by Ruxanda, Gheorghe and
Published in Romanian Journal of Economic Forecasting, 2008, volume 9 issue 3, 51-62.
| Requires a PDF viewer such as Xpdf
Economic time series are, in their vast majority, integrated series so, their modelling procedure stumbles upon the problem of spurious regression. When existent, cointegration is the simplest way of eliminating the illogical correlation established between time series due to the presence of trends. The analysis of macroeconomic time series through cointegration is a common fact. Modelling the Romanian M2 money demand through cointegration and vector error correction led to somewhat significant results being a starting point for future, more complex research.
spurious regression, cointegration, money demand, error correction mechanism