Impact of Exchange Rate Devaluation in Ethiopia and Forecasting Foreign Exchange Rate Using ARIMA Model

The exchange rate is an important indicator of economic activity that policymakers use to formulate policies. This paper is based on fitting historical nominal exchange rate time series expressed in terms of past values of itself plus current and lagged values of error term using the Autoregressive Integrated Moving Average (ARIMA) model over the period 1971-2020, resulting in the model (2,1,1) to forecast the next 15 years. The projection's performance is assessed using data from the preceding fifty years sourced from World Bank, and the forecast indicates that domestic currency depreciation will become more prevalent in order to stimulate exports. Moreover, the ARDL model is used to examine the impact of exchange rate, import, export, and inflation rate on RGDP from 1977 to 2018. The unit root test (Augmented Dickey- Fuller test) was used to verify the integration order of the variables. The five variables are used in the cointegration analysis. The bounds test of cointegration was used in this work. The results showed that the F-statistics value is greater than the upper boundaries at all levels of significance, indicating that individual variables have a long-term relationship with RGDP. This led to the estimation of the Error Correction Model (ECM), in which 25.1 percent of the disequilibrium is adjusted annually to bring the system back into equilibrium. Furthermore, nominal exchange rate and import has positive contribution to RGDP while inflation has adverse effect on RGDP in the long-run. Besides, inflation has a negative impact to RGDP in the short run due to its lag period effects.


INTRODUCTION 1.Background
Every nation has three economic goals to attain both in the short and in the long run, these are achieving economic growth, creating more employment and having no or minimum inflation simultaneously. In order to achieve these goals and make their countries better off, countries use monetary and fiscal policies as a strategy and let their nation's aggregate demand curve to shift either to the right-or left-side. The most common economic policy instruments used by governments to monitor and change their economies are fiscal and monitory measures. Fiscal policy is all about letting the government to collect taxes and spend it on public sectors like infrastructures, education and so on and which mainly focuses only on the domestic economy whereas, monetary policy deals with both domestic and international economy. Meaning, the government can use monetary policy and the exchange rate policy of devaluation in order to affect the domestic and international markets respectively (Fratzscher et al, Ethiopia has implemented several policy and structural changes on both micro and macro levels of the economy, such as the Structural Adjustment Program (SAP) which began after the fall of the Derg regime. As part of these general reform effort, on October 1, 1992, the Ethiopian Birr was devalued from its nominal level of 2.07 Birr per US dollar to 5.00 Birr per US dollar (Befikadu and Kibre, 1994).
Although devaluation is meant to improve the trade balance by promoting exports, there are theoretical distinctions in how exchange rates affect economic growth. Economic adjustment programs for less developed countries often focus on domestic absorption through restrictive monetary and fiscal policies, as well as expenditure switching (i.e., increased tradables production) through currency devaluation. If there is unutilized capacity, a nominal devaluation leads in higher output, according to classical theory. As a result, devaluation raises the relative prices of tradables and non-tradables, encouraging expenditure switching (Nguyen, 2014;Moller and Walker, 2015;Agus and Long, 2018).
However, the traditional view that devaluation-induced expenditure switching provides an important stimulus to economic growth has come under serious attack. Various theoretical arguments are advanced that point to contractionary effects of devaluations. Earlier studies mainly refer to adverse demand effects. The re-distribution of income in favour of profits and at the expense of wages, supposed to follow a devaluation, may depress economic activity because of the higher marginal propensity to consume of wage earners (Diaz Alejandro, 1963;Krugman, Taylor, 1978) and this theoretical foundation is substantiated by empirical findings of various scholars where exchange rate has adverse effect on economic growth since devaluation is likely to result in a higher domestic price level, aggregate demand and output may be reduced as a result of a negative real balance effect (Pigou effect) (Ahmad A. et al, 2018;Barguellil et al, 2018 ;Umaru and Davies, 2018).
As a result, these theoretical and empirical divergences have yet to be harnessed, and development practitioners are in desperate need of an empirical investigation. Therefore, the rationale of this paper is to pleat empirical information on the impact of foreign exchange rate of Ethiopian economic growth.

