Dynamic impact of foreign exchange trading volume on foreign exchange volatility

Kang, Jong Woo; Cabaero, Carlos Working Paper Dynamic impact of foreign exchange trading volume on foreign exchange volatility ADB Economics Working Paper Series, No. 768 Provided in Cooperation with: Asian Development Bank (ADB), Manila Suggested Citation: Kang, Jong Woo; Cabaero, Carlos (2025) : Dynamic impact of foreign exchange trading volume on foreign exchange volatility, ADB Economics Working Paper Series, No. 768, Asian Development Bank (ADB), Manila, https://doi.org/10.22617/WPS250025-2 This Version is available at: https://hdl.handle.net/10419/322304 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/3.0/igo/ ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org DYNAMIC IMPACT OF FOREIGN EXCHANGE TRADING VOLUME ON FOREIGN EXCHANGE VOLATILITY Jong Woo Kang and Carlos Cabaero ADB ECONOMICS WORKING PAPER SERIES NO. 768 February 2025 Dynamic Impact of Foreign Exchange Trading Volume on Foreign Exchange Volatility Foreign exchange (FX) trading volume is a key factor in volatility. This paper investigates the effect of trading volume on volatility using high-frequency data. Estimation results from econometric models reveal a significant impact of third-party trade volumes on the volatilities of original currency pairs. Though the United States dollar (USD) exerts sizeable effect through third-party channels, currency pairs without USD linkages also have impact, calling renewed attention to utilizing regional cooperation in mitigating volatility as compared with major FX trading partners. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 69 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. ASIAN DEVELOPMENT BANK The ADB Economics Working Paper Series presents research in progress to elicit comments and encourage debate on development issues in Asia and the Pacific. The views expressed are those of the authors and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they represent. ADB Economics Working Paper Series Dynamic Impact of Foreign Exchange Trading Volume on Foreign Exchange Volatility Jong Woo Kang and Carlos Cabaero No. 768 | February 2025 Jong Woo Kang ([email protected]) is the director of Regional Cooperation and Integration Division and Carlos Cabaero ([email protected]) is a consultant at the Economic Research and Development Impact Department, Asian Development Bank. Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) © 2025 Asian Development Bank 6 ADB Avenue, Mandaluyong City, 1550 Metro Manila, Philippines Tel +63 2 8632 4444; Fax +63 2 8636 2444 www.adb.org Some rights reserved. Published in 2025. ISSN 2313-6537 (print), 2313-6545 (PDF) Publication Stock No. WPS250025-2 DOI: http://dx.doi.org/10.22617/WPS250025-2 The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies ofthe Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The mention of specific companies or products of manufacturers does not imply that they are endorsed or recommended by ADB in preference to others of a similar nature that are not mentioned. By making any designation of or reference to a particular territory or geographic area inthis document, ADB does not intend to make any judgments as to the legal or other status of any territory or area. This publication is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) https://creativecommons.org/licenses/by/3.0/igo/. By using the content of this publication, you agree to be bound bytheterms of this license. For attribution, translations, adaptations, and permissions, please read the provisions andterms of use at https://www.adb.org/terms-use#openaccess. This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributed toanother source, please contact the copyright owner or publisher of that source for permission to reproduce it. ADB cannot be held liable for any claims that arise as a result of your use of the material. Please contact [email protected] if you have questions or comments with respect to content, or if you wish toobtain copyright permission for your intended use that does not fall within these terms, or for permission to use theADB logo. Corrigenda to ADB publications may be found at http://www.adb.org/publications/corrigenda. ABSTRACT Foreign exchange (FX) trading volume is a key factor in exchange rate volatility. Given the important role of volatility in economic growth and stability, this paper investigates the dynamic nature of exchange trading volume on exchange rate volatility using hourly high-frequency data. The estimation results from ordinary least squares, fixed effects and the general autoregressive conditional heteroskedasticity model point to a significant impact of third-party foreign exchange trade volumes on the FX volatilities of original currency pairs. The United States dollar (USD), as the dominant currency, exerts sizeable effect through this third-party channel and the magnitude of the foreign exchange trading volume turns out to be a crucial factor to this effect. However, third-party currency pairs without USD linkages also exert non-negligible impact, calling for renewed attention to the effectiveness of regional financial cooperation in mitigating exchange rate volatility as compared with major foreign exchange trading partners, not only through direct transaction mechanisms but through third party currency channels. Keywords: FX volatility, third party channel, GARCH model JEL codes: F31, G15, G18 I. Introduction The impact of exchange rate movement on economic growth, development, and stability is welldocumented. Much of the literature posits that exchange rates affect international trade performance, including through export/import competitiveness, volumes, and prices, as well as businesses’ external financing costs. Consequently, changes in exchange rates affect foreign and domestic consumption, productivity, and investment. Given these significant impacts, private and public entities alike must carefully monitor and prepare for exchange rate movements. The literature has shown the impact of exchange rates on multiple development and macroeconomic indicators. Eichengreen (2007) frames