The impact of oil price shocks on oil-dependent countries' currencies: The case of Azerbaijan and Kazakhstan

Czech, Katarzyna; Niftiyev, Ibrahim Article The impact of oil price shocks on oil-dependent countries' currencies: The case of Azerbaijan and Kazakhstan Journal of Risk and Financial Management Provided in Cooperation with: MDPI – Multidisciplinary Digital Publishing Institute, Basel Suggested Citation: Czech, Katarzyna; Niftiyev, Ibrahim (2021) : The impact of oil price shocks on oil-dependent countries' currencies: The case of Azerbaijan and Kazakhstan, Journal of Risk and Financial Management, ISSN 1911-8074, MDPI, Basel, Vol. 14, Iss. 9, pp. 1-13, https://doi.org/10.3390/jrfm14090431 This Version is available at: https://hdl.handle.net/10419/258535 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Journal of Risk and Financial Management Article The Impact of Oil Price Shocks on Oil-Dependent Countries’ Currencies: The Case of Azerbaijan and Kazakhstan Katarzyna Czech 1,* and Ibrahim Niftiyev 2,3   Citation: Czech, Katarzyna, and Ibrahim Niftiyev. 2021. The Impact of Oil Price Shocks on Oil-Dependent Countries’ Currencies: The Case of Azerbaijan and Kazakhstan. Journal of Risk and Financial Management 14: 431. https://doi.org/10.3390/jrfm14090431 Academic Editor: James M Hanly Received: 28 July 2021 Accepted: 4 September 2021 Published: 9 September 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1Department of Econometrics and Statistics, Institute of Economics and Finance, Warsaw University of Life Science, Nowoursynowska 166, 02-787 Warsaw, Poland 2Institute of Finance and International Economic Relations, Doctoral School in Economics, University of Szeged, Kálvária Sgt. 1, 6722 Szeged, Hungary; [email protected] 3Centre for Studies on European Economy (AIM), Azerbaijan State University of Economics (UNEC), Istiglaliyyet Str. 6, Baku 1001, Azerbaijan *Correspondence: [email protected] Abstract: The paper aims to assess the relationship between Azerbaijani and Kazakhstani exchange rates and crude oil prices volatility. The study applies the structural vector autoregressive (SVAR) model. The paper concentrates on Azerbaijan and Kazakhstan, the post-Soviet countries considered as some of the most oil-dependent countries in the Caspian Sea region. The impulse response functions suggest that the rise of crude oil prices is associated with the exchange rates decrease and thus with an Azerbaijani manat and Kazakhstani tenge appreciation against the U.S. dollar. Moreover, the results suggest that an oil price increase leads to the rise of Azerbaijani international reserves. However, the results are insignificant for the Kazakhstani foreign exchange reserves. Additionally, the study reveals a negative and significant relationship between crude oil prices and USD/KZT in both pre-crisis and the COVID-19 crisis periods. We reveal that the correlation has been stronger during the COVID-19 pandemic. However, the relationship is not significant in the case of the Azerbaijani manat. The USD/AZN exchange rate has been stable since 2017, and the first phase of the COVID-19 pandemic has not caused a change in the exchange rate and a weakening of the Azerbaijani currency, despite significant drops in crude oil prices. Keywords: crude oil; exchange rate; international reserves; Dutch disease; Caspian Sea region; Azerbaijan; Kazakhstan; USD/AZN; USD/KZT; the COVID-19 pandemic 1. Introduction Crude oil plays a significant role in the economic development of countries all over the world. Oil price volatility affects the exchange rates of oil-importing and oil-exporting countries. There are two main approaches in theory on this issue. The first approach is related to the wealth transfer and assumes that the oil prices affect exchange rates through a reallocation of wealth between oil-exporting and oil-importing countries (Golub 1983; Krugman 1983). An oil prices increase leads to a wealth transfer from oil importers to oil exporters. It may trigger the depreciation of the currencies of oil importers and appreciation of the oil exporters’ currencies when the oil importer demand for oil is inelastic (Golub 1983). According to Krugman (1983), the relationship between oil prices and exchange rates may be either positive or negative. The sign depends on the countries’ trade balance, trade elasticities, capital flows, and the scale of speculative transactions in the financial market. The second approach is related to the transmission channel and assumes that the oil prices affect exchange rates through the terms of trade (Chen and Rogoff 2003;Chen and Chen 2007). Chen and Chen (2007) show that, for G7 countries’ exchange rates related to the U.S. dollar, the increase in oil prices decreases the real exchange rates in the long run. This stems from the fact that the increase in real oil prices may increase the prices of tradable goods relative to non-tradable goods in both the domestic countries and the USA. J. Risk Financial Manag. 2021,14, 431. https://doi.org/10.3390/jrfm14090431 https://www.mdpi.com/journal/jrfm J. Risk Financial Manag. 