Objective of the study
The general objective of this seminar is to forecast the foreign exchange rate and assess its impact on Ethiopian economic growth, based on the foregoing theoretical and empirical foundations of the background setting. Specific objectives include: Forecasting a 15-year Ethiopian foreign exchange rate for the period 2021 to 2035. Assessing the impact of foreign exchange rate on Ethiopia's economic growth.

METHODOLOGY
This study had used yearly data from 1977 to 2018 for the macro variables that are supposed to analyze the impact of exchange rate on economic growth; and for the purpose of forecasting foreign exchange rate, the data spans from 1971 to 2020. The data was taken from World Bank and National Bank of Ethiopia. The parameters employed in the study to analyze the impact of foreign exchange rate on economic growth are, Real Gross Domestic Product (RGDP) which is a proxy for economic growth, and independent variables that are included in this study are nominal exchange rate, the total value import of goods and services, the total value of export items and the inflation rate. The estimation was performed using the auto regressive distributive lag (ARDL) model where as for the purpose of forecasting foreign exchange rate is auto regressive integrated moving average (ARIMA) model was employed. The data was analyzed with the help of EViews 10. The model of estimating the impact of exchange rate on economic growth is stated as follows: LnRGDP= β0 + β1 LnNEXR + β2 LnIMPORT + β3 LnEXPORT + β4 INFLATION+ ɛ (2.1) Where Β0, β1, β2, β3, β4 are coefficients. RGDP is the real gross domestic product that indicates Ethiopia's economic growth and is expressed as a dependent variable. The independent variable is: NEXR is the nominal exchange rate IMPORT is the total value of imported goods and service, EXPORT is the total value of goods and services of export GDP, INFLATION is the rate of inflation calculated using the consumer price index and ɛ is the error item.

ARIMA Models
The Box-Jenkins (ARIMA) model is in theory the most general class of models for forecasting time series and was first popularized by Box and Jenkins (1970). ARIMA (p, d, q) completely ignores independent variables and assumes that past values of the series plus previous error terms contain information for the purposes of forecasting. The integers refer to the Autoregressive (AR), Integrated (I) and Moving Average (MA) parts of the data set respectively. The models are applied in some cases on data which show evidence of non-stationarity which can be stationarized by transformations such as differencing and logging. The model takes into account historical data and decomposes it into AR process, where there is a memory of past events; an integrated process, which accounts for stationarity, making it easier to forecast; and MA of the forecast errors, such that the longer the historical data, the more accurate the forecasts will be, as it learns over time. The ARIMA models are applicable only to a stationary data series, where the mean, the variance, and the autocorrelation function remain constant through time.

