Short-term Relationship Between the NIS/$ Exchange Rate and Interest Rates in Israel and the US

11.9.2006
 
Short-term Relationship Between the NIS/$ Exchange Rate and Interest Rates in Israel and the US
 

The economic relationships between the exchange rate and domestic interest rates and the interest-rate differential––which are particularly relevant to investors, exports, importers and the managers of macroeconomic policy––are not unequivocal. On the one hand, according to the uncovered interest-rate parity (UIP), the interest-rate differential between the NIS and the dollar should act to weaken the NIS/$ exchange rate in line with the interest-rate gap. On the other hand, in a small and open economy like Israel, the exchange rate has a relatively strong effect on inflation and consequently on the domestic interest rate, which is affected by the Bank of Israel's monetary policy, and therefore on the interest-rate differential too. In other words, the exchange rate affects the interest rate and not vice versa.
An analysis of the short-term relationships between the NIS/$ exchange rate and the domestic interest rates, the dollar interest rates and the interest-rate differentials for various terms (one, two, three, six months, a year and five years) shows that a rise in the exchange rate pushes up long-term (five-year) interest rates as well as the interest-rate differential. Conversely though, interest rates have no clear effect on the exchange rate.
These were the findings of research by Dr. Ben Z. Schreiber of the Foreign Exchange Activity Department at the Bank of Israel, published today by the Bank of Israel as part of its Foreign Exchange Discussion Paper series. The research used a special procedure that allows the analysis of the NIS/$ exchange rate's comprehensive effect on interest rates for various term structures and vice versa.
It was found that changes in the interest-rate differential, particularly for the long term, were better at forecasting changes in the exchange rate. This finding is reflected in the significant reduction in the interest-rate differential through most of the sample period compared to the relatively small depreciation of the NIS/$ exchange rate for most of the sample period.
It was also found that a sharp rise in interest rate affected variance of the interest rates more than a sharp cut in interest did. This finding is apparently the result of several hikes in the interest rate during the sample period, compared to the more modest, gradual reductions in the interest during the same period.