#### DMCA

## RANGE UNIT ROOT TESTS (2003)

Citations: | 1 - 1 self |

### Citations

5091 | Probability and measure - Billingsley - 1995 |

3688 |
Density estimation for statistics and data analysis. Monographs on Statistics and Applied Probability 26
- Silverman
- 1986
(Show Context)
Citation Context ...ull hypothesis of a random walk with Nid(0, 1) errors. The estimates were obtained from 1000 replications and for different sample sizes using the Epanechnikov kernel (see for example Silverman, 1986 =-=[51]-=- and Hardle, 1990 [21]), which is optimal in the the mean-square error (MSE) sense 2 . 2 (n) The density of J 0 ˆfh(x) = 1 Pni=1 K nh was estimated as x−xi h ´ , with K(.) given by K(u) = 3 4 (1 − u... |

2364 | An Introduction to Probability Theory and - Feller - 1968 |

1823 |
Continuous martingales and Brownian motion. Third edition, Grundlehren der
- Revuz, Yor
- 1999
(Show Context)
Citation Context ...ocess Wt.on the interval [0, 1] 12The local time Lt(x) of a Brownian motion measures the amount of time spent by this process in the neighborhood of the point x up to time t (see Revuz and Yor, 1991 =-=[47]-=-). Z t 1 Lt(x) = lim 1(x − ɛ<Ws <x+ ɛ)ds. ɛ→0 2ɛ More details can be found in the Appendix. 0 Theorem 2 Let Lt(x) denote the local time defined as above. Under the null hypothesis H0 : xt = xt−1 + ɛt ... |

1685 |
Brownian Motion and Stochastic Calculus
- Karatzas, Shreve
- 1991
(Show Context)
Citation Context ...efollowinglemmas: Let M w t , max0≤s≤t Wt; and Yt = M w t − Wt. Lemma 1 (Lévy , 1948 [26]). The processes {|Wt| ; 0 ≤ t<∞} and {Yt; 0 ≤ t<∞} havethesamelaw. Proof: See Karatzas and Shreve (1988 p.210 =-=[24]-=-). Lemma 2 (Skorohod, 1961; Lévy, 1948[26]). The processes {(M w t , 0 ≤ t<∞} and {2Lt(0); 0 ≤ t<∞} havethesamelaw, given by: P {M w t <x} = r 2 πt Zx That is, M w t ∼ |N(0,t)| Proof: See Karatzas and... |

1217 | Testing the null hypothesis of stationarity against the alternative of a unit root: how sure are we that economic time series are non stationary - Kwiatkowski, Phillips, et al. - 1992 |

1056 | Modelling Extremal Events for Insurance and Finance - Embrechts, Klüppelberg, et al. - 1997 |

919 |
Estimating and testing linear models with multiple structural changes
- Bai, Perron
- 1998
(Show Context)
Citation Context ...anger, 1993 [14]). Alternative procedures for testing unit roots were proposed by Lo (1991 [28]), Kwiatowski et al. (1992 [25]), Stock (1994 [55]) and the recent contributions of Bai and Perron (1998 =-=[7]-=-). Yet, all these tests rely on assumptions which are too restrictive in practice. In fact, they were reported to have poor power performances when confronted to deviations from the standard linear co... |

894 |
The Great Crash, the Oil Price Shock, and the Unit Root Hypothesis. Econometrica 57(6): 1361--1401
- Perron
- 1989
(Show Context)
Citation Context ...too restrictive in practice. In fact, they were reported to have poor power performances when confronted to deviations from the standard linear context (see for example, Sims, 1988 [52]; Perron, 1989 =-=[38]-=-; Perron, 1990 [39]; Schotman and Van Dijk, 1991 [50]; Aparicio, 1995 [2]). The appropriate handling of such departures as level shifts, trend breaks and nonlinearities calls for the development of ro... |

764 |
Introduction to Statistical Time Series
- Fuller
- 1996
(Show Context)
Citation Context ...r for an overwhelming rejection of the null hypothesis of a unit root in favour of a stationary alternative. All the statistics are significant at the 1% level (the critical values, from Fuller (1976 =-=[16]-=-) are -29.5, for both PP and MPP test statistics, and -3.96 for the tratio ADF test statistic), while the values obtained for test statistics were -41.55, -37.57 and -6.61, respectively). In all cases... |

