Impact of Exchange Rate on Macroeconomic Indicators

The exchange rate is a pivotal pricing instrument that simultaneously impacts various components of the economy. Depreciation of nominal exchange rate is export promoting, which might be a desired export led growth policy, and especially critical to closing-down the widening current account imbalance. But negative effects resulting from high dollarization and high share of imported intermediate inputs can outweigh positive effect. The aim of this research is to quantify impact of change in nominal exchange rate and test contractionary depreciation hypothesis on Georgian economy using structural and Bayesian vector autoregression. According to the acquired results, appreciation of nominal exchange rate is expected to decrease inflation, monetary policy rate, interest rate on domestic currency loans and economic growth in the medium run; however, impact on economic growth in the short run is statistically not significant.


Introduction
During the last 10 years, exchange rate volatility increased worldwide as a result of global financial crisis (2008), regional turbulences and the decline in commodity prices including oil price (2014). In additional, several central banks, including Czech National Bank, People's Bank of China, ECB, started to use exchange rate as a monetary policy instrument to amplify economic growth and achieve their inflation target. Current episodes of exchange rate volatility confirmed that exchange rate is one of the important tools, which can affect country's economic situation, financial stability and social welfare. Taking into account current events, in-depth analysis of exchange rate effect on macroeconomic environment is becoming significant, in order to inform policy makers and implement optimal economic policies.
The aim of this research is to analyze whether exchange rate depreciation is expansionary or contractionary in Georgia and quantify its impact on macroeconomic variables that are important to policy makers. Using structural and Bayesian autoregression models the paper evaluates impact of nominal exchange rate movement on real GDP growth, inflation, monetary policy rate and interest rate on domestic currency loans.
Georgia, as a high dollarized economy with high share of imported inputs, is an interesting case to analyze contractionary depreciation hypothesis and will provide valuable insight for policy makers. In 2017 up to 60% of total loan portfolio and total deposits were in US dollars. In addition, imports used as inputs of production in agricultural, manufacturing, transportation and communication and financial intermediation sectors are as high as 36% in Georgia, which are therefore exposed to currency fluctuations. So, the nominal exchange rate depreciation could directly lead to a negative supply shocks. Moreover, exchange rate risk is one of the main issues that financial sector in Georgia faces. Banking sector is quite vulnerable with respect to exchange rate depreciation, because it increases credit risk (most of the FX borrowers are not hedged) and rises financial stability concerns. Therefore, this research will give valuable information regarding contractionary depreciation hypothesis in Georgia.

Literature review
In standard small-open economy models (Svensson, 2000), a nominal exchange rate depreciation has an expansionary effect on output through export promotion and import reduction. This impact is mostly valid When liabilities are denominated in foreign currency, while income is in local currency, exchange rate depreciation increases debt burden, deteriorated balance sheet position increases risk premium and subsequently causes drop in investment. Therefore, exchange rate depreciation is contractionary if this negative impact outweighs positive impact on net export.
Literature about contractionary depreciation includes paper by Diaz-Alejandro (1965), who examined Argentina's devaluation in 1959. He claims that contractionary depreciation happened due to income distribution in favor of people with low-propensity to consume. Another important paper in this subject is by Cooper (1971), who evaluated 24 periods of devaluation and concluded that improvement in external position can potentially be outweighed by higher debt service and income redistribution. As Bahmani-Oskooee and Miteza (2003) summarized, the main reasons behind contractionary depreciation is income redistribution towards people with low marginal propensity to consume, a drop in investment, increased debt service, decrease in real wealth, possible increase in interest rate, increase in imported input prices, wage indexation based on price levels and increase in cost of capital. However, there is also wide literature that could not confirm contractionary depreciation hypothesis. For example, Kruegel (1978) examined 22 major devaluations of different countries and could not find evidence of contractionary depreciation. Similar result was achieved by Gylfason (1987), who examined 32 devaluation periods.
Overall, evidence on the contractionary devaluation hypothesis is quite mixed and results depend on country characteristics and adopted methodologies of the different studies.

Methodology and Data
Effect of exchange rate differs from country to country, according to the structure of the economy, so it is crucial to take into consideration country's characteristics for implementation of correct exchange rate policy. To analyze impact of exchange rate on macroeconomic variables research is conducted using structural vector autoregression (SVAR) and Bayesian vector autoregression (BVAR). SVAR (proposed by Sims, 1980) is an extension of traditional vector autoregression (VAR) analysis. It combines economic theory with time series data to quantify impact and response of variables on different shocks or economic policies. Main advantage of SVAR analysis is that it needs a relatively low number of identifying assumptions. Reduced form model can be estimated using restrictions that are based on economic theory.
To identify structural shocks from reduced SVAR model Cholesky decomposition is used, which is one of commonly used approaches for this analysis. After recovering structural shocks, they can be used to Additionally, paper uses Bayesian vector autoregression to check the robustness of results attained in the structural model. Bayesian estimation used in this study is based on well-known paper by Litterman (1986), which partially solves problems of many parameter estimation. According to this methodology, if the database has a short series or is inaccurate, it is better to give more weight to prior probability distribution. By taking into account both database and the prior distribution, model can estimate more parameters and have a better forecasting power than the structural vector autoregression model. If the predefined distribution is correctly chosen some researchers conclude that the Bayesian autoregression gives more accurate results than the structural autoregression (Canova, 1995 In Cholesky decomposition, order of variables is important to reflect which variables react to other variables' exogenous shock in the same period. Growth rate of gross domestic product is located on the first place, because it does not adjust to the change of other variables in the same quarter. Respectively, inflation is on the following position, which is also quite rigid in the short run. Monetary policy rate is on the third place, because it can adjust to changes in GDP growth and inflation in the same quarter. Then follows interest rate on domestic currency loans, and on the last position is the exchange rate, because it reacts to changes to all other variables in the same period.
Based on LR test and Akaike information criterion two lags of endogenous variables have been chosen.
The model is covariance stationary, because all roots lie within the unit circle. Therefore, all components of the model are stationary and impulse response functions will be correctly estimated.

Conclusion
There is vast literature that evaluates impact of nominal exchange rate movement on macroeconomic environment and tests contractionary devaluation hypothesis. However, the evidence on contractionary effect of exchange rate depreciation is quite mixed. So, the question, whether nominal depreciations are expansionary or contractionary remains still open not only in Georgia but also worldwide. This is a very interesting issue in Georgia given its high level of dollarization and high share of imported inputs. This study aims to contribute to expand empirical evidence on this significant topic and using SVAR and BVAR models quantify impact of exchange rate on macroeconomic variables in Georgia.
According to the obtained results, nominal exchange rate appreciation leads to decline in inflation rate, monetary policy rate and interest rate of domestic currency loans, while impact on economic growth is negative in the medium run (1-2 years), but statistically insignificant in the short run. Therefore, it is possible to reject the hypothesis of contractionary impact of exchange rate depreciation in Georgia in the medium term, but in the short run, there is not enough evidence to reject this.