Jurnal Ekonomi & Studi Pembangunan Volume 24 Nomor 1, April 2023 Article Type: Research Paper Rupiah exchange rate: The determinants and impact of shocks on the economy Erida Pratiwik and Sucihatiningsih Dian Wisika Prajanti* Abstract: The repetition of policy dynamics on Quantitative Easing (QE) and interest rate by The Fed potentially cause fluctuations in the exchange rate, including in Indonesia. Therefore, this study aims to analyze the determinants and impacts of exchange rate shocks. Inflation (INF), Money Supply (LJUB), Open Market Operations (OPT), Foreign Exchange Reserves (LCD), Expected Inflation (LEHU) and Interest Rates (SB) were used to analyze the determinants of Exchange Rate (NT) through Auto Regressive Distributed Lag (ARDL). The impact of NT shocks was analyzed using Vector Auto Regressive (VAR) by LEHU, Residential Property Price Index (PIHPR), Stock Transactions (LTRANS), and Banking Credit Volume (VK). The Expected Inflation variable and incorporation of ARDL-VAR are novelties in this study. In the secondary time series data for 2014M1 – 2022M9 period, the ARDL results showed that INF and LJUB had positive effect on NT in both long and short run, while OPT, LCD and SB had negative effect. LEHU had negative effect in the short run, but positive in the long run. The speed of adjustment in the model was 49.86% per month. Shock of NT had impacted VK until 15 months, PIHPR at 7 months, LTRANS at 10 months, and LEHU at 14 months. Based on these results, it can be implied that the monetary authority must maintain stability of NT, especially by INF and LJUB transmission. Next, shock's impact must also be overcome, especially on VK. This research is only focused on monetary sector, further research will be refined with other macroeconomic variables. Keywords: Exchange Rate; Impact; Influence; Monetary; Rupiah JEL Classification: E31; E42; E50; F45 Introduction Monetary policy is a crucial aspect of a country's economy. The main goal to be achieved from monetary policy is to maintain exchange rate stability (Kabundi & Mlachila, 2019). Currently, monetary policy plays an important role in setting the price level, which is reflected by the inflation variable (Le et al., 2021). Even though the main success indicator of monetary policy is inflation, the role of other exogenous factors affecting exchange rates should not be ignored, especially in countries that adhere floating exchange rate regime. The stability of the exchange rate is important to maintain considering the vital role of this variable for the economy. Apart from reflecting the competitiveness of a domestic country's currency toward other countries, movement in exchange rates also has a strong potential impact on public economic behavior in allocating their resources. Exchange rate movements can be used as a reflection of cost, relative prices, and productivity of economic resources (Adewuyi et al., 2021). AFFILIATION: Department of Development Economics, Faculty of Economics and Business, Universitas Negeri Semarang, Central Java, Indonesia *CORRESPONDENCE: dianwisika@mail.unnes.ac.id THIS ARTICLE IS AVALILABLE IN: http://journal.umy.ac.id/index.php/esp DOI: 10.18196/jesp.v24i1.18016 CITATION: Pratiwik, E., & Prajanti, S. D. W. (2023). Rupiah exchange rate: The determinants and impact of shocks on the economy. Jurnal Ekonomi & Studi Pembangunan, 24(1), 100- 126. ARTICLE HISTORY Received: 24 Feb 2023 Revised: 03 May 2023 16 May 2023 Accepted: 19 May 2023 https://scholar.google.com/citations?user=wsXa7HwAAAAJ&hl=en&oi=ao https://scholar.google.co.id/citations?user=pJvOZncAAAAJ&hl=en https://ep.unnes.ac.id/web/ https://ep.unnes.ac.id/web/ https://ep.unnes.ac.id/web/ https://ep.unnes.ac.id/web/ mailto:dianwisika@mail.unnes.ac.id http://journal.umy.ac.id/index.php/esp http://dx.doi.org/10.18196/jesp.v24i1.18016 https://creativecommons.org/licenses/by-sa/4.0/ https://crossmark.crossref.org/dialog/?doi=10.18196/jesp.v24i1.18016&domain=pdf Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 101 Indonesia is one of the developing countries with fluctuating historical exchange rate data and has tended to depreciate over the last 10 years. Apart from the 1998 crisis, the largest exchange rate weakening of the Indonesian Rupiah (IDR) against Dollar of the United States (USD) occurred in 2014, following the control of the Quantitative Easing policy and the increase in interest rates by the Fed since May 2013 in order to economic recovery due to the subprime mortgage crisis that occurred since 2008. These policies raised expectations for the United States economy, causing investors to withdraw their funds (capital outflow) from emerging market countries. This condition resulted in USD strengthening, while the exchange rates of emerging market countries weakened, including Indonesia. The deteriorating condition of the exchange rate due to the aggressive policies of the Fed has been researched by Basri (2017), Dinata and Oktora (2020), Estrada et al. (2016), Le et al. (2021), Park et al. (2015), and Triggs et al. (2019). The results of these studies suggested various conclusions. The banking sector is the most affected by exchange rate fluctuations due to changes in Quantitative Easing and global interest rates (Basri, 2017; Estrada et al., 2016; Park et al., 2015; Triggs et al., 2019). Different findings