Page 26 - Azerbaijan State University of Economics

Vugar Rahimov, Nigar Jafarova: The Exchange Rate Pass-Through to Aggregate
Consumer Price Index and Its Components In Azerbaijan

Taylor (2000) also mentions the importance of monetary regime in the degree of the
ERPT. He finds out that the countries with inflation targeting regime experience
lower pass-through due to credibility of the CBs and low inflation rate environment.
Recently transition to the floating regime in sample countries makes Markov regime
switching model more appropriate for estimation. Considering the switch between
periods, such models enable to capture the relevant dynamic patterns. However, the
floating regime period covers only the recent short time span which makes it hard to
carry out such empirical assessment method.

4. DATA AND METHODOLOGY
In this paper, we try to assess the degree of the exchange rate pass-through to
domestic CPI and its main components. The full sample contains quarterly data for
2003Q1-2021Q2 on Azerbaijan. As a starting point, we employ a four variable VAR
model similar to those developed by Mccarthy (2000), Hahn (2003) and Ca’ Zorzi
(2007). Those variables include oil revenue, trading partners’ CPI (tp_cpi), nominal
effective exchange rate (neer) and domestic CPI (cpi). It would be of great
importance to include import price index as well, however data is not available on
that indicator. Oil revenue is calculated as the product of real price of oil and oil
production for a given country.

Oil prices are deflated using US CPI. The source for this indicator is the US Energy
Information Administration Database (EIA). Nominal effective exchange rate is a
weighted average of the bilateral nominal exchange rates vis-à-vis the trading
partners’ currency and is obtained from Bruegel database. Trading Partners’ CPI
(TP CPI) is derived from REER formula by dividing the product of NEER and
domestic CPI to REER. Domestic CPI and its components are the cumulative
consumer price index for which the base period is the end of 2002. The source for
aggregate CPI and its components (food, non-food and service CPI) are the State
Statistical Committee of Azerbaijan. All variables are seasonally adjusted through
Census-X-12 procedure and transformed into logarithmic form. A detailed
description of all series is presented in Appendix A1. According to unit root test
results [See Appendix A2, Table 2 for detailed information on unit roots], the
variables are non-stationary, so we run VAR model in first differences.

The existing literature employs different approaches to estimate the ERPT (Calpa
and Goldberg, 2005; Chabot and Khan, 2015; Choudri, 2005; Ca’ Zorzi et al., 2007;
Stulz, 2007). The choice of methodology for our paper is constrained with some
issues related to country-specific characteristics and time span.

26