738 |
Trends and Random Walks in Macroeconomic Time Series: Some Evidence and Implications
- Nelson, Plosser
- 1982
(Show Context)
Citation Context ...ents in time series are associated with the presence of unit roots in their data generating process (DGP). Such time series are said to be integrated. The pioneering work of Nelson and Plosser (1982) =-=[36]-=- led to the belief that many economic time series were best described in this way. This promted a large amount of research on unit root time series, covering both theoretical and empirical aspects. Th... |

624 |
Modelling Nonlinear Economic Relationships
- Granger, Terasvirta
- 1993
(Show Context)
Citation Context ...andard unit root model. To achieve this, we first propose a characterization of this “long wave” in terms of what we call LowFrequency Features (LFF hereafter). Following Granger and Terasvirta (1993 =-=[19]-=-) and Anderson and Vahid (1998 [1]), a feature is essentially any dominating statistical property exhibited by a time series. Features may refer to either the mean behavior or to higher-order moments ... |

599 | Testing for a Unit Root - Phillips, Perron - 1988 |

396 | Time series regression with a unit root - Phillips - 1987 |

241 |
Long-term memory in stock market prices
- Lo
- 1991
(Show Context)
Citation Context ...ike models, fool standard unit root tests (see for instance, Granger and Hallman 1991 [18], and Ermini and Granger, 1993 [14]). Alternative procedures for testing unit roots were proposed by Lo (1991 =-=[28]-=-), Kwiatowski et al. (1992 [25]), Stock (1994 [55]) and the recent contributions of Bai and Perron (1998 [7]). Yet, all these tests rely on assumptions which are too restrictive in practice. In fact, ... |

232 |
Tests for Unit Roots: A Monte Carlo Investigation
- Schwert
- 1989
(Show Context)
Citation Context ...ime series models impose, however, severe restrictions on the DGP’s of the data. For example, when the errors are negatively correlated, the DF test exhibits important size distortions (Schwert, 1989 =-=[57]-=-). Many real world time series exhibit also nonlinearities, outliers and structural breaks. All these features, which cannot be properly captured with random-walk-like models, fool standard unit root ... |

231 |
Recursive and sequential tests of the unit-root and trend-break hypotheses: Theory and international evidence
- Banerjee, Lumsdaine, et al.
- 1992
(Show Context)
Citation Context ...nce parameters such as the number of breaks and their timing, which has led several authors (see Zivot and Andrews, 1992 [59]; Perron and Vogelsang, 1992 [41]; Banerjee, 15Lumsdaine and Stock , 1992 =-=[8]-=-; and Stock, 1994 [55]) to propose recursive and sequential testing procedures in order to estimate these parameters. In the light of the previous difficulties, it may be interesting to analyse the po... |

211 |
Testing for a unit root in a time series with a changing mean
- Perron
- 1990
(Show Context)
Citation Context ...practice. In fact, they were reported to have poor power performances when confronted to deviations from the standard linear context (see for example, Sims, 1988 [52]; Perron, 1989 [38]; Perron, 1990 =-=[39]-=-; Schotman and Van Dijk, 1991 [50]; Aparicio, 1995 [2]). The appropriate handling of such departures as level shifts, trend breaks and nonlinearities calls for the development of robust unit root test... |

199 | The Great Crash, the Oil Price - PERRON - 1989 |

177 |
Processus stochastiques et mouvement brownien. Paris: GauthierVillars
- Lévy
- 1965
(Show Context)
Citation Context ...wnian path spends a lot of time close to 0, thus changing the sign relatively often. For the finalproof,werequirethefollowinglemmas: Let M w t , max0≤s≤t Wt; and Yt = M w t − Wt. Lemma 1 (Lévy , 1948 =-=[26]-=-). The processes {|Wt| ; 0 ≤ t<∞} and {Yt; 0 ≤ t<∞} havethesamelaw. Proof: See Karatzas and Shreve (1988 p.210 [24]). Lemma 2 (Skorohod, 1961; Lévy, 1948[26]). The processes {(M w t , 0 ≤ t<∞} and {2L... |

