Abstract
Several non-monetary (mainly supply) factors affect prices in the short-run. It is widely acknowledged that in countries (especially countries in transition), where the price level is highly volatile and seasonal, it is not expedient for central banks to use official inflation index while formulating monetary policy. For this reason, it is crucial for central banks to work out, study and follow the behavior of core inflation that enables to reflect long-run price movements. This paper presents the application of various methods of calculating core inflation to Armenian data (for 1996:1-2002:12). Each measure is calculated at monthly frequencies and evaluated by different criteria. The analysis shows that core inflation indices, calculated by trimming the distribution of prices at 10 or 15%, are the best and most effective indicators for monetary policy-makers in Armenia, since they capture inflation trends and are closely tied to monetary aggregates. However, the median seems to be the best predictor for forecasting inflation of all core inflation measures discussed in this paper