ARDL Model
Cointegration method is used in the analysis of long-term relationships between variables. Engle and Granger (1987), Johansen and Juselius (1990) and Johansen (1991) are the most commonly used tests for testing cointegration. The Autoregressive Distributed Lag (ARDL) model and the bounds testing approach which is developed by Pesaran and Shin (1999) and Pesaran, Shin, and Smith (2001) has some advantages over other conventional cointegration approaches. Unlike the other cointegration methods, there is no limiting assumption that all variables used in the ARDL model should be integrated of the same order. Therefore, I (0) and I (1) variables can be used together. However, as a limiting condition, no variable should be integrated of the second or higher order. With this approach, problems arising from non-stationary series are largely eliminated. In addition, the variables included in the analysis may have different lag lengths which is not possible in the VAR modelling. Another advantage of the ARDL model is that short and long-term parameters can be estimated together. By applying linear transformation to the model, it makes possible to obtain an Error Correction Model that combines short-term and long-term relationships without losing long-term information. Another important advantage is that it can be applied to small samples. It gives consistent and reliable results even in samples with limited observations. LnRGDP = α +β1 lnREXR + β2 lnIMPORT + β3 lnEXPORT + β4 INFLATION + εt (2.11) Where GDP is the Real Gross Domestic Product, REXR is real exchange rate, EXPORT is the exports of goods and services, IMPORT is the import of goods and services, INFLATION is the rate of inflation adjusted for European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.17, 2021 consumer price index. If we estimate the above equation by OLS or any other linear method, we obtain long-run effects of the explanatory variables on the explained variable (GDP). But the error correction modelling approach offers an opportunity to also estimate the short run effects. Moreover Pesaran et al., (2001) bounds testing approach has an advantage of estimating short-run and long-run effects in one step. Because of the mentioned advantages the ARDL model in equation (3.16) is estimated.
(2.12) The coefficients from λ1 to λ5 show the long-run relationship between the variables and the coefficients from α1i to α5i showing the dynamic short run relationships among the variables. For example, the short-run effects of exchange rate on real RGDP are inferred by the estimates of α2i. ∆ is the first difference operator, α0 is the constant and εt is the white noise error term.
The analysis of short and long-term dynamics with the ARDL bounds test approach requires a process consisting of several steps. In the first step, the above Model is estimated by OLS method and an F test is used to examine the long-run relationship between variables and test the coefficients of lagged variables together. The null hypothesis H0: λ1 = λ2 = λ3 = λ4 = λ5 = 0 indicates that there is no long-term relationship or cointegration between variables.
The alternative hypothesis states that the lagged coefficients are significant and there is a cointegration relationship among them. The sample value of the calculated F statistic is compared with the critical upper and lower limits created by Pesaran et al., (2001). If the sample value of the calculated F statistic is less than the table lower bound, the null hypothesis stating that there is no cointegration is not rejected. However, if the sample value of the calculated F statistic is greater than the upper bound of the table, the null hypothesis is rejected and the existence of a long-run relationship between the variables in the model is determined. The test is inconclusive if the calculated F statistic is between the upper and lower bound.
After determining the cointegration relationship, in the second step, appropriate lag lengths for the variables are determined by using model selection criteria such as Hannan Quinn Criteria, Akaike Information Criteria (AIC), Schwarz Criteria (SBC). In the third step, by using the information from model (3.17) the error correction model is estimated.
(2.13) The coefficients from α1 to α5 are the short-term dynamic coefficients that stabilize the model. ECM is the error correction term and its coefficient ф shows the speed of adjustment of the model to the long -term equilibrium after a short -term shock. This coefficient should be negative and statistically significant.
For testing the stability of the estimated model, CUSUM and CUSUMSQ tests which are developed by (Pesaran, M. H., Shin, 1999) and (Brown, Durbin, and Evans, 1975) are recommended. CUSUM and CUSUMSQ statistics are recursive estimates and they are marked against breakpoints. Visual inspection of recursive estimates provides information about structural breaks or stability of the model. If the CUSUM and CUSUMSQ statistics are within the critical limits drawn at the 5% significance level, the null hypothesis which states that the model is stable is not rejected.

Data base and analytical framework for ARIMA model
The data focuses on the data of nominal exchange rate movements of US Dollar in terms of Ethiopian birr covering from the period of 1971 to 2020 in annual basis. For this purpose, the data regarding the exchange rate was executed from the World Bank. The sample projection of the respective currencies for a lead time of fifteen years (from 2021 to 2035) were generated by applying the Box-Jenkin's ARIMA method. In univariate Box-Jenkin (UBJ) (Box and Jenkin, 1968; approach the structure of the past values of the variable is recognized and then the near future is extrapolated on the basis of the past. One of the benefits of Box-Jenkin's approach over other approaches is that it does not depend on any economic presumption and captures the slightest deviation in the data set easily (Makradakis and Hyndman, 1998). Box-Jenkin's methodology is based on the simple assumption of stationarity in the data set but most time series are hardly ever stationary and it is mandatory to convert the series to stationary (maximum up to second level) by differencing the series up to suitable level. If the non-stationarity component is added to a mixed ARMA model, then the general ARIMA (p, d, q) is obtained by having the form here under: European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.17, 2021 Ӧp(β) Wt = C+ ếq (β) et which will be non-stationary unless d = 0. The Box-Jenkin methodology has four steps. These are: Identification, estimation, diagnosis and forecasting. During identification step, the sample autocorrelation and partial autocorrelation functions were matched with their theoretical counterparts to identify the tentative model(s). in diagnosis testing of the autocorrelation coefficients of residuals was worked out so that the estimated ARIMA model is adequate. From the models' residuals of each variable, Ljung-Box Q statistics were calculated to examine the hypothesis of randomness and the suitability of ARIMA model to make the reliable forecast. To make forecast equation, equation (3.1) was worked out to get Yt and et by applying the relation Wt = (1-β) d Yt. The reliable estimates of Yt + -! (Designated by Yt (-!) at time t has been considered as conditional expectation of Yt+-! where t, is the forecast origin and -! is the lead time of estimates. Error term et completely dispersed once we forecast for more than q period ahead.