158 |
Smoothing Techniques with Implementation in S
- Härdle
- 1991
(Show Context)
Citation Context ...ndom walk with Nid(0, 1) errors. The estimates were obtained from 1000 replications and for different sample sizes using the Epanechnikov kernel (see for example Silverman, 1986 [51] and Hardle, 1990 =-=[21]-=-), which is optimal in the the mean-square error (MSE) sense 2 . 2 (n) The density of J 0 ˆfh(x) = 1 Pni=1 K nh was estimated as x−xi h ´ , with K(.) given by K(u) = 3 4 (1 − u2 )1(|u| ≤ 1). 110.0 ... |

156 |
Testing for a unit root in time series regression.” Biometrika 75
- Phillips, Perron
- 1988
(Show Context)
Citation Context ...on (1999 [11]) have reported empirical results obtained with the application of standard unit root tests, such as the Augmented DickeyFuller (ADF) test (1979 [12]), and the Phillips-Perron test (1988 =-=[44]-=-). They 26also have reported results obtained with a modification of the Phillips-Perron test (hereafter MPP), suggested by Stock (1990 [56]). Such a modified test is less prone to size distorsions i... |

125 |
Testing the null hypothesis of stationary against the alternative of a unit root
- Kwiatkowski, Phillips, et al.
- 1992
(Show Context)
Citation Context ...root tests (see for instance, Granger and Hallman 1991 [18], and Ermini and Granger, 1993 [14]). Alternative procedures for testing unit roots were proposed by Lo (1991 [28]), Kwiatowski et al. (1992 =-=[25]-=-), Stock (1994 [55]) and the recent contributions of Bai and Perron (1998 [7]). Yet, all these tests rely on assumptions which are too restrictive in practice. In fact, they were reported to have poor... |

93 |
Further Evidence on the Great Crash, the Oil Price Shock, and the Unit Root Hypothesis
- Zivot, DWK
- 1992
(Show Context)
Citation Context ...the critical values of standard unit root tests depend on the new unknown nuisance parameters such as the number of breaks and their timing, which has led several authors (see Zivot and Andrews, 1992 =-=[59]-=-; Perron and Vogelsang, 1992 [41]; Banerjee, 15Lumsdaine and Stock , 1992 [8]; and Stock, 1994 [55]) to propose recursive and sequential testing procedures in order to estimate these parameters. In t... |

78 |
asymptotic distribution functions for unit-root and cointegration tests
- MacKinnon
- 1994
(Show Context)
Citation Context ....1 10 20 30 40 50 60 70 80 Figure 15. Logarithm of the US/Finland real exchange rates deflated annual series from 1900 to 1987. Using the Mackinnon’s critical values for the ADF test (Mackinnon, 1994 =-=[31]-=-), the null hypothesis of a unit root is rejected at the 5% significance level (the ADF test statistic took the value -3.732041 while the 5% critical values was -3.4614). Alternatively, with a value o... |

75 |
Bayesian skepticism on unit root econometrics
- Sims
- 1988
(Show Context)
Citation Context ...umptions which are too restrictive in practice. In fact, they were reported to have poor power performances when confronted to deviations from the standard linear context (see for example, Sims, 1988 =-=[52]-=-; Perron, 1989 [38]; Perron, 1990 [39]; Schotman and Van Dijk, 1991 [50]; Aparicio, 1995 [2]). The appropriate handling of such departures as level shifts, trend breaks and nonlinearities calls for th... |

71 | Long Term Memory - Lo - 1991 |

59 |
The effects of additive outliers on tests for unit roots and cointegration
- Franses, Haldrup
- 1994
(Show Context)
Citation Context ...ess of the nominal size and thus will reject the unit-root hypothesis too often. The size distortion of the DF test in the presence of this type of outliers was quantified by Haldrup and Hanses (1994 =-=[20]-=-), who also demonstrated that the distribution of the parameter estimates changed dramatically when both the magnitude of the outliers and their frequency become large. Traditionally, the presence of ... |

