Annual Granger Lectures

2011

Peter Robinson, London School of Economics:
"Parametric inference on strong dependence", 6 October 2011, 1.00pm (A39 Sir Clive Granger Building)

2010

Hashem Pesaran, University of Cambridge, CIMF and USC:
"Aggregation in large dynamic panels", 25 May 2010, 11.45am (East Midlands Conference Centre)

2009

James Stock, Harvard University:
"Instrumental Variables Regression, GMM, and Weak Instruments in Time Series", 11 June 2009, 1.00pm (A48 Sir
Clive Granger Building) [slides]

2008

Peter Phillips, Yale University:
"Time Series Econometrics: Some Thoughts from the Toolroom", 4 June 2008, 1.00pm (A48 Sir Clive Granger
Building)

2007 (Inaugural Annual Granger Lecture)

Sir Clive Granger, University of California, San Diego:
"Time Series Analysis of Quantiles", 22 June 2007, 12.00pm (B63 Law & Social Sciences)
Including the opening of the Granger Centre for Time Series Econometrics

Conferences

Symposium on Structural Change, 14 December 2010

Sir Clive Granger Memorial Conference (joint with University of California San Diego), 24-25 May 2010

3rd Annual Conference: Recent Developments in Time Series Econometrics, 14-15 September 2009

3rd Meeting of the European Time Series Econometrics Research Network, 20-21 April 2009

2nd Annual Conference: Bootstrap and Numerical Methods in Time Series, 10-11 September 2008

1st Annual Conference: Conference in Honour of Paul Newbold, 21-22 September 2007
[Click to read the Tribute to Paul Newbold given at the conference]

Workshops

ESRC Econometric Study Group Workshop:
7 December 2007, 2.00-5.00pm (Room B326, 1st floor, Psychology Building, building no.29 on map)
2.00-2.30pmCoffee
2.30-3.30pmTim Vogelsang, Michigan State University
"Block Bootstrap HAC Robust Tests: The Sophistication of the Naive Bootstrap"
3.30-4.00pmCoffee
4.00-5.00pmHelle Bunzel, Iowa State University
"Testing for Breaks Using Alternating Observations"

Seminars

Stephen Wright, Brikbeck, University of London:
"Limits to Predictability: Univariate Bounds for Predictive Regressions with some Applications to Stock Returns"
24 February 2011, 3.00pm (A46 Trent Building)

Adam McCloskey, Boston University:
"Memory Parameter Estimation in the Presence of Level Shifts and Deterministic Trends"
29 June 2010, 4.00pm (A42 Sir Clive Granger Building)
Abstract: We propose estimators of the memory parameter of a time series that are robust to a wide variety of
random level shift processes, deterministic level shifts and deterministic time trends. The estimators are
simple trimmed versions of the popular log-periodogram regression estimator that employ certain sample
size-dependent, and in some cases, data-dependent trimmings which discard low-frequency components.
Regardless of whether the underlying long/short-memory process is contaminated by level shifts or
deterministic trends, our estimators are shown to be consistent and asymptotically normal with the same
limiting variance as the standard log-periodogram estimator. An extensive simulation study shows that our
estimators perform their intended purpose quite well, substantially decreasing both finite sample bias and
root mean-squared error in the presence of these contaminating components. Furthermore, we assess the
tradeoffs involved with their use when such components are not present but the underlying process
exhibits strong short-memory dynamics or is contaminated by noise. To balance the potential finite
sample biases involved in estimating the memory parameter, we recommend a particular version of our
estimators that performs well in a wide variety of circumstances. Finally, we apply our estimators to stock
market volatility and hydrological data to find that many of the time series typically thought to be long-
memory processes actually appear to be short-memory processes contaminated by level shifts or
deterministic trends.

Takashi Yamagata, University of York:
"Panels with Nonstationary Multifactor Error Structures"
29 June 2009, 4.00pm (A44 Sir Clive Granger Building)
Abstract: The presence of cross-sectionally correlated error terms invalidates much inferential theory of panel data
models. Recently, work by Pesaran (2006) has suggested a method which makes use of cross-sectional
averages to provide valid inference in the case of stationary panel regressions with a multifactor error
structure. This paper extends this work and examines the important case where the unobservable common
factors follow unit root processes. The extension to the I(1) processes is remarkable on two counts. Firstly,
it is of great interest to note that while intermediate results needed for deriving the asymptotic distribution of
the panel estimators differ between the I(1) and I(0) cases, the final results are surprisingly similar. This is
in direct contrast to the standard distributional results for I(1) processes that radically differ from those for
I(0) processes. Secondly, it is worth noting the significant extra technical demands required to prove the
new results. The theoretical findings are further supported for small samples via an extensive Monte Carlo
study. In particular, the results of the Monte Carlo study suggest that the cross-sectional average based
method is robust to a wide variety of data generation processes and has lower biases than the alternative
estimation methods considered in the paper.

Guillaume Chevillon, ESSEC Business School Paris:
"Inference in Models with Adaptive Learning, with an Application to the New Keynesian Phillips Curve"
23 June 2009, 4.00pm (A2 Law and Social Sciences Building)
Abstract: Replacing rational expectations by adaptive learning algorithms complicates the dynamics of economic
models. Identification of the structural parameters may improve relative to rational expectations, but it
deteriorates when learning converges. Learning also induces persistent dynamics, and this makes the
distribution of estimators and test statistics non-standard. We show that valid inference can be conducted
using the Anderson-Rubin statistic with appropriate choice of instruments. Application of this method to the
new Keynesian Phillips curve with US data provides evidence against standard versions of the model and
against least squares learning with a constant gain parameter.