Analytical result
The result has been elucidated in brief under the following sub heads:

Model identification
In this stage, the sample autocorrelation function with partial autocorrelation function were compared along with their theoretical counterparts (Pankratz, 1983) and it has been discovered that autoregressive (AR) and moving average (MA) process could not surpass the order 2. Keeping into consideration the principle of parsimony, we have applied all the four probable combination of ARIMA models depending on the values of p, d, q to choose the suitable order of ARIMA model. The possible combinations ware: {(1, 1, 1), (1, 1, 2), (2, 1, 1) and (2, 1, 2)} (see Annex table 3.2).

Estimation of different ordered ARIMA models
The estimation of different ordered ARIMA models is exhibited in table 4.3. We have picked these values of parameters as the final approximation in those cases where the sum of square (SSE) of disturbance terms was found to be least. In this paper the estimation was executed on the differenced data and to generate forecasts in consonance with the input data, we have performed the inverse of differencing before producing the predictions.
The estimation of the model was performed through maximum likelihood method (Box-Jenkin's and Reinsell, 1994, p.225 (2, 1, 1), AIC, SBC and SIGMASQ values are minimum which is 3.12, 3.24 and 1.1200 respectively compared to other competing models. Hence, the ARIMA (2, 1, 1) model was selected to forecast the exchange rate.     To summarize, this research aimed to assess the model's forecasting capacity in the setting of a developing country (Ethiopia) that operates under a flexible exchange rate regime. To develop US dollar in terms of Ethiopia birr projections, the Auto Regressive Integrated Moving Average (ARIMA) model proposed by Box-Jenkins was applied to stationary data with appropriate mixes of autoregressive and moving average processes. The model prediction indicated that the exchange rate devaluation will show an increasing trend over the next projected years. This forecast will give the government, policymakers, corporations, currency dealers, and others guidance on how to build policies in light of forecasts, as well as how to generate forecasts using the appropriate models.

ARDL Model Result and Estimation
This section begins with descriptive statistics of the variables. This is followed by analyzing nominal exchange trend in Ethiopia, while the time series property using test statistics of Augmented Dickey Fuller (ADF) to provide the basis for the analysis was also considered. Table 3.5 reports the descriptive values of all the variables employed and shows that the mean value of real domestic product, nominal exchange rate, import, export and inflation rate is 1.92E+10, 8.148669, 3.89E+09, 1.07E+09 and 9.460127.
The series that measures the highest level of discrepancy as shown in the standard deviation result is real GDP, while nominal exchange rate shows the lowest level.
Skewness indicates the rate of asymmetry or discrepancy of the variables. Accordingly, RGDP, NEXR, IMPORT, EXPORT, and INFLATION have long right tail. This is because the variables exhibit positive values.
Kurtosis measures the pawedness and flatness of the series. The result shows that only EXPORT is platykurtic relative to its normal distribution because its value is less than 3. While other variables have their kurtosis value greater than 3, this shows that the peak of their distributions are greater than normal, thus, referred to as leptokurtic distribution. Jarque-Bera statistical test indicates the variables that are normally distributed as it measures the differences in the skewness and kurtosis. The result shows that Jarque-Bera statistic rejects the null hypothesis of no normal distribution for all the variables. Thus, it is concluded that they are all normally distributed (p-value of all variables are significant).  ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.17, 2021 Residuals Diagnostic Test Before beginning the estimation procedure, a diagnosis test of the residual of several post estimation tests was required. As a result, the following tests were run, and the results are shown below. The Breusch-Godfrey serial correlation LM test was used to examine whether the data and variables were stable and cascaded using the residual test. The sequence correlation findings for the Breusch-Godfrey test are shown in Table 3.6. Because the P-Value was greater than 5% significance levels (0.6898 > 5%), we were unable to reject the null hypothesis of no sequence correlation, and we concluded that the model and data did not have categorization or sequence correlation. In table 3.7. the test result of Ramsey RESET test for model specification indicates that the p-value is larger than 0.05 implying there is no model specification error. The test for heteroscedasticity in the residuals is checked using Breusch-Pagan-Godfrey test for heteroscedasticity and the p-value is larger than 0.05 that is p-value = 0.9889 (table 4.8) which rejects the null hypothesis of "there is no constant variance among the residuals" indicating that the model has constant variance. Hence, the model is good to estimate. .4. displays the stability diagnosis test of the CUSUM test and the CUSUM square test for the variables considered. It is clear from the CUSUM and CUSUM of squares test that since the blue line does not exceed the 5% significance line (red line), the variables are very stable during the period under study. Therefore, we believe that all variables and data are well adapted.