2021,14, 431 2 of 13 If the domestic country is more dependent on imported oil than the USA, the increase of tradables to non-tradables prices ratio may exceed the ratio in the USA and thus leads to depreciation of the domestic currency against the U.S. dollar. According to Adam Smith and David Ricardo, countries rich with natural resources perform economically better than those without natural resources (Letiche 1960). This theory was questioned in the late 1970s when the so-called Dutch desease was found to have occurred. The term was introduced in 1977 by “The Economist” and described the crisis in the Netherland driven by discoveries of vast gas deposits in the North Sea in 1959 (Badeeb et al. 2017). Corden and Neary (1982) and Corden (1984) were the first researchers that presented the Dutch disease phenomenon. The phenomenon applies to all negative effects driven by resource discoveries such as factor reallocations and deindustralisation (Magud and Sosa 2013) . Furthermore, Dutch disease is often characterised by the substantial appreciation of the resource-rich country’s currency. The paper focuses on the relationship between oil price and exchange rates of oil net-exporting countries. Lizardo and Mollick (2010) find that oil price changes significantly affect net oil-exporter currencies’ value. They show that the currencies of Canada, Mexico, and Russia appreciated against the U.S. dollar due to rising oil prices. Akram (2004) examines the relationship between oil prices and Norwegian exchange rates. He reports a significant impact of an oil price increase on the Norwegian currency’s appreciation. Chen et al. (2016), based on the monthly data for 16 OECD countries, show that the U.S. dollar exchange rates response to oil price shocks depend on whether oil price changes result from changes in supply or demand. Volkov and Yuhn (2016) find that the reaction of exchange rates to oil price changes depends more on financial market efficiency than on the relevance of oil revenues in the economy. Malik and Umar (2019) find that oil price shocks driven by demand changes and risk significantly affect exchange rates volatility. They conduct their research on currencies of major oil-exporting and oil-importing countries such as the Brazilian Real, Canadian Dollar, Chinese Yuan, Indian Rupee, Japanese Yen, Mexican Peso, and Russian Ruble. Moreover, they show that the strength of the relationship between oil price shocks and exchange rates has significantly increased since the global financial crisis. Reboredo (2012) shows that the relationship between oil prices and exchange rates is stronger for oil-exporting countries than oil-importing ones. Furthermore, he reveals that the co-movement is rather weak in pre-crisis periods and stronger in crisis periods. Driven by the global financial crisis (2008–2009), sharp oil price decreases in the international commodity market (2014–2015), the COVID-19 pandemic outbreak, and conflicts over oil prices, the oil revenue decreases have challenged the monetary system of oil-dependent countries. This paper focuses on the exchange rate of oil-exporting countries from the Caspian Sea region. The region is one of the oldest oil-producing areas in the world. Kazakhstan, Russia, Azerbaijan, Iran, and Turkmenistan share the shores of the Caspian Sea. The paper concentrates on Azerbaijan and Kazakhstan, the post-Soviet countries considered as some of the region’s most oil-dependent countries (Czech 2018). Bresser-Pereira (2013) argue that the exchange rate affects the national economy more than other macroeconomic variables such as the interest rate, profit rate, wage rate, and inflation. According to a Center for Economic and Social Development (CESD 2019) study on Azerbaijan, Russia, and Kazakhstan, oil price shocks in 2015 substantially hampered these countries’ economic growth, mainly through the exchange rate channel. Aleksandrova (2016) has noted that, during the collapse of oil prices, the economies of Azerbaijan, Kazakhstan, Turkmenistan, and Russia underperformed from a macroeconomic perspective compared to during the oil boom period. J. Risk Financial Manag. 2021,14, 431 3 of 13 Oil prices heavily impact the leading macroeconomic indicators in Azerbaijan (Humbatova et al. 2019;Niftiyev and Namazova 2020). Several papers study the Dutch disease phenomenon in Azerbaijan (Gahramanov and Fan 2002;Bayramov and Conway 2010; Hasanov 2013). Oil price fluctuations affect the Azerbaijani economy mainly through the exchange rate channel. Hasanov (2010) studies the impact of oil price changes on the value of Azerbaijani currency and finds that the real effective exchange rate of the Azerbaijani manat appreciates by approximately 0.7 percentage points when there is a one percentage point increase in oil prices. Like other post-Soviet resource-rich economies, Kazakhstan experienced a transition from a command economy and built its economy on extractive industriies since gaining independence. Kazakhstan achieved fast growth and improved competitiveness around 2010 (Macerinskiene and Sakhanova 2011), implemented substantial financial sector reforms (Akimov and Dollery 2008), and reduced poverty as government revenue increased thanks to