrics Nonlinear time series analysis Gerald P. Dwyer Trinity College, Dublin January 2016 Outline 1 Nonlinearity Does nonlinearity matter? Nonlinear models Tests for nonlinearity Forecasting
More informationAnalytical derivates of the APARCH model
Analytical derivates of the APARCH model Sébastien Laurent Forthcoming in Computational Economics October 24, 2003 Abstract his paper derives analytical expressions for the score of the APARCH model of
More informationA Functional Central Limit Theorem for an ARMA(p, q) Process with Markov Switching
Communications for Statistical Applications and Methods 2013, Vol 20, No 4, 339 345 DOI: http://dxdoiorg/105351/csam2013204339 A Functional Central Limit Theorem for an ARMAp, q) Process with Markov Switching
More informationEmpirical 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 informationInternational Journal of Pure and Applied Mathematics Volume 21 No , THE VARIANCE OF SAMPLE VARIANCE FROM A FINITE POPULATION
International Journal of Pure and Applied Mathematics Volume 21 No. 3 2005, 387-394 THE VARIANCE OF SAMPLE VARIANCE FROM A FINITE POPULATION Eungchun Cho 1, Moon Jung Cho 2, John Eltinge 3 1 Department
More informationDEPARTMENT OF ECONOMICS
ISSN 0819-64 ISBN 0 7340 616 1 THE UNIVERSITY OF MELBOURNE DEPARTMENT OF ECONOMICS RESEARCH PAPER NUMBER 959 FEBRUARY 006 TESTING FOR RATE-DEPENDENCE AND ASYMMETRY IN INFLATION UNCERTAINTY: EVIDENCE FROM
More informationMOMENT CONVERGENCE RATES OF LIL FOR NEGATIVELY ASSOCIATED SEQUENCES
J. Korean Math. Soc. 47 1, No., pp. 63 75 DOI 1.4134/JKMS.1.47..63 MOMENT CONVERGENCE RATES OF LIL FOR NEGATIVELY ASSOCIATED SEQUENCES Ke-Ang Fu Li-Hua Hu Abstract. Let X n ; n 1 be a strictly stationary
More informationSingle Equation Linear GMM with Serially Correlated Moment Conditions
Single Equation Linear GMM with Serially Correlated Moment Conditions Eric Zivot November 2, 2011 Univariate Time Series Let {y t } be an ergodic-stationary time series with E[y t ]=μ and var(y t )
More informationOn the robustness of cointegration tests when series are fractionally integrated
On the robustness of cointegration tests when series are fractionally integrated JESUS GONZALO 1 &TAE-HWYLEE 2, 1 Universidad Carlos III de Madrid, Spain and 2 University of California, Riverside, USA
More informationAR, MA and ARMA models
AR, MA and AR by Hedibert Lopes P Based on Tsay s Analysis of Financial Time Series (3rd edition) P 1 Stationarity 2 3 4 5 6 7 P 8 9 10 11 Outline P Linear Time Series Analysis and Its Applications For
More informationChange detection problems in branching processes
Change detection problems in branching processes Outline of Ph.D. thesis by Tamás T. Szabó Thesis advisor: Professor Gyula Pap Doctoral School of Mathematics and Computer Science Bolyai Institute, University
More information5: MULTIVARATE STATIONARY PROCESSES
5: MULTIVARATE STATIONARY PROCESSES 1 1 Some Preliminary Definitions and Concepts Random Vector: A vector X = (X 1,..., X n ) whose components are scalarvalued random variables on the same probability
More informationNonstationary Panels
Nonstationary Panels Based on chapters 12.4, 12.5, and 12.6 of Baltagi, B. (2005): Econometric Analysis of Panel Data, 3rd edition. Chichester, John Wiley & Sons. June 3, 2009 Agenda 1 Spurious Regressions
More information1 Linear Difference Equations
ARMA Handout Jialin Yu 1 Linear Difference Equations First order systems Let {ε t } t=1 denote an input sequence and {y t} t=1 sequence generated by denote an output y t = φy t 1 + ε t t = 1, 2,... with
More informationResponse surface models for the Elliott, Rothenberg, Stock DF-GLS unit-root test
Response surface models for the Elliott, Rothenberg, Stock DF-GLS unit-root test Christopher F Baum Jesús Otero Stata Conference, Baltimore, July 2017 Baum, Otero (BC, U. del Rosario) DF-GLS response surfaces
More informationDiscussion Papers in Economics
Discussion Papers in Economics No. 14/19 Specification Testing in Nonstationary Time Series Models Jia Chen, Jiti Gao, Degui Li and Zhengyan Lin Department of Economics and Related Studies University of