foreign exchange rates as a vital facilitating condition for economic growth. Development experiences in high-growth economies, such as in East Asia, and developing economies demonstrate that competitive exchange rates are critical in jumpstarting growth. An efficient exchange rate mechanism encourages efficient redeployment of resources to productive sectors, thus unlocking gains in productivity. Studies covering multiple economies across the globe affirm this, particularly that undervaluation of a currency against foreign counterparts is often accompanied by gross domestic product (GDP) growth (Rodrik 2008, Seraj and Coskuner 2021). In a study focusing on India, Shaik and Rao (2020) state that the depreciation of the Indian rupee is associated with an increase in the country’s foreign exchange reserves and in real GDP. Zhao (2020) shows that exchange rates have a direct effect on the prices and costs of commodities, also impacting exports to foreign markets. Further research shows that exchange rates impact productivity and foreign tourism. Fluctuations in the exchange rate may influence economic policy, particularly in economies adopting inflation-targeting regimes, while harming economic growth. Exchange rates have significant impact on international trade, as noted. Exchange rates have a sizable effect on the import and export costs of products. When an economy’s currency is valued higher, products from foreign markets become cheaper, thus encouraging greater importation. Conversely, undervalued currencies facilitate lower prices for an economy’s commodities and lead to greater product exportation. Thus, movements in the exchange rate also affect an economy’s trade balance; a higher exchange rate moves towards a negative trade account balance, while a lower exchange rate leads to a positive balance, although the growing complexities associated with deepening global value chains tend to compound this linear relationship. The strength of this relationship has been widely studied in the literature, with varying results. Research on developed and developing countries (Kang 2016) after the global financial crisis show that the effect of currency devaluation on export growth postcrisis has not been as 2 strong as before the crisis. Further research posits that an increase in the number of economies that engage in deeper currency devaluation may lead to sluggish growth in international trade. Thus, great importance has been placed in understanding exchange rate movements, particularly through the concept of exchange rate volatility. Exchange rate volatility is defined as the risk associated with unexpected movements in exchange rates (Ozturk 2006). The representative indicator of exchange rate volatility is the degree of variance in an exchange rate over a certain period. Price movements originate from an events-based approach, such as political-economic news and announcements, that informs decisions of economic actors, both public and private. Other factors like comparative inflation and interest rate differentials between economies likewise lead to exchange rate volatility. Aside from macroeconomics and international trade, exchange rate volatility is consequential for firms and traders, which could be exposed to sizeable exchange rate risks in their financing costs and financial management. Volatile foreign exchange markets lead to greater market uncertainty, which impacts costs and revenues for firms and in turn informs hedging and investment strategies. Furthermore, a volatile exchange rate increases uncertainty in the foreign exchange market and discourages risk-averse traders from engaging in investments, leading to changes in investor portfolio flows (Flores-Sosa, Aviles-Ochoa, and Merigo 2023). Given the influence of exchange rates on economic growth and trade, maintaining a stable and predictable exchange rate has been a constant priority of sovereign financial authorities across the globe. Both the sources and implications of exchange rate volatility have been key areas of economic research. Findings generally show that higher exchange rate volatility leads to higher costs for risk-averse traders, as well as lower foreign trade. This is due to changes in the value of the exchange rate upon the agreement of a contract versus its actual payment and implementation. When exchange rates become volatile, uncertainty rises in predicting the costs and thereby the profits from transactions, which disincentivizes trade (Hooper and Kolhagen 1978). Despite this, other researchers have been less definite about the impact of exchange rate volatility on international trade. De Grauwe (1988), for example, posits that dominance of income effects over substitution effects may lead to a positive relationship between volatility and trade. In this theory, an increase in exchange rate volatility could raise the marginal utility of export revenue in the eyes of sufficiently risk-averse exporters and could induce increased exports. De Grauwe thus suggests that the effect of exchange rate volatility depends on the degree of risk aversion of market players. 3 Numerous efforts have been made to study exchange rate volatility. As it is not a directly observable phenomenon, extensive research has been carried out to predict movements in the exchange rates, as well as to identify causes, and possible indicators of exchange rate volatility. The literature identifies economic fundamentals, such as inflation, interest rates, and balance of payments as sources of exchange rate volatility, especially as these factors themselves have become more volatile since the 1980s. Furthermore, factors such as capital account liberalization, technological innovation, and currency speculation all contribute to increased cross-border flows and trade volumes, adding to exchange rate volatility (Hook and Boon 2000). One of the key factors affecting exchange rate volatility is foreign exchange trading volume, which is used as a measure of the state of the foreign exchange market. Foreign exchange trading is essential to engaging in international trade as it allows traders and firms alike to convert domestic currency