increased oil extraction and mineral revenues (Agrawal 2007). Economic prosperity was resource-based, as mineral exports increased and high oil prices provided a golden opportunity to recover from the collapsed economy and policy errors (Pomfret 2005). As in the case of Azerbaijan, this led to a situation where oil prices extensively influenced the national economy, which indicates a dependency on external economic conditions and oil-led economic growth (Petrenko et al. 2016). Gronwald et al. (2009) documented a strong, negative, and significant relationship between oil price decreases and macroeconomic variables in the Kazakhstani economy. Similar results were found in the paper of Kose and Baimaganbetov (2015). As a result, the Dutch disease hypothesis in the Kazakhstani economy was intensively studied by Kutan and Wyzan (2005), Egert and Leonard (2008), and Aliev (2015). Dikkaya and Doyar (2017) illustrated a one-way causality from oil prices to the exchange rate and unidirectional causality from the exchange rate to GDP in Kazakhstan. In this case, the authors evaluated the freely floating exchange rate regime that the Kazakhstani government had started to apply since 2015 to optimise and stabilise monetary policy. Moreover, Kretzschmar and Nurmakhanova (2010) find that oil prices directly impacted the exchange rate in Kazakhstan but indirectly influenced real GDP. The paper aims to assess the relationship between Azerbaijani and Kazakhstani exchange rates, international reserves, and crude oil prices. Our contribution is that by applying the structural vector autoregressive models, we assessed the reaction of Azerbaijani and Kazakhstani exchange rates to oil price and international reserves changes. There is still little research in the literature on the effect of oil price changes on the Caspian Sea region’s most oil-dependent economies. In addition, similarly to Reboredo (2012), we used Spearman’s rank correlation coefficient to test the relationship between oil price and exchange rate changes during the pre-crisis and crisis period. However, in contrast to most similar studies, we conducted the research, not for the global financial crisis 2008–2009 but for another crisis period, i.e., the COVID-19 pandemic. The paper is organised as follows. Section 2describes methods and data applied in the study. Section 3presents research results. Section 4summarises and concludes. 2. Materials and Methods The initial graphical analysis was carried out based on the monthly data from January 2000 to May 2020. Monthly spot exchange rates of Kazakhstan tenge to U.S. dollar (USD/KZT) and Azerbaijani manat to U.S. dollar (USD/AZM) were applied as the country’s official nominal exchange rates data. Monthly WTI crude oil prices (in USD per bushel) were used as the oil prices in the model. Data were obtained from Refinitiv Datastream. Additionally, the analysis includes monthly data about the international reserves, excluding gold held by the central bank in Azerbaijan and Kazakhstan. Data on international reserves were obtained from Macrobond. We checked data stationarity based on the Augmented Dickey–Fuller test. J. Risk Financial Manag. 2021,14, 431 4 of 13 The quantitative analysis was carried out for monthly data from December 2015, i.e., when both countries officially changed the currency system to floating, until May 2020. The end of the time frame results from the availability of data on foreign exchange reserves. We applied the structural vector autoregressive (SVAR) model to study the relationship between the crude oil prices and the value of Azerbaijani and Kazakhstani currencies. Following Sims (1980), the SVAR models have been applied by many researchers. Aleisa and Diboo ˆ glu (2002), Benhabib et al. (2014), Brahmasrene et al. (2014), Obioma, and Eke Obioma and Charles (2015) among others use the SVAR model to study the relationship between oil prices and exchange rates of oil-exporting and oil-importing countries. The general form of the SVAR model is as follows: AXt=δ1Xt−1+δ2Xt−2+. . . +δpXt−p+But, (1) where Xt is a vector of model variables, A , δ , B are the model’s coefficients, and ut is a serially uncorrelated error term. The reduced form corresponding to the structural from (1) was obtained by pre multiplying with A−1: AXt=A−1δ1Xt−1+A−1δ2Xt−2+. . . +A−1δpXt−p+A−1But. (2) Model (2) can be represented by: Xt=D1Xt−1+D2Xt−2+. . . +DpXt−p+εt, (3) where Di=A−1δi , i= 1, . . . , p , and εt=A−1But . To identify the structural form parameters, restrictions were placed on the parameter matrices. It was assumed that the shocks affect a subset of variables directly within the current period, while another subset of variables is affected with a time lag only. Sims (1980) proposed recursive (triangular) identification, one of the most popular identification schemes. In the recursive identification, we assumed that the shocks enter the equations successively so that the additional shocks of the equation do not affect the variable explained by the first equation simultaneously (Lütkepohl and Krätzig 2004) . The order of endogenous variables determines the connection between innovations. Usually, we