2009•RePEc: Research Papers in EconomicsRequires access

Design and evaluation of core inflation measures for Turkey

Oğuz Atuk, M. Utku Özmen

Open publisher page 4 citations

Abstract

This study analyzes core inflation concept for Turkey from the perspective of design and evaluation. To our best knowledge, this is the first study to evaluate the performance of the core inflation measures for Turkey. We first define and calculate new core inflation indicators using the disaggregated price data for the time period of January 1994 to June 2007. The core inflation measures developed in this paper include weighted median, trimmed mean and volatility-based measures along with the permanent exclusion-based methods. Later, we evaluate the performance of the core inflation measures with regards to pre-defined criteria including unbiasedness, reduced volatility, trend tracking ability and predictive ability. Considering the significant shift in the inflation process in Turkey, we split the sample into two as pre- and post-2003 when evaluating the performance of the core inflation measures. Overall, no single core inflation indicator stands out alone as �the core� and therefore different core measures should be followed simultaneously to get a better understanding of the real trend of inflation and to identify various shocks hitting the inflation. However, limited influence estimators and volatility-based core measures perform relatively better compared to permanent exclusion-based methods. The study also provides an extensive analysis of the distribution of price changes at various aggregation levels.

About this research paper

What this paper is about

This study analyzes core inflation concept for Turkey from the perspective of design and evaluation. To our best knowledge, this is the first study to evaluate the performance of the core inflation measures for Turkey. We first define and calculate new core inflation indicators using the disaggregated price data for the time period of January 1994 to June 2007. The core inflation measures developed in this paper include weighted median, trimmed mean and volatility-based measures along with the permanent exclusion-based methods. Later, we evaluate the performance of the core inflation measures with regards to pre-defined criteria including unbiasedness, reduced volatility, trend tracking ability and predictive ability. Considering the significant shift in the inflation process in Turkey, we split the sample into two as pre- and post-2003 when evaluating the performance of the core inflation measures. Overall, no single core inflation indicator stands out alone as �the core� and therefore different core measures should be followed simultaneously to get a better understanding of the real trend of inflation and to identify various shocks hitting the inflation. However, limited influence estimators and volatility-based core measures perform relatively better compared to permanent exclusion-based methods. The study also provides an extensive analysis of the distribution of price changes at various aggregation levels.

Why it matters

OpenAlex reports 4 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

This study analyzes core inflation concept for Turkey from the perspective of design and evaluation. To our best knowledge, this is the first study to evaluate the performance of the core inflation measures for Turkey. We first define and calculate new core inflation indicators using the disaggregated price data for the time period of January 1994 to June 2007. The core inflation measures developed in this paper include weighted median, trimmed mean and volatility-based measures along with the permanent exclusion-based methods. Later, we evaluate the performance of the core inflation measures with regards to pre-defined criteria including unbiasedness, reduced volatility, trend tracking ability and predictive ability. Considering the significant shift in the inflation process in Turkey, we split the sample into two as pre- and post-2003 when evaluating the performance of the core inflation measures. Overall, no single core inflation indicator stands out alone as �the core� and therefore different core measures should be followed simultaneously to get a better understanding of the real trend of inflation and to identify various shocks hitting the inflation. However, limited influence estimators and volatility-based core measures perform relatively better compared to permanent exclusion-based methods. The study also provides an extensive analysis of the distribution of price changes at various aggregation levels.

Key concepts: Core inflation, Econometrics, Inflation (cosmology), Economics, Volatility (finance), Core (optical fiber), Estimator, Truncated mean

Related papers

Back to paper searchBrowse research topicsOriginal source
Design and evaluation of core inflation measures for Turkey — Research Paper | ScholarLens