49 | A Functional Central Limit Theorem For Weakly Dependent Sequences of Random Variables", Annals of Probability 12 - Herrndorf - 1984 |

45 | Nonlinear transformations of integrated time series - Granger, Hallman - 1991 |

45 | Theoretical Probability for Applications - Port - 1994 |

38 |
Spurious rejections by Dickey–Fuller tests in the presence of a break under the null
- Leybourne, Mills, et al.
- 1998
(Show Context)
Citation Context ...he oil crisis. In brief, if the permanent break is not explicitly taken into account standard unit root tests tend to find too many unit roots. However, as shown by Leybourne, Mills and Newbold (1998 =-=[27]-=-), it is also possible to reach the opposite conclusion when the break’s location appears at the beginning of the sample, that is that an I(0) time series be interpreted as I(1). Moreover, the critica... |

33 |
A Class of Tests for Integration and Cointegration
- Stock
- 1999
(Show Context)
Citation Context ...(ADF) test (1979 [12]), and the Phillips-Perron test (1988 [44]). They 26also have reported results obtained with a modification of the Phillips-Perron test (hereafter MPP), suggested by Stock (1990 =-=[56]-=-). Such a modified test is less prone to size distorsions in the presence of serial correlation in the first differences of the series. These three unit roots tests concur for an overwhelming rejectio... |

26 | Testing for a unit root in the presence of a variance shift - Hamori, Tokihisa - 1997 |

25 |
Two simple procedures for testing for a unit root when there are additive outliers
- Vogelsang
- 1999
(Show Context)
Citation Context ...This series, which contains a total of n =88observations (from 1900 to 1987), was constructed using the Gross Domestic Product (GDP) deflator. Previous analyses on this series done by Vogelsang (1999 =-=[58]-=-), Franses and Haldrup (1994 [20]), Perron and Vogelsang (1992 [41]), and Perron and Rodriguez (2000 [40]), point to the presence of an AO at date 1918 together with IO’s that produce temporary change... |

23 |
A Bayesian analysis of the unit root in real exchange rates
- Schotman, Dijk
- 1991
(Show Context)
Citation Context ...ted to have poor power performances when confronted to deviations from the standard linear context (see for example, Sims, 1988 [52]; Perron, 1989 [38]; Perron, 1990 [39]; Schotman and Van Dijk, 1991 =-=[50]-=-; Aparicio, 1995 [2]). The appropriate handling of such departures as level shifts, trend breaks and nonlinearities calls for the development of robust unit root tests. Certainly, the rejection of the... |

22 |
The Structure of Unemployment
- Papell, Murray, et al.
- 2000
(Show Context)
Citation Context ...esearch has supported the evidence of the existence of unit roots in the unemployment rate series of most countries (see for instance Mitchell, 1993 [35]; Arrufat et al.,1999 [6]; Papell et al., 2000 =-=[37]-=-). Yet the low power of standard unit root tests on stationary alternatives close to the nonstationary border contradicts such reports. One among many controversial study cases is the quarterly Montev... |

22 | In functionals for time series - Martin, Yohai - 1986 |

19 |
Exchange Rates and Inflation
- DORNBUSCH
- 1988
(Show Context)
Citation Context ...sitory. As an example, one such implication is the Purchasing Power Parity (PPP) hypothesis, which asserts that fluctuations in the real exchange rates of any countries are stationary (Dornbush, 1988 =-=[13]-=-). The existence of unit roots in time series is investigated by means of unit root tests. The application of standard unit root tests, such as the Dickey-Fuller (DF hereafter) test (Dickey and Fuller... |

17 |
Robust Rank Tests of the Unit Root Hypothesis
- Hasan, Koenker
- 1997
(Show Context)
Citation Context ...aussian distribution (Hoek et al., 1995 [23]; Rothenberg and Stock, 1997 [49]). The use of nonparametric statistics is another avenue of research in robust unit root testing. Hasan and Kroenker (1997 =-=[22]-=-) applied rank-based methods.to this problem and reported improved power performances on time series corrupted by a few large observations. The RUR procedure that we propose also belongs to the nonpar... |