place the variable by the timeline of its occurrence. A variable that is assumed to occur first is set first in the vector of endogenous variables. Here, we placed oil prices as a first variable. Then, we assumed that oil shocks affect foreign exchange reserves, and USD/AZN and USD/KZT exchange rates are affected both by oil prices and international reserves shocks. The order of variables in the model was set based on Aizenman and Riera-Crichton (2008), Kasman and Ayhan (2008), Aizenman et al. (2012), and Habib et al. (2016). The assumptions can be summarised in Equation (4) that links the reduced-form errors to the structural shocks:   εOIL εIR εER  =  1 0 0 g(IR,OIL)1 0 g(ER,OIL)g(ER,IR)1    uOIL uIR uER  , (4) where OIL is the price of WTI crude oil, I.R. is the international reserves (excluding gold) of the central bank in Azerbaijan (AIR), and Kazakhstan (KIR) and E.R. are the USD/AZN and USD/KZT exchange rates for Azerbaijan and Kazakhstan, respectively. We studied the relationship between oil prices, international reserves, and the selected exchange rates based on the impulse response functions and forecast error variance decompositions. An impulse response function depicts the effect of an analysed shock of one variable on the other variables in the model. The forecast error variance decomposition analysis reflects the relative importance of each shock to one variable in affecting another one. We applied the Cholesky decomposition with OIL, I.R., and E.R. ordering. J. Risk Financial Manag. 2021,14, 431 5 of 13 Furthermore, we supplemented our study with a quantitative analysis of the correlation between oil price and exchange rate changes in the pre-crisis and crisis period. Similarly to Reboredo (2012), we used Spearman’s rank correlation coefficient to test the strength and significance of the analysed relationship. As a pre-crisis period, we set the period from December 2015 to December 2019. As a crisis period, we set the period of the first phase of the COVID-19 pandemic, i.e., from January 2020 to September 2020. We applied daily data on WTI crude oil changes and USD/AZN and USD/KZT spot exchange rate returns. Data came from Refinitiv Datastream. Due to the lack of daily data on the countries’ international reserves, the analysis related to the COVID-19 crisis period does not include that variable. 3. Results Oil is one of the most important energy sources in the world. Figure 1presents the monthly crude oil prices from January 2000 until May 2020. Oil prices have recorded significant increases between January 2007 and July 2008 and between January 2009 and May 2011. In the period 2011–2014, oil prices maintained a high level. The situation changed rapidly in 2014. Between August 2014 and January 2016, oil prices plunged more than 75 per cent. Oil exporters encountered many challenges after the 2014 oil price collapse. It led to broad-based and adverse effects on their economies. Most of them experienced a sharp fall in private consumption and investment. Nearly 70 per cent of oil-exporting emerging markets and developing countries registered a significant decline in the GDP growth rate (World Bank 2018). From February 2018 until January 2020, crude oil prices ranged from 33 to 74 USD/Bbl. Another considerable drop in oil prices could be observed in March and April 2020, which resulted from the COVID-19 pandemic. At that time, oil prices fell to around 20 USD/Bbl, reaching levels even lower than in 2009 and 2016. J. Risk Financial Manag. 2021, 14, x FOR PEER REVIEW 5 of 13 countries’ international reserves, the analysis related to the COVID-19 crisis period does not include that variable. 3. Results Oil is one of the most important energy sources in the world. Figure 1 presents the monthly crude oil prices from January 2000 until May 2020. Oil prices have recorded significant increases between January 2007 and July 2008 and between January 2009 and May 2011. In the period 2011–2014, oil prices maintained a high level. The situation changed rapidly in 2014. Between August 2014 and January 2016, oil prices plunged more than 75 per cent. Oil exporters encountered many challenges after the 2014 oil price collapse. It led to broad-based and adverse effects on their economies. Most of them experienced a sharp fall in private consumption and investment. Nearly 70 per cent of oil-exporting emerging markets and developing countries registered a significant decline in the GDP growth rate (World Bank 2018). From February 2018 until January 2020, crude oil prices ranged from 33 to 74 USD/Bbl. Another considerable drop in oil prices could be observed in March and April 2020, which resulted from the COVID-19 pandemic. At that time, oil prices fell to around 20 USD/Bbl, reaching levels even lower than in 2009 and 2016. Figure 1. Monthly crude WTI oil prices in the period from January 2000 until May 2020 (USD/Bbl). Oil prices changes have a significant impact on the oil-exporting countries economy. The paper shows the impact of oil price changes on oil-exporting countries’ exchange rates and their international reserves. It should be stressed that most of the adverse effects of oil price decline