More informationEconometric Methods for Panel Data
Based on the books by Baltagi: Econometric Analysis of Panel Data and by Hsiao: Analysis of Panel Data Robert M. Kunst robert.kunst@univie.ac.at University of Vienna and Institute for Advanced Studies
More informationSUMMARY OF RESULTS ON PATH SPACES AND CONVERGENCE IN DISTRIBUTION FOR STOCHASTIC PROCESSES
SUMMARY OF RESULTS ON PATH SPACES AND CONVERGENCE IN DISTRIBUTION FOR STOCHASTIC PROCESSES RUTH J. WILLIAMS October 2, 2017 Department of Mathematics, University of California, San Diego, 9500 Gilman Drive,
More informationAsymptotic Properties of White s Test for Heteroskedasticity and the Jarque-Bera Test for Normality
Asymptotic Properties of White s est for Heteroskedasticity and the Jarque-Bera est for Normality Carlos Caceres Nuffield College, University of Oxford Bent Nielsen Nuffield College, University of Oxford
More informationAn almost sure invariance principle for additive functionals of Markov chains
Statistics and Probability Letters 78 2008 854 860 www.elsevier.com/locate/stapro An almost sure invariance principle for additive functionals of Markov chains F. Rassoul-Agha a, T. Seppäläinen b, a Department
More informationARTICLE IN PRESS Statistics and Probability Letters ( )
Statistics and Probability Letters ( ) Contents lists available at ScienceDirect Statistics and Probability Letters journal homepage: www.elsevier.com/locate/stapro The functional central limit theorem
More informationResponse surface models for the Elliott, Rothenberg, Stock DF-GLS unit-root test
Response surface models for the Elliott, Rothenberg, Stock DF-GLS unit-root test Christopher F Baum Jesús Otero UK Stata Users Group Meetings, London, September 2017 Baum, Otero (BC, U. del Rosario) DF-GLS
More informationLONG TERM DEPENDENCE IN STOCK RETURNS
LONG TERM DEPENDENCE IN STOCK RETURNS John T. Barkoulas Department of Economics Boston College Christopher F. Baum Department of Economics Boston College Keywords: Stock returns, long memory, fractal dynamics,
More informationThe autocorrelation and autocovariance functions - helpful tools in the modelling problem
The autocorrelation and autocovariance functions - helpful tools in the modelling problem J. Nowicka-Zagrajek A. Wy lomańska Institute of Mathematics and Computer Science Wroc law University of Technology,
More informationA Primer on Asymptotics
A Primer on Asymptotics Eric Zivot Department of Economics University of Washington September 30, 2003 Revised: October 7, 2009 Introduction The two main concepts in asymptotic theory covered in these
More informationEconometrics. Week 11. Fall Institute of Economic Studies Faculty of Social Sciences Charles University in Prague
Econometrics Week 11 Institute of Economic Studies Faculty of Social Sciences Charles University in Prague Fall 2012 1 / 30 Recommended Reading For the today Advanced Time Series Topics Selected topics
More informationEconometrics of financial markets, -solutions to seminar 1. Problem 1
Econometrics of financial markets, -solutions to seminar 1. Problem 1 a) Estimate with OLS. For any regression y i α + βx i + u i for OLS to be unbiased we need cov (u i,x j )0 i, j. For the autoregressive
More informationThis note introduces some key concepts in time series econometrics. First, we
INTRODUCTION TO TIME SERIES Econometrics 2 Heino Bohn Nielsen September, 2005 This note introduces some key concepts in time series econometrics. First, we present by means of examples some characteristic
More informationLM TEST FOR THE CONSTANCY OF REGRESSION COEFFICIENT WITH MOVING AVERAGE INNOVATION
Sankhyā : he Indian Journal of Statistics 22, Volume 64, Series B, Pt.2, pp 24-233 LM ES FOR HE CONSANCY OF REGRESSION COEFFICIEN WIH MOVING AVERAGE INNOVAION By MEIHUI GUO and C.C. SHEN National Sun Yat-Sen
More informationTitle. Description. var intro Introduction to vector autoregressive models
Title var intro Introduction to vector autoregressive models Description Stata has a suite of commands for fitting, forecasting, interpreting, and performing inference on vector autoregressive (VAR) models
More information