resulted from the research of Dinata and Oktora (2020) and Le et al. (2021), the findings of these studies stated that the impact of exchange rate movements does not only occur in the banking sector, but other sectors are also affected, particularly regarding asset price movements and changes in consumption behavior. Figure 1. Statistics of IDR/USD in 2003 - 2022 Source: Central Bank of Indonesia, 2023 (data processed) Various studies regarding the effect of monetary variables on exchange rates have been carried out and found several inconsistent results. Related to inflation, research from Bato et al. (2017) and Chandrarin et al. (2022) stated that inflation has a plus coefficient effect on the exchange rate, whereas the results of research from Fauji (2016) and Wicesa et al. (2021) showed the opposite result. Next, the influence of the money supply is positive (Funashima, 2020; Yulianti, 2014). However, research from Alawiyah et al. (2019) and Ghosh and Bhadury (2018) showed the opposite result. The open market operations has 0 2000 4000 6000 8000 10000 12000 14000 16000 Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 102 a negative effect on the exchange rate (Carli & Gomis-Porqueras, 2021; Rocheteau et al., 2018). Meanwhile, the research results of Cassola and Koulischer (2019) dan Kirana (2017) showed that the effect of open market operations on the exchange rate is positive. Furthermore, the monetary policy instrument by increase in interest rates resulted in strengthening of the exchange rate or appreciation (Elias et al., 2022; Kilian & Zhou, 2022). The research results from Sa’adah (2020) and Yung (2021) precisely showed conflicting results, where an increase in interest rates causes depreciation of the exchange rate. The condition of international economic flows as reflected by the position of foreign exchange reserves also has been studied by various previous studies. The higher foreign exchange reserves have been proven to cause the appreciation of exchange rate (Chanda et al., 2020; Ito & McCauley, 2020). Discordant results were shown from the research findings of Gupta et al. (2014) and Hammoudeh et al. (2022). The exchange rate as a reflection of the condition country's currency competitiveness is also influenced by expectations of future prices or expectations of inflation. Research from Herkenhoff and Sauré (2021) and Lee and Kim (2019) showed that expected inflation have a positive effect on the exchange rate. The expected inflation variable is rarely used in research, so conflicting results have not been found. The findings of different research results show about determinants of exchange rate especially related to the monetary sector and international trade are important to discuss, considering that to intervent the exchange rate movements the main tool used by the state will be realized through policies in the monetary sector and international competitiveness. Recently, The Fed seems to have re-imposed QE controls and expanded interest rates to overcome post-crisis economic overheating due to the pandemic (Covid-19). From March to December 2022, the Fed has raised interest rates 7 times with an accumulation of 425 basis points, from 0% to 4.25%. Central Bank of Indonesia responded that condition by increasing the Bank Indonesia 7 Day Repo Rate (BI 7DRR) by 200 basis points from July (3.5%) to December 2022 (5.5%). This economic turmoil is predicted to continue, so that the rupiah exchange rate is also predicted to fluctuate and tend to depreciate (Sunaryati & Munandar, 2023). Historical data on the deteriorating exchange rate as a outcome of QE policy and The Fed's interest rate also economy condition which is predicted continue to worsen for several years after the Covid-19 pandemic crisis must be responded to with appropriate policies so that the negative impact can be overcome immediately. Identification of the factors that work on exchange rate movements as well as analysis of the sectors most affected will assist in formulating these policies. Monetary policy as responsible for exchange rate stability must be properly formulated. Therefore, this research objective is to analyze the determinants of exchange rate, also the impact of exchange rate shocks on the economy. determinants exchange rate carried out using Auto Regressive Distributed Lag (ARDL). Next, to analyze the impact of exchange rate shocks, Vector Auto Regression (VAR) will be used. The use of ARDL combined with VAR is a novelty in this study. In addition, this study also includes Expected Inflation as a variable that rarely used to estimate exchange rate movements, even though this variable has an important role in policy making. The research will be focused 2014 – 2022 period considering there was a massive movement Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 103 in the rupiah exchange rate in that period. Apart from contributing theoretically to the literature study regarding the influence factors and impacts of exchange rate shocks, the findings of this study also contribute practically in the form of detailed policy recommendations. The importance of this research is considering the exchange rate not only as a volatile variable but also its role as monetary policies main goal. Policies are compiled and established with the aim of being a solution of the