← volver

2000 Documento de Trabajo #89

Estructura de tasas de interés en Chile: ¿Qué tan buen predictor de crecimiento e inflación? (Interest rates structure in Chile: how well does it predict growth and inflation?)

Based upon Granger causality and Pesaran-Shin’s generalized impulse-response functions, this paper studies the link between the term structure and economic growth, and the link between the term structure and actual and expected percent changes of the Consumer Price Index (CPI) in Chile. Our measure of economic growth is the percent variation in the Monthly Indicator of Economic Activity of the Central Bank of Chile (IMACEC). The slope and the level of the term structure are captured, respectively, by the spread of long and short-term interest rates -in real and nominal terms-, and the short real rate. Our main findings can be summarized as follows. First, when considering a short-term horizon, the relationship between the percent change in the IMACEC, the spread of nominal and real rates, and the level of the short real rate is, in general, weak in statistical terms. For a longer-term horizon, the strongest relationship is that observed between the level of the short real rate and the 12-month percent variation in the IMACEC. Second, the link between inflation and the nominal spread of interest rates is weak, especially for a long-term horizon. Moreover, changes in the CPI seem to have more predictive power to explain changes in the spread of nominal interest rates than vice versa. We believe that one explanation may be the price inertia existing in the Chilean economy due to indexation. Indeed, it is possible that changes in expected inflation transmit to the spread of nominal interest rates faster than they do to actual inflation.

Viviana Fernández

Keywords: generalized impulse-response., Granger causality