book review: new introduction to multiple time series analysis

5
This article was downloaded by: [University of Illinois at Urbana-Champaign] On: 03 October 2014, At: 23:24 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Econometric Reviews Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/lecr20 Book Review: New Introduction to Multiple Time Series Analysis Òscar Jordà a a University of California – Davis , Davis , California , USA Published online: 23 Nov 2009. To cite this article: Òscar Jordà (2009) Book Review: New Introduction to Multiple Time Series Analysis, Econometric Reviews, 29:2, 243-246, DOI: 10.1080/07474930903472868 To link to this article: http://dx.doi.org/10.1080/07474930903472868 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Upload: oscar

Post on 16-Feb-2017

214 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Book Review: New Introduction to Multiple Time Series Analysis

This article was downloaded by: [University of Illinois at Urbana-Champaign]On: 03 October 2014, At: 23:24Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Econometric ReviewsPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/lecr20

Book Review: New Introduction to Multiple Time SeriesAnalysisÒscar Jordà aa University of California – Davis , Davis , California , USAPublished online: 23 Nov 2009.

To cite this article: Òscar Jordà (2009) Book Review: New Introduction to Multiple Time Series Analysis, Econometric Reviews,29:2, 243-246, DOI: 10.1080/07474930903472868

To link to this article: http://dx.doi.org/10.1080/07474930903472868

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Page 2: Book Review: New Introduction to Multiple Time Series Analysis

Econometric Reviews, 29(2):243–246, 2010Copyright © Taylor & Francis Group, LLCISSN: 0747-4938 print/1532-4168 onlineDOI: 10.1080/07474930903472868

BOOK REVIEW: NEW INTRODUCTION TO MULTIPLETIME SERIES ANALYSIS

Òscar Jordà

University of California – Davis, Davis, California, USA

Lütkepohl, Helmut (2006). New Introduction to Multiple Time Series Analysis.Berlin and Heidelberg: Springer.

Helmut Lütkepohl’s New Introduction of Multiple Time Series Analysis isdestined to become a classic that will sit, dog-eared and profuselyhighlighted, on the shelves of many practitioners and students of timeseries analysis. While very few advanced graduate-level textbooks go beyondthe first edition, New Introduction to Multiple Time Series Analysis is thede facto third edition of Introduction to Multiple Time Series Analysis,first published in 1991. (The New adjective indicates a more substantialrevision with additional material This influential book has been favorablyreviewed at least three times previously in leading econometric journals(Baillie, 1993 in the Journal of Applied Econometrics; Jolliffee, 1993 inTechnometrics; and Kilian, 2006 in Econometric Theory).1

Kilian’s (2006) review of the New book is the most recent andextensive, with commentaries on target audience, contents, and generaloverview. In order to avoid repeating what is already a very thorough andcompetent review, let me concentrate on the “hidden nuggets” I havefound tremendously useful in my own research and teaching, and whichmake the New book a worthy and significant extension of the earlierIntroduction book.

First off, this book is primarily about specification, estimation andinference of discrete-time, linear, vector time-series models and a fewextensions. As expected when nearly 800 pages are dedicated to such a

1The citation count of all three editions reported by scholar.google.com is above 1,000. JamesHamilton’s Time Series Analysis, arguably one of the most successful graduate time series textbooks,has a citation count of almost 1,400 adjusted for topic overlap (over 4,000 citations overall).

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 2

3:24

03

Oct

ober

201

4

Page 3: Book Review: New Introduction to Multiple Time Series Analysis

244 Book Review

narrow topic, virtually all the results are derived in painstaking detail:Where else would one find together the definitions of the elimination,duplication and commutation matrices; a presentation of multivariateunit root asymptotics; and a discussion on simulation techniques formultivariate time series?

The first of my favorite “nuggets” is the discussion on how tocompute the standard errors of impulse responses in structural vectorautoregressions (SVARs). The New book contains the derivation of thematrix algebra formulae for structural vector autoregressions identifiedwith short-run zero restrictions (the ubiquitous Cholesky decomposition ofthe residual covariance matrix already available in the Introduction book);long-run zero restrictions (a la Blanchard and Quah, 1989); but moreimportantly, the derivations based on the A-, B -, and AB-models describedin Amisano and Giannini (1997), which form the basis of the routinesavailable in the popular software econometric packages EViews, RATS, andStata, for example.

The level of detail in the derivations is such that it would not be aparticular burden to construct one’s own version, for example, of an SVARroutine with a traditional matrix language package such as Mata, Matlabor GAUSS (hence the usefulness of the very extensive matrix algebraappendix). In this regard, it is worth pointing out that the datasets inthe book are freely available from the website www.jmulti.de, which alsocontains the java application JMulTi. JMulTi provides a simple, menu-driven interface that runs the GAUSS engine in the background for thecalculations. While not a replacement for a general purpose econometricpackage, users will find a broad collection of traditional univariate andmultivariate time series analysis tools along with other less conventionaloptions (e.g., smooth transition regression models and nonparametricestimators), often with a bootstrap option available.

In the short span of one year, a recent paper by Chari et al. (2005) hasalready generated 37 citations in scholar.google.com – mostly a literatureof defenders and attackers of the main premise of the paper: a critiqueon the inability of short-order vector autoregressions to capture essentialproperties of real business cycle (RBC) models. Chari et al. (2005) arguethat impulse responses estimated by a short-order VAR grossly misrepresentthe dynamics of an RBC type economy because the standard RBC modelhas an infinite order autoregressive reduced-form representation.