concern primary countries that highly depend on oil income. The paper focuses on Azerbaijan and Kazakhstan, the economies representing the former Soviet Union states of the Caspian Sea region and are among the world’s top 15 oil-dependent countries (Czech 2018). Figure 2 presents the volatility of USD/AZN and USD/KZT exchange rates from January 2000 until May 2020. It can be observed that in the period from January 2000 to January 2015, the value of the Kazakhstani tenge and Azerbaijani manat to the U.S. dollar remained at a stable level. The situation changed rapidly in 2015 when the Azerbaijani and Kazakhstani currencies depreciated against the USD dollar. The USD/KZT exchange rate increased from 180 to 334, and USD/AZN rose from 0.79 to 1.57. A sharp depreciation of the Kazakhstani tenge against the U.S. dollar in 2015 was the main reason to introduce a floating exchange rate and shift the monetary policy in Kazakhstan to an inflation-tar- geting regime. As far as Azerbaijan is concerned, it suffered double devaluation in 2015. Similarly to Kazakhstan, the central bank in Azerbaijan decided to switch the exchange rate regime from a currency peg to a floating exchange rate in December 2015. Figure 2 shows a sharp increase in the USD/AZN and USD/KZT exchange rates in 2015. It illustrates the sharp decline in the value of both countries’ currencies. 0 20 40 60 80 100 120 140 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Figure 1. Monthly crude WTI oil prices in the period from January 2000 until May 2020 (USD/Bbl). Oil prices changes have a significant impact on the oil-exporting countries economy. The paper shows the impact of oil price changes on oil-exporting countries’ exchange rates and their international reserves. It should be stressed that most of the adverse effects of oil price decline concern primary countries that highly depend on oil income. The paper focuses on Azerbaijan and Kazakhstan, the economies representing the former Soviet Union states of the Caspian Sea region and are among the world’s top 15 oil-dependent countries (Czech 2018). Figure 2presents the volatility of USD/AZN and USD/KZT exchange rates from January 2000 until May 2020. It can be observed that in the period from January 2000 to January 2015, the value of the Kazakhstani tenge and Azerbaijani manat to the U.S. dollar remained at a stable level. The situation changed rapidly in 2015 when the Azerbaijani and Kazakhstani currencies depreciated against the USD dollar. The USD/KZT exchange rate J. Risk Financial Manag. 2021,14, 431 6 of 13 increased from 180 to 334, and USD/AZN rose from 0.79 to 1.57. A sharp depreciation of the Kazakhstani tenge against the U.S. dollar in 2015 was the main reason to introduce a floating exchange rate and shift the monetary policy in Kazakhstan to an inflation-targeting regime. As far as Azerbaijan is concerned, it suffered double devaluation in 2015. Similarly to Kazakhstan, the central bank in Azerbaijan decided to switch the exchange rate regime from a currency peg to a floating exchange rate in December 2015. Figure 2shows a sharp increase in the USD/AZN and USD/KZT exchange rates in 2015. It illustrates the sharp decline in the value of both countries’ currencies. Moreover, Figure 2indicates that the USD/KZT exchange rate has become more volatile since the transition to a floating currency regime; however, in the case of USD/AZN, the exchange rate still seems to be stable, mainly since 2017. Azerbaijan is not, however, a unique case. Levy-Yeyati and Sturzenegger (2005) point out that many countries that theoretically have a floating exchange rate often intervene in the foreign exchange market. In practice, their exchange rate resembles a fixed exchange rate, not a floating one. J. Risk Financial Manag. 2021, 14, x FOR PEER REVIEW 6 of 13 Moreover, Figure 2 indicates that the USD/KZT exchange rate has become more volatile since the transition to a floating currency regime; however, in the case of USD/AZN, the exchange rate still seems to be stable, mainly since 2017. Azerbaijan is not, however, a unique case. Levy-Yeyati and Sturzenegger (2005) point out that many countries that theoretically have a floating exchange rate often intervene in the foreign exchange market. In practice, their exchange rate resembles a fixed exchange rate, not a floating one. Figure 2. Monthly USD/AZN and USD/KZT exchange rates in the period from January 2000 until May 2020. The central bank of a country with a fixed currency regime must constantly monitor the supply and demand of the currency and manage the cash flow to avoid spikes in the demand or supply of its currency. These spikes can cause the currency to deviate from the pegged price. This means that the central bank must maintain large foreign exchange reserves to counteract exchange rate fluctuations. Figure 3 depicts the changes in international reserves excluding gold held by the central bank in Azerbaijan and Kazakhstan. In Azerbaijan, the substantial decrease in the reserves is observed in 2014 and 2015, when the falling oil prices triggered a deficit in the balance of payments of Azerbaijan and led to a substantial decrease in the international reserves, which in turn brought about a currency devaluation and a move to a floating exchange rate. From 2016 to 2020, the level of foreign exchange reserves, similarly to the Azerbaijani manat exchange rate, was characterised by low volatility. According to Levy-Yeyati and Sturzenegger (2005), if neither the nominal exchange rate nor reserves move, then the exchange rate regime that the country is implementing is not apparent. As far as Kazakhstan is concerned, both the exchange rates and reserves have been quite volatile since 2016. This may be associated with the dirty floating exchange rate regime (Levy-Yeyati and Sturzenegger 2005). 