The exchange rate (ER) acts like a barometer for the economy, demonstrating the relative weakness or strength of the currency in developing economies Kohrt et al. (2024). The ER has a significant impact on the citizens of any country. Changes in the ER directly affect the prices of necessary goods and services (Nor, 2015). Recently, shifts in the global economy are due to fluctuations in the ER, making it unpredictable (Kayani et al., 2023). This fluctuation is due to the divergence of monetary and fiscal policies and adaptation by the emerging and developed economies (Devereux and Lane, 2003). Furthermore, the trade balance (TB) influences exchange rates through the demand and supply dynamics of foreign currency (Zhu et al., 2025; Lioudis, 2023). The ER is an important policy tool for assessing the economic policies and economic growth of a country. The fluctuations in ER can pose risks to export and import pricing and the overall economic growth (Iqbal et al., 2024).

Following the breakdown of the Bretton Woods system in 1971, many developed economies switched from fixed to floating exchange rates. Due to this transformation, significant changes in both real and nominal exchange rates are anticipated. It is a debatable phenomenon that changes in exchange rates affect trade and price stability. From Pakistan’s perspective, in 1970, the government changed its policy, and all private institutions were nationalized, but there was no significant impact on the economy. Thus, in 1980, the government of Pakistan revised policies in favor of liberalization, deregulation, and decentralization. This period also saw a revision of the ER policy, with Pakistan adopting a managed floating ER in 1982, leading to a 20% depreciation of the Rupee. Moreover, in 1988, Pakistan signed the “Medium Term Stand by Extended Fund Facility (EFF)” agreement with the International Monetary Fund, which included “provisions for devaluation, trade liberalization, tariff reductions, and financial reforms such as the deregulation of the interest rate structure” (Stockman, 1983). Throughout a decade of instability, Pakistan underwent a series of adjustments and stabilizations, facing the lowest GDP in South Asia during this period.

Pakistan’s economy has mostly faced trade deficits and experienced currency depreciation, except in 1951 and 1971. The main reason for Pakistan’s trade deficit is that the exports are less than imports. Pakistan follows the conception-based behavior, and the exports are mostly primary goods. As a result, our import bills are high, and we are always in deficit (Sajid et al., 2024; Iqbal et al., 2023a; Iqbal et al., 2023b). Many developing economies have focused on outward-oriented policies in foreign trade, which enhanced the level of exports from emerging markets (Razmi, 2007). Amraoui et al. (2025) investigated the impact of ER volatility on trade and investment using the NARDL model. Empirical evidence shows that ER appreciation reduces exports and FDI while boosting imports due to lag effects. In the long run, exports are observed to have a significantly larger asymmetric impact, while imports display asymmetric effects. This study highlights that ER management, export diversification, and industrial upgradation enhance trade behavior. Jabeen et al. (2022) examined the effects of news on the volatility of the Pakistan ER. They have used time-series data and employed the GARCH model. The evidence shows that foreign and domestic macroeconomic news announcements significantly affect the Pakistani currency (rupees). Additionally, the evidence shows that domestic news in Pakistan and the US immediately affected the rupee–dollar ER relative to other currencies, particularly the U.S. dollar–rupee exchange rate ($/Rs). Thus, the outcomes suggest that the response of the Pakistani rupee ER is consistent with models of ER determination and the monetary authority’s response function.

Pakistan and Japan have had strong economic relations and mutual common interests through trade and commerce for 40 years. After the World War II, Japan’s economy was rebuilt, and its restoration was the priority for them. Among the many common elements between Pakistan and Japan, there is a strong fight for economic growth and recovery. Since its inception in 1947, Pakistan has struggled with a lack of industrial infrastructure, a weak agribusiness sector, and challenging economic conditions. Similarly, Japan’s economy was severely damaged during World War II. Both countries faced similar distressing financial circumstances. Japan may have faced more intense monetary circumstances than Pakistan. However, Japan had previously advanced in education and skill development. The Japanese economy experienced severe physical distortion following the annihilation of the Second World War. The transformation of the economy’s essential design toward outward-looking, growth-oriented policies and the foundation of a modern political framework is a serious issue. In contrast to Japan, labor planning was never an integral part of Pakistan’s economic planning since the beginning of the 6th Five-Year Plan. Thus, Pakistan has become one of the least skilled nation, with more than 60% of its population lacking access to education. In any case, Japan has more advantages in skilled human resources and a robust industrial base. Signs of modernization in Japan began to appear before 1890. The country achieved nearly 100% literacy by 1920. This progress has led to the development of well-trained labor force, the establishment of an efficient society, appropriate preparation, and difficult working conditions, all of which facilitated Japanese’s rapid entry into modern economic development (Ryoshin, 1986). Khalid et al. (2025a, 2025b, 2025c) studied the asymmetric impact of exchange rates on trade flows between Turkey and Germany using the NARDL model. The findings show that 39 exporting and 54 importing industries found the long-run relationship. This study suggests that the government should adopt an industry-specific trade policy that supports the positive trade flows.