Figure 3.4: CUSUM square and CUSUM tests
The next section reports the empirical results and interpretations of the estimated model that apply Bound testing approach for cointegration. The results of unit-root test that has been performed by applying test of stationarity, that is, ADF for the application of ARDL. After unit-root analysis, there is a need to scrutinize the response of exchange rate to output along with the combination of import, export and inflation variables. European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.17, 2021  Constant & trend -5.2205*** -8.7591*** I (0) Note: *, **, *** indicate the rejection of null hypothesis of unit-root at 10%, 5% and 1% level of significance, respectively. Table 3.9 reveals the result of the unit root; it shows that variables such as nominal inflation rate was integrated at order zero, while real gross domestic product, exchange rate, import and export were found to be stationary at first difference. The result of the unit root provides the basis for the study to use autoregressive distributed lag for both short-and long-run estimation of the model. Table 4.10. reveals the lag selection criterion suggested by LR, FPE, AIC, SC, HQ. The result shows that the optimum number of lag suitable for this analysis is 2. The suggestion is taken into consideration when analyzing ARDL model.  Table 3.11 shows that the F-statistic (11.682) is beyond all the significance levels. It, therefore, indicates clearly the long-run relationship among the variables. 3.74 5.06 After determining the existence of long-run relationship among the series by applying the bounds testing approach, there is a need to define the long-run coefficients of ARDL estimation. The derived results that reflect the reaction of real output to the series of explanatory variables along with "nominal exchange rate" are reported in Table (3.12).
European Journal of Business and Management www.iiste.org ISSN 2222-1905(Paper) ISSN 2222-2839(Online) Vol.13, No.17, 2021  Note: ***, ** denote the significance level at 1% and 5% level of significance, respectively. The theoretical justification of estimated results requires brief explanation of previous literature. On theoretical ground, currency devaluation leads to increase in net exports that may result in expansion of output (Dornbush's, 1988;Keynesian, 1931). On empirical side, some studies (Miteza, 2006;and Kalyoncu, et al, 2008;Muhammad N. and Ejaz G., 2017) follow the expansionary output hypothesis that is consistent with conventional literature. The long-run result of this study is in line with that of theoretical and empirical literature. The short-run findings of the analysis evaluate that currency devaluation (nominal exchange rate) has no significance relevance to affect economic growth.
The model is estimated in log transformed form, with calculating coefficient values of "nominal exchange rate" of 0.1426, which is significant at the 5% level. A 1 percent increase in nominal exchange rate will increase the real gross domestic product by 0.1426 percent in the long-run. This indicates that exchange rate has a long-run expansionary effect on economic growth. This finding is consistent with the findings of other empirical findings of (Jones and Olken, 2008;Gala, 2007;Rodrik, 2014). When managed carefully, the exchange rate can be a critical policy tool for stimulating export growth since Ethiopia's export-to-GDP is one of the lowest ratios in the world i.e., 3.89% of GDP (IMF 2018). The manufacturing sector has been associated with strong economic growth in developing countries and still remains at the heart of catch-up development. Countries with competitive or undervalued currencies tend to experience more rapid manufacturing growth yet Ethiopia's manufactured exports remain below 10% of total exports (AfDB 2017, p 7). In developing countries, a sizable devaluation facilitates the expansion of the manufacturing sector by encouraging firms to enter new export markets and new export sectors (Freund and Pierola, 2012).
Although the coefficients of imports and exports are both positive, the variable export is insignificant in terms of influencing economic growth. At a 1% significance level, the coefficient of import is statistically significant. In the long run, a 1% increase in imports will result in a 0.532 percent increase in real GDP. This positive relationship between import and real GDP of Ethiopia was consistent with the previous empirical studies of (Adenutsi, 2008;Charles R. et al, 2019). This empirical study's result confirms the positive contribution for import-dominant sectors, which have direct growth impacts on Ethiopia's overall economic growth. In the case of Ethiopia, the major part of the economy is dependent on the import of raw material, intermediate or capital goods, particularly oil, machinery and such other materials which have significant contribution to the growth of real GDP. Some conditions that are often associated with official imports to developing countries, Ethiopia inclusive, might be directly favorable to initiating higher levels of industrial performance as well as economic growth.
Another explanatory variable that is negatively related is the rate of inflation, which indicates that a 1% increase in the rate of inflation in the long run leads to a 0.27 percent reduction in real GDP in the case of Ethiopia. The outcome is consistent with the findings of Anthony O. et al (2020). Table 4.11 (b) explains the short-run relationship between Ethiopia's economic growth and the exchange rate, which shows whether the exchange rate is deterministic or not. To begin with, the significance of the error correction mechanism (ECM) result and the negative sign of the coefficient support the establishment of cointegration among variables in this study. This coefficient is -0.251, implying that approximately 25.1 percent of the previous year's disequilibrium is corrected in the current year. As a result, in the long run, the error correction model (ECM) adjusts slowly to changes.
The main explanatory variable, "nominal exchange rate," has the expected sign and is positively related to RGDP; however, it is not significant, implying that it has no significant effect on RGDP in the short run. Furthermore, export has no significant effect on RGDP in the short run, whereas inflation rate (INFLATION) and its lag period (INFLATION (-1)) have a significant and negative effect on real gross domestic product in the short run. In the short run, the lag period of variable import (IMPORT (-1)) has a significant and positive effect on RGDP, whereas the instantaneous period of import has no significant effect on RGDP.