0.6 0.8 1 1.2 1.4 1.6 1.8 100 150 200 250 300 350 400 450 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 USD/AZN USD/KZT USD/KZT USD/AZN Figure 2. Monthly USD/AZN and USD/KZT exchange rates in the period from January 2000 until May 2020. The central bank of a country with a fixed currency regime must constantly monitor the supply and demand of the currency and manage the cash flow to avoid spikes in the demand or supply of its currency. These spikes can cause the currency to deviate from the pegged price. This means that the central bank must maintain large foreign exchange reserves to counteract exchange rate fluctuations. Figure 3depicts the changes in international reserves excluding gold held by the central bank in Azerbaijan and Kazakhstan. In Azerbaijan, the substantial decrease in the reserves is observed in 2014 and 2015, when the falling oil prices triggered a deficit in the balance of payments of Azerbaijan and led to a substantial decrease in the international reserves, which in turn brought about a currency devaluation and a move to a floating exchange rate. From 2016 to 2020, the level of foreign exchange reserves, similarly to the Azerbaijani manat exchange rate, was characterised by low volatility. According to Levy-Yeyati and Sturzenegger (2005), if neither the nominal exchange rate nor reserves move, then the exchange rate regime that the country is implementing is not apparent. As far as Kazakhstan is concerned, both the exchange rates and reserves have been quite volatile since 2016. This may be associated with the dirty floating exchange rate regime (Levy-Yeyati and Sturzenegger 2005). J. Risk Financial Manag. 2021,14, 431 7 of 13 J. Risk Financial Manag. 2021, 14, x FOR PEER REVIEW 7 of 13 Figure 3. Monthly international reserves excluding gold in Azerbaijan and Kazakhstan, in the period from January 2000 until May 2020 (in billion U.S. dollars). The paper applies the SVAR model to study the relationship between oil prices (OIL), international reserves excluding gold held by the central bank in Azerbaijan (AIR) and Kazakhstan (KIR) and USD/AZN and USD/KZT exchange rates. The model is built based on monthly data from December 2015, i.e., when both countries officially changed the currency system to floating, until May 2020. To avoid any spurious inferences, the variables were tested for stationarity. The most frequently applied unit root test is based on the augmented Dickey–Fuller (ADF) test (Dickey and Fuller 1979). The ADF test results are presented in Table 1. Table 1. Augmented Dickey–Fuller test results. Variables Intercept Intercept and Trend Level First Difference Level First Difference USD/AZN −0.04 −14.77 *** −1.05 −14.84 *** USD/KZT 1.37 −11.67 *** −0.53 −11.99 *** AIR −1.52 −6.02 *** −1.83 −6.02 *** KIR −1.52 −14.85 *** −0.31 −15.06 *** OIL −2.27 −12.46 *** −2.05 −12.49 *** Note: *** H0 of a unit root is rejected at the 1%, ** 5%, and * 10% significance level. The Augmented Dickey–Fuller test was applied to the level and first difference. The ADF tests indicate that the hypothesis of a non-stationary level cannot be rejected for all analysed series. However, the results for the first difference show that the null hypothesis of a unit root is rejected at the 1% significance level. Based on the ADF test, the analysed series needs to be differenced once to generate a stationary series (are integrated of the order one). Thus, all data are expressed as the first difference of their logarithm value. The optimal lag length of the VAR estimation was selected based on the Akaike Information Criterion (AIC). The VAR Model (4) for USD/AZN obtained the lowest AIC for 3 lags (AIC = −7.26). The VAR model for Kazakhstani exchange rates has the lowest AIC for 1 lag (AIC = −8.45). The Lagrange Multiplier test results suggest that the residuals are not serially correlated. In the VAR model for the Azerbaijani exchange rate, the L.M. statistics for three lag equal 9.20 with the corresponding p-value of 0.42. In the VAR model for the Kazakhstani exchange rate, the L.M. statistics for one lag equals 8.62 with the corresponding p-value of 0.47. Therefore, based on the following results, we cannot reject the null hypothesis that there is no serial correlation in the residuals. Moreover, the estimated 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Azerbaijani international reserves Kazakhstani international reserves Kazakhstan IR Azerbaijan IR Figure 3. Monthly international reserves excluding gold in Azerbaijan and