This study focused on testing the depreciation concept that currency depreciation improves the trade balance (TB) of Pakistan involved in trade with Japan. In this context, we used a new approach introduced by Bahmani-Oskooee and Ratha (2004). Before explaining this concept, we have demonstrated three basic approaches of the ER that influence the TB. The Marshall-Lerner condition in economic theory states that a depreciation of a country’s currency will expand its TB. If the sum of the absolute value of the elasticities of export and import demand is more than one. If the Marshall-Lerner condition holds, then depreciation has a positive effect on a country’s TB in the long run (Ogbonna, 2018). The second method is the J-curve effect, which shows the association between the ER and the TB. This phenomenon shows that due to ER depreciation, TB initially worsens in the short run and expands in the long run because the quantity effect is dominant. Magee (1973) noted that TB does not improve immediately during the short run. However, once existing contracts are fulfilled and new ER and prices are implemented, the TB starts to improve. This phenomenon is called the J-Curve effect. Bahmani-Oskooee and Malixi (1992) used a trade dataset to investigate J-curve behavior and found that it did not confirm or support the J-curve effect in Korea. Bahmani-Oskooee and Hegerty (2010) conducted a survey on J- and S-curve at the industry-specific level. They accounted for the presence of J- and S-curve behavior across many industries. Khalid et al. (2024) examined the impact of ER misalignment on Pakistan’s economic growth. The outcome of the study indicates that undervaluation of currency stimulates economic growth while overvaluation slows down the economic growth of Pakistan. The study also suggests that overvaluation of the Pakistani rupee promotes economic growth up to 9.34%, while undervaluation results in 10.025%, highlighting that Pakistan should avoid both options. Bahmani-Oskooee and Arize (2020) investigated the asymmetric behavior of the ER on trade flows. For the estimation purpose, we used commodity-level data and employed the NARDL model. The evidence found asymmetric behavior in the majority of industries. Additionally, the J-curve was found in the following African countries: Zambia, Algeria, Morocco, Tanzania, Cameroon, and Ethiopia. Baek (2013) considered the impact of the exchange rate on TB between Korea and Japan using the commodity-level data. The evidence shows that Korean exports are very sensitive to changes in the bilateral (ER) in the short run but less responsive in the long run. Mustafa et al. (2015) also found a J-curve results in the trade between Brazil and the US. In the short run, TB is not significantly effective, but in the long run, the TB improves. Baek and Kim (2025) investigated the asymmetric impact of the ER on the trade flows between Korea and Vietnam. For this purpose, they used data from 2000 to 2024 and investigated the S-curve hypothesis. The evidence shows that, in total, 37 industries exhibit asymmetric S-curve behavior due to currency appreciation and depreciation. Similar results were also reported by Khalid et al. (2024) during their study of the asymmetric impact of the ER on trade flows between Turkey and Germany. They used data from the 114 industries and applied the NARDL model. Their findings show that the ER appreciation and depreciation have different impacts on the industry’s TB.