17 |
Testing for unit roots and persistence in OECD unemployment rates
- Mitchell
- 1993
(Show Context)
Citation Context ...authors. For example, a great bulk of econometric research has supported the evidence of the existence of unit roots in the unemployment rate series of most countries (see for instance Mitchell, 1993 =-=[35]-=-; Arrufat et al.,1999 [6]; Papell et al., 2000 [37]). Yet the low power of standard unit root tests on stationary alternatives close to the nonstationary border contradicts such reports. One among man... |

17 | Searching for Additive Outliers in Nonstationary Time Series
- Perron, Rodriguez
- 2003
(Show Context)
Citation Context ...ross Domestic Product (GDP) deflator. Previous analyses on this series done by Vogelsang (1999 [58]), Franses and Haldrup (1994 [20]), Perron and Vogelsang (1992 [41]), and Perron and Rodriguez (2000 =-=[40]-=-), point to the presence of an AO at date 1918 together with IO’s that produce temporary changes at dates 1917, 1932, 1949 and 1957. 270.6 0.5 0.4 0.3 0.2 0.1 10 20 30 40 50 60 70 80 Figure 15. Logar... |

16 |
An outlier robust unit root test with an application to the extended Nelson-Plosser data
- Lucas
- 1995
(Show Context)
Citation Context ...owed in Arranz and Escribano (1998 [5]), who proposed filtering the contaminated series prior to standard unit-root testing. Single-stage robust unit-root tests were first proposed by Lucas (1995 a,b =-=[29,30]-=-) and by Franses and Lucas (1997 [15]) using M-estimators with high breakdown point and efficiency, instead of OLS estimators. However, these tests were really conceived for dealing with fat-tailed di... |

11 |
The limit distribution of level crossings of a random walk, and a simple unit root test,” Econometric Theory
- BURRIDGE, GUERRE
- 1996
(Show Context)
Citation Context ...bustizing cointegration tests. The key idea behind it the fact that the average number of level crossings for a unit root time series is smaller than for 2a stationary one (Burridge and Guerre, 1996 =-=[9]-=-). The RUR test outperforms standard unit root tests in several aspects. First, because of its invariance to monotonic transformations of the series and its robustness to the presence of additive outl... |

11 | Some Generalization of the Algebra of of I(1) Processes - Ermini, Granger - 1993 |

10 | Extremal theory for stochastic processes. The Annals of Probability - Leadbetter, Rootzén - 1988 |

7 |
Classical and Bayesian Aspects of Robust Unit Root Inference
- Hoek, Lucas, et al.
- 1995
(Show Context)
Citation Context ...s. Alternatively, some authors have followed a likelihood-based approach where inference is made about a particular fat-tailed distribution rather than on the Gaussian distribution (Hoek et al., 1995 =-=[23]-=-; Rothenberg and Stock, 1997 [49]). The use of nonparametric statistics is another avenue of research in robust unit root testing. Hasan and Kroenker (1997 [22]) applied rank-based methods.to this pro... |

6 |
1995! Nonlinear Cointegration and Some New Tests for Comovements+ Working paper 95-15, University of California at
- Aparicio, Granger
- 1981
(Show Context)
Citation Context ...since its test statistic is constructed from the running ranges of the series. The RUR test is a natural follow-up of the methodology proposed in Aparicio (1995 [2]) and in Aparicio and Granger (1995 =-=[4]-=-) for robustizing cointegration tests. The key idea behind it the fact that the average number of level crossings for a unit root time series is smaller than for 2a stationary one (Burridge and Guerr... |

6 |
Time series regression with a unit root
- B
- 1987
(Show Context)
Citation Context ...show that this test is either robust or invariant to a number of departures from this null hypothesis. J (n) 0 = 1 √ n nX t=1 1(∆R (x) t > 0). (1) The following set of assumptions from Phillips (1987 =-=[43]-=-) on the model errors ɛt will be needed in the subsequent results. These assumptions allows trading an increasing degree of temporal dependence against a decreasing degree of heteroskedasticity (and v... |