The New book throws light on this debate. It contains an entire sectiondedicated to infinite order vector autoregressive processes and theirproperties. This section discusses vector autoregressive moving average(VARMA) models and the usually ignored cointegrated VARMA models.The New book’s discussion of the results in Lewis and Reinsel (1985) aboutthe asymptotic properties of finite-order VAR estimates of infinite-orderprocesses is particularly illuminating for the Chari et al. (2005) controversy.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 2

3:24

03

Oct

ober

201

4

Page 4: Book Review: New Introduction to Multiple Time Series Analysis

Book Review 245

It shows that if the lag length k of the VAR grows with the sample size atrate k3/T → 0, the VAR(k) provides consistent and asymptotically normalestimates of the first k coefficients of the infinite-order VAR process.Because the impulse response coefficients at horizon k are a polynomialfunction of the first k coefficients of the autoregressive representation,a VAR(k) provides consistent estimates of the first k coefficients of theimpulse response function. This result suggests that the Chari et al. (2005)critique refers to, at best, a very specific and pathological small-sampleresult and not a general intrinsic deficiency of short-order VARs.

Bayesian econometrics but more importantly, Markov ChainMonte Carlo (MCMC) estimation techniques, are quickly becomingcommonplace in econometrics. MCMC methods allow simpler numericalmaximization of complicated likelihood problems (for example dataaugmentation methods for latent variable problems) and computationof associated inferential statistics. This trend is clearly visible in therecent work on the estimation of dynamic, stochastic, general equilibrium(DSGE) models (see Lubik and Schorfheide, 2004), and time-varyingparameter VARs (see Cogley and Sargent, 2001, 2005), for two recentmultivariate time-series examples. Hence, the New book’s discussion ofBayesian techniques interspersed in chapters 5 (estimation of VARs)and 7 (estimation of I(1) systems) is very welcome and timely, andnicely complements another “nugget” rarely discussed in the literature,estimation of VARs with constraints.

Discussions of GARCH models in general time series textbooks are rare(except Hamilton, 1994) and are usually relegated to more specializedbooks (for example Tsay, 2005). Multivariate GARCH models are theexclusive realm of journal articles so it is with great delight that the finalof the “nuggets” that I discuss here is the detailed presentation of thesemodels in the New book. Rather than focusing on the plethora of GARCHspecifications that have littered the financial econometrics landscape, theNew book narrowly focuses on the central elements of estimation andinference of what is ostensibly, a rather difficult model to handle inpractice.

No review is complete without raising some criticism if only becauseit highlights what the New book does really well. In this respect the lastpart of the book (part V, Time Series Topics) is less detailed than therest. This part includes the chapter on GARCH models just discussedbut also a chapter on periodic VAR processes and intervention modelsand a chapter on state-space models. The latter I found less completeand less linked to the rest of the book than the topic would seem towarrant. Of course, it is also easy to criticize a book for the topics itdoes not include. So it is perhaps more useful to cast these as a wish-list for the fourth edition. Two topics that come to mind, especially sincethey are not completely foreign to Helmut Lütkepohl’s own research

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 2

3:24

03

Oct

ober

201

4

Page 5: Book Review: New Introduction to Multiple Time Series Analysis

246 Book Review

are: structural change – testing and modeling (see for example Krolzig’s,1997 monograph on Markov-switching vector-autoregressions); and factormodels, which have a natural connection with the forecasting and state-space chapters. However, at 800 pages it is hard to argue for morecoverage. At a sales price of $70.00 for the paperback version, it is hard toargue that you can get more bang for your buck.

REFERENCES

Amisano, G., Giannini, C. (1997). Topics in Structural VAR Econometrics. 2nd ed. Berlin and NewYork: Springer-Verlag.

Baillie, R. T. (1993). Book review: Introduction to multiple time series analysis by HelmutLütkepohl. Journal of Applied Econometrics 8:325–331.

Blanchard, O. J., Quah, D. (1989). The dynamic effects of aggregate demand and supplydisturbances. American Economic Review 79(4):655–673.

Chari, V. V., Kehoe, P. J., McGrattan, E. R. (2005). A critique of structural VARs using businesscycle theory. The Federal Reserve Bank of Minneapolis Staff Report no. 364.

Cogley, T., Sargent, T. J. (2001). Evolving Post World War II U.S. inflation dynamics. NBERMacroeconomics Annual 16:331–373.

Cogley, T., Sargent, T. J. (2005). The conquest of U.S. inflation: Learning and robustness to modeluncertainty. Review of Economic Dynamics 8:528–563.

Hamilton, J. D. (1994). Time Series Analysis. Princeton, New Jersey: Princeton University Press.Jolliffee, I. T. (1993). Book review: Introduction to multiple time series analysis by Helmut

Lütkepohl. Technometrics 35(1):88–89.Kilian, L. (2006). Book review: New introduction to multiple time series analysis by Helmut

Lütkepohl. Econometric Theory 22(5):961–967.Krolzig, H.-M. (1997). Markov-Switching Autoregressions. Modelling, Statistical Inference and Application

to Business Cycle Analysis. Lecture Notes in Economics and Mathematical Systems. Vol. 454.Berlin and New York: Springer-Verlag.

Lewis, R. A., Reinsel, G. C. (1985). Prediction of multivariate time series by autoregressive modelfitting. Journal of Multivariate Analysis 16(33):393–411.

Lubik, T. A., Schorfheide, F. (2004). Testing for indeterminacy: An application to U.S. monetarypolicy. American Economic Review 94(1):190–217.

Lütkepohl, H. (1991). Introduction to Multiple Time Series Analysis. Princeton, New Jersey: PrincetonUniversity Press.

Tsay, R. S. (2005). Analysis of Financial Time Series. Hoboken, New Jersey: John Wiley and Sons.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 2

3:24

03

Oct

ober

201

4