Cointegration in a historical perspective
Introduction
The Journal of Econometrics and Econometrica are the two journals that contain the most cited papers in the econometrics discipline. These citation classics have in common that they mainly concern econometric techniques for the analysis of time series variables. By far the best cited time series econometrics paper is Engle and Granger (1987). The Nobel-worthy concept of cointegration had already been introduced in Granger (1981), but the Econometrica paper in 1987 meant an explosive take-off for this novel idea. Many academics and practitioners resorted to the use of the cointegration technique, and theoretical developments in the area covered quite some space in econometrics conferences all over the world. A glance at the programs of the Econometric Society meetings in the 80s and 90s of the previous century, which can be found in back issues of Econometrica, reveals that a large number of sessions were dedicated to just “Cointegration”. Even today there still are workshops and sessions during conferences where new developments in cointegration are being discussed.
It is of course intriguing to ask the question of why the concept of cointegration became that important and even deserved Nobel Prize recognition. A substantial part of its success undoubtedly is attributable to the elegance of the concept and the fact that it combines various streams of literature into one single framework. Another part of the success could be due to favourable circumstances at the time cointegration was discovered and put forward. In the present paper we do indeed argue that cointegration could have become such an important research and application area partly also because it appeared at just the right time. Our argument draws upon the discussion in Gladwell (2008), where the success factors of Microsoft and The Beatles are studied. Here we will argue that part of the success of cointegration can be found in the combination of four external factors that were prominent when the concept first appeared. First, in the early 1980s large macroeconomic models were losing to simple time series models in terms of forecasting, although people felt that such ARIMA models were lacking economic substance. Second, due to the “discovery” of stochastic trends in macroeconomic time series a few years earlier, there was a sense of urgency for developing new statistical tools to analyze such data in a correct way. In a sense, when cointegration entered the stage, theoretical and applied econometricians were ready for it. Third, large enough samples of macroeconomic data were becoming available, so it started to become a meaningful exercise to explore the presence of long-run equilibrium relationships. Fourth, but certainly not least important, the computing facilities and software needed to do the calculations involved in cointegration analysis became available to a wider audience at that time, so the methods could be widely applied.
The outline of our paper is as follows. In the next section we give a few facts and figures on the publication itself. In Section 3 we will compose our argument. In Section 4, we make an attempt to forecast when the next breakthrough, like cointegration, will happen, and what it will look like.
Section snippets
Some facts and figures
This section presents a few facts and figures to indicate how important and influential the paper of Engle and Granger (1987) has been and still is.
Table 1 presents the ten most cited papers (as documented in December 2008) that have appeared in Econometrica. The paper of White (1980) is a clear winner, but the second most cited paper is Engle and Granger (1987).
Why did it fly?
The text that accompanies the announcement of the Nobel Prize in 20031 clearly outlines what the concept of cointegration is, how one can estimate and interpret the parameters in the error correction model, how forecasts can be improved when cointegration is imposed, how important it is for empirical data which often have unit roots, and how cointegration unifies the literatures on economic theory (equilibrium across
What will happen next?
Now we have seen that cointegration could fly not only due to its particular relevance to the econometrics discipline but also due to five favourable circumstances, we are tempted to put it into an even more historical perspective and to make a prediction of what might happen next.
Acknowledgement
We thank Malcolm Gladwell for encouragement.
References (31)
- et al.
Modeling the diffusion of scientific publications
Journal of Econometrics
(2007) Five alternative methods of estimating long-run equilibrium relationships
Journal of Econometrics
(1994)Some properties of time series data and their use in econometric model specification
Journal of Econometrics
(1981)- et al.
Seasonal integration and cointegration
Journal of Econometrics
(1990) - et al.
Testing structural hypotheses in a multivariate cointegration analysis of the PPP and UIP for UK
Journal of Econometrics
(1992) - et al.
Testing the null hypothesis of stationarity against the alternative of a unit root: how sure are we that economic time series have a unit root?
Journal of Econometrics
(1992) - et al.
Unit root tests in panel data: asymptotic and finite-sample properties
Journal of Econometrics
(2002) - et al.
Trends and random walks in macroeconomic time series
Journal of Monetary Economics
(1982) - et al.
Estimating long-run relationships from dynamic heterogeneous panels
Journal of Econometrics
(1995) Understanding spurious regressions in econometrics
Journal of Econometrics
(1986)
Time series analysis and simultaneous equations models
Journal of Econometrics
A new product growth model for consumer durables
Management Science
On the econometrics of the Bass diffusion model
Journal of Business & Economic Statistics
Time Series Analysis: Forecasting and Control
Econometric modelling of the aggregate time-series relationship between consumers’ expenditure and income in the United Kingdom
Economic Journal
Cited by (5)
30 years of cointegration and dynamic factor models forecasting and its future with big data: Editorial
2021, International Journal of ForecastingCitation Excerpt :Cointegration and common trends are important concepts that changed the way we model and forecast the future evolution of multivariate non-stationary variables, avoiding misleading spurious relations.2 The fact that linear combinations of integrated time series could be stationary was advanced by Box and Tiao (1977) but they did not develop the idea further; see Boswijk, Franses, and van Dijk (2010) for a historical perspective on cointegration. There is an extremely large list of works implementing cointegration to forecast macroeconomic and financial variables, and there is no point in attempting to list them.
Time Series Analysis Framework for Forecasting the Construction Labor Costs
2021, KSCE Journal of Civil EngineeringForecasting Construction Material Prices Using Vector Error Correction Model
2018, Journal of Construction Engineering and ManagementEnjoyable Econometrics
2018, Enjoyable Econometrics