The S-curve trajectory is the extension of the J-curve and is based on the cross-correlation function. The cross-correlation function suggests that there may be a positive correlation of the future TB with the current exchange rate, while a negative correlation of the past TB with the current ER. Thus, plotting these positive and negative coefficient values becomes an S-curve pattern. Its mechanism shows that in the initial stage of the economy, an increase in the ER leads to a decrease in TB due to a lag effect called the price effect, while in the long run. It improves TB, called the quantity effect or volume effect. After a certain point, as TB improves, the domestic currency will appreciate, making local goods for foreigners expensive, and they will decrease the demand for Pakistani currency. Thus, TB will decrease automatically, and the shape becomes an S-curve. Limited literature is available on this phenomenon because it was introduced recently. Previously, we discussed the J-curve outcome, in which the ER depreciates, and TB worsens in the short run but improves in the long run. Over time, the TB improves, and the currency of a country becomes appreciated, which means the demand for the currency rises. For foreign investors or consumers, this country’s product starts to become expensive, then they decrease the demand for the currency, and TB becomes worse sloping, and it becomes the S-curve shape. The S-curve behavior was examined in 20 African economies using the correlation function (Bahmani-Oskooee et al., 2008a, 2008b). The evidence indicates that the S-curve pattern was observed in eight countries out of 20 (Bahmani-Oskooee and Fariditavana, 2014) decomposed the ER into positive & negative values. In data analysis using the cross-correlation function, we found an asymmetric S-curve behavior when the ER changes. Bahmani-Oskooee and Ratha (2008) also studied the S-curve for the commodity trade between the US and China, revealing that 50% of industries follow the S-curve shape.

In the present study, we reinvestigated the influence of the ER on TB using the S-curve pattern. We used commodity-level data from 1980 to 2020 and applied the cross-correlation function. Ahmad et al. (2023) investigated the asymmetric behavior of the ER on TB, whereas we examined the symmetric impact of the ER on TB in the context of bilateral trade between Pakistan and Japan. This study will have important policy implications for those countries whose trade is always in deficit. Moreover, the existing literature on ER and trade dynamics mostly focused on Pakistan’s trade behavior with major trading partners such as the US, China, and the EU. Conversely, the present study focused on the trade behavior between Pakistan and Japan. Japan is characterized by advanced technology, skilled labor, and a quality-sensitive market. Furthermore, we have formulated the following hypothesis:

The remaining sections of the manuscript are prepared as follows: Section 2 reviews the literature; Section 3 focuses on the data source for variable construction; Section 4 contains the results and their interpretation, while Section 5 explains the conclusion and policy recommendations.

In this section, we have reported selected literature studies based on areas such as ER behavior, TB, and J- and S-curve patterns.

Jabeen et al. (2022) found that foreign and domestic macroeconomic news announcements significantly affect the Pakistani currency (rupees). Additionally, the evidence shows that Pakistan and US domestic news immediately affected the rupee per dollar as compared to other currencies over rupees ($/Rs). Thus, the outcomes of the study suggested that the response of the Pakistani rupee ER seems to be consistent with the models of ER determination and the monetary authority’s response function. Khalid et al. (2025a, 2025b, 2025c) investigate the asymmetric impact of exchange rates on trade flows between Turkey and Germany using the NARDL model. This study suggests that the government should adopt an industry-specific trade policy that supports the positive trade flows. The existing study explores the influence of the ER on TB using the S-curve method. There are many studies on ER and trade balance, but the present study is a little different. The existing study demonstrated the S-curve behavior using the disaggregate commodity-level data and employed the cross-correlation function. Here, we present the most relevant literature supporting our study. Ullah and Nobanee (2025) investigate the energy- and financial-related factors influencing the ER in Pakistan. They used data from 2006 to 2021 and employed the ARDL model. The findings indicate that FDI strengthens the ER in both the short- and long-run. Energy-related factors are positively associated with the exchange rate, while the stock market decline is negatively associated with it (Gupta and Varshney, 2021). Investigating the influence of the ER on trade flows between India and the US using data from six importing industries. The study employs the E-GARCH model to measure volatility and also utilizes the ARDL model. The findings show that the ER influences the importing industries. The purchasing power parity theory shows that the ER equals the ratio of domestic and foreign prices. ER variation reflects inflation in both countries. If domestic inflation is higher than foreign inflation, then the domestic currency depreciate. Another theory, the interest parity theory, demonstrates that ER moments are driven by differences in interest rates across countries. Capital flows to the countries with higher returns, impacting the ER. Khalid et al. (2025a, 2025b, 2025c) found that third-country ER changes clearly affect Turkish and German trade, particularly in the short run, and these effects fluctuate across many export and import industries. In nearly half of industries, these differences persist over the long run, which suggests that trade policies should also reflect ER movements in third economies.

