TY  - JOUR
AU  - Christoffersen,Peter F.
AU  - Diebold,Francis X.
TI  - Cointegration and Long-Horizon Forecasting
JF  - National Bureau of Economic Research Technical Working Paper Series
VL  - No. 217
PY  - 1997
Y2  - October 1997
UR  - http://www.nber.org/papers/t0217
L1  - http://www.nber.org/papers/t0217.pdf
N1  - Author contact info:
Peter Christoffersen
Peter Christoffersen
Professor of Finance
Rotman School of Management
University of Toronto
105 St. George Street 447
Toronto, ON, M5S 3E6, Canada
Tel: 416-946-5511
E-Mail: peter.christoffersen@rotman.utoronto.ca

Francis X. Diebold
Department of Economics
University of Pennsylvania
3718 Locust Walk
Philadelphia, PA  19104-6297
Tel: 215/898-1507
Fax: 212/573-4217
E-Mail: fdiebold@sas.upenn.edu
AB  - We consider the forecasting of cointegrated variables, and we show that at long horizons" nothing is lost by ignoring cointegration when forecasts are evaluated using standard multivariate" forecast accuracy measures. In fact, simple univariate Box-Jenkins forecasts are just as accurate. " Our results highlight a potentially important deficiency of standard forecast accuracy" measures they fail to value the maintenance of cointegrating relationships among" variables and we suggest alternatives that explicitly do so.
ER  -