4 |
Order Statistics
- Galambos
- 1984
(Show Context)
Citation Context ... results in a sequence of running ranges. Formally, for a given time series xt, the terms x1,i =min{x1,··· ,xi} and xi,i =max{x1,··· ,xi} are called the i-th extremes (see for instance Galambos, 1984 =-=[17]-=-). The sequence of ranges for xt is then defined as R (x) i = xi,i − x1,i, fori =1, 2, 3, ··· ,n,withndenoting the sample size. Basically, a process defined by a sequence of ranges is an integrated ju... |

4 |
Influence functionals for time series." The Annals of Statistics
- Martin, Yohai
- 1986
(Show Context)
Citation Context ...quence. In particular, it is known that the presence of AO’s leads to a downward bias of the OLS parameter estimates in a stationary AR(1) process (Bustos and Yohai, 1986 [10]; Martin and Yohai, 1986 =-=[34]-=-), and that unit-root inferences are sensitive to the presence of extreme observations. For example, the DF test will have an actual size in excess of the nominal size and thus will reject the unit-ro... |

4 | Unit Root Tests Based - Lucas - 1995 |

3 |
Nonlinear Modelling and Analysis under Long-Range Dependence with an Application to Positive Time
- Aparicio
- 1995
(Show Context)
Citation Context ...r performances when confronted to deviations from the standard linear context (see for example, Sims, 1988 [52]; Perron, 1989 [38]; Perron, 1990 [39]; Schotman and Van Dijk, 1991 [50]; Aparicio, 1995 =-=[2]-=-). The appropriate handling of such departures as level shifts, trend breaks and nonlinearities calls for the development of robust unit root tests. Certainly, the rejection of the unit root hypothesi... |

3 |
Robust Estimates for ARIMA Models
- Bustos, Yohai
- 1986
(Show Context)
Citation Context ...ortant in magnitude or in frequence. In particular, it is known that the presence of AO’s leads to a downward bias of the OLS parameter estimates in a stationary AR(1) process (Bustos and Yohai, 1986 =-=[10]-=-; Martin and Yohai, 1986 [34]), and that unit-root inferences are sensitive to the presence of extreme observations. For example, the DF test will have an actual size in excess of the nominal size and... |

3 |
Distribution of the Estimators for AR Time Series with a Unit Root
- Dickey, Fuller
(Show Context)
Citation Context ...existence of unit roots in time series is investigated by means of unit root tests. The application of standard unit root tests, such as the Dickey-Fuller (DF hereafter) test (Dickey and Fuller, 1979 =-=[12]-=-), has been an important step in the construction of a useful parametric model for many economic time series. In a one-sided DF test, the null hypothesis of a unit root in a series xt, sayH0 : (1−B)(x... |

3 |
Nonstationary Time Series and Level Shifts with an Application to Purchasing Power Parity
- Perron, Volgelsang
- 1992
(Show Context)
Citation Context ...e literature on testing for unit roots in the presence of both known and unknown break points is large (see Maddala and Kim, 1998 [32] for a review). Perron (1989 [38]) and Perron and Vogelsang (1992 =-=[41]-=-) reported evidence that structural breaks can make an I(0) time series behave locally as I(1) and, as a result, these breaks are able to fool standard unit root tests (this is shown by the simulation... |

2 | 2000), “Synchronicity between Macroeconomic Time Series: an Exploratory Analysis - Aparicio, Escribano, et al. |

2 | Segmented Trends and Nonstationary Time - Rappoport, Reichlin - 1989 |

1 |
Detecting Common Components using Canonical Correlations
- Anderson, Vahid
- 1998
(Show Context)
Citation Context ...this, we first propose a characterization of this “long wave” in terms of what we call LowFrequency Features (LFF hereafter). Following Granger and Terasvirta (1993 [19]) and Anderson and Vahid (1998 =-=[1]-=-), a feature is essentially any dominating statistical property exhibited by a time series. Features may refer to either the mean behavior or to higher-order moments of the series, such as heteroskeda... |