into foreign currency and vice versa to be able to transact with external markets. Traders often study movement in the forex market and carry out trades based on their valuation of various currencies to make a profit. Thus, fluctuations in forex trading volume are also used to study and predicate the degree of foreign exchange volatility between currencies. This places a clear impetus in understanding the relationship between foreign exchange volatility and trading volumes. Foreign exchange trades can also be influenced by the firm’s motives to minimize losses from exchange rate volatility in determining the timing of exchanges for foreign borrowing or repayment, and the incentives to hedge against exchange rate volatility risks. The relationship between foreign exchange volatility and foreign exchange trading volumes has been explored in the literature. Foreign exchange trading volume is regarded as a proxy for unobservable market conditions, such as relative liquidity and privately informed trading (Gargano, Riddiough, and Sarno 2018). Volatility tends to move in conjunction with trading volumes, in that a steep increase in trading volumes often coincides with more volatile foreign exchange currencies (Figure on page 4). Various theoretical explanations aim to explain this. Copeland (1976, 1997) presented the model of “sequential information arrival,” wherein trading participants react to information on the financial market individually. Their reaction to the arrival of the news thereby shifts their demand curve for a particular currency. These trades then act as a proxy for traders’ changing demand for a particular currency, and thus coincide with increased volatility in the foreign exchange market. Another explanation for this phenomenon is the “mixture of distributions hypothesis” proposed by Clark (1973). Under this theory, volatility and volume are determined by a common, unobservable factor that reflects the arrival of new information in the foreign exchange trading market. How traders internalize this information changes the pricing of a particular currency, thus encouraging a higher number of trades. These trades therefore signify 4 disagreements between traders on the pricing of a particular foreign exchange currency and more volatile price movement. Figure: USD-JPY Volatility and Trading Volumes for May 2023 USD-JPY = US dollar-Japanese yen currency pair. Notes: Values above are the daily average of the hourly volatility rates. The hourly volatility rates are calculated as the absolute sum of the 5-minute interval price change of a currency pair within an hour, reflected as a percentage of the exchange rate. Sources: Asian Development Bank calculations using data from Bloomberg and the CLS FX databases. Sensoy and Serdengecty (2019) explore the viability of the Mixed Distribution hypothesis by investigating the relationship between USD-TRY (Turkish lira) volatility and foreign exchange trade volumes by currency trade and counterpart. Their results used a generalized method of moments framework to establish a positive contemporaneous relationship between USD-TRY exchange volatility and trade volumes in the spot market. The research further showed that the dispersion of trader beliefs on the future USD-TRY exchange rate significantly increases the positive relationship between volatility and trading volume, strengthening the hypothesis that the joint movement of two variables are explained by trade uncertainty and disagreement in foreign exchange rate predictions. Galati (2000) likewise explores the effect of local currencies in developing economies using ordinary least squares (OLS) regression with a general autoregressive conditional heteroskedasticity (GARCH) component, distinguishing expected and unexpected changes in trade volume. The results showed a positive significant relationship between foreign exchange volatility and volumes for four out of the six currencies examined. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0 2 4 6 8 10 12 Billions (USD-JPY) Average daily volatility Average of volatility Average of volume, USD-JPY 11 ln _𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖1 = 𝛽𝛽𝑜𝑜+ 𝛽𝛽1ln _𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑒𝑒𝑖𝑖1𝑡𝑡 + 𝛽𝛽2ln _𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑒𝑒𝑖𝑖2𝑡𝑡 + 𝛽𝛽3ln _𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑒𝑒𝑖𝑖3𝑡𝑡 + 𝛽𝛽4𝑣𝑣𝑣𝑣𝑙𝑙𝑙𝑙𝑒𝑒𝑑𝑑𝑣𝑣𝑜𝑜𝑣𝑣𝑣𝑣𝑡𝑡𝑖𝑖𝑣𝑣𝑖𝑖𝑡𝑡𝑣𝑣𝑖𝑖1𝑡𝑡−1 +𝛽𝛽5𝑣𝑣𝑣𝑣𝑙𝑙𝑙𝑙𝑒𝑒𝑑𝑑𝑣𝑣𝑜𝑜𝑣𝑣𝑣𝑣𝑡𝑡𝑖𝑖𝑣𝑣𝑖𝑖𝑡𝑡𝑣𝑣𝑖𝑖1𝑡𝑡−2 + 𝛽𝛽6𝑣𝑣𝑣𝑣𝑒𝑒𝑡𝑡𝑑𝑑𝑣𝑣𝑣𝑣 + 𝛽𝛽7𝑤𝑤𝑒𝑒𝑑𝑑𝑤𝑤𝑒𝑒𝑡𝑡𝑑𝑑𝑣𝑣𝑣𝑣 + 𝛽𝛽8𝑣𝑣ℎ𝑣𝑣𝑢𝑢𝑡𝑡𝑑𝑑𝑣𝑣𝑣𝑣 + 𝛽𝛽9𝑓𝑓𝑢𝑢𝑣𝑣𝑑𝑑𝑣𝑣𝑣𝑣 +𝜀𝜀𝑖𝑖𝑡𝑡 where i and t correspond to the currency pair and hour, respectively. We then use a fixed effects model, wherein the hour of each trading day has an unobserved effect on its respective foreign exchange volatility. This model is used to observe the relationship between volatility and trading volume considering any effects brought about by the hours within a trading day. ln _𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖𝑡𝑡 = 𝛽𝛽𝑜𝑜+ 𝛽𝛽1ln _𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑒𝑒𝑖𝑖1𝑡𝑡 + 𝛽𝛽2ln_𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑒𝑒𝑖𝑖2𝑡𝑡 + 𝛽𝛽3ln_𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑒𝑒𝑖𝑖3𝑡𝑡 + 𝛽𝛽4𝑣𝑣𝑣𝑣𝑙𝑙𝑙𝑙𝑒𝑒𝑑𝑑_𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖1𝑡𝑡−1 +𝛽𝛽5𝑣𝑣𝑣𝑣𝑙𝑙𝑙𝑙𝑒𝑒𝑑𝑑_𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖1𝑡𝑡−2 + 𝛽𝛽6𝑣𝑣𝑣𝑣𝑒𝑒𝑡𝑡𝑑𝑑𝑣𝑣𝑣𝑣 + 𝛽𝛽7𝑤𝑤𝑒𝑒𝑑𝑑𝑤𝑤𝑒𝑒𝑡𝑡𝑑𝑑𝑣𝑣𝑣𝑣 + 𝛽𝛽8𝑣𝑣ℎ𝑣𝑣𝑢𝑢𝑡𝑡𝑑𝑑𝑣𝑣𝑣𝑣 + 𝛽𝛽9𝑓𝑓𝑢𝑢𝑣𝑣𝑑𝑑𝑣𝑣𝑣𝑣 +𝛽𝛽10ℎ𝑣𝑣𝑣𝑣𝑢𝑢𝑡𝑡 + 𝜀𝜀𝑡𝑡 where i and t correspond to the currency pair and hour, respectively. We further adopt the model by Epaphra (2017), wherein the ARCH and GARCH models are used to identify volatility clustered in foreign exchange rates. It is well-established in the literature that foreign exchange rates tend to behave like financial data, and thus are also examinable by models that aim to account for time-related effects in volatility. For this analysis, the adopted GARCH model employs the following estimation procedure. First, volatility and trade volumes are log-differenced to satisfy the condition of non-stationarity, as confirmed by the Augmented Dickey Fuller-Test. Second, the basic