Backus et al. (1994) constructed a stochastic growth model for two countries to examine the features of short-term ER volatility on TB in 11 OECD countries. They provide a dynamic general equilibrium model with endogenous factors for both variables. It demonstrates that two major drivers of economic fluctuations, namely government purchases and productivity shocks, have a considerable impact on the behavior of the S-curve between two variables. A series of positive shocks can boost output at a single point in time, resulting in better trade terms. However, the consumption pattern may rise by a smaller proportion than the output due to productive shocks. Investment expenditure may grow faster than consumption if capital is reallocated to more productive uses. As a result, investment and consumption growth may exceed production growth, resulting in an early trade imbalance. The short-run dynamic relationship between TB and ER is characterized by an S-curve pattern in which the transient boom in investment and consumer expenditure slows down as time passes, and the trade deficit turns into a surplus. The monetary approach to the ER depicts that the ER is determined by the demand and supply of money of the two countries. Its mechanism shows that increasing the domestic money supply leads to an increase in price, which supports currency depreciation. According to the purchasing power parity theory, the ER will be adjusted if the costs of goods are the same across countries. To achieve an effective comparison, the price differences are eliminated based on the law of one price. If Pakistan rupees rise faster than the Japanese yen, the rupees will lose their value at trade fairs.

The S-curve behavior was investigated in the United States and Hong Kong (Bahmani-Oskooee and Hegerty, 2010). The results show that the S-curve exists in 34 out of 104 industries, including both small and large, durable and non-durable. The S-curve was tested (Bahmani-Oskooee and Xu, 2014), which indicated that 39 out of 74 industries supported S-curve behavior, and suggested that depreciation has a positive influence on the future TB. The portfolio theory of ER holds that the ER is determined through investors’ decisions. If Pakistan offers higher returns than Japan, investors will definitely buy rupees, and the rupee will appreciate.

Mohamud and Mohamud (2025) investigated the impact of the ER on TB in Somalia. They used data from 1990 to 2022 and employed the Bayesian VAR model. Their findings show that, in the long run, TB and inflation have a negative impact on the exchange rate, while foreign direct investment has a positive influence. Another report (Bahmani-Oskooee and Wang, 2007) examined the behavior of the ER of Australia using the J-curve effect. The 108 industries were analyzed, and the J-curve effect was observed in the majority of the industries. Furthermore, another report (Bahmani-Oskooee and Ratha, 2007c) showed S-curve behavior in Japanese commodity trade involving trade with 12 trading partners. The S-curve behavior in trade was reported for the US and China based on data from 104 industries (Bahmani-Oskooee and Ratha, 2009a, 2009b). The evidence found that S-curves exist in 50% of the industries. The trade pattern of the United States and India from 1973 to 2004 was reported (Bahmani-Oskooee and Ratha, 2010). They found that the S-curve pattern is supported in 15 out of 27 industries.

To my knowledge, existing studies investigating ER and trade flows the majority of studies often uses aggregate-level data and analyzes the impact of ER on trade flows in the context of the S-curve. The evidence shows that the studies available on the J-curve using the ARDL or nonlinear ARDL model investigate the impact of the ER on trade flows. The studies on S-curves show that only the cross-correlation function has been used with no regression because only two variables are based on it. In this study, we investigated the effect of the real ER on TB of Pakistan involving trade with Japan using data at the disaggregate level. This study provides significant policy insights for those low-income economies that are always in trade deficits.

In keeping with the methodology of Bahmani-Oskooee and Ratha (2010), who examined the S-curve pattern for China at the commodity level. Regarding the S-curve pattern, the majority of studies available are by Mohsin Bahmani-Oskooee, a well-known economist in this field. We also plan to investigate the S-curve behavior between Pakistan and Japan at the commodity level to prevent aggregation bias. The study used 15 2-digit industries and time-series data spanning the years 1980 to 2020. All commodity-level data were collected from the World Integrated Trade Solution (WITS) website, while using other data sources like World Development Indicators and International Financial Statistics for other variables, such as Gross Domestic Product, exchange rate, and Consumer Price Index. Since Pakistan is the reported country, we defined the variables from Pakistan’s context. Thus, TB is our dependent variable, which is defined as the difference between exports and imports, divided by the GDP of Pakistan TB = Xi − Mi GDP Pak , as the same procedure was also used by Bahmani-Oskooee and Xu (2010, 2013, 2014). To normalize the data, we have divided the export–import data by the GDP of the host country. The data have been taken in nominal terms, with Xi denoting Pakistani exports of industry I to Japan, Mi denoting imports of the same industry from Japan, and GDP_Pak representing Pakistan’s gross domestic product. The real ER is indirectly defined in this study, where RER equals [( CP I PAK ÷ CP I JPN ) × NER ]. Here, the nominal ER is represented by NER, expressed as the amount of rupees using the bilateral Consumer Price Index of that commodity per Yen.