1 |
Cointegration Testing using First Differences of Ranges: An Exploratory Analysis. Working Paper 98-08, Statistics and Econometric Series. Universidad Carlos III de
- Aparicio, Escribano
- 1998
(Show Context)
Citation Context ...inimum is reached. The behavior of the running ranges can be used to assess serial dependence in a single time series as well as the relationship between two time series (Aparicio and Escribano, 1998 =-=[3]-=-). One important finding is that the range sequence R (x) i for stationary time series is stochastically bounded 1 , whereas it is not for nullrecurrent time series such as integrated time series or t... |

1 |
Detrending Procedures and Cointegration Testing: ECM Tests under
- Arranz, Escribano
- 1998
(Show Context)
Citation Context ...ns in the sample, by removing them with the inclusion of a dummy variable in the model, or by treating them as missing observations. This two-stage approach was followed in Arranz and Escribano (1998 =-=[5]-=-), who proposed filtering the contaminated series prior to standard unit-root testing. Single-stage robust unit-root tests were first proposed by Lucas (1995 a,b [29,30]) and by Franses and Lucas (199... |

1 |
Díaz Cafferata A.J. and Figueras. Unit-roots in Spacial Unemployment in Argentina. Testing in the presence of Structural Breaks. Anales de la Asociación Argentina de Economía Política. XXXIIIa reunión Anual
- Arrufat, M
- 1988
(Show Context)
Citation Context ...eat bulk of econometric research has supported the evidence of the existence of unit roots in the unemployment rate series of most countries (see for instance Mitchell, 1993 [35]; Arrufat et al.,1999 =-=[6]-=-; Papell et al., 2000 [37]). Yet the low power of standard unit root tests on stationary alternatives close to the nonstationary border contradicts such reports. One among many controversial study cas... |

1 |
Unit roots in the presence of adrupt governmental interventions with an application to brazilian data
- Catri, García, et al.
- 1999
(Show Context)
Citation Context ...attempt to stop the process of soaring inflation. 80 60 40 20 0 74 76 78 80 82 84 86 88 90 92 Figure 14. Monthly Brazilian inflation rate from January 1974 to June 1993. Cati, Garcia and Perron (1999 =-=[11]-=-) have reported empirical results obtained with the application of standard unit root tests, such as the Augmented DickeyFuller (ADF) test (1979 [12]), and the Phillips-Perron test (1988 [44]). They 2... |

1 |
Outlier robust cointegration tests
- Franses, Lucas
(Show Context)
Citation Context ... who proposed filtering the contaminated series prior to standard unit-root testing. Single-stage robust unit-root tests were first proposed by Lucas (1995 a,b [29,30]) and by Franses and Lucas (1997 =-=[15]-=-) using M-estimators with high breakdown point and efficiency, instead of OLS estimators. However, these tests were really conceived for dealing with fat-tailed distributions of the model errors and, ... |

1 |
and Halmman J.J. Nonlinear transformations of integrated time series
- Granger
- 1991
(Show Context)
Citation Context ...earities, outliers and structural breaks. All these features, which cannot be properly captured with random-walk-like models, fool standard unit root tests (see for instance, Granger and Hallman 1991 =-=[18]-=-, and Ermini and Granger, 1993 [14]). Alternative procedures for testing unit roots were proposed by Lo (1991 [28]), Kwiatowski et al. (1992 [25]), Stock (1994 [55]) and the recent contributions of Ba... |

1 |
Cointegration and structural change. Cambridge university press
- Maddala
- 1998
(Show Context)
Citation Context ...see for instance Perron, 1990 [39], and Malliaropulos, 2000 [33] ). The literature on testing for unit roots in the presence of both known and unknown break points is large (see Maddala and Kim, 1998 =-=[32]-=- for a review). Perron (1989 [38]) and Perron and Vogelsang (1992 [41]) reported evidence that structural breaks can make an I(0) time series behave locally as I(1) and, as a result, these breaks are ... |

1 |
A note on nonstationary, structural breaks and the Fisher effect
- Maliiaropulos
- 2000
(Show Context)
Citation Context ...nditional mean which affects the standard inferential procedures and often makes constant coefficient models to perform poorly in practice (see for instance Perron, 1990 [39], and Malliaropulos, 2000 =-=[33]-=- ). The literature on testing for unit roots in the presence of both known and unknown break points is large (see Maddala and Kim, 1998 [32] for a review). Perron (1989 [38]) and Perron and Vogelsang ... |