Kazakhstan, in the period from January 2000 until May 2020 (in billion U.S. dollars). The paper applies the SVAR model to study the relationship between oil prices (OIL), international reserves excluding gold held by the central bank in Azerbaijan (AIR) and Kazakhstan (KIR) and USD/AZN and USD/KZT exchange rates. The model is built based on monthly data from December 2015, i.e., when both countries officially changed the currency system to floating, until May 2020. To avoid any spurious inferences, the variables were tested for stationarity. The most frequently applied unit root test is based on the augmented Dickey–Fuller (ADF) test (Dickey and Fuller 1979). The ADF test results are presented in Table 1. Table 1. Augmented Dickey–Fuller test results. Variables Intercept Intercept and Trend Level First Difference Level First Difference USD/AZN −0.04 −14.77 *** −1.05 −14.84 *** USD/KZT 1.37 −11.67 *** −0.53 −11.99 *** AIR −1.52 −6.02 *** −1.83 −6.02 *** KIR −1.52 −14.85 *** −0.31 −15.06 *** OIL −2.27 −12.46 *** −2.05 −12.49 *** Note: *** H0of a unit root is rejected at the 1%, ** 5%, and * 10% significance level. The Augmented Dickey–Fuller test was applied to the level and first difference. The ADF tests indicate that the hypothesis of a non-stationary level cannot be rejected for all analysed series. However, the results for the first difference show that the null hypothesis of a unit root is rejected at the 1% significance level. Based on the ADF test, the analysed series needs to be differenced once to generate a stationary series (are integrated of the order one). Thus, all data are expressed as the first difference of their logarithm value. The optimal lag length of the VAR estimation was selected based on the Akaike Information Criterion (AIC). The VAR Model (4) for USD/AZN obtained the lowest AIC for 3 lags (AIC = − 7.26). The VAR model for Kazakhstani exchange rates has the lowest AIC for 1 lag (AIC = − 8.45). The Lagrange Multiplier test results suggest that the residuals are not serially correlated. In the VAR model for the Azerbaijani exchange rate, the L.M. statistics for three lag equal 9.20 with the corresponding p-value of 0.42. In the VAR model for the Kazakhstani exchange rate, the L.M. statistics for one lag equals 8.62 with the corresponding p-value of 0.47. Therefore, based on the following results, we cannot reject the null hypothesis that there is no serial correlation in the residuals. Moreover, the J. Risk Financial Manag. 2021,14, 431 8 of 13 estimated model is stable. All of the roots have a modulus of less than one and lie inside the unit A.R. roots circle. We applied the impulse response functions to assess how and to what extent oil price shocks influence Azerbaijani and Kazakhstani international reserves and the USD/AZN and USD/KZT exchange rates. Figure 4displays the Azerbaijani and Kazakhstani international reserves’ response to a positive oil price shock over the 30-month range. The figure includes a black line that depicts the mean reaction function and red lines that reflect the confidence interval of two standard deviations around the mean. The results suggest that oil price increases lead to a change of international reserves (excluding gold) both in Azerbaijan and in Kazakhstan. As far as the Azerbaijani international reserves are concerned, their reaction to the oil price shock is positive and significant from the second month after the shock. For the first month, the confidence interval consists of zero. The reaction of Kazakhstani foreign exchange reserves is statistically insignificant in the entire analysed period after the shock. The standard error band crosses the zero axes. J. Risk Financial Manag. 2021, 14, x FOR PEER REVIEW 8 of 13 model is stable. All of the roots have a modulus of less than one and lie inside the unit A.R. roots circle. We applied the impulse response functions to assess how and to what extent oil price shocks influence Azerbaijani and Kazakhstani international reserves and the USD/AZN and USD/KZT exchange rates. Figure 4 displays the Azerbaijani and Kazakhstani international reserves’ response to a positive oil price shock over the 30-month range. The figure includes a black line that depicts the mean reaction function and red lines that reflect the confidence interval of two standard deviations around the mean. The results suggest that oil price increases lead to a change of international reserves (excluding gold) both in Azerbaijan and in Kazakhstan. As far as the Azerbaijani international reserves are concerned, their reaction to the oil price shock is positive and significant from the second month after the shock. For the first month, the confidence interval consists of zero. The reaction of Kazakhstani foreign exchange reserves is statistically insignificant in the entire analysed period after the shock. The standard error band crosses the zero axes. - .02 .00 .02 .04 .06 .08 .10 .12 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Azerbaijani international reserves -.025 -.020 -.015 -.010 -.005 .000 .005 .010 .015 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Kazakhstani international reserves Figure 4. The impulse