regression equation is fitted as follows for each currency pair: 𝑑𝑑𝑣𝑣𝑤𝑤_𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑖𝑖𝑡𝑡 = 𝛽𝛽𝑜𝑜+ 𝛽𝛽2dln_tradevol𝑖𝑖1𝑡𝑡 + 𝛽𝛽3dln_tradevol𝑖𝑖2𝑡𝑡 + 𝛽𝛽4dln_tradevol𝑖𝑖3𝑡𝑡 +𝜀𝜀𝑖𝑖𝑡𝑡 where i and t correspond to the currency pair and hour, respectively. The model consists of the log-differenced trade volumes within each tripartite relationship as regressors. Following the Box-Jenkins method for specifying models, the correlogram, Autocorrelation (ACF) and Partial Autocorrelation (PACF) of the basic regressions are tested to identify the appropriate amount of Autoregressive (AR) and Moving Average (MA) components to add to the equation. The following tests are vital to ensure that there is no serial autocorrelation in the time series data, that is, that there are no correlations between the error terms of different periods within the time series. Autoregressive Moving Average (ARMA) components are added and removed based on repeated testing of the ACF and PACF of the model. When the ARMA 12 model is satisfied, the Autoregressive Conditional Heteroskedasticity (ARCH) test is applied on the squared residuals of the model to see if there are ARCH effects in the data. The presence of an ARCH effect in the ARMA model would show that there is a tendency for the data to exhibit periods of high volatility, then followed by even higher levels of volatility, and vice-versa. This means that the model also needs to be controlled for its conditional variance. For this analysis, the GARCH (general ARCH) model is utilized to capture this conditional variance. The GARCH model is an extension of the ARCH model, which models the conditional variance of time series data as a linear function of past squared observations. GARCH builds upon this by allowing more flexibility by modelling the conditional variance as a linear function of past squared observations, as well as past conditional variances. Thus, the GARCH model adds in both short-term and longterm memory in the volatility analysis. The standard GARCH (1,1) model is then fitted to the equation, and the ARMA components are modified until the residuals are not autocorrelated through the ACF and PACF test. The final equation for the model is as follows: where i and t respectively correspond to the currency pair and hour, x refers to the number of lags, z refers to a white noise process with mean zero and variance 1, and αo refers to the constant in the variance. III. Estimation Results 3.1. OLS estimation The OLS estimation firmly establishes the significant positive impact of a currency pair’s trade volumes on the currency pair’s foreign exchange volatility. The model also demonstrates the potency of USD-related trade volume as a determinant of volatility across multiple currency pairs. Table 5 models foreign exchange volatility in the USD-JPY-AUD trilateral, which shows that bilateral exchange rate volatility is significantly affected by the trading volume of corresponding currency pairs. A 1% increase in AUD-USD trading volume leads to a 0.3% increase in AUD-USD 13 exchange rate volatility, the same in AUD-JPY trading volumes leads to a 0.09% increase in the volatility of corresponding currency exchange rate, and the same in USD-JPY leads to a 0.33% increase in USD-JPY exchange rate volatility. AUD-JPY volatility, however, is also affected by trades of non-corresponding currencies, i.e., the trading volumes of AUD-USD and USD-J P Y. Interestingly, AUD-JPY volatility is more strongly influenced by AUD-USD trading volume. These phenomena attest to the crucial role of USD in affecting the exchange volatility of non-USD related currency pairs. While the 1 hour lagged dependent variable shows significant, positive impact on exchange rate volatility, the 2-hour lagged dependent variable has negative impact, indicating short-lived persistency in exchange rate volatility. Table 6 likewise models the volatility in the USD-JPY-NZD trilateral. This model also presents significant and positive impacts of corresponding trading volumes on the exchange rate volatility of currency pairs similar to the results for USD-JPY-AUD trilateral. In this case, the impact of trades of non-corresponding currency pairs turns out to be even stronger, through USD linked trade channels. For example, NZD-JPY volatility is more strongly affected by USD-JPY trading volume than the trading volume of its own corresponding currency pair (0.14% vs. 0.09%). In both trilaterals, non-USD-related currency pair volatilities are more affected by the trades of third currency pairs, in particular USD-related ones. Table 5: Regressions Results for Volatility, Trilateral Relationship (USD-JPY-AUD) Regressors Volatility AUDUSD AUDJPY USDJPY Constant -5.60 *** -3.22 *** -5.42 *** Volume (AUDUSD) 0.30 *** 0.10 ** -0.02 Volume (AUDJPY) 0.06 0.09 ** 0.03 Volume (USDJPY) 0.01 0.07 ** 0.33 *** Tuesday 0.01 0.04 -0.01 Wednesday 0.04 0.05 -0.01 Thursday 0.07 * 0.16 *** 0.04 Friday 0.01 0.03 -0.02 Lagged Dependent Variable (1 hour) 0.20 *** 0.21 *** 0.29 *** Lagged Dependent Variable (2 hours) -0.09 *** -0.11 *** -0.11 *** R-squared 0.60 0.51 0.58 Notes: * for significance below 10%; ** for significance below 5%; *** for significance below 1%. Sources: Authors’ calculations using data from the Bloomberg and CLS FX Database 14 Table 6: Regressions for Volatility, by Trilateral Relationship (USD-JPY-NZD) Regressors Volatility NZDJPY NZDUSD USDJPY Constant -4.36 *** -6.58 *** -6.86 *** Volume (NZDJPY) 0.09 *** 0.12 *** 0.10 Volume (NZDUSD) 0.09 *** 0.15 *** 0.03 Volume (USDJPY) 0.14 *** 0.14 *** 0.37 *** Tuesday 0.04 0.02 0.01 Wednesday 0.1 ** 0.11 -0.00 Thursday 0.15 *** 0.88 ** 0.05 Friday 0.04 0.05 -0.22 Lagged Dependent Variable (1 hour) 0.15 *** 0.10 ** 0.20 *** Lagged Dependent Variable (2 hours) -0.11 *** -0.08 *** -0.12 *** R-squared 0.58 0.58 0.61 Notes: * for significance below 10%; ** for significance below 5%; *** for significance below 1% Sources: Authors’ calculations using data from Bloomberg and CLS FX Database. The OLS models from the two trilateral relationships affirm the strength of the USD-related trade flows in influencing foreign exchange