problem or improving the quality of a condition. At the macroeconomic level, particularly in the monetary sector, policymaking requires complex stages. When the Central Bank as the policy maker for the monetary sector has established a policy, it will continuously affect other sectors. The exchange rate as a reflection of a country's monetary power whose movements are displayed in real time on the foreign exchange market, is a variable that has great power (impactful) in influencing the balance of economy. The position of monetary policy instruments is vital in controlling the rupiah movement. Even though a country adheres to floating exchange rate regime, the intervention of monetary institutions in transmissions that affect exchange rates must still be carried out for the sake of economic stability and public expectations of future conditions. However, exchange rate movements cannot be completely controlled, so the impact of these movements is a consequence that must be addressed. By knowing the factors that determine the exchange rate, it will help in overcoming the impact of shock that occur on the exchange rate. This topic will be shown in this research with 5 main parts, starting from introduction to explain the research background, research gaps, data, research aims, research importance and novelty. Then, the literature review is written to show some of previous studies and hypotheses. The research continued at research method part to explain the analysis technique and the procedure. The answer of hypotheses and research aims are answered in the result and discussion. Finally, research concludes and explained the policies recommendation, which also acknowledged the limitation in the conclusion. Various literature reviews regarding exchange rates have been carried out to develop research hypotheses. The exchange rate reflects the circumstances of domestic purchasing power also the international competitiveness of a country's currency. When the domestic economy is currently in high inflation, the purchasing power of money in the country becomes weak (theory of money quantity). In addition, on the international trade flow, there will be more Rupiah that must be converted to USD. This condition means that the higher level of domestic inflation, the more depreciated the exchange rate will be Proof that inflation causes exchange rate depreciation had been carried out by Bato et al. (2017), and Chandrarin et al. (2022). H1: Inflation has a positive effect on the Exchange Rate. Inflation is a success indicator of monetary policy. When inflation increase exceeds the target, the monetary authority (Central Bank) will control it through monetary policy instruments. The causes of inflation can be traced from production side (cost push inflation) or by consumption side (demand pull inflation), both of them are related to the Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 104 price level and purchasing power. To overcome this, the volume of money supply has a crucial role. The higher volume of money supply, it means that easier for people to consume. This condition will promote the higher quantity of demand (goods and services). If the supply side is unable to equal with the level of demand because economic conditions not yet possible to improve production input factors, then the equilibrium will shift, the price level will rise. If this price increase continues, it is called inflation (demand pull inflation), and in the long run it has the potential to cause exchange rate depreciation. It can be concluded that high volume of money supply involve a depreciation in the exchange rate (Funashima, 2020; Gong et al., 2022). H2: The Money Supply has a positive effect on the Exchange Rate. The movement of money supply represents purchasing power and is influenced by inflation expectations. If the public currently faced high money supply, expectations for future prices of domestic products will also increase. This has resulted in people preferring to buy imported products because they are cheaper than domestic ones. The higher interest in imported products, the quantity of imports will increase, so the trade balance will decrease and result in exchange rate depreciation. The research result from Herkenhoff and Sauré (2021) and Lee and Kim (2019) validated that higher expected inflation led the exchange rate depreciation. H3: Expected Inflation have a positive effect on Exchange Rates. Controlling volume of money supply is a strategic step the Central Bank takes to manage inflation. Instruments that can be responded quickly by financial institutions and the general public in dealing with this matter are interest rates. The higher level of interest rate will press the public's interest in applying for bank credit loans (liquidity preference theory). This happens because the rate of return that must be given in the future is higher than the condition when interest rates are slight. Moreover, people will be more interested in saving because the value of money that will be received in the future will be higher. The higher saving cause the smaller of money supply (the volume of money supply decreases). Public interest in declining consumption also suppresses inflation, due to the low level of demand for goods and services (Andrieș et al., 2017). At a fixed level of supply, this shift in demand causes the equilibrium to shift to the lower left, the price level decreases. If this condition happen continuously, deflation will occur. The price of domestic goods and services becomes cheaper. It has the potential to encourage consumption interest in domestic goods to be higher so that the level of imports can be reduced. The lower the import, the less rupiah (domestic currency) that must be converted to USD, the stronger the exchange rate (Ulm & Hambuckers, 2022). H4: Interest Rates have a negative effect on Exchange Rates. Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 105 A decrease in imports will increase foreign exchange reserves because there is no deficit on trade balance (Aulia & Masbar, 2016). High foreign exchange reserves indicate that the ability of the monetary authority to control exchange rate is getting stronger, it will encourage exchange rate appreciation (Chanda et al., 2020; Ito & McCauley, 2020; Uz Akdogan, 2020). H5: Foreign Exchange Reserves have a negative effect on the Exchange Rate. The interest rate is the instrument taken by the Central Bank to restrain money supply through the transmission of financial institution responses. This requires a time lag, while the exchange rate changes every second and within a certain time requires fast action to prevent the depreciation rate from worsening. To handle this condition, the Central Bank will use the open market operation instrument because it can directly intervene in the foreign exchange (forex) market on a daily basis. When the rupiah is depreciating (high money supply), open market operations to encourage exchange rate strengthening will be carried out (Rocheteau et al., 2018), one of which is through the issuance of Bank Indonesia Certificates (BIC). When BI issues BIC, Commercial Banks and Intermediaries have the right to purchase them. These Commercial Banks and Intermediary Institutions will provide a certain amount of funds (according to the BIC nominal) to Bank Indonesia. This condition reabsorbed money supply, increasing the power to supply rupiah in the forex market. At a constant level of demand for the rupiah, while the supply of rupiah increases (liquidity injection), the equilibrium (exchange rate) in the market will decrease (appreciate). This condition indicates that open market injection operations affect the appreciated exchange rate, while absorption of that instrument will result in exchange rate depreciation (Carli & Gomis-Porqueras, 2021). H6: Open Market Operations have a negative effect on Exchange Rates. Significant exchange rate movements (due to shocks or some phenomena) lead to changes in economic variables, especially macroeconomics. Shocks that occur in exchange rates and lead to depreciation have the potential to reduce bank credit volume (Fabiani et al., 2022). This can be traced from one of the indicators of exchange rate depreciation: the high domestic inflation marked by large money supply volumes. To seek an appreciation of the exchange rate, policies will be taken that are able to reduce the volume of money supply, the example is increasing interest rates. This condition will encourage an increase in saving, but on the other hand it will suppress credit volume (Yun & Cho, 2022). H7: If there is a shock of Exchange Rate, the Credit Volume will decrease. The decline in credit volume indicates that the level of public consumption will decrease. This is a separate signal for the property sector, where a decline in consumption indicates low public purchasing power so that the level of demand for the property sector will also Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 106 decrease. To cope with the decline in the level of demand, prices from property supplies will be lowered (Anastasia & Hidayat, 2019; Dąbrowski & Wróblewska, 2020). H8: If there is a shock of Exchange Rate, Property Prices will decrease. Lower interest rates in emerging market countries tend to generate negative sentiment for foreign investors. Low interest rates decrease foreign investors expectations of the rate of return on capital, resulting in an outflow of funds. This condition will be exploited by domestic investors to dominate the stock exchange. Therefore, when the exchange rate depreciates, the volume of stock transactions appears to have increased (Ding, 2021; Huang et al., 2021; Sa’adah, 2020). H9: If there is a shock of Exchange Rate, Stock Transaction Volume will increase. A shock to the exchange rate also has the potential to make expectations of future prices more expensive. This occurs because a weak exchange rate will make production costs and output prices (especially with imported raw materials) more expensive. Such conditions will push future general prices (inflation expectations) higher than at present (Janah & Pujiati, 2018). Exchange rate shocks make people expect future prices to be more expensive than the current price. It can be state that the response of expected inflation to exchange rate shocks is positive or increases (Anderl & Caporale, 2022; Lee & Kim, 2019). H10: If there is a shock of Exchange Rate, Expected Inflation will increase. Research Method Research objectives and methods are two closely related aspects. Auto Regressive Distributed Lag (ARDL) was used in this study to analyze the variables that influence (determinants) exchange rates. To attain this purpose, the monetary sector's determinants, namely Inflation (INF), Money Supply (LJUB), Open Market Operations (OPT), Foreign Exchange Reserves (LCD), Interest Rates (SB), and Expected 3-month General Prices or Expected Inflation (LEHU), are being broken down. The Rupiah Exchange Rate (NT) toward the United States Dollar (USD) has a role as the dependent variable. Next, the research objective is the impact of exchange rate shocks on the economy, executed using Vector Auto Regressive (VAR), by positioning the Exchange Rate variable (NT) as a vector as well as the Inflation Expectation variable (LEHU), Changes in Residential Property Price Index (PIHPR), Stock Transaction Volume (LTRANS), and General Banking Credit Volume (VK) as response variables. This research was executed using a quantitative approach to secondary time series data for the period January 2014 to September 2022 Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 107 in Indonesia obtained from the official websites of Central Bank of Indonesia (BI), the Central Bureau of Statistics (BPS) and the Financial Services Authority (OJK). Figure 2 Research Design Table 1 Attribute of Variables Code Variable Specification Measurement Source NT Exchange Rate Last mont of JISDOR kurs Rupiah/USD. IDR/USD BI INF Inflation Consumer Price Index approach. % BI LJUB Money Supply (Log) Broad total of money supply (M2) Rp Trilion ➔% BPS OPT Open Market Operation Absorbtion or injection from rupiah and forex intervence Rp Trilion BI LCD Foreign Exchange Reserves (Log) Amount of monetary external assets (IRFCL concept). USD Milion ➔% BI SB Interest Rate Central Bank of Indonesia 7 Days Repo Rate (BI7DRR). % BI LEHU Expected Inflation (Log) Index of general price expectation on the next 3 months. Base Point ➔% BI PIHPR Residential Property Index Average of Residential Property Index by 15 Cities in Indonesia. Base Point BI LTRANS Stock Transaction Volume (Log) Amount of stock transaction value in Indonesia stock exchange. Rp Bilion ➔% BPS VK General Banking Credit Volume Amount of loan realization by general commercial bank. Rp Bilion OJK Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 108 In the ARDL method, hypothesis testing was done by looking at the direction of the statistical coefficient (positive or negative), if the result is contrary to the established hypothesis, it means the null hypothesis (H0) is accepted. Next, the level of significance decided from the probability (prob) value of the statistical results of the data processing. If the p value is <0.05 then the result is that the effect has a significant probability. Models with BLUE (Best Linear Unbiased Estimation) criteria in ARDL were obtained through some of the stages, starting from stationarity tests (with the Augmented Dickey Fuller approach unit root test), classic assumption tests (Autocorrelation and Multicollinearity), Bound testing, model selection (Akaike Information Criteria approach), ARDL estimation, and model stability tests (CUSUM and CUSUM Square Test). The mathematical model used was: 𝑁𝑇𝑡 = ∑ 𝛼1. 𝑁𝑇𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼2. 𝐼𝑁𝐹𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼3. 𝐿𝐽𝑈𝐵𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼4. 𝑂𝑃𝑇𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼5. 𝐿𝐶𝐷𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼6. 𝐿𝐸𝐻𝑈𝑡−𝑖 + ∑ 𝛼7. 𝑆𝐵𝑡−𝑖 𝑛 𝑖=0 + 𝐶𝑡 𝑛 𝑖=0 + 𝜀𝑡 (1) However, if the Bound test results showed cointegration (the value of I(1) is greater than I(0)) then the model that must be used was ARDL-ECM, namely ARDL which was divided into long-run and short-run, with the following mathematical model: ∆𝑁𝑇𝑡 = ∑ 𝛽1. ∆𝐼𝑁𝐹𝑡−𝑖 + 𝑛 𝑖=0 ∑ 𝛽2. ∆𝐿𝐽𝑈𝐵𝑡−𝑖 + 𝑛 𝑖=0 ∑ 𝛽3. ∆𝑂𝑃𝑇𝑡−𝑖 + 𝑛 𝑖=0 ∑ 𝛽4. ∆𝐿𝐶𝐷𝑡−𝑖 + 𝑛 𝑖=0 ∑ 𝛽5. ∆𝐿𝐸𝐻𝑈𝑡−𝑖 + 𝑛 𝑖=0 ∑ 𝛽6. ∆𝑆𝐵𝑡−𝑖 + 𝑛 𝑖=0 𝐶𝑜𝑖𝑛𝑡𝐸𝑞. (𝑁𝑇 − (𝛼1. 𝐼𝑁𝐹 + 𝛼2. 𝐿𝐽𝑈𝐵𝑡−1 + 𝛼3. 𝑂𝑃𝑇𝑡−1 + 𝛼4. 𝐿𝐶𝐷𝑡−1 + 𝛼5. 𝐿𝐸𝐻𝑈𝑡−1 + 𝛼6. 𝑆𝐵𝑡−1 + 𝐶𝑡 + 𝜀𝑡 (2) Where: ∑ 𝑖s the influence period of the variable,𝑛𝑖 𝛽1 to 𝛽6 is the short-run coefficient of the variable, 𝐶𝑜𝑖𝑛𝑡𝐸𝑞 is the speed of adjustment of the model, 𝛼1 to 𝛼6 is the long-run coefficient of each variable, C is the constant of the model , and ε is the error term. Next, the VAR method was carried out procedurally starting from the stationarity test, choosing the optimum lag, VAR stability test, causality test (Granger approach), VAR model estimation, analysis of Impulse Response Function (IRF), and finished with Forecast Error Variance Decomposition (FEVD) analysis. The VAR mathematical model used in this study, was: 𝑁𝑇𝑡 = ∑ 𝛼1. 𝑁𝑇𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼2. 𝐿𝐸𝐻𝑈𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼3. 𝑃𝐼𝐻𝑃𝑅𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼4. 𝐿𝑇𝑅𝐴𝑁𝑆𝑡−𝑖 𝑛 𝑖=0 + ∑ 𝛼5. 