response function of Azerbaijani and Kazakhstani international reserves to oil price shocks. Figure 5 presents the USD/AZN and USD/KZT exchange rates response to positive oil price shock over the 30-month range with ± 2 standard error bands. The results suggest that the oil price increase is associated with a decrease in USD/AZN and USD/KZT exchange rates. The decline occurs in the first four to five months, stabilising at lower than the initial level. It means that the rise in oil prices leads to both Azerbaijani manat and Kazakhstani tenge appreciation against the U.S. dollar. In the case of the USD/KZT exchange rate, its response to oil price shock is statistically significant, while the reaction of USD/AZN is significant from the fifth month after the shock. During the first four months after an oil price shock, the confidence interval contains zero. -.035 -.030 -.025 -.020 -.015 -.010 -.005 .000 .005 .010 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 USD/AZN -.035 -.030 -.025 -.020 -.015 -.010 -.005 .000 246810 12 14 16 18 20 22 24 26 28 30 USD/KZT Figure 5. Impulse response function of USD/AZN and USD/KZT exchange rates to oil price shocks. The analysis of impulse response functions shows that the rise in oil prices leads to the appreciation of both Kazakhstani and Azerbaijani currencies against the U.S. dollar. Figure 4. The impulse response function of Azerbaijani and Kazakhstani international reserves to oil price shocks. Figure 5presents the USD/AZN and USD/KZT exchange rates response to positive oil price shock over the 30-month range with ± 2 standard error bands. The results suggest that the oil price increase is associated with a decrease in USD/AZN and USD/KZT exchange rates. The decline occurs in the first four to five months, stabilising at lower than the initial level. It means that the rise in oil prices leads to both Azerbaijani manat and Kazakhstani tenge appreciation against the U.S. dollar. In the case of the USD/KZT exchange rate, its response to oil price shock is statistically significant, while the reaction of USD/AZN is significant from the fifth month after the shock. During the first four months after an oil price shock, the confidence interval contains zero. J. Risk Financial Manag. 2021, 14, x FOR PEER REVIEW 8 of 13 model is stable. All of the roots have a modulus of less than one and lie inside the unit A.R. roots circle. We applied the impulse response functions to assess how and to what extent oil price shocks influence Azerbaijani and Kazakhstani international reserves and the USD/AZN and USD/KZT exchange rates. Figure 4 displays the Azerbaijani and Kazakhstani international reserves’ response to a positive oil price shock over the 30-month range. The figure includes a black line that depicts the mean reaction function and red lines that reflect the confidence interval of two standard deviations around the mean. The results suggest that oil price increases lead to a change of international reserves (excluding gold) both in Azerbaijan and in Kazakhstan. As far as the Azerbaijani international reserves are concerned, their reaction to the oil price shock is positive and significant from the second month after the shock. For the first month, the confidence interval consists of zero. The reaction of Kazakhstani foreign exchange reserves is statistically insignificant in the entire analysed period after the shock. The standard error band crosses the zero axes. - .02 .00 .02 .04 .06 .08 .10 .12 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Azerbaijani international reserves -.025 -.020 -.015 -.010 -.005 .000 .005 .010 .015 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Kazakhstani international reserves Figure 4. The impulse response function of Azerbaijani and Kazakhstani international reserves to oil price shocks. Figure 5 presents the USD/AZN and USD/KZT exchange rates response to positive oil price shock over the 30-month range with ± 2 standard error bands. The results suggest that the oil price increase is associated with a decrease in USD/AZN and USD/KZT exchange rates. The decline occurs in the first four to five months, stabilising at lower than the initial level. It means that the rise in oil prices leads to both Azerbaijani manat and Kazakhstani tenge appreciation against the U.S. dollar. In the case of the USD/KZT exchange rate, its response to oil price shock is statistically significant, while the reaction of USD/AZN is significant from the fifth month after the shock. During the first four months after an oil price shock, the confidence interval contains zero. -.035 -.030 -.025 -.020 -.015 -.010 -.005 .000 .005 .010 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 USD/AZN -.035 -.030 -.025 -.020 -.015 -.010 -.005 .000 246810 12 14 16 18 20 22 24 26 28 30 USD/KZT Figure 5. Impulse response function of USD/AZN and USD/KZT exchange rates to oil price shocks. The analysis of impulse response functions shows that the rise in oil prices leads to the appreciation of both Kazakhstani and Azerbaijani currencies against the U.S. dollar. Figure 5. Impulse response function of USD/AZN and USD/KZT exchange rates to oil price shocks.