Karim Abadir, Imperial College London:
"Beyond Co-Integration: Modelling Co-Movements in Macro and Finance"
26 February 2009, 5.00pm (A44 Sir Clive Granger Building)
Abstract: Macroeconomic and aggregate financial series share an unconventional type of nonlinear dynamics.
Existing techniques (like co-integration) model these dynamics incompletely, hence generating seemingly
paradoxical results. To avoid this, we provide a methodology to disentangle the long-run relation between
variables from their own dynamics, and illustrate with two applications. First, in the forward-premium puzzle,
adding a component quantifying the persistent nonlinear dynamics of exchange rates yields substantial
predictability and makes the forward-premium term insignificant. Second, S&P 500 grows in a pattern of
momentum followed by reversal, forming long cycles around a trend given by GDP, a stable non-breaking
relation since WWII.

Isaac Miller, University of Missouri:
"Cointegrating Regressions with Messy Regressors: Missingness, Mixed Frequency, and Nonclassical
Measurement Error"
17 February 2009, 4.00pm (A14 Pope Building)
Abstract: We consider a cointegrating regression in which the integrated regressors are messy in the sense that they
contain data that may be mismeasured, missing, observed at mixed frequencies, or have other irregularities
that cause the econometrician to observe them with possibly nonstationary noise. We motivate the notion
of messy data with a nontechnical example using linear interpolation. Even with such a straightforward DGP,
we show that the resulting noise is mildly nonstationary. We adopt a unified theoretical approach to avoid
strict distributional assumptions and to allow for such nonstationarity. Least squares estimation of the
cointegrating vector is consistent under general conditions, even though the estimator is neither
asymptotically normal nor unbiased. In order to allow valid statistical inference, we construct a canonical
cointegrating regression (CCR) using standard consistent nonparametric variance estimators, and we show
that least squares estimation of the CCR provides consistent and asymptotically normal estimation – even
with nonstationary disturbances. We briefly examine large- and small-sample properties of the estimator
when linear interpolation is the specific driver behind the messiness.

Yongcheol Shin, University of Leeds:
"Bootstrap-based Bias Corrected Within Estimation of Threshold Regression Models in Dynamic Panels"
12 June 2008, 2.00pm (A40 Sir Clive Granger Building)
Abstract: Recently, Shin (2007) proposes new estimation procedure to analyse asymmetric threshold effects in a
threshold autoregressive model in dynamic panels with unobserved individual effects when the number of
time periods is fixed by combining time series techniques on nonlinear threshold modelling with the
existing FD-GMM estimation techniques. This paper follows the bias corrected within estimator approach
introduced by Everaert and Pozzi (2006) in linear dynamic panels and advances a bias corrected
estimation procedure for the threshold dynamic panel data model based on an iterative bootstrap. Monte
Carlo simulation exercises confirm the validity of our proposed approach. In an application to the dynamic
threshold version of Tobin’s Q investment function using the company panel data set examined by many
researches and using di¤erent variables resepctively as a transition variable, we are able to find strong
evidence in favor of nonlinear dynamic threshold effects.

Brendan Beare, Nuffield College, University of Oxford:
"Unit Root Testing with Unstable Volatility"
17 April 2008, 4.00pm (A40 Sir Clive Granger Building)
Abstract: It is known that unit root test statistics may not have the usual asymptotic properties when the variance of
innovations is unstable. In particular, persistent changes in volatility can cause the size of unit root tests
to differ from the nominal level. In this paper we propose a class of modified unit root test statistics that
are robust to the presence of unstable volatility. The modification is achieved by purging heteroskedasticity
from the data using a kernel estimate of volatility prior to the application of standard tests. This approach
delivers test statistics that achieve standard asymptotics under the null hypothesis of a unit root. When the
data are homoskedastic, the local power of unit root tests is unchanged by our modification. We use Monte
Carlo simulations to compare the finite sample performance of our modified tests with that of existing
methods of correcting for unstable volatility.

Fabrizio Iacone, University of York:
"Testing for a Break in Trend when the Order of Integration is Unknown"
25 June 2007, 4.15pm (C38 Sir Clive Granger Building)
Abstract: We present a test for a break in a linear trend when the order of integration (d) is unknown, and potentially
fractional. The limit distribution of the test statistic is well known (the supremum of a Brownian bridge),
and does not depend on d. Small sample properties are investigated in a Monte Carlo exercise. We apply
this test to the US real GDP over the period 1929-2006, finding no evidence of a break.

Visitors

David Harris, Monash University, May 2011

Adam McCloskey, Boston University, June-July 2010

Giullaume Chevillon, ESSEC Business School Paris, June 2009

Giuseppe Cavaliere, University of Bologna, June 2009

Paulo Rodrigues, University of Algarve, June 2008

Jens Mehrhoff, Deutsche Bundesbank, October-December 2007

Sir Clive Granger, University of California, San Diego, June 2007

David Harris, University of Melbourne, December 2006

ANNUAL GRANGER
LECTURES

CONFERENCES

WORKSHOPS

SEMINARS

VISITORS