𝑉𝐾𝑡−𝑖 𝑛 𝑖=0 + 𝜀𝑡 (3) The research hypothesis regarding the impact of Exchange Rate (ER) shocks was proven by examining each variable's movement response (up or down) in the IRF analysis. Meanwhile, the significance of the effect between variables was seen by comparing the t statistic with the t table. The stationarity test is crucial in research with time series data. This test serves as a step that is able to show the stability of data changes. Data that is not stationary has the potential to produce an invalid model. Then, the basic assumption that must be met in Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 109 ARDL is that data cannot be stationary at the 2nd Difference level. In addition, the stationarity test also serves as a determination of the proper use of the model in the VAR method. When the data is stationary at levels, then the VAR model can be used, but data that is only stationary at the 1st Difference level must be continued with a cointegration test and estimated using the Vector Error Correction Model (VECM). Result and Discussion Determinants of Exchange Rate Analysis of the exchange rate determinants using ARDL began with the data stationarity test. The result of test using the Augmented Dickey Fuller (ADF) unit root test showed that the Exchange Rate (NT) and Expected Inflation (LEHU) variables were stationary at the level (probability value less than 0.05), while at the 1st difference all variables were stationary. These results can be interpreted that all variables can be used in the ARDL method, where the main requirement was that the data must be stationary at the level and/or 1st difference. Table 2 Result of Stationary Testing on ARDL Data by ADF Test Prob. Value Variable NT INF LJUB OPT LCD LEHU SB Level 0.0371* 0.1511 0.7489 0.9998 0.3380 0.0011* 0.5602 1st Difference 0.0000* 0.0000* 0.0000* 0.0000* 0.0000* 0.0000* 0.0000* *p-value <0.05 Next, a classical assumption test was performed to ensure that the data produces a BLUE model. The normality test is not required to do on data with more than 30 (n) observations, because it fulfills the assumptions of Central Limit Theorem. The heteroscedasticity test also not required considering that the study location (cross section) is only 1. Therefore, the classic assumption test that was carried out were autocorrelation and multicollinearity. The results shown in Table 3 can be interpreted that variable formulation of ARDL model clear of classical assumption problems. Table 3 Result of Classical Asumption Testing Autocorrelation Multicolinearity Method Threshold Result Variable Measurement Threshold Result LM Test Prob. Chi Square <0.05 0.0712 (No Autocor- relation) INF Variance Inflation Factor (VIF) Centered VIF <10 3.2977 No Multico- linearity LJUB 2.3091 OPT 1.8261 LCD 1.3253 LEHU 6.1620 SB 7.3349 Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 110 The Bound test is a cointegration test approach used to ensure that all variables in the model have a long-run connection. The results of test in Table 4 can be concluded that a long-run connection among variables occurred, because the I(1) Bound value was greater than the I(0) Bound value. After passing the stationarity test, classical assumption, and bound test, data processing can proceed to the ARDL estimation step. Table 4 Result of Bound Test Significance Level I (0) Bound I (1) Bound 10% 2.12 3.23 5% 2.45 3.61 2.5% 2.75 3.99 1% 3.15 4.43 The selected ARDL model (based on the AIC Graph) is 1,0,1,0,2,4,0 orde (has the smallest AIC value). However, because the results of the Bound Test reveal that cointegration occurred, the model must be further broken down into short run and long run (ARDL- ECM). The results are showed in Table 5. Figure 3 Result of Model Selection Test In the short run, the Inflation (INF) variable had a significant positive effect on Exchange Rate (NT). It means that the increase of inflation affects the depreciation in the exchange rate. In the long run, % increase in Inflation (INF) of 1% led to depreciation of the Exchange Rate (NT) by 12,616.80 IDR/USD. The long-run coefficient of the Inflation (INF) variable was greater than the short-run coefficient, meaning that if inflation consistently increased in the long run, the coefficient of depreciation experienced by the exchange rate will also increase. These results proved that statistically, Hypothesis 1 was accepted. The Money Supply (LJUB) variable in the short run and long run had a significant positive effect. If the Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 111 volume of money supply increases, the exchange rate will depreciate, but the depreciation rate in the long run was smaller than in the short run. This can be observed from the magnitude of the variable coefficient. These results can be concluded that Hypothesis 2 of the study could be accepted. Expected Inflation (LEHU) in the short-run (in the previous 3 periods) had a significant negative effect on the Exchange Rate (NT), while in the long-run the effect was actually a significant positive one. When the expected inflation in the previous 3 periods were at a good level (high index), the exchange rate for the next 3 months (current period) will appreciate. Table 5 Result of ARDL Estimation Short-run Long-run Variable Coefficient S.E Prob. Variable Coefficient S.E Prob. ∆INF 6,291.57 2,998.78 0.0388** INF 12,616.80 6,082.73 0.0411** ∆LJUB 12,109.77 2,020.69 0.0000* LJUB 7,567.23 832.86 0.0000* ∆OPT -0.81 0.26 0.0025* OPT -1.62 0.45 0.0005* ∆LCD -7,456.24 1,205.07 0.0000* LCD -9,738.13 1,698.19 0.0000* ∆LCDt-1 -1,894.48 1,192.56 0.1158 ∆LEHU 665.91 470.10 0.1602 LEHU 2,948.81 930.17 0.0021* ∆LEHUt-1 827.50 600.92 0.1721 ∆LEHUt-2 -313.75 592.73 0.5979 ∆LEHUt-3 -1,096.21 453.25 0.0177** ∆SB -106.41 50.40 0.0376** SB -213.39 101.28 0.0380** CointEq(- 1) -0.4986 0.08 0.0000* C -18,888.31 7,075.49 0.0091* *significance at 