volatilities, and show the possibility of a third regional currency impacting exchange rate volatility. Another point worth noting is that the size of the foreign exchange trading market, especially in comparison to other currency pairs within each trilateral relationship, could drive the strength of the effect of a particular currency trading volume. As an example, the USD-JPY trading market is significantly larger than NZD-JPY and NZD-USD markets, which leads to the USD-JPY trading volume having the strongest effect on currency pair volatilities among the three bilateral pairs. The fact that non-USD currency pairs AUD-JPY and NZD-JPY show different dynamics, in that the former is affected strongest by USD-AUD trades while the latter is affected strongest by USD-JPY trades, also partly reflects the relatively bigger size of the USD-AUD exchange market than the USD-NZD. 3.2. Fixed effect estimation The above findings are further supported by the time-fixed effect model estimations for both trilateral relationships. Table 7 summarizes the impact of trading volumes on the USD-JPY-AUD trilateral (see Appendix 2 for the full model). The results show that the volatility of each currency pair is affected significantly by the trading volume of corresponding currency pairs. The increase in USD-JPY trading volume accounts for a 0.29% increase in USD-JPY exchange rate volatility. At the same time, however, the volatility of country pair exchange rates is also significantly affected by the trade of non-corresponding country pairs. This third currency trading channel works strongly for the AUD-JPY volatility, which is more strongly affected by USD-AUD trading 15 volume than by its own corresponding trading volume. This result is consistent with OLS estimation results. Moreover, the inclusion of the time-fixed effect variable broadly establishes a significant relationship between volatility and specific hours within the trading day, although this is not as prevalent in the case of USD-JPY volatility. Furthermore, the 2-hour lagged volatility turns out to be no longer significant while the positive impact of 1-hour lagged volatility is largely maintained, and the impact of trading in the middle of the week is only significant in one of the three currency pairs. Table 7: Fixed-Effect Regressions on USD-JPY-AUD Trilateral Currency Trading Volumes USDJPY (Major Pair) USDAUD (Major Pair) AUDJPY (Minor Pair) USDJPY 0.29%*** 0.08%** 0.01% USDAUD 0.06%*** 0.33%*** 0.03% AUDJPY 0.11%*** 0.17%*** 0.09%*** Notes: * for significance below 10%; ** for significance below 5%; *** for significance below 1%. Major pair refers to currency pairs involving widely used currencies paired with the USD. Minor pairs refer to widely used currencies, excluding USD. Sources: Authors’ calculations using data from Bloomberg and CLS FX Database. Somewhat different results are estimated for the case of USD-JPY-NZD, as illustrated by Table 8 (see Appendix 3 for the full model). Exchange rate volatility is affected by the corresponding currency pairs significantly. However, the magnitude of impact turns out to be strongest from the non-corresponding currency pairs except for USD-JPY volatility. Consistent with OLS estimation results, NZD-JPY volatility seems to be more strongly affected through the USD-linked third currency trading channel, i.e., USD-JPY in our model. Different from OLS estimation results, USDNZD volatility is also more strongly affected by the USD-linked third currency trading channel, i.e., USD-JPY instead of its own corresponding pair. USD-JPY exchange rate volatility is still affected by USD-JPY trading volume the most, with 1% increase in USD-JPY trading volumes leading to 0.43% increase in USD-JPY exchange rate volatility. In the case of USD-JPY volatility, this direct effect is dominant, and the indirect impact of other currency pairs turn out to be insignificant. The time-fixed effect variable in the USD-JPY-NZD trilateral are mostly significantly correlated with volatility, although it is worth noting that not all hours have a significant effect on USD-JPY volatility. The inclusion of the time-fixed effect also renders the 2-hour lagged volatility as insignificant, while in trading day variables, Wednesdays and Thursdays are significant for NZDJPY volatility, and Wednesdays for NZD-USD volatility. 16 Table 8: Fixed-Effect Regression on USD-JPY-NZD Trilateral Currency Trading Volumes USDJPY (Major Pair) USDNZD (Major Pair) NZDJPY (Minor Pair) USDJPY 0.43%*** 0.01% 0.002% USDNZD 0.17%*** 0.13*** 0.10*** NZDJPY 0.22%*** 0.09*** 0.08%*** Notes: * for significance below 10%; ** for significance below 5%; *** for significance below 1% Sources: Authors’ calculations using data from Bloomberg and CLS FX Database. Overall, the weaker influence of NZD-related trades on the volatility of their own corresponding currency pair may be attributed to the AUD-USD and AUD-JPY trading markets being larger than the NZD-JPY and NZD-USD markets. AUD-USD volume, for example, is second only to USDJPY trading volumes in terms of size. Regardless, the fixed effect models show that third currency pair trading volumes can have an impact on exchange rate volatility. While this third-party currency pair channel is stronger through USD related trades, non-USD related third party currency trade can also exert significant impact on the exchange volatility, as the estimation results indicate. 3.3. GARCH estimation GARCH analysis is used to further explore the effects of trading volumes on foreign exchange volatility, controlling for time-specific components. The differenced time series are then converted into log variables and fitted with their respective log-differenced trading volumes for each tripartite relationship, as well as an ARMA component to control for time-specific properties. An ARCH test is then performed to investigate possible heteroskedasticity within each time series. Table 9 summarizes the results of the ARCH test on each currency volatility. The test shows that foreign exchange volatilities for the AUD-USD, AUDJPY and NZD-JPY ARMA models show heteroskedasticity, and thus necessitate the addition of an ARCH component to capture further time effects. In contrast, NZD-USD and USD-JPY for both AUS and NZD tripartite ARMA models are already sufficiently explained by the ARMA model, and do not necessarily need an additional ARCH component. Nevertheless, to analyze the models on a uniform ground, the GARCH (1,1) component is adopted all models. 