| Abstract: | Summary This paper examines the question how much independent variation exists among observed market interest rates in the Netherlands. Therefore a particular form of factor analysis, viz. principal component analysis, has been applied to a time series of monthly data for the period 1962–1970. Ten money market rates as well as ten capital market rates are analysed separately. By doing this we have found that on the money market about 96% of total variation is explained by the first component; an additional percentage of 2 is explained by the second principal component. For the capital market these percentages are 90 and 8. Similar results are obtained when subperiods are studied or when the money and capital market are pooled together.The main insights to be gained by this analysis are the following. First our results suggest that the first component identifies the true interest rate. The second principal component, which is highly correlated with the rate of return on shares, reflects the risk aspect of the rate of interest. The third component seemsto be related to the rate of inflation. The second results is that our analysis shows that the use of many different interest rates in macro models has only a limited economic meaning. The this conclusion is that the usual textbook distinction between money and capital markets does not show up in the principal components obtained.Ik ben veel dank verschuldigd aan de heren P. M. Cambeen en R. L. Coenen, medewerkers op de Studiedienst van de Nederlandsche Bank N.V., die mij behulpzaam zijn geweest bij het uitvoeren van de berekeningen voor dit onderzoek. Vanzelfsprekend komen eventuele tekortkomingen geheel voor miju rekening. |