0.01; **significance at 0.05 Meanwhile, expected inflation that was continuously high in the long run will actually result in exchange rate depreciation. The coefficient of expected inflation effect was in accordance with the established Hypothesis 3, but the short-run coefficient was not appropriate. This happened because rising expected inflation in the short-run will lower consumption because people are on the alert for future price increases. Consumption of products with high prices, such as imported products, is also affected. If the interest in imported products decreases, the volume of imports will also run in harmony, so the exchange rate is getting strenghten in the short run. In the theory of aggregate demand-supply, if a country does not adhere to a sticky price system, then in the long run the condition of economy will return to a balance point even though the price level is higher than before. This is due to economic improvements (policies) that encourage stability on the aggregate demand side, for example through increasing wages or reducing tax rates. The Interest Rate (SB) variable statistically had a significant negative effect in the short and long run. From these statistical results, it can be said that an increase of interest rates will encourage exchange rate appreciation, and the amount of appreciation will improve in the long-run. The results of research on the interest rate variable in this topic were Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 112 statistically in accordance with the established Hypothesis 4. Furthermore, the Foreign Exchange Reserves (LCD) variable produced a significant negative effect coefficient for the long-run and short-run, but the foreign exchange reserved in the previous period had an insignificant influence probability. This condition can be interpreted that the higher of foreign exchange reserves held by the monetary authorities, the more exchange rate will appreciate. In the long-run, if the foreign exchange reserves increase, the exchange rate will appreciate significantly. This result is in accordance with the hypothesis that had been set (Hypothesis 5). Next, Open Market Operations (OPT) had a negative influence coefficient with a significant probability for both the short-run and the long-run. These results indicated that the exchange rate will appreciate when the open market operation intervenes (increases the volume of rupiah in the foreign exchange market). From these results, it can be concluded that Hypothesis 6 could be accepted. The magnitude of the exchange rate appreciation in the long-run was stronger when compared to the short-run. -30 -20 -10 0 10 20 30 2015 2016 2017 2018 2019 2020 2021 2022 CUSUM 5% Significance -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 2015 2016 2017 2018 2019 2020 2021 2022 CUSUM of Squares 5% Significance Figure 4 Result of ARDL Stability Test The short-run model had a speed of adjustment of -0.4986. That is, the short-run model will reach a long-run balance at a rate of 49.86% per month. Next, the stability test of the ARDL model using CUSUM and CUSUM Square Test shows that the model was proven to be stable and can be used for the long-run because the CUSUM value was within the 5% significance (red line) area. From the data processing that had been done, the ARDL model produced in this study is: ∆𝑁𝑇𝑡 = 6,291.57 . ∆𝐼𝑁𝐹𝑡 + 12,109.77 . ∆𝐿𝐽𝑈𝐵𝑡 − 0.81 . ∆𝑂𝑃𝑇𝑡 − 7,456.24 . ∆𝐿𝐶𝐷𝑡 − 1,894.48 . ∆𝐿𝐶𝐷𝑡−1 + 665.91 . ∆𝐿𝐸𝐻𝑈𝑡 + 827.50 . ∆𝐿𝐸𝐻𝑈𝑡−1 − 313.75 . ∆𝐿𝐸𝐻𝑈𝑡−2 − 1,096.21 . ∆𝐿𝐸𝐻𝑈𝑡−3 − 106.41 . ∆𝑆𝐵𝑡 + 49.86%. (𝑁𝑇 − (12,616.80 . 𝐼𝑁𝐹 + 7,567.23 . 𝐿𝐽𝑈𝐵𝑡−1 − 1.62 . 𝑂𝑃𝑇𝑡−1 − 9,738.13 . 𝐿𝐶𝐷𝑡−1 + 2,948.81 . 𝐿𝐸𝐻𝑈𝑡−1 − 213.39. 𝑆𝐵𝑡−1 − 18,888.31 + 𝜀𝑡 (4) Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 113 Impacts of Exchange Rate Shocks The Impact analysis on shocks of Exchange Rate (NT) to the variables of Expected Inflation (LEHU), Residential Property Index (PIPR), Stock Transaction Volume (LTRANS), and General Banking Credit Volume (VK) using the VAR method also began with a stationarity test. The unit root test result (Table 6) showed that every variables were stationary at the levels. This condition allowed the use of the VAR model and did not need to do the cointegration test. Table 6 Result of Stationary Testing on VAR Data by ADF Test Prob. Value Variable NT LEHU PIHPR LTRANS VK Level 0.0294** 0.0078* 0.0269** 0.0000* 0.0103* 1st Difference 0.0000* 0.0000* 0.0000* 0.0000* 0.7623 *p-value <0.01; **p-value <0.05 Table 7 Result of Lag Optimum Selection Lag LogL LR FPE AIC SC HQ 0 -1,895.08 NA 71,100,000,000 39.17687 39.30958 39.23053 1 -1,522.54 698.9942 54,980,033* 32.01108* 32.80738* 32.33306* 2 -1,506.25 28.88703 66,065,494 32.19064 33.65053 32.78095 3 -1,487.45 31.3899 75,860,083 32.31858 34.44205 33.17721 4 -1,456.37 48.71270 68,237,698 32.19309 34.98014 33.32003 5 -1,432.56 34.84605 72,238,543 32.21776 35.6684 33.61303 6 -1,419.06 18.37882 96,145,436 32.45476 36.56899 34.11835 7 -1,396.78 28.01927 109,000,000 32.51089 37.2887 34.4428 8 -1,374.19 26.07925 126,000,000 32.56065 38.00205 34.76089 *selected optimum lag After the stationarity test, the next step was to specify the optimum lag which was selected from the smallest value of AIC. Table 7 shows that the optimum lag in this model was lag 1. The stability test of the VAR model was carried out by looking at