Consumer Inflation Expectations in Turkey

Abstract

The expectations obtained from surveys play an important role as leading indicators for the application of the monetary policies. The ability to measure inflation expectations is an integral part of central bank policy especially for central banks that are implementing inflation-targeting regime. A forward-looking perspective is essential to the success of inflation targeting. Therefore, a central bank which has the primary objective of price stability is interested in inflation expectations. Qualitative data on inflation expectations obtained from surveys can be quantified into numerical indicators of the expected rates of price change. This paper presents the results of different quantification methods such as Carlson-Parkin method, balance method, regression method put into action in order to estimate Turkish consumer inflation predictions based on monthly consumer surveys. Carlson-Parkin method quantifies qualitative survey data on expectations assuming aggregate expectations are normally distributed. In order to capture non-normality, stable distributions are also considered. The quantification techniques are compared with each other as well. The regression method is found to be the closest one to realizations. Therefore, expectations via this method is used for all the remaining analyses. Actual inflation and inflation expectations are found to have a cointegration relation. Unbiasedness assumption under REH is rejected within VECM. After rejecting a rational model of the formation of inflation expectations, hybrid model of expectations formation is considered. The "pure" backward and forward looking expectations hypotheses are rejected. As a final result, there exists the strong backward looking nature of expectations in the long run.