CONCLUSION
A competitive exchange rate would help to create an environment conducive to manufacturing-led structural transformation, sustained growth acceleration, and improved external balance. Empirical analysis using Ethiopian data provides fairly strong evidence to support the proposition that devaluation can stimulate exports, improve the external balance, boost economic growth, and promote industrialization-driven structural transformation. In other words, a birr devaluation will play an important role in promoting expenditure switching and long-term development (i.e., the growth of efficient export-oriented and import-substituting industries). However, devaluation would involve key macroeconomic trade-offs including the higher cost of imported capital equipment and increase in external debt stock and servicing when expressed in local currency.
Using the Autoregressive Distributed Lag (ARDL) Model, this study examines the impact of the Ethiopian exchange rate on economic development between 1977 and 2018. The findings show that the nominal exchange rate has a statistically significant positive impact on Ethiopian economic growth. Long-term coefficients show that imports have a large positive effect on real GDP, whereas inflation has a significant negative effect on real GDP. Despite the fact that devaluation benefits the economy, the exchange rate has the disadvantage of feeding inflation, which reduces real GDP. In addition, the ARIMA model was used to forecast the exchange rate. This forecasting is important for monetary policy formulation. The forecast will provide guidance to the government, policymakers, corporations, currency dealers, and others on how to build policies based on forecasts.

Recommendation
A more depreciated nominal exchange rate would be a major step forward, but it is not a panacea. Because exports are also constrained by binding capacity and the business environment constraints, supply side factors must be addressed concurrently. The high cost of doing business is one of the major impediments to Ethiopia's export development that ultimately hampers economic growth. As a result, a one-time exchange rate adjustment will be insufficient to address the structural bottlenecks that impede exports, and necessitating a more comprehensive policy package to improve supply response to devaluation. This would include efforts to overcome significant obstacles in the still-challenging business climate (e.g., trade logistics, access to capital and land, infrastructure shortfall, cumbersome customs procedures, skills gap), as real GDP growth and diversification are critical.
One of the government's macroeconomic challenges is price stability. Because devaluation may cause inflation to rise. The IMF estimates inflation pass-through of devaluation is around 0.43% for every 1% devaluation, so for every 1% devaluation (to the US$), inflation can be expected to rise 0.43% in the short term as import costs rise. Hence, the government must prioritize maintaining a single-digit inflation rate. As a result, the policy implications of using fiscal and monetary measures concurrently must be investigated.