17 Table 9: ARCH Test on Currency Volatilities (24 lags) Volatility F statistic Obs R2 Prob F(24,478) Prob Chi2 (24) AUDUSD 3.56 76.21 0 0 AUDJPY 2.71 60.35 0 0 NZDUSD 0.56 13.78 0.96 0.95 NZDJPY 1.65 38.32 0.03 0.03 USDJPY (AUD) 0.69 17 0.86 0.85 USDJPY (NZD) 0.68 16.6 0.87 0.86 Sources: Authors’ calculations using data from Bloomberg and CLS FX Database. The individual GARCH (1,1) models for each currency pair all show that most currency trade volumes exert a positive significant effect on the respective currency volatilities. The time-specific effects vary per model. Results from the USD-JPY-AUD tripartite analysis (see Table 10) show that USD-AUD and AUDJPY volatilities all have positive correlations with the various currency pair trading volumes. In the case of AUD-USD, the trading volume with the strongest effect is its own pair (0.33), followed by USD-JPY (0.10) and AUD-JPY (0.05). Likewise, AUD-JPY volatility is most influenced by USDJPY (0.19), followed by AUD-JPY (0.11) and USD-AUD (0.08). In contrast with these currency pairs, USD-JPY volatility is only influenced by the USD-JPY trading volume (0.38). It is noteworthy that these relationships hold even with the inclusion of time series variables in the model. All three of the volatility pairs are positively correlated with the 24-hour AR component at similar levels, which suggest that volatility values tend to be above the mean volatility if its 24-hour lagged value was also above the mean. Since the coefficients register at around 0.33–0.34, the past value’s influence on present values, though significant, are not very strong. The 1-hour MA component for all three volatilities are significant, large and negatively correlated with present volatility values. The MA (1) coefficients for AUD-USD and AUD-JPY are both higher in magnitude than in the USD-JPY; this means that the error terms of 1-hour lagged values tend to exert a stronger influence on the preceding two currency pairs. A negative correlation suggests that a higher error term or shock from the previous hour tends to pull current volatility value lower (and vice versa). The MA (2) components are also negatively correlated with current volatility values, though the magnitude of their coefficients is much lower. This implies that the shocks or error terms from 1hour lagged values have a stronger influence over current volatility than 2-hour lagged values. The MA (2) coefficient magnitude of USD-JPY volatility is also marginally higher than the other pairs, suggesting that the USD-JPY currency is more influenced by its past 2-hour error terms. The ARCH and GARCH components of the model offer an idea of whether volatility values are 18 more influenced by short-term or long-term shocks. The ARCH (1) coefficients of USD-JPY and AUD-USD are both insignificant, while the magnitude of AUD-JPY’s, though significant, is very small. This suggests that that short-term shocks do not have a persistent effect on current volatility values for all currency pairs. Meanwhile, GARCH (1) components are all significant and large in magnitude across all currency pairs, which implies that older volatility values have a more longterm and persistent effect on current volatility. Table 10: GARCH Model Estimations for USD-JPY-AUD Regressors Volatility AUDUSD AUDJPY USDJPY Constant -0.00 *** -0.00 *** -0.00 ** Volume (AUDUSD) 0.33 *** 0.08 *** 0.02 Volume (AUDJPY) 0.05 *** 0.11 *** 0.03 Volume (USDJPY) 0.10 *** 0.19 *** 0.38 *** AR (24) 0.35 *** 0.34 *** 0.34 *** MA (1) -0.75 *** -0.76 *** -0.68 *** MA (2) -0.19 *** -0.14 *** -0.29 *** ARCH (1) -0.02 -0.01 * 0.00 GARCH (1) 0.85 *** 0.87 *** 0.92 *** R-squared 0.53 0.47 0.56 Aike info criterion -0.03 -0.19 0.13 Notes: 24-hour lagged volatility values are in logarithmic form, while all trade volume variables are in logarithmic form, differenced by 1 hour. The ARMA model of ARMA-GARCH model of AR (24) MA (1) MA (2) ARCH (1) GARCH (1) are used for all models. Sources: Authors’ calculations using data from Bloomberg and CLS FX Database. Results from the USD-JPY-NZD triparty (see Table 11) reflect varied patterns from the previous tripartite analysis. NZD-USD volatility is affected by all trading volumes. The magnitude of the values is more even than the currencies in the previous analysis, wherein NZD-USD volatility is most affected by its own volume (0.16), followed by USD-JPY (0.14), and NZD-JPY (0.10). NZDJPY volatility is also influenced by all trading volumes, with the largest influence from USD-JPY (0.21), followed by USD-NZD (0.10), and NZD-JPY (0.06). USD-JPY volatility is heavily influenced by its own trading volume (0.31), then distantly followed by NZD-JPY (0.05). The AR (1) component of NZD-JPY is positively significant at a low magnitude (0.24), which means that its volatility is mildly influenced by its 1-hour lagged value. The AR (24) coefficients for all volatilities are likewise positively significant, wherein USD-JPY (0.31) and USD-NZD (0.35) have higher magnitude than NZD-JPY (0.24). This would show that the 24-hour lagged volatilities of the preceding currencies have a slightly larger impact on current values. The MA (1) coefficients of all three currencies are negatively significant. The analysis shows that the MA (1) component of 19 NZD-JPY is highest (-0.99), followed by NZD-USD (-0.89), and USD-JPY (-0.61). This would imply that current volatility values are variedly influenced by immediate past error terms and shocks. The MA (2) components are also negatively correlated for USD-NZD and USD-JPY but have vastly different magnitudes. The analysis would show that shocks from a 2-hour window still have a strong effect on USD-NZD but have significantly tapered for USD-JP Y. The ARCH and GARCH components are both significant for NZD-USD, with a larger magnitude for the GARCH coefficient