EXCHANGE RATE UNCERTAINTY AND INFLATION

Hasan ERSEL (Yapı Kredi Bank)
Muhammet MERCAN(Yapi Kredi Invest)

Economic decisions are taken in an uncertain environment, i.e., agents can not be certain about the consequences of their actions. Therefore they try to gather information either by observing the environment they are operating in or through communicating with other agents. Shannon in his celebrated article, Shannon (1948), proposed the following approach to measure uncertainty:

Let x_i denote a particular information and X={x₁,….x_n} be the set of all possible information. xi is a random variable with the probability of occurrence (i.e. the probability of this particular information to reach the decision-maker) pi. Then the uncertainty that the messages convey to a particular decision-maker can be measured by:

Where k is a constant and n is the number of different information.

H measure is calculated by using daily US$ buying rate of the Central Bank of the Republic of Turkey for 1990-2002 period. It is assumed that the representative decision-maker has a limited memory (20 working days) to calculate the H-index. It is also assumed the decision-makers are not infinitely sensitive to exchange rate movements. Thus the H values are clustered in symmetric thick indifferences curves around no change point.

This measure is used, first, to describe the pattern of uncertainty created by the exchange rate movements. The findings indicate that, on average uncertainty is higher under the floating exchange rate regime of post 2001 February period when compared with the rest of the period examined. Secondly it is also shown that an uncertainty variable based on H measure, when introduced in an inflation equation, has a positive and statistically significant coefficient. Based on this finding it is argued that the Central Bank in implementing inflation targeting strategy should be sensitive to exchange rate fluctuations.