1 |
Acourseintimeseries.Outliers,Influential observations and missing data
- Peña
- 2001
(Show Context)
Citation Context ...kdown effects on unemployment or sales data collection and processing, to name a few. Outliers can also appear as a result of misspecified estimated relationships or omitted variables (see Peña, 2001 =-=[42]-=-). Outliers are often classified into two groups: additive outliers (AO) and innovation outliers (IO), of which the former ones have the most insidious effects on classical inference. In both cases, s... |

1 |
Segmented trends and nonstationary time series
- Springer
- 1999
(Show Context)
Citation Context ...estion the validity of the standard test outcome. 30Appendix Proof of Theorem 1 The proof of Theorem 1 will be established by using the methodology of subsampling (see Politis, Romano and Wolf, 1999 =-=[45]-=-, for further details). Let Ψ (n) 0 = 1 n nX t=1 1(R (x) t > 0) (5) denotes the frequency of jumps, corresponding to “arrivals” of new maxima or minima, in a sample {x1,...,xn}, where1(.) represents t... |

1 |
Modelización y desestacionalización de la tasa de desempleo de Montevideo. Documento de trabajo, instituto de Esatadística de la Facultad de Ciencias Económicas y Administración
- Rodriguez
- 1998
(Show Context)
Citation Context ...employment rate series, from 1981 to 2001. Spremolla (1998 [54]) found a fractional root in this series, which suggested a stationary ARFIMA model, thus opposing a previous finding by Rodriguez (1998 =-=[48]-=-) of a single unit root, using the DF test. It is then natural to ask whether our RUR test is powerful enough to reject the null hypothesis of unit root in this case. We found J0 =1.0660, a value smal... |

1 |
Inference in a Nearly Integrated Autorregresive Model with Non-Normal Innovations
- Rothenberg, Stock
- 1997
(Show Context)
Citation Context ...ve followed a likelihood-based approach where inference is made about a particular fat-tailed distribution rather than on the Gaussian distribution (Hoek et al., 1995 [23]; Rothenberg and Stock, 1997 =-=[49]-=-). The use of nonparametric statistics is another avenue of research in robust unit root testing. Hasan and Kroenker (1997 [22]) applied rank-based methods.to this problem and reported improved power ... |

1 | Evidencia empírica para Uruguay. Trabajo monográfico NC 501600, Facultad de Ciencias Económicas y Administración - Spremolla - 1998 |

1 |
Histérisis o persistencia en el desempleo. Trabajo monográfico NC 501600, Facultad de Ciencias Económicas y Administración
- Spremolla
- 1998
(Show Context)
Citation Context ...bservations of this series are plotted in Figure 16 below. 2818 16 14 12 10 8 6 4 10 20 30 40 50 60 70 80 Figure 16. Quarterly Montevideo unemployment rate series, from 1981 to 2001. Spremolla (1998 =-=[54]-=-) found a fractional root in this series, which suggested a stationary ARFIMA model, thus opposing a previous finding by Rodriguez (1998 [48]) of a single unit root, using the DF test. It is then natu... |

1 |
Deciding Between I(1) and I(0) Time Series
- Stock
- 1994
(Show Context)
Citation Context ... instance, Granger and Hallman 1991 [18], and Ermini and Granger, 1993 [14]). Alternative procedures for testing unit roots were proposed by Lo (1991 [28]), Kwiatowski et al. (1992 [25]), Stock (1994 =-=[55]-=-) and the recent contributions of Bai and Perron (1998 [7]). Yet, all these tests rely on assumptions which are too restrictive in practice. In fact, they were reported to have poor power performances... |

1 | 2002), “Range Unit Root Tests”, Working - Aparicio, Escribano, et al. |

1 | Extremal Values: Theory and - Lindgren, Rootzén - 1987 |

1 | 2000), “A Note on Nonstationarity, Structural Breaks and the Fisher E¤ect - Maliiaropulos |

1 | 2003), “E¢cient Tests for Unit Roots with Prediction Errors - Sanchez |