For the depreciation of the rupee to have a positive influence on the industry. The net exports must have positive values, as observed in the contemporaneous correlation coefficient where K = 0. Net exports will increase if the Japanese yen strengthens and the Pakistani rupee weakens. In the short term, trade theory predicts that the rupee’s devaluation will make Pakistani products more affordable for foreign investors, leading to increased demand for the Pakistani currency and an improvement in TB. But over time, Pakistani currency means Pakistani products will be expensive for foreigners, which will decrease their demand. Then again, TB in Pakistan is declining, which reflects the S-curve pattern. Before estimating the S-curve, a few steps must be taken. To remove spurious outcomes from the existing dataset, we must first detrend the data using the HP filter method. Next, we must calculate the cross-correlation, which shows that the relationship between the current ER and past values of TB is negative, while the current ER with future values of TB will be positive. Plotting these lags and leads values becomes the S-curve pattern. We determined the cross-correlation function by examining earlier research and came up with the following formula:

Where the mean values of all data points for the real ER and net exports are shown by (RER¯) and (TB¯), respectively. To obtain the S-curve, we first determine the cross-correlation between the ER and TB. For this to happen, the lead (future) and lags (past) values must both be positive. If [1 2 3 4 5….] are acceptable values for k. When k takes negative values (−1, −2, −3, −4, −5….), it signifies the COR between the current ER and past trade values, while (1, 2, 3, 4, 5….) values show the correlation between the ER and future TB. The S-curve shape can be obtained by plotting the created COR coefficients of leads and lags.

This study investigates the dynamic association between the real ER and TB using the cross-correlation function based on time-series data of 15 industries at the disaggregated level in Pakistan and Japan. The S-curve behavior shows that the contemporaneous correlation should be negative, indicating that TB deteriorates in the short run due to currency depreciation, while the lead correlation should steadily turn positive, demonstrating the improvement of TB in the long run.

Figure 1 shows the pattern of the ER change between Pakistan and Japan. It depicts the response of the ER when currencies appreciate and depreciate. The real exchange rate’s downward trend indicates that the Pakistani currency has appreciated, while the upward trend indicates that the Pakistani rupee has depreciated. It displays mean-reverting behavior rather than a secular trend.

Figure 2 shows the disaggregated evidence of the S-curve for trade between Pakistan and Japan. The results show the substantial heterogeneity across industries. A limited number of industries demonstrated S-curve patterns, as indicated by plotted correlation values, lags, and leads. Lags values consistent with short-run effects showing that, due to the increase in the exchange rates, the TB decreases initially, dominating the price effect, and the response is very slow due to expenditure rigidity. The lead values indicate that TB increases in the long run due to steady capacity and demand adjustments. Additionally, positive signs of the lead values and negative signs for past lags are displayed in the cross-correlation values. As a result, the data indicate that improvements in TB do not positively covariate with Pakistan’s ER policy devaluation. Contrary to popular belief and economic theory, currency depreciation increases a nation’s exports. But in simple terms, we cannot say that due to currency depreciation, the TB will improve; actually, it depends on the elasticity of the product. If the product is more elastic and an increase in the ER leads to a decrease in export prices for foreign consumers, then the quantity of exports rises, increasing export revenue. Thus, this has a positive impact on TB. If the product is inelastic, an increase in the ER results in no increase in export earnings. Pakistan has a large trade deficit with Japan because its economy is dependent on Japanese imports. Our results thus contradict existing literature. We do not care whether the S-curve exists; we are just exploring the behavior. In this instance, all we were looking for was a pattern or shape in the curves. We merely observe how the ER responds to TB. The existence of the S-curve indicates that depreciation has a favorable effect on Pakistan’s TB. Conversely, its absence indicates a negative effect. Figure 2 provides a graphical representation and shows that out of 15 industries, only three have the S-curve validated. Based on the plot of the COR coefficients for the five leads and lag values, the S-curve exists in the following industries: (i) coffee, tea, cocoa, spices, and manufacturing; (ii) special transactions. According to the industry, textile fibers are not manufactured. Thus, the empirical results show that currency depreciation has a positive impact only on these three industries, whereas ER depreciation does not have a positive impact on all remaining industries included in the sample.