the modulus value of the Auto Regressive (AR) Root. The model can be said to be stable if the modulus value < 1. Table 8 shows that the modulus generated by AR Root ranges from 0.06 – 0.99. From these results, it can be avowed that the VAR model used was stable. Table 8 Result of VAR Stability Model Root Modulus 0.998110 0.99811 0.761425 + 0.042978i 0.762637 0.761425 + 0.042978i 0.762637 0.356666 – 0.371321i 0.514869 0.356666 + 0.371321i 0.514869 0.412662 – 0.232676i 0.473739 0.412662 + 0.232676i 0.473739 -0.31297 0.312969 -0.19053 0.190532 -0.06789 0.067893 Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 114 The Granger causality test showed that all variables did not have two-way connection (cause and effect), but there was one-way connection from Exchange Rate (NT) to Stock Transaction Volume (LTANS), then Banking Credit Volume (VK) to Exchange Rate, also Banking Credit Volume to Stock Transaction Volume. This conclusion was obtained from the F statistic p-value which was smaller than the 5% significance level (0.05). After passing the stationarity test, choosing the optimum lag, and Granger causality test, the VAR model can be estimated. The estimation results of the VAR model at lag 1 can be interpreted that current condition (period) of Exchange Rate (NTt) is formed from the positive conditions of the Exchange Rate itself in the previous period (NTt-1), as well as the positive conditions of the Expected Inflation, Residential Property Prices Index, and Banking Credit Volume variables in the previous period (LEHUt-1, PIHPRt-1, and VKt-1). In addition, the negative condition of the Stock Transaction Volume in the previous period (LTRANSt-1)also had an impact. Among the 5 variables in the VAR model, which contributed significantly to the current condition of the Exchange Rate were the Exchange Rate itself, and Expected Inflation, and Banking Credit Volume. Table 9 Result of Granger Causality Test Variable F. Statistics Prob.Value NT LEHU PIHPR LTRANS VK NT - 0.7881 0.8068 0.0278** 0.1540 LEHU 0.6489 - 0.3933 0.4048 0.3179 PIHPR 0.9957 0.7156 - 0.5657 0.5054 LTRANS 0.1449 0.9051 0.9494 - 0.0700*** VK 0.0023* 0.8349 0.7959 0.0002* - *p-value <0.01; **p-va;ue <0.05; ***p-value <0,1 After passing the stationarity test, choosing the optimum lag, and Granger causality test, the VAR model can be estimated. The estimation results of the VAR model at lag 1 can be interpreted that current condition (period) of Exchange Rate (NTt) was formed from the positive conditions of the Exchange Rate itself in the previous period (NTt-1), as well as the positive conditions of the Expected Inflation, Residential Property Prices Index, and Banking Credit Volume variables in the previous period (LEHUt-1, PIHPRt-1, and VKt-1). In addition, the negative condition of the Stock Transaction Volume in the previous period (LTRANSt-1)also had an impact. Among the 5 variables in the VAR model that contributed significantly to the current condition of the Exchange Rate were the Exchange Rate itself, Expected Inflation, and Banking Credit Volume. From the data processing that has been done, the formed of VAR model is: 𝑁𝑇𝑡 = 0.74 . 𝑁𝑇𝑡−1 + 527.14 . 𝐿𝐸𝐻𝑈𝑡−1 + 470.37 . 𝑃𝐼𝐻𝑃𝑅𝑡−1 − 11.93 . 𝐿𝑇𝑅𝐴𝑁𝑆𝑡−1 + 0.74 . 𝑉𝐾𝑡−1 + 𝜀𝑡 (5) After estimating the VAR model, the impact of shocks on Exchange Rate (NT) can be seen by looking at the responses of other variables to the shocks that occur to them (Figure 4). The Banking Credit Volume (VK) variable shows a negative (decreasing) response to exchange rate shocks. These results are in accordance with the established hypothesis Pratiwik & Prajanti Rupiah exchange rate: The determinants and impact of shocks on the economy Jurnal Ekonomi & Studi Pembangunan, 2023 | 115 (Hypothesis 7). Banking Credit Volume response had been moving for 15 months, after that, the reponse looked stable and did not approach point 0, so it can be interpreted that Exchange Rate shocks had a long-run impact on Banking Credit Volume. Next, the Residential Property Prices Index (PIHPR) response to Exchange Rate shocks was negative (decreasing). From these results, it can be concluded that Hypothesis 8 was proven. This response lasted for 7 months, then in 8th to 24th month it looked close to point 0. It means that shocks in the Exchange Rate did not leave a long-run impact on the Residential Property Price Index variable. Table 10 Result of VAR Estimation Variable NT LEHU PIHPR LTRANS VK NT (-1) (S.E) [t stat] 0.737444 (0.06887) [10.7080]* 0.000006 (0.000012) [0.49946] 0.000002 (0.000006) [0.29499] 0.00001 (0.000044) [0.22635] -16.8799 (11.2122) [-1.50549] LEHU (-1) (S.E) [t stat] 527.1357 (264.585) [1.99231]* 0.880359 (0.04801) [18.338]* -0.017861 (0.02297) [-0.77771] 0.468425 (0.168160) [2.78553]* 12,321.85 (43,076) [0.28605] PIHPR (-1) (S.E) [t stat] 470.3657 (889.31) [0.52891] 0.064927 (0.16136) [0.40238] 0.633724 (0.07719) [0.78737] 0.553937 (0.5622) [0.98003] -62,568.12 (144,785) [-0.43215] LTRANS (-1) (S.E) [t stat] -11.93128 (105.362) [-0.11324] 0.048801 (0.019120) [2.55274]* 0.007201 (0.00915) [0.78737] 0.766728 (0.06697) [11.4496]* 14,076 (17,153.6) [0.82059] VK (-1) (S.E) [t stat] 0.737444 (0.06887) [10.7080]* 0.000006 (0.000012) [0.49946] 0.000002 (0.000006) [0.29499] 0.00001 (0.000044) [0.22635] -16.8799 (11.2122) [-1.50549] R Square 0.8590 0.5037 0.4202 0.6376 0.9957 *t-stat