A study was carried out using Chilean government statistics on interest rates and data from the performance of the Chilean stock market between 1989 and 1997 to evaluate their predictive value in relation to the Chilean gross national product. The best predictive functions for the Chilean national product, in descending order, included: (a) intermediate term interest rates, (b) differential "spread" between short and long term interest rates, and (c) behavior of the stock market. In contrast with the United States, the "spread" contains less information concerning the projected national product than do short, medium and long-term interest rates. To obtain the best projections for stock market performance, the general stock market index (IGPA) needs to be calculated on an 18-month basis, rather than as typically done on a monthly or quarterly basis. In any case, the predictive power of the IGPA is low compared with that of intermediate term interest rates. We found, in contrast to previous studies reported for Chile, that predictions of gross national product based on the stock market were reliable only over the short term (less than one year), while predictions made from the observed behavior of medium-term interest rates spanned a two-year horizon. In particular, the intermediate term interest rate had a greater potential for prediction of economic activity over a six-month period than stock market-based predictions, which were only good over four months. Finally, the models proposed were parsimonious and efficient, and continued to be valid when applied to data outside the present study (post 1997).
Keywords:
Financial markets, Predictive value, Intermediate term interest rates, Differential, Behavior of the stock market
Zúñiga J., S., & Soria B., K. (2000). Financial markets and product prediction: Chilean evidence 1989-1997. Estudios De Administración, 7(1), 67–93. https://doi.org/10.5354/0719-0816.2000.56758