The previous literature argues that the absence of S-curve behavior in most industries indicates low ER elasticity, structural trade asymmetries, limited substitutability, and the dominance of non-price factors in Pakistan and Japan commodity trade. Thus, ER movements alone are insufficient to make the dynamic trade adjustments projected by the S-curve hypothesis at the industry level. Bahmani-Oskooee and Fariditavana (2014) also argue that due to low ER elasticity, contracts between countries, structural trade imbalance, and aggregation bias reduce the response of ER movements, making the S-curve less likely at the industry level. Similar studies also found that ER depreciation is not favorable for correcting TB in Pakistan. They found that minority industries exhibit S-curve behavior because traders react differently during currency appreciation and depreciation (Ahmad et al., 2023; Iqbal et al., 2023a, 2023b). Additionally, they reported that the effect of the ER on the TB depends on the elasticity of the product.

The empirical results based on plotted data demonstrated that currency depreciation is not a favorable policy instrument for correcting Pakistan’s trade deficit with Japan. As a substitute for relying on ER adjustments, the Government of Pakistan and policymakers should emphasize improving non-price competitiveness by attracting export quality, promoting product diversification, upgrading industrial capacity, and strengthening trade facilitation mechanisms. The main aim of policies to upgrade technology and deepen integration into Japanese value chains is likely to be more effective in enlightening Pakistan’s bilateral trade performance than ER depreciation alone.

Exchange rates serve as a barometer of the economy, reflecting the relative strength or weakness of the currency in developing economies. Pakistan’s economy suffers from a persistent trade deficit, and policymakers have used currency depreciation to address it. In this study, we explored the impact of the ER on TB in Pakistan, focusing on trade with Japan. We investigated how the trader responds when the ER changes. This study used trade data at the disaggregated level to capture the industry-specific impact of ER changes on TB, using the cross-correlation function. The results of the study show that the S-curve pattern is observed only in three industries out of 15. The S-curve exists in these industries, such as (i) coffee, tea, cocoa, spices and manufacturing, (ii) special transactions. According to the industry, textile fibers are not manufactured. Thus, the empirical results showed that currency depreciation has a positive impact only on these three industries, whereas ER depreciation does not have a positive impact on all remaining industries included in the sample. The existing literature suggests that TB may improve through appreciation or depreciation. It depends upon the elasticity of the product. If the product is more elastic and an increase in the ER leads to a decrease in export prices for foreign consumers, then the quantity of exports rises, increasing export revenue. Thus, this has a positive impact on TB. If the product is inelastic, an increase in the ER leads to no rise in export earnings.

The literature argues that the absence of S-curve behavior in most industries indicates low ER elasticity, structural trade asymmetries, limited substitutability, and the dominance of non-price factors in commodity trade between Pakistan and Japan. Thus, ER movements alone are insufficient to make the dynamic trade adjustments projected by the S-curve hypothesis at the industry level. Bahmani-Oskooee and Fariditavana (2014) also argued that due to low ER elasticity, contracts between countries, structural trade imbalance, and aggregation bias reduce the response of ER movements, making the S-curve less likely at the industry level. The evidence indicates that the trader’s response is different during currency appreciation and depreciation. In this sample, include large, durable, and non-durable goods, and small-scale industries. S-curve patterns found in these sectors of the economy align with accepted economic theories, most notably the Product Life Cycle Theory. Observed S-curve patterns may result from the confluence of factors, including product life-cycle dynamics, technological advancements, market maturity, global economic trends, innovation, and trade dynamics. Therefore, it is observed in the industries where there is the S-curve effect that J-curve effect also occurs, initially, the exchange rate increases the trade balance will deteriorate in the short run, and then gradually the trade balance will improve in the long run.