implying that much previous volatility values have a more persistent effect on current values. In comparison, only the ARCH coefficient is significant for USD-JPY and NZD-JPY, which means that these volatility values are more influenced by recent shocks. Table 11: GARCH Model Estimations for USD-JPY-NZD Regressors Volatility NZDJPY NZDUSD USDJPY Constant -0.00 -0.00 ** -0.00 Volume (NZDJPY) 0.06 *** 0.10 *** 0.05 *** Volume (NZDUSD) 0.10 *** 0.16 *** 0.01 Volume (USDJPY) 0.21 *** 0.14 *** 0.31 *** AR (1) 0.24 *** NA NA AR (24) 0.24 *** 0.35 *** 0.31 *** MA (1) -0.99 *** -0.89 *** -0.61 *** MA (2) NA -0.8 *** -0.34 *** ARCH Residual 0.23 ** -0.01 *** 0.10 ** GARCH -0.03 0.58 ** 0.10 R-squared 0.46 0.54 0.58 Aike info criterion -0.01 0.23 0.24 Notes: 24-hour lagged volatility values are in logarithmic form, while all trade volume variables are in logarithmic form, differenced by 1 hour. NA implies that a particular ARMA component was omitted in the analysis for the currency pair. The omission of these components for particular models was done to ensure that the individual model exhibited no serial autocorrelation to provide a clearer analysis. Sources: Authors’ calculations using data from the Bloomberg and CLS FX Database. While the GARCH analysis for both tripartite relationships shows that time-related coefficients have varying effects on each individual time-series, a pattern emerges that, for most of the currency pairs, all trading volumes have a positive and significant correlation with their trading volumes, controlling for these time elements (see Tables 12 and 13). While NZD-JPY and AUDJPY volatility are both more influenced by USD-JPY trading volumes, other currency pairs such as USD-AUD and USD-NZD are most strongly determined by their own currency pair volatilities. In all these four cases, the third-party currency also has a significant influence on their respective volatilities. As expected, USD-JPY volatility is influenced mostly by its own trading volume; in the USD-JPY-AUD tripartite, it is in fact the sole influencer to USD-JPY volatility. The USD-JPY-NZD 20 tripartite, however, suggests that even small currency pairs like NZD-JPY also have a significant, if minute, effect on USD-JPY volatility. Table 12: Percentage Effects to FX volatility, by Foreign Exchange Trading Volume Pair (USD, JPY, AUD) Currency Trading Volumes USDJPY (Major Pair) USDAUD (Major Pair) AUDJPY (Minor Pair) USDJPY 0.38%*** 0.02% 0.03% USDAUD 0.10%*** 0.33%*** 0.05%*** AUDJPY 0.19%*** 0.08%*** 0.11%*** Notes: Data is taken from GARCH (1,1) regressions on matching each individual currency with trading volume pairs. *** denotes significance at 0.01, ** denotes significance at 0.05, * denotes significance at 0.10. Major pair refers to currency pairs involving widely used currencies paired with the USD. Minor pairs refer to widely used currencies, excluding USD. Sources: Authors’ calculations from data from Bloomberg and CLS Database. Table 13: Percentage Effects to Foreign Exchange Volatility by Trading Volume Pair (USD, JPY, NZD) Currency Trading Volumes USDJPY (Major Pair) USDNZD (Major Pair) NZDJPY (Minor Pair) USDJPY 0.31%*** 0.01% 0.05%*** USDNZD 0.14%*** 0.16*** 0.10*** NZDJPY 0.21%*** 0.10*** 0.06%*** Notes: Data is taken from GARCH (1,1) regressions on matching each individual currency with trading volume pairs. *** denotes significance at 0.01, ** denotes significance at 0.05, * denotes significance at 0.10 Sources: Asian Development Bank calculations from data from Bloomberg and CLS Database. IV. Conclusions US dollar dominance as a global reserve currency has significant implications for the impact of USD-related currency trade volumes on exchange rate volatility. This paper has investigated whether third-party foreign exchange trade volumes can have any significant effect on the foreign exchange volatilities of original currency pairs despite the crucial role of the US dollar as the dominant currency. While the study demonstrates the significant effect of exchange trade volumes on the volatility of the corresponding currency pairs, third party currency trade volumes also exert significant impact on the volatility of the corresponding currency pairs. This third party trading channel is quite strong for the USD-related currency trades. However, non-USD-related third party trades also significantly affect the volatility of corresponding currency pairs. These findings are investigated through three types of models that explore the relationship between foreign exchange volatility and trade volumes. The OLS regressions point out that, individually, most currency volatilities are affected by all currency volumes in the tripartite relationships, including the indirect, third-party currency trading channels. The time-fixed effect model adds that, even 27 Appendix 3: Time-Fixed Effect Model for USD-JPY-NZD Trilateral (NZDJPY) (NZDUSD) (USDJPY) VARIABLES Volatility Volatility Volatility Volume (NZDJPY) 0.0782*** 0.102*** 0.00283 (0.0140) (0.0175) (0.0150) Volume (NZDUSD) 0.0909*** 0.132*** 0.0118 (0.0151) (0.0189) (0.0164) Volume (USDJPY) 0.219*** 0.168*** 0.433*** (0.0238) (0.0296) (0.0277) Lagged Dependent Variable (1 hour) 0.111*** 0.118*** 0.195*** (0.0376) (0.0379) (0.0375) Lagged Dependent Variable (2 hours) 0.0239 0.0212 0.00274 (0.0352) (0.0358) (0.0351) Friday 0.000770 0.0184 -0.0400 (0.0317) (0.0397) (0.0344) Thursday 0.0820** 0.0399 -0.00216 (0.0324) (0.0397) (0.0347) Tuesday -0.00363 -0.00785 -0.0208 (0.0288) (0.0359) (0.0314) Wednesday 0.0509* 0.0667* -0.0323 (0.0300) (0.0376) (0.0324) 1.hour -0.181** -0.579*** 0.476*** (0.0721) (0.0918) (0.0744) 2.hour -0.384*** -0.665*** 0.0732 (0.0703) (0.0880) (0.0778) 3.hour -0.453*** -0.604*** -0.0361 (0.0800) (0.100) (0.0880) 4.hour -0.455*** -0.512*** -0.312*** (0.0814) (0.102) (0.0881) 5.hour -0.569*** -0.669*** -0.430*** (0.0786) (0.0993) (0.0845) 6.hour -0.624*** -0.737*** -0.359*** (0.0752) (0.0952) (0.0809) 7.hour -0.681*** -0.825*** -0.250*** (0.0735) (0.0928) (0.0802) 8.hour -0.578*** -0.687*** -0.181** (0.0742) (0.0930) (0.0818) 9.hour -0.439*** -0.472*** -0.123 (0.0802) (0.100) (0.0880) 10.hour -0.387*** -0.539*** -0.0399 (0.0813) (0.103) (0.0877) 11.hour -0.405*** -0.528*** -0.222*** (0.0807) (0.102) (0.0859) 12.hour -0.544*** -0.635*** -0.163** (0.0775) (0.0974) (0.0824) 13.hour -0.532*** -0.645*** -0.144* (0.0747) (0.0943) (0.0801) 14.hour -0.503*** -0.529*** -0.0699 (0.0757) (0.0954) (0.0823) 15.hour -0.487*** -0.462*** 0.130 (0.0839) (0.105) (0.0917) Continued on the next page 28 (NZDJPY) (NZDUSD) (USDJPY) VARIABLES Volatility Volatility Volatility 16.hour -0.501*** -0.568*** -0.150 (0.0855) (0.108) (0.0937) 17.hour -0.522*** -0.618*** -0.188* (0.0901) (0.114) (0.0982) 18.hour -0.741*** -0.842*** -0.400*** (0.0863) (0.109) (0.0927) 19.hour -0.621*** -0.822*** -0.251*** (0.0785) (0.0992) (0.0846) 20.hour -0.557*** -0.729*** -0.126 (0.0733) (0.0923) (0.0794) 21.hour -0.508*** -0.638*** -0.103 (0.0717) (0.0899) (0.0779) 22.hour -0.430*** -0.602*** -0.194** (0.0727) (0.0913) (0.0786) 23.hour -0.677*** -0.885*** -0.385*** (0.0724) (0.0905) (0.0790) Constant -5.397*** -5.509*** -7.882*** (0.455) (0.569) (0.512) Observations 546 546 546 R-squared 0.696 0.687 0.752 Notes: Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Sources: ADB calculations using Bloomberg and CLS database. 29 REFERENCES Clark. (1973). A subordinated stochastic process model with finite variance for speculative. Econometrica. Copeland. (1976). “A model of asset trading under the assumption of sequential information arrival”. Journal of Finance, 1149-1168. Copeland. (1997). “A probability model of asset trading”. Journal of Financial and Quantitative Analysis. De Grauwe, P. (1988). Exchange rate variability and the slowdown in growth of international trade. Retrieved from International Monetary Fund: https://www.jstor.org/stable/3867277 Eichengreen, B. (2007). The Real Exchange Rate and Economic Growth. Social and Economic Studies. Epaphra, M. (2017). Modeling exchange rate volatility: Application of the GARCH and EGARCH models. Journal of Mathematical Finanace. Flores-Sosa, M., Aviles-Ochoa, E., and Merigo , J. (2020). Exchange rate and volatility: A bibliometric review. Retrieved from Wiley Online Library: https://onlinelibrary.wiley.com/doi/epdf/10.1002/ijfe.2223 Galati, G. (2000). Trading volumes, volatility and spreads in FX markets:. Bank for International Settlements. Gargano , A., Riddiough, S., and Sarno, L. (2018). The value of volume in foreign exchange. retrieved from http://www.olsendata.com/data_products/client_papers/papers/201804AntonioGargano-ValueofVolume.pdf Hooper, P. and Kohlhagen, Steven. (1978). The effect of exchange rate uncertainty on the prices and volume of international trade. Journal of International Economics, 483-511. Kang, J. W. (2016). International Trade and Exchange Rate. Retrieved from Asian Develoment Bank: https://www.adb.org/publications/international-trade-and-exchange-rate Khemiri, R. (2012). Volume and volatility in foreign exchange market microstructure: a Markov switching approach. Mercado, R., Jacildo, R., and Basu Das, S. (2022). US dollar dominance in Asia’s trade invoicing. Asian Development Bank. Mougoue, M., and Aggarwal, R. (2011). Trading volume and exchange rate volatility: Evidence for the sequential arrival of information hypothesis. Journal of Banking and Finance, 2690-2703. Ozturk, I. (2006). Exchange Rate Volatility and Trade: A Literature Survey. International Journal of Applied Econometrics and Quantitative Studies, 85-102. Park, D., Tian, S., Qureshi, I., and Villaruel, M. (2022). Impact of Monetary Policy Uncertainty . Asian Development Bank. 30 Rajbongshi, G., and Suresh, P. S. (2023). Exchange rate volatility: A review of global experience. IUP Journal of Applied Economic, 52-72. Rodrik, D. (2008). The real exchange rate and economic growth. Harvard University. Retrieved from https://muse.jhu.edu/pub/11/article/261232/summary S.L., H., and Boon, H. (2000). Real exchange rate volatility and Malaysian export its major trading partners. University Putra Malaysia. Sensoy, A., and Serdengecti, S. (2019). Intraday volume-volatility nexus in the FX markets: Evidence from an emerging market. International Review of Financial Analysis. Seraj, M., and Coskuner, C. (2021). Real exchange rate effect on economic growth: comparison of fundamental equilibrium exchange rate and Balassa–Samuelson based Rodrik approach. Journal of Applied Economics, 541-554. Retrieved from https://www.tandfonline.com/doi/full/10.1080/15140326.2021.1977083 Shaik, K., and Rao, B. (2020). Does exchange rate have any impact on economic growth in India? An empirical analysis. Theoretical and Applied Economics, 223-234. Retrieved from https://store.ectap.ro/articole/1485.pdf Vienna Initiative. (2024). Vienna Initiative. Retrieved from Vienna Initiative: https://www.viennainitiative.com/about/ Zhao, Y. (2020, January 2020). The influence and impact of the exchange rate on the economy. Retrieved from https://www.researchgate.net/publication/347413240_The_Influence_and_Impact_of_the _Exchange_Rate_on_the_Economy Zhao, Y. (2020). The influence and impact of the exchange rate on the economy . Retrieved from E3S Web of Conferences: https://www.e3sconferences.org/articles/e3sconf/pdf/2020/74/e3sconf_ebldm2020_03007.pdf ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org DYNAMIC IMPACT OF FOREIGN EXCHANGE TRADING VOLUME ON FOREIGN EXCHANGE VOLATILITY Jong Woo Kang and Carlos Cabaero ADB ECONOMICS WORKING PAPER SERIES NO. 768 February 2025 Dynamic Impact of Foreign Exchange Trading Volume on Foreign Exchange Volatility Foreign exchange (FX) trading volume is a key factor in volatility. This paper investigates the effect of trading volume on volatility using high-frequency data. Estimation results from econometric models reveal a significant impact of third-party trade volumes on the volatilities of original currency pairs. Though the United States dollar (USD) exerts sizeable effect through third-party channels, currency pairs without USD linkages also have impact, calling renewed attention to utilizing regional cooperation in mitigating volatility as compared with major FX trading partners. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 69 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance.